| 1 |
<!-- doc/src/sgml/ecpg.sgml --> |
| 2 |
|
| 3 |
<chapter id="ecpg"> |
| 4 |
<title><application>ECPG</application> — Embedded <acronym>SQL</acronym> in C</title> |
| 5 |
|
| 6 |
<indexterm zone="ecpg"><primary>embedded SQL</primary><secondary>in C</secondary></indexterm> |
| 7 |
<indexterm zone="ecpg"><primary>C</primary></indexterm> |
| 8 |
<indexterm zone="ecpg"><primary>ECPG</primary></indexterm> |
| 9 |
|
| 10 |
<para> |
| 11 |
This chapter describes the embedded <acronym>SQL</acronym> package |
| 12 |
for <productname>PostgreSQL</productname>. It was written by |
| 13 |
Linus Tolke (<email>[email protected]</email>) and Michael Meskes |
| 14 |
(<email>[email protected]</email>). Originally it was written to work with |
| 15 |
<acronym>C</acronym>. It also works with <acronym>C++</acronym>, but |
| 16 |
it does not recognize all <acronym>C++</acronym> constructs yet. |
| 17 |
</para> |
| 18 |
|
| 19 |
<para> |
| 20 |
This documentation is quite incomplete. But since this |
| 21 |
interface is standardized, additional information can be found in |
| 22 |
many resources about SQL. |
| 23 |
</para> |
| 24 |
|
| 25 |
<sect1 id="ecpg-concept"> |
| 26 |
<title>The Concept</title> |
| 27 |
|
| 28 |
<para> |
| 29 |
An embedded SQL program consists of code written in an ordinary |
| 30 |
programming language, in this case C, mixed with SQL commands in |
| 31 |
specially marked sections. To build the program, the source code (<filename>*.pgc</filename>) |
| 32 |
is first passed through the embedded SQL preprocessor, which converts it |
| 33 |
to an ordinary C program (<filename>*.c</filename>), and afterwards it can be processed by a C |
| 34 |
compiler. (For details about the compiling and linking see <xref linkend="ecpg-process"/>.) |
| 35 |
Converted ECPG applications call functions in the libpq library |
| 36 |
through the embedded SQL library (ecpglib), and communicate with |
| 37 |
the PostgreSQL server using the normal frontend-backend protocol. |
| 38 |
</para> |
| 39 |
|
| 40 |
<para> |
| 41 |
Embedded <acronym>SQL</acronym> has advantages over other methods |
| 42 |
for handling <acronym>SQL</acronym> commands from C code. First, it |
| 43 |
takes care of the tedious passing of information to and from |
| 44 |
variables in your <acronym>C</acronym> program. Second, the SQL |
| 45 |
code in the program is checked at build time for syntactical |
| 46 |
correctness. Third, embedded <acronym>SQL</acronym> in C is |
| 47 |
specified in the <acronym>SQL</acronym> standard and supported by |
| 48 |
many other <acronym>SQL</acronym> database systems. The |
| 49 |
<productname>PostgreSQL</productname> implementation is designed to match this |
| 50 |
standard as much as possible, and it is usually possible to port |
| 51 |
embedded <acronym>SQL</acronym> programs written for other SQL |
| 52 |
databases to <productname>PostgreSQL</productname> with relative |
| 53 |
ease. |
| 54 |
</para> |
| 55 |
|
| 56 |
<para> |
| 57 |
As already stated, programs written for the embedded |
| 58 |
<acronym>SQL</acronym> interface are normal C programs with special |
| 59 |
code inserted to perform database-related actions. This special |
| 60 |
code always has the form: |
| 61 |
<programlisting> |
| 62 |
EXEC SQL ...; |
| 63 |
</programlisting> |
| 64 |
These statements syntactically take the place of a C statement. |
| 65 |
Depending on the particular statement, they can appear at the |
| 66 |
global level or within a function. |
| 67 |
</para> |
| 68 |
|
| 69 |
<para> |
| 70 |
Embedded |
| 71 |
<acronym>SQL</acronym> statements follow the case-sensitivity rules of |
| 72 |
normal <acronym>SQL</acronym> code, and not those of C. Also they allow nested |
| 73 |
C-style comments as per the SQL standard. The C part of the |
| 74 |
program, however, follows the C standard of not accepting nested comments. |
| 75 |
Embedded <acronym>SQL</acronym> statements likewise use SQL rules, not |
| 76 |
C rules, for parsing quoted strings and identifiers. |
| 77 |
(See <xref linkend="sql-syntax-strings"/> and |
| 78 |
<xref linkend="sql-syntax-identifiers"/> respectively. Note that |
| 79 |
ECPG assumes that <varname>standard_conforming_strings</varname> |
| 80 |
is <literal>on</literal>.) |
| 81 |
Of course, the C part of the program follows C quoting rules. |
| 82 |
</para> |
| 83 |
|
| 84 |
<para> |
| 85 |
The following sections explain all the embedded SQL statements. |
| 86 |
</para> |
| 87 |
</sect1> |
| 88 |
|
| 89 |
<sect1 id="ecpg-connect"> |
| 90 |
<title>Managing Database Connections</title> |
| 91 |
|
| 92 |
<para> |
| 93 |
This section describes how to open, close, and switch database |
| 94 |
connections. |
| 95 |
</para> |
| 96 |
|
| 97 |
<sect2 id="ecpg-connecting"> |
| 98 |
<title>Connecting to the Database Server</title> |
| 99 |
|
| 100 |
<para> |
| 101 |
One connects to a database using the following statement: |
| 102 |
<programlisting> |
| 103 |
EXEC SQL CONNECT TO <replaceable>target</replaceable> <optional>AS <replaceable>connection-name</replaceable></optional> <optional>USER <replaceable>user-name</replaceable></optional>; |
| 104 |
</programlisting> |
| 105 |
The <replaceable>target</replaceable> can be specified in the |
| 106 |
following ways: |
| 107 |
|
| 108 |
<itemizedlist> |
| 109 |
<listitem> |
| 110 |
<simpara> |
| 111 |
<literal><replaceable>dbname</replaceable><optional>@<replaceable>hostname</replaceable></optional><optional>:<replaceable>port</replaceable></optional></literal> |
| 112 |
</simpara> |
| 113 |
</listitem> |
| 114 |
|
| 115 |
<listitem> |
| 116 |
<simpara> |
| 117 |
<literal>tcp:postgresql://<replaceable>hostname</replaceable><optional>:<replaceable>port</replaceable></optional><optional>/<replaceable>dbname</replaceable></optional><optional>?<replaceable>options</replaceable></optional></literal> |
| 118 |
</simpara> |
| 119 |
</listitem> |
| 120 |
|
| 121 |
<listitem> |
| 122 |
<simpara> |
| 123 |
<literal>unix:postgresql://localhost<optional>:<replaceable>port</replaceable></optional><optional>/<replaceable>dbname</replaceable></optional><optional>?<replaceable>options</replaceable></optional></literal> |
| 124 |
</simpara> |
| 125 |
</listitem> |
| 126 |
|
| 127 |
<listitem> |
| 128 |
<simpara> |
| 129 |
an SQL string literal containing one of the above forms |
| 130 |
</simpara> |
| 131 |
</listitem> |
| 132 |
|
| 133 |
<listitem> |
| 134 |
<simpara> |
| 135 |
a reference to a character variable containing one of the above forms (see examples) |
| 136 |
</simpara> |
| 137 |
</listitem> |
| 138 |
|
| 139 |
<listitem> |
| 140 |
<simpara> |
| 141 |
<literal>DEFAULT</literal> |
| 142 |
</simpara> |
| 143 |
</listitem> |
| 144 |
</itemizedlist> |
| 145 |
|
| 146 |
The connection target <literal>DEFAULT</literal> initiates a connection |
| 147 |
to the default database under the default user name. No separate |
| 148 |
user name or connection name can be specified in that case. |
| 149 |
</para> |
| 150 |
|
| 151 |
<para> |
| 152 |
If you specify the connection target directly (that is, not as a string |
| 153 |
literal or variable reference), then the components of the target are |
| 154 |
passed through normal SQL parsing; this means that, for example, |
| 155 |
the <replaceable>hostname</replaceable> must look like one or more SQL |
| 156 |
identifiers separated by dots, and those identifiers will be |
| 157 |
case-folded unless double-quoted. Values of |
| 158 |
any <replaceable>options</replaceable> must be SQL identifiers, |
| 159 |
integers, or variable references. Of course, you can put nearly |
| 160 |
anything into an SQL identifier by double-quoting it. |
| 161 |
In practice, it is probably less error-prone to use a (single-quoted) |
| 162 |
string literal or a variable reference than to write the connection |
| 163 |
target directly. |
| 164 |
</para> |
| 165 |
|
| 166 |
<para> |
| 167 |
There are also different ways to specify the user name: |
| 168 |
|
| 169 |
<itemizedlist> |
| 170 |
<listitem> |
| 171 |
<simpara> |
| 172 |
<literal><replaceable>username</replaceable></literal> |
| 173 |
</simpara> |
| 174 |
</listitem> |
| 175 |
|
| 176 |
<listitem> |
| 177 |
<simpara> |
| 178 |
<literal><replaceable>username</replaceable>/<replaceable>password</replaceable></literal> |
| 179 |
</simpara> |
| 180 |
</listitem> |
| 181 |
|
| 182 |
<listitem> |
| 183 |
<simpara> |
| 184 |
<literal><replaceable>username</replaceable> IDENTIFIED BY <replaceable>password</replaceable></literal> |
| 185 |
</simpara> |
| 186 |
</listitem> |
| 187 |
|
| 188 |
<listitem> |
| 189 |
<simpara> |
| 190 |
<literal><replaceable>username</replaceable> USING <replaceable>password</replaceable></literal> |
| 191 |
</simpara> |
| 192 |
</listitem> |
| 193 |
</itemizedlist> |
| 194 |
|
| 195 |
As above, the parameters <replaceable>username</replaceable> and |
| 196 |
<replaceable>password</replaceable> can be an SQL identifier, an |
| 197 |
SQL string literal, or a reference to a character variable. |
| 198 |
</para> |
| 199 |
|
| 200 |
<para> |
| 201 |
If the connection target includes any <replaceable>options</replaceable>, |
| 202 |
those consist of |
| 203 |
<literal><replaceable>keyword</replaceable>=<replaceable>value</replaceable></literal> |
| 204 |
specifications separated by ampersands (<literal>&</literal>). |
| 205 |
The allowed key words are the same ones recognized |
| 206 |
by <application>libpq</application> (see |
| 207 |
<xref linkend="libpq-paramkeywords"/>). Spaces are ignored before |
| 208 |
any <replaceable>keyword</replaceable> or <replaceable>value</replaceable>, |
| 209 |
though not within or after one. Note that there is no way to |
| 210 |
write <literal>&</literal> within a <replaceable>value</replaceable>. |
| 211 |
</para> |
| 212 |
|
| 213 |
<para> |
| 214 |
Notice that when specifying a socket connection |
| 215 |
(with the <literal>unix:</literal> prefix), the host name must be |
| 216 |
exactly <literal>localhost</literal>. To select a non-default |
| 217 |
socket directory, write the directory's pathname as the value of |
| 218 |
a <varname>host</varname> option in |
| 219 |
the <replaceable>options</replaceable> part of the target. |
| 220 |
</para> |
| 221 |
|
| 222 |
<para> |
| 223 |
The <replaceable>connection-name</replaceable> is used to handle |
| 224 |
multiple connections in one program. It can be omitted if a |
| 225 |
program uses only one connection. The most recently opened |
| 226 |
connection becomes the current connection, which is used by default |
| 227 |
when an SQL statement is to be executed (see later in this |
| 228 |
chapter). |
| 229 |
</para> |
| 230 |
|
| 231 |
<para> |
| 232 |
Here are some examples of <command>CONNECT</command> statements: |
| 233 |
<programlisting> |
| 234 |
EXEC SQL CONNECT TO [email protected]; |
| 235 |
|
| 236 |
EXEC SQL CONNECT TO tcp:postgresql://sql.mydomain.com/mydb AS myconnection USER john; |
| 237 |
|
| 238 |
EXEC SQL BEGIN DECLARE SECTION; |
| 239 |
const char *target = "[email protected]"; |
| 240 |
const char *user = "john"; |
| 241 |
const char *passwd = "secret"; |
| 242 |
EXEC SQL END DECLARE SECTION; |
| 243 |
... |
| 244 |
EXEC SQL CONNECT TO :target USER :user USING :passwd; |
| 245 |
/* or EXEC SQL CONNECT TO :target USER :user/:passwd; */ |
| 246 |
</programlisting> |
| 247 |
The last example makes use of the feature referred to above as |
| 248 |
character variable references. You will see in later sections how C |
| 249 |
variables can be used in SQL statements when you prefix them with a |
| 250 |
colon. |
| 251 |
</para> |
| 252 |
|
| 253 |
<para> |
| 254 |
Be advised that the format of the connection target is not |
| 255 |
specified in the SQL standard. So if you want to develop portable |
| 256 |
applications, you might want to use something based on the last |
| 257 |
example above to encapsulate the connection target string |
| 258 |
somewhere. |
| 259 |
</para> |
| 260 |
|
| 261 |
<para> |
| 262 |
If untrusted users have access to a database that has not adopted a |
| 263 |
<link linkend="ddl-schemas-patterns">secure schema usage pattern</link>, |
| 264 |
begin each session by removing publicly-writable schemas |
| 265 |
from <varname>search_path</varname>. For example, |
| 266 |
add <literal>options=-c search_path=</literal> |
| 267 |
to <literal><replaceable>options</replaceable></literal>, or |
| 268 |
issue <literal>EXEC SQL SELECT pg_catalog.set_config('search_path', '', |
| 269 |
false);</literal> after connecting. This consideration is not specific to |
| 270 |
ECPG; it applies to every interface for executing arbitrary SQL commands. |
| 271 |
</para> |
| 272 |
</sect2> |
| 273 |
|
| 274 |
<sect2 id="ecpg-set-connection"> |
| 275 |
<title>Choosing a Connection</title> |
| 276 |
|
| 277 |
<para> |
| 278 |
SQL statements in embedded SQL programs are by default executed on |
| 279 |
the current connection, that is, the most recently opened one. If |
| 280 |
an application needs to manage multiple connections, then there are |
| 281 |
three ways to handle this. |
| 282 |
</para> |
| 283 |
|
| 284 |
<para> |
| 285 |
The first option is to explicitly choose a connection for each SQL |
| 286 |
statement, for example: |
| 287 |
<programlisting> |
| 288 |
EXEC SQL AT <replaceable>connection-name</replaceable> SELECT ...; |
| 289 |
</programlisting> |
| 290 |
This option is particularly suitable if the application needs to |
| 291 |
use several connections in mixed order. |
| 292 |
</para> |
| 293 |
|
| 294 |
<para> |
| 295 |
If your application uses multiple threads of execution, they cannot share a |
| 296 |
connection concurrently. You must either explicitly control access to the connection |
| 297 |
(using mutexes) or use a connection for each thread. |
| 298 |
</para> |
| 299 |
|
| 300 |
<para> |
| 301 |
The second option is to execute a statement to switch the current |
| 302 |
connection. That statement is: |
| 303 |
<programlisting> |
| 304 |
EXEC SQL SET CONNECTION <replaceable>connection-name</replaceable>; |
| 305 |
</programlisting> |
| 306 |
This option is particularly convenient if many statements are to be |
| 307 |
executed on the same connection. |
| 308 |
</para> |
| 309 |
|
| 310 |
<para> |
| 311 |
Here is an example program managing multiple database connections: |
| 312 |
<programlisting><![CDATA[ |
| 313 |
#include <stdio.h> |
| 314 |
|
| 315 |
EXEC SQL BEGIN DECLARE SECTION; |
| 316 |
char dbname[1024]; |
| 317 |
EXEC SQL END DECLARE SECTION; |
| 318 |
|
| 319 |
int |
| 320 |
main() |
| 321 |
{ |
| 322 |
EXEC SQL CONNECT TO testdb1 AS con1 USER testuser; |
| 323 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 324 |
EXEC SQL CONNECT TO testdb2 AS con2 USER testuser; |
| 325 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 326 |
EXEC SQL CONNECT TO testdb3 AS con3 USER testuser; |
| 327 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 328 |
|
| 329 |
/* This query would be executed in the last opened database "testdb3". */ |
| 330 |
EXEC SQL SELECT current_database() INTO :dbname; |
| 331 |
printf("current=%s (should be testdb3)\n", dbname); |
| 332 |
|
| 333 |
/* Using "AT" to run a query in "testdb2" */ |
| 334 |
EXEC SQL AT con2 SELECT current_database() INTO :dbname; |
| 335 |
printf("current=%s (should be testdb2)\n", dbname); |
| 336 |
|
| 337 |
/* Switch the current connection to "testdb1". */ |
| 338 |
EXEC SQL SET CONNECTION con1; |
| 339 |
|
| 340 |
EXEC SQL SELECT current_database() INTO :dbname; |
| 341 |
printf("current=%s (should be testdb1)\n", dbname); |
| 342 |
|
| 343 |
EXEC SQL DISCONNECT ALL; |
| 344 |
return 0; |
| 345 |
} |
| 346 |
]]></programlisting> |
| 347 |
|
| 348 |
This example would produce this output: |
| 349 |
<screen> |
| 350 |
current=testdb3 (should be testdb3) |
| 351 |
current=testdb2 (should be testdb2) |
| 352 |
current=testdb1 (should be testdb1) |
| 353 |
</screen> |
| 354 |
</para> |
| 355 |
|
| 356 |
<para> |
| 357 |
The third option is to declare an SQL identifier linked to |
| 358 |
the connection, for example: |
| 359 |
<programlisting> |
| 360 |
EXEC SQL AT <replaceable>connection-name</replaceable> DECLARE <replaceable>statement-name</replaceable> STATEMENT; |
| 361 |
EXEC SQL PREPARE <replaceable>statement-name</replaceable> FROM :<replaceable>dyn-string</replaceable>; |
| 362 |
</programlisting> |
| 363 |
Once you link an SQL identifier to a connection, you execute dynamic SQL |
| 364 |
without an AT clause. Note that this option behaves like preprocessor |
| 365 |
directives, therefore the link is enabled only in the file. |
| 366 |
</para> |
| 367 |
<para> |
| 368 |
Here is an example program using this option: |
| 369 |
<programlisting><![CDATA[ |
| 370 |
#include <stdio.h> |
| 371 |
|
| 372 |
EXEC SQL BEGIN DECLARE SECTION; |
| 373 |
char dbname[128]; |
| 374 |
char *dyn_sql = "SELECT current_database()"; |
| 375 |
EXEC SQL END DECLARE SECTION; |
| 376 |
|
| 377 |
int main(){ |
| 378 |
EXEC SQL CONNECT TO postgres AS con1; |
| 379 |
EXEC SQL CONNECT TO testdb AS con2; |
| 380 |
EXEC SQL AT con1 DECLARE stmt STATEMENT; |
| 381 |
EXEC SQL PREPARE stmt FROM :dyn_sql; |
| 382 |
EXEC SQL EXECUTE stmt INTO :dbname; |
| 383 |
printf("%s\n", dbname); |
| 384 |
|
| 385 |
EXEC SQL DISCONNECT ALL; |
| 386 |
return 0; |
| 387 |
} |
| 388 |
]]></programlisting> |
| 389 |
|
| 390 |
This example would produce this output, even if the default connection is testdb: |
| 391 |
<screen> |
| 392 |
postgres |
| 393 |
</screen> |
| 394 |
</para> |
| 395 |
</sect2> |
| 396 |
|
| 397 |
<sect2 id="ecpg-disconnect"> |
| 398 |
<title>Closing a Connection</title> |
| 399 |
|
| 400 |
<para> |
| 401 |
To close a connection, use the following statement: |
| 402 |
<programlisting> |
| 403 |
EXEC SQL DISCONNECT <optional><replaceable>connection</replaceable></optional>; |
| 404 |
</programlisting> |
| 405 |
The <replaceable>connection</replaceable> can be specified |
| 406 |
in the following ways: |
| 407 |
|
| 408 |
<itemizedlist> |
| 409 |
<listitem> |
| 410 |
<simpara> |
| 411 |
<literal><replaceable>connection-name</replaceable></literal> |
| 412 |
</simpara> |
| 413 |
</listitem> |
| 414 |
|
| 415 |
<listitem> |
| 416 |
<simpara> |
| 417 |
<literal>CURRENT</literal> |
| 418 |
</simpara> |
| 419 |
</listitem> |
| 420 |
|
| 421 |
<listitem> |
| 422 |
<simpara> |
| 423 |
<literal>ALL</literal> |
| 424 |
</simpara> |
| 425 |
</listitem> |
| 426 |
</itemizedlist> |
| 427 |
|
| 428 |
If no connection name is specified, the current connection is |
| 429 |
closed. |
| 430 |
</para> |
| 431 |
|
| 432 |
<para> |
| 433 |
It is good style that an application always explicitly disconnect |
| 434 |
from every connection it opened. |
| 435 |
</para> |
| 436 |
</sect2> |
| 437 |
|
| 438 |
</sect1> |
| 439 |
|
| 440 |
<sect1 id="ecpg-commands"> |
| 441 |
<title>Running SQL Commands</title> |
| 442 |
|
| 443 |
<para> |
| 444 |
Any SQL command can be run from within an embedded SQL application. |
| 445 |
Below are some examples of how to do that. |
| 446 |
</para> |
| 447 |
|
| 448 |
<sect2 id="ecpg-executing"> |
| 449 |
<title>Executing SQL Statements</title> |
| 450 |
|
| 451 |
<para> |
| 452 |
Creating a table: |
| 453 |
<programlisting> |
| 454 |
EXEC SQL CREATE TABLE foo (number integer, ascii char(16)); |
| 455 |
EXEC SQL CREATE UNIQUE INDEX num1 ON foo(number); |
| 456 |
EXEC SQL COMMIT; |
| 457 |
</programlisting> |
| 458 |
</para> |
| 459 |
|
| 460 |
<para> |
| 461 |
Inserting rows: |
| 462 |
<programlisting> |
| 463 |
EXEC SQL INSERT INTO foo (number, ascii) VALUES (9999, 'doodad'); |
| 464 |
EXEC SQL COMMIT; |
| 465 |
</programlisting> |
| 466 |
</para> |
| 467 |
|
| 468 |
<para> |
| 469 |
Deleting rows: |
| 470 |
<programlisting> |
| 471 |
EXEC SQL DELETE FROM foo WHERE number = 9999; |
| 472 |
EXEC SQL COMMIT; |
| 473 |
</programlisting> |
| 474 |
</para> |
| 475 |
|
| 476 |
<para> |
| 477 |
Updates: |
| 478 |
<programlisting> |
| 479 |
EXEC SQL UPDATE foo |
| 480 |
SET ascii = 'foobar' |
| 481 |
WHERE number = 9999; |
| 482 |
EXEC SQL COMMIT; |
| 483 |
</programlisting> |
| 484 |
</para> |
| 485 |
|
| 486 |
<para> |
| 487 |
<literal>SELECT</literal> statements that return a single result |
| 488 |
row can also be executed using |
| 489 |
<literal>EXEC SQL</literal> directly. To handle result sets with |
| 490 |
multiple rows, an application has to use a cursor; |
| 491 |
see <xref linkend="ecpg-cursors"/> below. (As a special case, an |
| 492 |
application can fetch multiple rows at once into an array host |
| 493 |
variable; see <xref linkend="ecpg-variables-arrays"/>.) |
| 494 |
</para> |
| 495 |
|
| 496 |
<para> |
| 497 |
Single-row select: |
| 498 |
<programlisting> |
| 499 |
EXEC SQL SELECT foo INTO :FooBar FROM table1 WHERE ascii = 'doodad'; |
| 500 |
</programlisting> |
| 501 |
</para> |
| 502 |
|
| 503 |
<para> |
| 504 |
Also, a configuration parameter can be retrieved with the |
| 505 |
<literal>SHOW</literal> command: |
| 506 |
<programlisting> |
| 507 |
EXEC SQL SHOW search_path INTO :var; |
| 508 |
</programlisting> |
| 509 |
</para> |
| 510 |
|
| 511 |
<para> |
| 512 |
The tokens of the form |
| 513 |
<literal>:<replaceable>something</replaceable></literal> are |
| 514 |
<firstterm>host variables</firstterm>, that is, they refer to |
| 515 |
variables in the C program. They are explained in <xref |
| 516 |
linkend="ecpg-variables"/>. |
| 517 |
</para> |
| 518 |
</sect2> |
| 519 |
|
| 520 |
<sect2 id="ecpg-cursors"> |
| 521 |
<title>Using Cursors</title> |
| 522 |
|
| 523 |
<para> |
| 524 |
To retrieve a result set holding multiple rows, an application has |
| 525 |
to declare a cursor and fetch each row from the cursor. The steps |
| 526 |
to use a cursor are the following: declare a cursor, open it, fetch |
| 527 |
a row from the cursor, repeat, and finally close it. |
| 528 |
</para> |
| 529 |
|
| 530 |
<para> |
| 531 |
Select using cursors: |
| 532 |
<programlisting> |
| 533 |
EXEC SQL DECLARE foo_bar CURSOR FOR |
| 534 |
SELECT number, ascii FROM foo |
| 535 |
ORDER BY ascii; |
| 536 |
EXEC SQL OPEN foo_bar; |
| 537 |
EXEC SQL FETCH foo_bar INTO :FooBar, DooDad; |
| 538 |
... |
| 539 |
EXEC SQL CLOSE foo_bar; |
| 540 |
EXEC SQL COMMIT; |
| 541 |
</programlisting> |
| 542 |
</para> |
| 543 |
|
| 544 |
<para> |
| 545 |
For more details about declaring a cursor, see <xref |
| 546 |
linkend="ecpg-sql-declare"/>; for more details about fetching rows from a |
| 547 |
cursor, see <xref linkend="sql-fetch"/>. |
| 548 |
</para> |
| 549 |
|
| 550 |
<note> |
| 551 |
<para> |
| 552 |
The ECPG <command>DECLARE</command> command does not actually |
| 553 |
cause a statement to be sent to the PostgreSQL backend. The |
| 554 |
cursor is opened in the backend (using the |
| 555 |
backend's <command>DECLARE</command> command) at the point when |
| 556 |
the <command>OPEN</command> command is executed. |
| 557 |
</para> |
| 558 |
</note> |
| 559 |
</sect2> |
| 560 |
|
| 561 |
<sect2 id="ecpg-transactions"> |
| 562 |
<title>Managing Transactions</title> |
| 563 |
|
| 564 |
<para> |
| 565 |
In the default mode, statements are committed only when |
| 566 |
<command>EXEC SQL COMMIT</command> is issued. The embedded SQL |
| 567 |
interface also supports autocommit of transactions (similar to |
| 568 |
<application>psql</application>'s default behavior) via the <option>-t</option> |
| 569 |
command-line option to <command>ecpg</command> (see <xref |
| 570 |
linkend="app-ecpg"/>) or via the <literal>EXEC SQL SET AUTOCOMMIT TO |
| 571 |
ON</literal> statement. In autocommit mode, each command is |
| 572 |
automatically committed unless it is inside an explicit transaction |
| 573 |
block. This mode can be explicitly turned off using <literal>EXEC |
| 574 |
SQL SET AUTOCOMMIT TO OFF</literal>. |
| 575 |
</para> |
| 576 |
|
| 577 |
<para> |
| 578 |
The following transaction management commands are available: |
| 579 |
|
| 580 |
<variablelist> |
| 581 |
<varlistentry id="ecpg-transactions-exec-sql-commit"> |
| 582 |
<term><literal>EXEC SQL COMMIT</literal></term> |
| 583 |
<listitem> |
| 584 |
<para> |
| 585 |
Commit an in-progress transaction. |
| 586 |
</para> |
| 587 |
</listitem> |
| 588 |
</varlistentry> |
| 589 |
|
| 590 |
<varlistentry id="ecpg-transactions-exec-sql-rollback"> |
| 591 |
<term><literal>EXEC SQL ROLLBACK</literal></term> |
| 592 |
<listitem> |
| 593 |
<para> |
| 594 |
Roll back an in-progress transaction. |
| 595 |
</para> |
| 596 |
</listitem> |
| 597 |
</varlistentry> |
| 598 |
|
| 599 |
<varlistentry id="ecpg-transactions-exec-sql-prepare-transaction"> |
| 600 |
<term><literal>EXEC SQL PREPARE TRANSACTION </literal><replaceable class="parameter">transaction_id</replaceable></term> |
| 601 |
<listitem> |
| 602 |
<para> |
| 603 |
Prepare the current transaction for two-phase commit. |
| 604 |
</para> |
| 605 |
</listitem> |
| 606 |
</varlistentry> |
| 607 |
|
| 608 |
<varlistentry id="ecpg-transactions-exec-sql-commit-prepared"> |
| 609 |
<term><literal>EXEC SQL COMMIT PREPARED </literal><replaceable class="parameter">transaction_id</replaceable></term> |
| 610 |
<listitem> |
| 611 |
<para> |
| 612 |
Commit a transaction that is in prepared state. |
| 613 |
</para> |
| 614 |
</listitem> |
| 615 |
</varlistentry> |
| 616 |
|
| 617 |
<varlistentry id="ecpg-transactions-exec-sql-rollback-prepared"> |
| 618 |
<term><literal>EXEC SQL ROLLBACK PREPARED </literal><replaceable class="parameter">transaction_id</replaceable></term> |
| 619 |
<listitem> |
| 620 |
<para> |
| 621 |
Roll back a transaction that is in prepared state. |
| 622 |
</para> |
| 623 |
</listitem> |
| 624 |
</varlistentry> |
| 625 |
|
| 626 |
<varlistentry id="ecpg-transactions-exec-sql-autocommit-on"> |
| 627 |
<term><literal>EXEC SQL SET AUTOCOMMIT TO ON</literal></term> |
| 628 |
<listitem> |
| 629 |
<para> |
| 630 |
Enable autocommit mode. |
| 631 |
</para> |
| 632 |
</listitem> |
| 633 |
</varlistentry> |
| 634 |
|
| 635 |
<varlistentry id="ecpg-transactions-exec-sql-autocommit-off"> |
| 636 |
<term><literal>EXEC SQL SET AUTOCOMMIT TO OFF</literal></term> |
| 637 |
<listitem> |
| 638 |
<para> |
| 639 |
Disable autocommit mode. This is the default. |
| 640 |
</para> |
| 641 |
</listitem> |
| 642 |
</varlistentry> |
| 643 |
</variablelist> |
| 644 |
</para> |
| 645 |
</sect2> |
| 646 |
|
| 647 |
<sect2 id="ecpg-prepared"> |
| 648 |
<title>Prepared Statements</title> |
| 649 |
|
| 650 |
<para> |
| 651 |
When the values to be passed to an SQL statement are not known at |
| 652 |
compile time, or the same statement is going to be used many |
| 653 |
times, then prepared statements can be useful. |
| 654 |
</para> |
| 655 |
|
| 656 |
<para> |
| 657 |
The statement is prepared using the |
| 658 |
command <literal>PREPARE</literal>. For the values that are not |
| 659 |
known yet, use the |
| 660 |
placeholder <quote><literal>?</literal></quote>: |
| 661 |
<programlisting> |
| 662 |
EXEC SQL PREPARE stmt1 FROM "SELECT oid, datname FROM pg_database WHERE oid = ?"; |
| 663 |
</programlisting> |
| 664 |
</para> |
| 665 |
|
| 666 |
<para> |
| 667 |
If a statement returns a single row, the application can |
| 668 |
call <literal>EXECUTE</literal> after |
| 669 |
<literal>PREPARE</literal> to execute the statement, supplying the |
| 670 |
actual values for the placeholders with a <literal>USING</literal> |
| 671 |
clause: |
| 672 |
<programlisting> |
| 673 |
EXEC SQL EXECUTE stmt1 INTO :dboid, :dbname USING 1; |
| 674 |
</programlisting> |
| 675 |
</para> |
| 676 |
|
| 677 |
<para> |
| 678 |
If a statement returns multiple rows, the application can use a |
| 679 |
cursor declared based on the prepared statement. To bind input |
| 680 |
parameters, the cursor must be opened with |
| 681 |
a <literal>USING</literal> clause: |
| 682 |
<programlisting> |
| 683 |
EXEC SQL PREPARE stmt1 FROM "SELECT oid,datname FROM pg_database WHERE oid > ?"; |
| 684 |
EXEC SQL DECLARE foo_bar CURSOR FOR stmt1; |
| 685 |
|
| 686 |
/* when end of result set reached, break out of while loop */ |
| 687 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 688 |
|
| 689 |
EXEC SQL OPEN foo_bar USING 100; |
| 690 |
... |
| 691 |
while (1) |
| 692 |
{ |
| 693 |
EXEC SQL FETCH NEXT FROM foo_bar INTO :dboid, :dbname; |
| 694 |
... |
| 695 |
} |
| 696 |
EXEC SQL CLOSE foo_bar; |
| 697 |
</programlisting> |
| 698 |
</para> |
| 699 |
|
| 700 |
<para> |
| 701 |
When you don't need the prepared statement anymore, you should |
| 702 |
deallocate it: |
| 703 |
<programlisting> |
| 704 |
EXEC SQL DEALLOCATE PREPARE <replaceable>name</replaceable>; |
| 705 |
</programlisting> |
| 706 |
</para> |
| 707 |
|
| 708 |
<para> |
| 709 |
For more details about <literal>PREPARE</literal>, |
| 710 |
see <xref linkend="ecpg-sql-prepare"/>. Also |
| 711 |
see <xref linkend="ecpg-dynamic"/> for more details about using |
| 712 |
placeholders and input parameters. |
| 713 |
</para> |
| 714 |
</sect2> |
| 715 |
</sect1> |
| 716 |
|
| 717 |
<sect1 id="ecpg-variables"> |
| 718 |
<title>Using Host Variables</title> |
| 719 |
|
| 720 |
<para> |
| 721 |
In <xref linkend="ecpg-commands"/> you saw how you can execute SQL |
| 722 |
statements from an embedded SQL program. Some of those statements |
| 723 |
only used fixed values and did not provide a way to insert |
| 724 |
user-supplied values into statements or have the program process |
| 725 |
the values returned by the query. Those kinds of statements are |
| 726 |
not really useful in real applications. This section explains in |
| 727 |
detail how you can pass data between your C program and the |
| 728 |
embedded SQL statements using a simple mechanism called |
| 729 |
<firstterm>host variables</firstterm>. In an embedded SQL program we |
| 730 |
consider the SQL statements to be <firstterm>guests</firstterm> in the C |
| 731 |
program code which is the <firstterm>host language</firstterm>. Therefore |
| 732 |
the variables of the C program are called <firstterm>host |
| 733 |
variables</firstterm>. |
| 734 |
</para> |
| 735 |
|
| 736 |
<para> |
| 737 |
Another way to exchange values between PostgreSQL backends and ECPG |
| 738 |
applications is the use of SQL descriptors, described |
| 739 |
in <xref linkend="ecpg-descriptors"/>. |
| 740 |
</para> |
| 741 |
|
| 742 |
<sect2 id="ecpg-variables-overview"> |
| 743 |
<title>Overview</title> |
| 744 |
|
| 745 |
<para> |
| 746 |
Passing data between the C program and the SQL statements is |
| 747 |
particularly simple in embedded SQL. Instead of having the |
| 748 |
program paste the data into the statement, which entails various |
| 749 |
complications, such as properly quoting the value, you can simply |
| 750 |
write the name of a C variable into the SQL statement, prefixed by |
| 751 |
a colon. For example: |
| 752 |
<programlisting> |
| 753 |
EXEC SQL INSERT INTO sometable VALUES (:v1, 'foo', :v2); |
| 754 |
</programlisting> |
| 755 |
This statement refers to two C variables named |
| 756 |
<varname>v1</varname> and <varname>v2</varname> and also uses a |
| 757 |
regular SQL string literal, to illustrate that you are not |
| 758 |
restricted to use one kind of data or the other. |
| 759 |
</para> |
| 760 |
|
| 761 |
<para> |
| 762 |
This style of inserting C variables in SQL statements works |
| 763 |
anywhere a value expression is expected in an SQL statement. |
| 764 |
</para> |
| 765 |
</sect2> |
| 766 |
|
| 767 |
<sect2 id="ecpg-declare-sections"> |
| 768 |
<title>Declare Sections</title> |
| 769 |
|
| 770 |
<para> |
| 771 |
To pass data from the program to the database, for example as |
| 772 |
parameters in a query, or to pass data from the database back to |
| 773 |
the program, the C variables that are intended to contain this |
| 774 |
data need to be declared in specially marked sections, so the |
| 775 |
embedded SQL preprocessor is made aware of them. |
| 776 |
</para> |
| 777 |
|
| 778 |
<para> |
| 779 |
This section starts with: |
| 780 |
<programlisting> |
| 781 |
EXEC SQL BEGIN DECLARE SECTION; |
| 782 |
</programlisting> |
| 783 |
and ends with: |
| 784 |
<programlisting> |
| 785 |
EXEC SQL END DECLARE SECTION; |
| 786 |
</programlisting> |
| 787 |
Between those lines, there must be normal C variable declarations, |
| 788 |
such as: |
| 789 |
<programlisting> |
| 790 |
int x = 4; |
| 791 |
char foo[16], bar[16]; |
| 792 |
</programlisting> |
| 793 |
As you can see, you can optionally assign an initial value to the variable. |
| 794 |
The variable's scope is determined by the location of its declaring |
| 795 |
section within the program. |
| 796 |
You can also declare variables with the following syntax which implicitly |
| 797 |
creates a declare section: |
| 798 |
<programlisting> |
| 799 |
EXEC SQL int i = 4; |
| 800 |
</programlisting> |
| 801 |
You can have as many declare sections in a program as you like. |
| 802 |
</para> |
| 803 |
|
| 804 |
<para> |
| 805 |
The declarations are also echoed to the output file as normal C |
| 806 |
variables, so there's no need to declare them again. Variables |
| 807 |
that are not intended to be used in SQL commands can be declared |
| 808 |
normally outside these special sections. |
| 809 |
</para> |
| 810 |
|
| 811 |
<para> |
| 812 |
The definition of a structure or union also must be listed inside |
| 813 |
a <literal>DECLARE</literal> section. Otherwise the preprocessor cannot |
| 814 |
handle these types since it does not know the definition. |
| 815 |
</para> |
| 816 |
</sect2> |
| 817 |
|
| 818 |
<sect2 id="ecpg-retrieving"> |
| 819 |
<title>Retrieving Query Results</title> |
| 820 |
|
| 821 |
<para> |
| 822 |
Now you should be able to pass data generated by your program into |
| 823 |
an SQL command. But how do you retrieve the results of a query? |
| 824 |
For that purpose, embedded SQL provides special variants of the |
| 825 |
usual commands <command>SELECT</command> and |
| 826 |
<command>FETCH</command>. These commands have a special |
| 827 |
<literal>INTO</literal> clause that specifies which host variables |
| 828 |
the retrieved values are to be stored in. |
| 829 |
<command>SELECT</command> is used for a query that returns only |
| 830 |
single row, and <command>FETCH</command> is used for a query that |
| 831 |
returns multiple rows, using a cursor. |
| 832 |
</para> |
| 833 |
|
| 834 |
<para> |
| 835 |
Here is an example: |
| 836 |
<programlisting> |
| 837 |
/* |
| 838 |
* assume this table: |
| 839 |
* CREATE TABLE test1 (a int, b varchar(50)); |
| 840 |
*/ |
| 841 |
|
| 842 |
EXEC SQL BEGIN DECLARE SECTION; |
| 843 |
int v1; |
| 844 |
VARCHAR v2; |
| 845 |
EXEC SQL END DECLARE SECTION; |
| 846 |
|
| 847 |
... |
| 848 |
|
| 849 |
EXEC SQL SELECT a, b INTO :v1, :v2 FROM test; |
| 850 |
</programlisting> |
| 851 |
So the <literal>INTO</literal> clause appears between the select |
| 852 |
list and the <literal>FROM</literal> clause. The number of |
| 853 |
elements in the select list and the list after |
| 854 |
<literal>INTO</literal> (also called the target list) must be |
| 855 |
equal. |
| 856 |
</para> |
| 857 |
|
| 858 |
<para> |
| 859 |
Here is an example using the command <command>FETCH</command>: |
| 860 |
<programlisting> |
| 861 |
EXEC SQL BEGIN DECLARE SECTION; |
| 862 |
int v1; |
| 863 |
VARCHAR v2; |
| 864 |
EXEC SQL END DECLARE SECTION; |
| 865 |
|
| 866 |
... |
| 867 |
|
| 868 |
EXEC SQL DECLARE foo CURSOR FOR SELECT a, b FROM test; |
| 869 |
|
| 870 |
... |
| 871 |
|
| 872 |
do |
| 873 |
{ |
| 874 |
... |
| 875 |
EXEC SQL FETCH NEXT FROM foo INTO :v1, :v2; |
| 876 |
... |
| 877 |
} while (...); |
| 878 |
</programlisting> |
| 879 |
Here the <literal>INTO</literal> clause appears after all the |
| 880 |
normal clauses. |
| 881 |
</para> |
| 882 |
|
| 883 |
</sect2> |
| 884 |
|
| 885 |
<sect2 id="ecpg-variables-type-mapping"> |
| 886 |
<title>Type Mapping</title> |
| 887 |
|
| 888 |
<para> |
| 889 |
When ECPG applications exchange values between the PostgreSQL |
| 890 |
server and the C application, such as when retrieving query |
| 891 |
results from the server or executing SQL statements with input |
| 892 |
parameters, the values need to be converted between PostgreSQL |
| 893 |
data types and host language variable types (C language data |
| 894 |
types, concretely). One of the main points of ECPG is that it |
| 895 |
takes care of this automatically in most cases. |
| 896 |
</para> |
| 897 |
|
| 898 |
<para> |
| 899 |
In this respect, there are two kinds of data types: Some simple |
| 900 |
PostgreSQL data types, such as <type>integer</type> |
| 901 |
and <type>text</type>, can be read and written by the application |
| 902 |
directly. Other PostgreSQL data types, such |
| 903 |
as <type>timestamp</type> and <type>numeric</type> can only be |
| 904 |
accessed through special library functions; see |
| 905 |
<xref linkend="ecpg-special-types"/>. |
| 906 |
</para> |
| 907 |
|
| 908 |
<para> |
| 909 |
<xref linkend="ecpg-datatype-hostvars-table"/> shows which PostgreSQL |
| 910 |
data types correspond to which C data types. When you wish to |
| 911 |
send or receive a value of a given PostgreSQL data type, you |
| 912 |
should declare a C variable of the corresponding C data type in |
| 913 |
the declare section. |
| 914 |
</para> |
| 915 |
|
| 916 |
<table id="ecpg-datatype-hostvars-table"> |
| 917 |
<title>Mapping Between PostgreSQL Data Types and C Variable Types</title> |
| 918 |
<tgroup cols="2"> |
| 919 |
<thead> |
| 920 |
<row> |
| 921 |
<entry>PostgreSQL data type</entry> |
| 922 |
<entry>Host variable type</entry> |
| 923 |
</row> |
| 924 |
</thead> |
| 925 |
|
| 926 |
<tbody> |
| 927 |
<row> |
| 928 |
<entry><type>smallint</type></entry> |
| 929 |
<entry><type>short</type></entry> |
| 930 |
</row> |
| 931 |
|
| 932 |
<row> |
| 933 |
<entry><type>integer</type></entry> |
| 934 |
<entry><type>int</type></entry> |
| 935 |
</row> |
| 936 |
|
| 937 |
<row> |
| 938 |
<entry><type>bigint</type></entry> |
| 939 |
<entry><type>long long int</type></entry> |
| 940 |
</row> |
| 941 |
|
| 942 |
<row> |
| 943 |
<entry><type>decimal</type></entry> |
| 944 |
<entry><type>decimal</type><footnote id="ecpg-datatype-table-fn"><para>This type can only be accessed through special library functions; see <xref linkend="ecpg-special-types"/>.</para></footnote></entry> |
| 945 |
</row> |
| 946 |
|
| 947 |
<row> |
| 948 |
<entry><type>numeric</type></entry> |
| 949 |
<entry><type>numeric</type><footnoteref linkend="ecpg-datatype-table-fn"/></entry> |
| 950 |
</row> |
| 951 |
|
| 952 |
<row> |
| 953 |
<entry><type>real</type></entry> |
| 954 |
<entry><type>float</type></entry> |
| 955 |
</row> |
| 956 |
|
| 957 |
<row> |
| 958 |
<entry><type>double precision</type></entry> |
| 959 |
<entry><type>double</type></entry> |
| 960 |
</row> |
| 961 |
|
| 962 |
<row> |
| 963 |
<entry><type>smallserial</type></entry> |
| 964 |
<entry><type>short</type></entry> |
| 965 |
</row> |
| 966 |
|
| 967 |
<row> |
| 968 |
<entry><type>serial</type></entry> |
| 969 |
<entry><type>int</type></entry> |
| 970 |
</row> |
| 971 |
|
| 972 |
<row> |
| 973 |
<entry><type>bigserial</type></entry> |
| 974 |
<entry><type>long long int</type></entry> |
| 975 |
</row> |
| 976 |
|
| 977 |
<row> |
| 978 |
<entry><type>oid</type></entry> |
| 979 |
<entry><type>unsigned int</type></entry> |
| 980 |
</row> |
| 981 |
|
| 982 |
<row> |
| 983 |
<entry><type>character(<replaceable>n</replaceable>)</type>, <type>varchar(<replaceable>n</replaceable>)</type>, <type>text</type></entry> |
| 984 |
<entry><type>char[<replaceable>n</replaceable>+1]</type>, <type>VARCHAR[<replaceable>n</replaceable>+1]</type></entry> |
| 985 |
</row> |
| 986 |
|
| 987 |
<row> |
| 988 |
<entry><type>name</type></entry> |
| 989 |
<entry><type>char[NAMEDATALEN]</type></entry> |
| 990 |
</row> |
| 991 |
|
| 992 |
<row> |
| 993 |
<entry><type>timestamp</type></entry> |
| 994 |
<entry><type>timestamp</type><footnoteref linkend="ecpg-datatype-table-fn"/></entry> |
| 995 |
</row> |
| 996 |
|
| 997 |
<row> |
| 998 |
<entry><type>interval</type></entry> |
| 999 |
<entry><type>interval</type><footnoteref linkend="ecpg-datatype-table-fn"/></entry> |
| 1000 |
</row> |
| 1001 |
|
| 1002 |
<row> |
| 1003 |
<entry><type>date</type></entry> |
| 1004 |
<entry><type>date</type><footnoteref linkend="ecpg-datatype-table-fn"/></entry> |
| 1005 |
</row> |
| 1006 |
|
| 1007 |
<row> |
| 1008 |
<entry><type>boolean</type></entry> |
| 1009 |
<entry><type>bool</type><footnote><para>declared in <filename>ecpglib.h</filename> if not native</para></footnote></entry> |
| 1010 |
</row> |
| 1011 |
|
| 1012 |
<row> |
| 1013 |
<entry><type>bytea</type></entry> |
| 1014 |
<entry><type>char *</type>, <type>bytea[<replaceable>n</replaceable>]</type></entry> |
| 1015 |
</row> |
| 1016 |
</tbody> |
| 1017 |
</tgroup> |
| 1018 |
</table> |
| 1019 |
|
| 1020 |
<sect3 id="ecpg-char"> |
| 1021 |
<title>Handling Character Strings</title> |
| 1022 |
|
| 1023 |
<para> |
| 1024 |
To handle SQL character string data types, such |
| 1025 |
as <type>varchar</type> and <type>text</type>, there are two |
| 1026 |
possible ways to declare the host variables. |
| 1027 |
</para> |
| 1028 |
|
| 1029 |
<para> |
| 1030 |
One way is using <type>char[]</type>, an array |
| 1031 |
of <type>char</type>, which is the most common way to handle |
| 1032 |
character data in C. |
| 1033 |
<programlisting> |
| 1034 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1035 |
char str[50]; |
| 1036 |
EXEC SQL END DECLARE SECTION; |
| 1037 |
</programlisting> |
| 1038 |
Note that you have to take care of the length yourself. If you |
| 1039 |
use this host variable as the target variable of a query which |
| 1040 |
returns a string with more than 49 characters, a buffer overflow |
| 1041 |
occurs. |
| 1042 |
</para> |
| 1043 |
|
| 1044 |
<para> |
| 1045 |
The other way is using the <type>VARCHAR</type> type, which is a |
| 1046 |
special type provided by ECPG. The definition on an array of |
| 1047 |
type <type>VARCHAR</type> is converted into a |
| 1048 |
named <type>struct</type> for every variable. A declaration like: |
| 1049 |
<programlisting> |
| 1050 |
VARCHAR var[180]; |
| 1051 |
</programlisting> |
| 1052 |
is converted into: |
| 1053 |
<programlisting> |
| 1054 |
struct varchar_var { int len; char arr[180]; } var; |
| 1055 |
</programlisting> |
| 1056 |
The member <structfield>arr</structfield> hosts the string |
| 1057 |
including a terminating zero byte. Thus, to store a string in |
| 1058 |
a <type>VARCHAR</type> host variable, the host variable has to be |
| 1059 |
declared with the length including the zero byte terminator. The |
| 1060 |
member <structfield>len</structfield> holds the length of the |
| 1061 |
string stored in the <structfield>arr</structfield> without the |
| 1062 |
terminating zero byte. When a host variable is used as input for |
| 1063 |
a query, if <literal>strlen(arr)</literal> |
| 1064 |
and <structfield>len</structfield> are different, the shorter one |
| 1065 |
is used. |
| 1066 |
</para> |
| 1067 |
|
| 1068 |
<para> |
| 1069 |
<type>VARCHAR</type> can be written in upper or lower case, but |
| 1070 |
not in mixed case. |
| 1071 |
</para> |
| 1072 |
|
| 1073 |
<para> |
| 1074 |
<type>char</type> and <type>VARCHAR</type> host variables can |
| 1075 |
also hold values of other SQL types, which will be stored in |
| 1076 |
their string forms. |
| 1077 |
</para> |
| 1078 |
</sect3> |
| 1079 |
|
| 1080 |
<sect3 id="ecpg-special-types"> |
| 1081 |
<title>Accessing Special Data Types</title> |
| 1082 |
|
| 1083 |
<para> |
| 1084 |
ECPG contains some special types that help you to interact easily |
| 1085 |
with some special data types from the PostgreSQL server. In |
| 1086 |
particular, it has implemented support for the |
| 1087 |
<type>numeric</type>, <type>decimal</type>, <type>date</type>, <type>timestamp</type>, |
| 1088 |
and <type>interval</type> types. These data types cannot usefully be |
| 1089 |
mapped to primitive host variable types (such |
| 1090 |
as <type>int</type>, <type>long long int</type>, |
| 1091 |
or <type>char[]</type>), because they have a complex internal |
| 1092 |
structure. Applications deal with these types by declaring host |
| 1093 |
variables in special types and accessing them using functions in |
| 1094 |
the pgtypes library. The pgtypes library, described in detail |
| 1095 |
in <xref linkend="ecpg-pgtypes"/> contains basic functions to deal |
| 1096 |
with those types, such that you do not need to send a query to |
| 1097 |
the SQL server just for adding an interval to a time stamp for |
| 1098 |
example. |
| 1099 |
</para> |
| 1100 |
|
| 1101 |
<para> |
| 1102 |
The follow subsections describe these special data types. For |
| 1103 |
more details about pgtypes library functions, |
| 1104 |
see <xref linkend="ecpg-pgtypes"/>. |
| 1105 |
</para> |
| 1106 |
|
| 1107 |
<sect4 id="ecpg-special-types-timestamp-date"> |
| 1108 |
<title>timestamp, date</title> |
| 1109 |
|
| 1110 |
<para> |
| 1111 |
Here is a pattern for handling <type>timestamp</type> variables |
| 1112 |
in the ECPG host application. |
| 1113 |
</para> |
| 1114 |
|
| 1115 |
<para> |
| 1116 |
First, the program has to include the header file for the |
| 1117 |
<type>timestamp</type> type: |
| 1118 |
<programlisting> |
| 1119 |
#include <pgtypes_timestamp.h> |
| 1120 |
</programlisting> |
| 1121 |
</para> |
| 1122 |
|
| 1123 |
<para> |
| 1124 |
Next, declare a host variable as type <type>timestamp</type> in |
| 1125 |
the declare section: |
| 1126 |
<programlisting> |
| 1127 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1128 |
timestamp ts; |
| 1129 |
EXEC SQL END DECLARE SECTION; |
| 1130 |
</programlisting> |
| 1131 |
</para> |
| 1132 |
|
| 1133 |
<para> |
| 1134 |
And after reading a value into the host variable, process it |
| 1135 |
using pgtypes library functions. In following example, the |
| 1136 |
<type>timestamp</type> value is converted into text (ASCII) form |
| 1137 |
with the <function>PGTYPEStimestamp_to_asc()</function> |
| 1138 |
function: |
| 1139 |
<programlisting> |
| 1140 |
EXEC SQL SELECT now()::timestamp INTO :ts; |
| 1141 |
|
| 1142 |
printf("ts = %s\n", PGTYPEStimestamp_to_asc(ts)); |
| 1143 |
</programlisting> |
| 1144 |
This example will show some result like following: |
| 1145 |
<screen> |
| 1146 |
ts = 2010-06-27 18:03:56.949343 |
| 1147 |
</screen> |
| 1148 |
</para> |
| 1149 |
|
| 1150 |
<para> |
| 1151 |
In addition, the DATE type can be handled in the same way. The |
| 1152 |
program has to include <filename>pgtypes_date.h</filename>, declare a host variable |
| 1153 |
as the date type and convert a DATE value into a text form using |
| 1154 |
<function>PGTYPESdate_to_asc()</function> function. For more details about the |
| 1155 |
pgtypes library functions, see <xref linkend="ecpg-pgtypes"/>. |
| 1156 |
</para> |
| 1157 |
</sect4> |
| 1158 |
|
| 1159 |
<sect4 id="ecpg-type-interval"> |
| 1160 |
<title>interval</title> |
| 1161 |
|
| 1162 |
<para> |
| 1163 |
The handling of the <type>interval</type> type is also similar |
| 1164 |
to the <type>timestamp</type> and <type>date</type> types. It |
| 1165 |
is required, however, to allocate memory for |
| 1166 |
an <type>interval</type> type value explicitly. In other words, |
| 1167 |
the memory space for the variable has to be allocated in the |
| 1168 |
heap memory, not in the stack memory. |
| 1169 |
</para> |
| 1170 |
|
| 1171 |
<para> |
| 1172 |
Here is an example program: |
| 1173 |
<programlisting> |
| 1174 |
#include <stdio.h> |
| 1175 |
#include <stdlib.h> |
| 1176 |
#include <pgtypes_interval.h> |
| 1177 |
|
| 1178 |
int |
| 1179 |
main(void) |
| 1180 |
{ |
| 1181 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1182 |
interval *in; |
| 1183 |
EXEC SQL END DECLARE SECTION; |
| 1184 |
|
| 1185 |
EXEC SQL CONNECT TO testdb; |
| 1186 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 1187 |
|
| 1188 |
in = PGTYPESinterval_new(); |
| 1189 |
EXEC SQL SELECT '1 min'::interval INTO :in; |
| 1190 |
printf("interval = %s\n", PGTYPESinterval_to_asc(in)); |
| 1191 |
PGTYPESinterval_free(in); |
| 1192 |
|
| 1193 |
EXEC SQL COMMIT; |
| 1194 |
EXEC SQL DISCONNECT ALL; |
| 1195 |
return 0; |
| 1196 |
} |
| 1197 |
</programlisting> |
| 1198 |
</para> |
| 1199 |
</sect4> |
| 1200 |
|
| 1201 |
<sect4 id="ecpg-type-numeric-decimal"> |
| 1202 |
<title>numeric, decimal</title> |
| 1203 |
|
| 1204 |
<para> |
| 1205 |
The handling of the <type>numeric</type> |
| 1206 |
and <type>decimal</type> types is similar to the |
| 1207 |
<type>interval</type> type: It requires defining a pointer, |
| 1208 |
allocating some memory space on the heap, and accessing the |
| 1209 |
variable using the pgtypes library functions. For more details |
| 1210 |
about the pgtypes library functions, |
| 1211 |
see <xref linkend="ecpg-pgtypes"/>. |
| 1212 |
</para> |
| 1213 |
|
| 1214 |
<para> |
| 1215 |
No functions are provided specifically for |
| 1216 |
the <type>decimal</type> type. An application has to convert it |
| 1217 |
to a <type>numeric</type> variable using a pgtypes library |
| 1218 |
function to do further processing. |
| 1219 |
</para> |
| 1220 |
|
| 1221 |
<para> |
| 1222 |
Here is an example program handling <type>numeric</type> |
| 1223 |
and <type>decimal</type> type variables. |
| 1224 |
<programlisting> |
| 1225 |
#include <stdio.h> |
| 1226 |
#include <stdlib.h> |
| 1227 |
#include <pgtypes_numeric.h> |
| 1228 |
|
| 1229 |
EXEC SQL WHENEVER SQLERROR STOP; |
| 1230 |
|
| 1231 |
int |
| 1232 |
main(void) |
| 1233 |
{ |
| 1234 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1235 |
numeric *num; |
| 1236 |
numeric *num2; |
| 1237 |
decimal *dec; |
| 1238 |
EXEC SQL END DECLARE SECTION; |
| 1239 |
|
| 1240 |
EXEC SQL CONNECT TO testdb; |
| 1241 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 1242 |
|
| 1243 |
num = PGTYPESnumeric_new(); |
| 1244 |
dec = PGTYPESdecimal_new(); |
| 1245 |
|
| 1246 |
EXEC SQL SELECT 12.345::numeric(4,2), 23.456::decimal(4,2) INTO :num, :dec; |
| 1247 |
|
| 1248 |
printf("numeric = %s\n", PGTYPESnumeric_to_asc(num, 0)); |
| 1249 |
printf("numeric = %s\n", PGTYPESnumeric_to_asc(num, 1)); |
| 1250 |
printf("numeric = %s\n", PGTYPESnumeric_to_asc(num, 2)); |
| 1251 |
|
| 1252 |
/* Convert decimal to numeric to show a decimal value. */ |
| 1253 |
num2 = PGTYPESnumeric_new(); |
| 1254 |
PGTYPESnumeric_from_decimal(dec, num2); |
| 1255 |
|
| 1256 |
printf("decimal = %s\n", PGTYPESnumeric_to_asc(num2, 0)); |
| 1257 |
printf("decimal = %s\n", PGTYPESnumeric_to_asc(num2, 1)); |
| 1258 |
printf("decimal = %s\n", PGTYPESnumeric_to_asc(num2, 2)); |
| 1259 |
|
| 1260 |
PGTYPESnumeric_free(num2); |
| 1261 |
PGTYPESdecimal_free(dec); |
| 1262 |
PGTYPESnumeric_free(num); |
| 1263 |
|
| 1264 |
EXEC SQL COMMIT; |
| 1265 |
EXEC SQL DISCONNECT ALL; |
| 1266 |
return 0; |
| 1267 |
} |
| 1268 |
</programlisting> |
| 1269 |
</para> |
| 1270 |
</sect4> |
| 1271 |
|
| 1272 |
<sect4 id="ecpg-special-types-bytea"> |
| 1273 |
<title>bytea</title> |
| 1274 |
|
| 1275 |
<para> |
| 1276 |
The handling of the <type>bytea</type> type is similar to |
| 1277 |
that of <type>VARCHAR</type>. The definition on an array of type |
| 1278 |
<type>bytea</type> is converted into a named struct for every |
| 1279 |
variable. A declaration like: |
| 1280 |
<programlisting> |
| 1281 |
bytea var[180]; |
| 1282 |
</programlisting> |
| 1283 |
is converted into: |
| 1284 |
<programlisting> |
| 1285 |
struct bytea_var { int len; char arr[180]; } var; |
| 1286 |
</programlisting> |
| 1287 |
The member <structfield>arr</structfield> hosts binary format |
| 1288 |
data. It can also handle <literal>'\0'</literal> as part of |
| 1289 |
data, unlike <type>VARCHAR</type>. |
| 1290 |
The data is converted from/to hex format and sent/received by |
| 1291 |
ecpglib. |
| 1292 |
</para> |
| 1293 |
|
| 1294 |
<note> |
| 1295 |
<para> |
| 1296 |
<type>bytea</type> variable can be used only when |
| 1297 |
<xref linkend="guc-bytea-output"/> is set to <literal>hex</literal>. |
| 1298 |
</para> |
| 1299 |
</note> |
| 1300 |
</sect4> |
| 1301 |
</sect3> |
| 1302 |
|
| 1303 |
<sect3 id="ecpg-variables-nonprimitive-c"> |
| 1304 |
<title>Host Variables with Nonprimitive Types</title> |
| 1305 |
|
| 1306 |
<para> |
| 1307 |
As a host variable you can also use arrays, typedefs, structs, and |
| 1308 |
pointers. |
| 1309 |
</para> |
| 1310 |
|
| 1311 |
<sect4 id="ecpg-variables-arrays"> |
| 1312 |
<title>Arrays</title> |
| 1313 |
|
| 1314 |
<para> |
| 1315 |
There are two use cases for arrays as host variables. The first |
| 1316 |
is a way to store some text string in <type>char[]</type> |
| 1317 |
or <type>VARCHAR[]</type>, as |
| 1318 |
explained in <xref linkend="ecpg-char"/>. The second use case is to |
| 1319 |
retrieve multiple rows from a query result without using a |
| 1320 |
cursor. Without an array, to process a query result consisting |
| 1321 |
of multiple rows, it is required to use a cursor and |
| 1322 |
the <command>FETCH</command> command. But with array host |
| 1323 |
variables, multiple rows can be received at once. The length of |
| 1324 |
the array has to be defined to be able to accommodate all rows, |
| 1325 |
otherwise a buffer overflow will likely occur. |
| 1326 |
</para> |
| 1327 |
|
| 1328 |
<para> |
| 1329 |
Following example scans the <literal>pg_database</literal> |
| 1330 |
system table and shows all OIDs and names of the available |
| 1331 |
databases: |
| 1332 |
<programlisting> |
| 1333 |
int |
| 1334 |
main(void) |
| 1335 |
{ |
| 1336 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1337 |
int dbid[8]; |
| 1338 |
char dbname[8][16]; |
| 1339 |
int i; |
| 1340 |
EXEC SQL END DECLARE SECTION; |
| 1341 |
|
| 1342 |
memset(dbname, 0, sizeof(char)* 16 * 8); |
| 1343 |
memset(dbid, 0, sizeof(int) * 8); |
| 1344 |
|
| 1345 |
EXEC SQL CONNECT TO testdb; |
| 1346 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 1347 |
|
| 1348 |
/* Retrieve multiple rows into arrays at once. */ |
| 1349 |
EXEC SQL SELECT oid,datname INTO :dbid, :dbname FROM pg_database; |
| 1350 |
|
| 1351 |
for (i = 0; i < 8; i++) |
| 1352 |
printf("oid=%d, dbname=%s\n", dbid[i], dbname[i]); |
| 1353 |
|
| 1354 |
EXEC SQL COMMIT; |
| 1355 |
EXEC SQL DISCONNECT ALL; |
| 1356 |
return 0; |
| 1357 |
} |
| 1358 |
</programlisting> |
| 1359 |
|
| 1360 |
This example shows following result. (The exact values depend on |
| 1361 |
local circumstances.) |
| 1362 |
<screen> |
| 1363 |
oid=1, dbname=template1 |
| 1364 |
oid=11510, dbname=template0 |
| 1365 |
oid=11511, dbname=postgres |
| 1366 |
oid=313780, dbname=testdb |
| 1367 |
oid=0, dbname= |
| 1368 |
oid=0, dbname= |
| 1369 |
oid=0, dbname= |
| 1370 |
</screen> |
| 1371 |
</para> |
| 1372 |
</sect4> |
| 1373 |
|
| 1374 |
<sect4 id="ecpg-variables-struct"> |
| 1375 |
<title>Structures</title> |
| 1376 |
|
| 1377 |
<para> |
| 1378 |
A structure whose member names match the column names of a query |
| 1379 |
result, can be used to retrieve multiple columns at once. The |
| 1380 |
structure enables handling multiple column values in a single |
| 1381 |
host variable. |
| 1382 |
</para> |
| 1383 |
|
| 1384 |
<para> |
| 1385 |
The following example retrieves OIDs, names, and sizes of the |
| 1386 |
available databases from the <literal>pg_database</literal> |
| 1387 |
system table and using |
| 1388 |
the <function>pg_database_size()</function> function. In this |
| 1389 |
example, a structure variable <varname>dbinfo_t</varname> with |
| 1390 |
members whose names match each column in |
| 1391 |
the <literal>SELECT</literal> result is used to retrieve one |
| 1392 |
result row without putting multiple host variables in |
| 1393 |
the <literal>FETCH</literal> statement. |
| 1394 |
<programlisting> |
| 1395 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1396 |
typedef struct |
| 1397 |
{ |
| 1398 |
int oid; |
| 1399 |
char datname[65]; |
| 1400 |
long long int size; |
| 1401 |
} dbinfo_t; |
| 1402 |
|
| 1403 |
dbinfo_t dbval; |
| 1404 |
EXEC SQL END DECLARE SECTION; |
| 1405 |
|
| 1406 |
memset(&dbval, 0, sizeof(dbinfo_t)); |
| 1407 |
|
| 1408 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT oid, datname, pg_database_size(oid) AS size FROM pg_database; |
| 1409 |
EXEC SQL OPEN cur1; |
| 1410 |
|
| 1411 |
/* when end of result set reached, break out of while loop */ |
| 1412 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1413 |
|
| 1414 |
while (1) |
| 1415 |
{ |
| 1416 |
/* Fetch multiple columns into one structure. */ |
| 1417 |
EXEC SQL FETCH FROM cur1 INTO :dbval; |
| 1418 |
|
| 1419 |
/* Print members of the structure. */ |
| 1420 |
printf("oid=%d, datname=%s, size=%lld\n", dbval.oid, dbval.datname, dbval.size); |
| 1421 |
} |
| 1422 |
|
| 1423 |
EXEC SQL CLOSE cur1; |
| 1424 |
</programlisting> |
| 1425 |
</para> |
| 1426 |
|
| 1427 |
<para> |
| 1428 |
This example shows following result. (The exact values depend on |
| 1429 |
local circumstances.) |
| 1430 |
<screen> |
| 1431 |
oid=1, datname=template1, size=4324580 |
| 1432 |
oid=11510, datname=template0, size=4243460 |
| 1433 |
oid=11511, datname=postgres, size=4324580 |
| 1434 |
oid=313780, datname=testdb, size=8183012 |
| 1435 |
</screen> |
| 1436 |
</para> |
| 1437 |
|
| 1438 |
<para> |
| 1439 |
Structure host variables <quote>absorb</quote> as many columns |
| 1440 |
as the structure as fields. Additional columns can be assigned |
| 1441 |
to other host variables. For example, the above program could |
| 1442 |
also be restructured like this, with the <varname>size</varname> |
| 1443 |
variable outside the structure: |
| 1444 |
<programlisting> |
| 1445 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1446 |
typedef struct |
| 1447 |
{ |
| 1448 |
int oid; |
| 1449 |
char datname[65]; |
| 1450 |
} dbinfo_t; |
| 1451 |
|
| 1452 |
dbinfo_t dbval; |
| 1453 |
long long int size; |
| 1454 |
EXEC SQL END DECLARE SECTION; |
| 1455 |
|
| 1456 |
memset(&dbval, 0, sizeof(dbinfo_t)); |
| 1457 |
|
| 1458 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT oid, datname, pg_database_size(oid) AS size FROM pg_database; |
| 1459 |
EXEC SQL OPEN cur1; |
| 1460 |
|
| 1461 |
/* when end of result set reached, break out of while loop */ |
| 1462 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1463 |
|
| 1464 |
while (1) |
| 1465 |
{ |
| 1466 |
/* Fetch multiple columns into one structure. */ |
| 1467 |
EXEC SQL FETCH FROM cur1 INTO :dbval, :size; |
| 1468 |
|
| 1469 |
/* Print members of the structure. */ |
| 1470 |
printf("oid=%d, datname=%s, size=%lld\n", dbval.oid, dbval.datname, size); |
| 1471 |
} |
| 1472 |
|
| 1473 |
EXEC SQL CLOSE cur1; |
| 1474 |
</programlisting> |
| 1475 |
</para> |
| 1476 |
</sect4> |
| 1477 |
|
| 1478 |
<sect4 id="ecpg-variables-nonprimitive-c-typedefs"> |
| 1479 |
<title>Typedefs</title> |
| 1480 |
<indexterm> |
| 1481 |
<primary>typedef</primary> |
| 1482 |
<secondary>in ECPG</secondary> |
| 1483 |
</indexterm> |
| 1484 |
|
| 1485 |
<para> |
| 1486 |
Use the <literal>typedef</literal> keyword to map new types to already |
| 1487 |
existing types. |
| 1488 |
<programlisting> |
| 1489 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1490 |
typedef char mychartype[40]; |
| 1491 |
typedef long serial_t; |
| 1492 |
EXEC SQL END DECLARE SECTION; |
| 1493 |
</programlisting> |
| 1494 |
Note that you could also use: |
| 1495 |
<programlisting> |
| 1496 |
EXEC SQL TYPE serial_t IS long; |
| 1497 |
</programlisting> |
| 1498 |
This declaration does not need to be part of a declare section; |
| 1499 |
that is, you can also write typedefs as normal C statements. |
| 1500 |
</para> |
| 1501 |
|
| 1502 |
<para> |
| 1503 |
Any word you declare as a <literal>typedef</literal> cannot be used as |
| 1504 |
an SQL keyword in <literal>EXEC SQL</literal> commands later in the same |
| 1505 |
program. For example, this won't work: |
| 1506 |
<programlisting> |
| 1507 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1508 |
typedef int start; |
| 1509 |
EXEC SQL END DECLARE SECTION; |
| 1510 |
... |
| 1511 |
EXEC SQL START TRANSACTION; |
| 1512 |
</programlisting> |
| 1513 |
ECPG will report a syntax error for <literal>START |
| 1514 |
TRANSACTION</literal>, because it no longer |
| 1515 |
recognizes <literal>START</literal> as an SQL keyword, |
| 1516 |
only as a typedef. |
| 1517 |
(If you have such a conflict, and renaming the typedef |
| 1518 |
seems impractical, you could write the SQL command |
| 1519 |
using <link linkend="ecpg-dynamic">dynamic SQL</link>.) |
| 1520 |
</para> |
| 1521 |
|
| 1522 |
<note> |
| 1523 |
<para> |
| 1524 |
In <productname>PostgreSQL</productname> releases before v16, use |
| 1525 |
of SQL keywords as typedef names was likely to result in syntax |
| 1526 |
errors associated with use of the typedef itself, rather than use |
| 1527 |
of the name as an SQL keyword. The new behavior is less likely to |
| 1528 |
cause problems when an existing ECPG application is recompiled in |
| 1529 |
a new <productname>PostgreSQL</productname> release with new |
| 1530 |
keywords. |
| 1531 |
</para> |
| 1532 |
</note> |
| 1533 |
</sect4> |
| 1534 |
|
| 1535 |
<sect4 id="ecpg-variables-nonprimitive-c-pointers"> |
| 1536 |
<title>Pointers</title> |
| 1537 |
|
| 1538 |
<para> |
| 1539 |
You can declare pointers to the most common types. Note however |
| 1540 |
that you cannot use pointers as target variables of queries |
| 1541 |
without auto-allocation. See <xref linkend="ecpg-descriptors"/> |
| 1542 |
for more information on auto-allocation. |
| 1543 |
</para> |
| 1544 |
|
| 1545 |
<para> |
| 1546 |
<programlisting> |
| 1547 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1548 |
int *intp; |
| 1549 |
char **charp; |
| 1550 |
EXEC SQL END DECLARE SECTION; |
| 1551 |
</programlisting> |
| 1552 |
</para> |
| 1553 |
</sect4> |
| 1554 |
</sect3> |
| 1555 |
</sect2> |
| 1556 |
|
| 1557 |
<sect2 id="ecpg-variables-nonprimitive-sql"> |
| 1558 |
<title>Handling Nonprimitive SQL Data Types</title> |
| 1559 |
|
| 1560 |
<para> |
| 1561 |
This section contains information on how to handle nonscalar and |
| 1562 |
user-defined SQL-level data types in ECPG applications. Note that |
| 1563 |
this is distinct from the handling of host variables of |
| 1564 |
nonprimitive types, described in the previous section. |
| 1565 |
</para> |
| 1566 |
|
| 1567 |
<sect3 id="ecpg-variables-nonprimitive-sql-arrays"> |
| 1568 |
<title>Arrays</title> |
| 1569 |
|
| 1570 |
<para> |
| 1571 |
Multi-dimensional SQL-level arrays are not directly supported in ECPG. |
| 1572 |
One-dimensional SQL-level arrays can be mapped into C array host |
| 1573 |
variables and vice-versa. However, when creating a statement ecpg does |
| 1574 |
not know the types of the columns, so that it cannot check if a C array |
| 1575 |
is input into a corresponding SQL-level array. When processing the |
| 1576 |
output of an SQL statement, ecpg has the necessary information and thus |
| 1577 |
checks if both are arrays. |
| 1578 |
</para> |
| 1579 |
|
| 1580 |
<para> |
| 1581 |
If a query accesses <emphasis>elements</emphasis> of an array |
| 1582 |
separately, then this avoids the use of arrays in ECPG. Then, a |
| 1583 |
host variable with a type that can be mapped to the element type |
| 1584 |
should be used. For example, if a column type is array of |
| 1585 |
<type>integer</type>, a host variable of type <type>int</type> |
| 1586 |
can be used. Also if the element type is <type>varchar</type> |
| 1587 |
or <type>text</type>, a host variable of type <type>char[]</type> |
| 1588 |
or <type>VARCHAR[]</type> can be used. |
| 1589 |
</para> |
| 1590 |
|
| 1591 |
<para> |
| 1592 |
Here is an example. Assume the following table: |
| 1593 |
<programlisting> |
| 1594 |
CREATE TABLE t3 ( |
| 1595 |
ii integer[] |
| 1596 |
); |
| 1597 |
|
| 1598 |
testdb=> SELECT * FROM t3; |
| 1599 |
ii |
| 1600 |
------------- |
| 1601 |
{1,2,3,4,5} |
| 1602 |
(1 row) |
| 1603 |
</programlisting> |
| 1604 |
|
| 1605 |
The following example program retrieves the 4th element of the |
| 1606 |
array and stores it into a host variable of |
| 1607 |
type <type>int</type>: |
| 1608 |
<programlisting> |
| 1609 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1610 |
int ii; |
| 1611 |
EXEC SQL END DECLARE SECTION; |
| 1612 |
|
| 1613 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT ii[4] FROM t3; |
| 1614 |
EXEC SQL OPEN cur1; |
| 1615 |
|
| 1616 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1617 |
|
| 1618 |
while (1) |
| 1619 |
{ |
| 1620 |
EXEC SQL FETCH FROM cur1 INTO :ii ; |
| 1621 |
printf("ii=%d\n", ii); |
| 1622 |
} |
| 1623 |
|
| 1624 |
EXEC SQL CLOSE cur1; |
| 1625 |
</programlisting> |
| 1626 |
|
| 1627 |
This example shows the following result: |
| 1628 |
<screen> |
| 1629 |
ii=4 |
| 1630 |
</screen> |
| 1631 |
</para> |
| 1632 |
|
| 1633 |
<para> |
| 1634 |
To map multiple array elements to the multiple elements in an |
| 1635 |
array type host variables each element of array column and each |
| 1636 |
element of the host variable array have to be managed separately, |
| 1637 |
for example: |
| 1638 |
<programlisting> |
| 1639 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1640 |
int ii_a[8]; |
| 1641 |
EXEC SQL END DECLARE SECTION; |
| 1642 |
|
| 1643 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT ii[1], ii[2], ii[3], ii[4] FROM t3; |
| 1644 |
EXEC SQL OPEN cur1; |
| 1645 |
|
| 1646 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1647 |
|
| 1648 |
while (1) |
| 1649 |
{ |
| 1650 |
EXEC SQL FETCH FROM cur1 INTO :ii_a[0], :ii_a[1], :ii_a[2], :ii_a[3]; |
| 1651 |
... |
| 1652 |
} |
| 1653 |
</programlisting> |
| 1654 |
</para> |
| 1655 |
|
| 1656 |
<para> |
| 1657 |
Note again that |
| 1658 |
<programlisting> |
| 1659 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1660 |
int ii_a[8]; |
| 1661 |
EXEC SQL END DECLARE SECTION; |
| 1662 |
|
| 1663 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT ii FROM t3; |
| 1664 |
EXEC SQL OPEN cur1; |
| 1665 |
|
| 1666 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1667 |
|
| 1668 |
while (1) |
| 1669 |
{ |
| 1670 |
/* WRONG */ |
| 1671 |
EXEC SQL FETCH FROM cur1 INTO :ii_a; |
| 1672 |
... |
| 1673 |
} |
| 1674 |
</programlisting> |
| 1675 |
would not work correctly in this case, because you cannot map an |
| 1676 |
array type column to an array host variable directly. |
| 1677 |
</para> |
| 1678 |
|
| 1679 |
<para> |
| 1680 |
Another workaround is to store arrays in their external string |
| 1681 |
representation in host variables of type <type>char[]</type> |
| 1682 |
or <type>VARCHAR[]</type>. For more details about this |
| 1683 |
representation, see <xref linkend="arrays-input"/>. Note that |
| 1684 |
this means that the array cannot be accessed naturally as an |
| 1685 |
array in the host program (without further processing that parses |
| 1686 |
the text representation). |
| 1687 |
</para> |
| 1688 |
</sect3> |
| 1689 |
|
| 1690 |
<sect3 id="ecpg-variables-nonprimitive-sql-composite"> |
| 1691 |
<title>Composite Types</title> |
| 1692 |
|
| 1693 |
<para> |
| 1694 |
Composite types are not directly supported in ECPG, but an easy workaround is possible. |
| 1695 |
The |
| 1696 |
available workarounds are similar to the ones described for |
| 1697 |
arrays above: Either access each attribute separately or use the |
| 1698 |
external string representation. |
| 1699 |
</para> |
| 1700 |
|
| 1701 |
<para> |
| 1702 |
For the following examples, assume the following type and table: |
| 1703 |
<programlisting> |
| 1704 |
CREATE TYPE comp_t AS (intval integer, textval varchar(32)); |
| 1705 |
CREATE TABLE t4 (compval comp_t); |
| 1706 |
INSERT INTO t4 VALUES ( (256, 'PostgreSQL') ); |
| 1707 |
</programlisting> |
| 1708 |
|
| 1709 |
The most obvious solution is to access each attribute separately. |
| 1710 |
The following program retrieves data from the example table by |
| 1711 |
selecting each attribute of the type <type>comp_t</type> |
| 1712 |
separately: |
| 1713 |
<programlisting> |
| 1714 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1715 |
int intval; |
| 1716 |
varchar textval[33]; |
| 1717 |
EXEC SQL END DECLARE SECTION; |
| 1718 |
|
| 1719 |
/* Put each element of the composite type column in the SELECT list. */ |
| 1720 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT (compval).intval, (compval).textval FROM t4; |
| 1721 |
EXEC SQL OPEN cur1; |
| 1722 |
|
| 1723 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1724 |
|
| 1725 |
while (1) |
| 1726 |
{ |
| 1727 |
/* Fetch each element of the composite type column into host variables. */ |
| 1728 |
EXEC SQL FETCH FROM cur1 INTO :intval, :textval; |
| 1729 |
|
| 1730 |
printf("intval=%d, textval=%s\n", intval, textval.arr); |
| 1731 |
} |
| 1732 |
|
| 1733 |
EXEC SQL CLOSE cur1; |
| 1734 |
</programlisting> |
| 1735 |
</para> |
| 1736 |
|
| 1737 |
<para> |
| 1738 |
To enhance this example, the host variables to store values in |
| 1739 |
the <command>FETCH</command> command can be gathered into one |
| 1740 |
structure. For more details about the host variable in the |
| 1741 |
structure form, see <xref linkend="ecpg-variables-struct"/>. |
| 1742 |
To switch to the structure, the example can be modified as below. |
| 1743 |
The two host variables, <varname>intval</varname> |
| 1744 |
and <varname>textval</varname>, become members of |
| 1745 |
the <structname>comp_t</structname> structure, and the structure |
| 1746 |
is specified on the <command>FETCH</command> command. |
| 1747 |
<programlisting> |
| 1748 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1749 |
typedef struct |
| 1750 |
{ |
| 1751 |
int intval; |
| 1752 |
varchar textval[33]; |
| 1753 |
} comp_t; |
| 1754 |
|
| 1755 |
comp_t compval; |
| 1756 |
EXEC SQL END DECLARE SECTION; |
| 1757 |
|
| 1758 |
/* Put each element of the composite type column in the SELECT list. */ |
| 1759 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT (compval).intval, (compval).textval FROM t4; |
| 1760 |
EXEC SQL OPEN cur1; |
| 1761 |
|
| 1762 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1763 |
|
| 1764 |
while (1) |
| 1765 |
{ |
| 1766 |
/* Put all values in the SELECT list into one structure. */ |
| 1767 |
EXEC SQL FETCH FROM cur1 INTO :compval; |
| 1768 |
|
| 1769 |
printf("intval=%d, textval=%s\n", compval.intval, compval.textval.arr); |
| 1770 |
} |
| 1771 |
|
| 1772 |
EXEC SQL CLOSE cur1; |
| 1773 |
</programlisting> |
| 1774 |
|
| 1775 |
Although a structure is used in the <command>FETCH</command> |
| 1776 |
command, the attribute names in the <command>SELECT</command> |
| 1777 |
clause are specified one by one. This can be enhanced by using |
| 1778 |
a <literal>*</literal> to ask for all attributes of the composite |
| 1779 |
type value. |
| 1780 |
<programlisting> |
| 1781 |
... |
| 1782 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT (compval).* FROM t4; |
| 1783 |
EXEC SQL OPEN cur1; |
| 1784 |
|
| 1785 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1786 |
|
| 1787 |
while (1) |
| 1788 |
{ |
| 1789 |
/* Put all values in the SELECT list into one structure. */ |
| 1790 |
EXEC SQL FETCH FROM cur1 INTO :compval; |
| 1791 |
|
| 1792 |
printf("intval=%d, textval=%s\n", compval.intval, compval.textval.arr); |
| 1793 |
} |
| 1794 |
... |
| 1795 |
</programlisting> |
| 1796 |
This way, composite types can be mapped into structures almost |
| 1797 |
seamlessly, even though ECPG does not understand the composite |
| 1798 |
type itself. |
| 1799 |
</para> |
| 1800 |
|
| 1801 |
<para> |
| 1802 |
Finally, it is also possible to store composite type values in |
| 1803 |
their external string representation in host variables of |
| 1804 |
type <type>char[]</type> or <type>VARCHAR[]</type>. But that |
| 1805 |
way, it is not easily possible to access the fields of the value |
| 1806 |
from the host program. |
| 1807 |
</para> |
| 1808 |
</sect3> |
| 1809 |
|
| 1810 |
<sect3 id="ecpg-variables-nonprimitive-sql-user-defined-base-types"> |
| 1811 |
<title>User-Defined Base Types</title> |
| 1812 |
|
| 1813 |
<para> |
| 1814 |
New user-defined base types are not directly supported by ECPG. |
| 1815 |
You can use the external string representation and host variables |
| 1816 |
of type <type>char[]</type> or <type>VARCHAR[]</type>, and this |
| 1817 |
solution is indeed appropriate and sufficient for many types. |
| 1818 |
</para> |
| 1819 |
|
| 1820 |
<para> |
| 1821 |
Here is an example using the data type <type>complex</type> from |
| 1822 |
the example in <xref linkend="xtypes"/>. The external string |
| 1823 |
representation of that type is <literal>(%f,%f)</literal>, |
| 1824 |
which is defined in the |
| 1825 |
functions <function>complex_in()</function> |
| 1826 |
and <function>complex_out()</function> |
| 1827 |
in <xref linkend="xtypes"/>. The following example inserts the |
| 1828 |
complex type values <literal>(1,1)</literal> |
| 1829 |
and <literal>(3,3)</literal> into the |
| 1830 |
columns <literal>a</literal> and <literal>b</literal>, and select |
| 1831 |
them from the table after that. |
| 1832 |
|
| 1833 |
<programlisting> |
| 1834 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1835 |
varchar a[64]; |
| 1836 |
varchar b[64]; |
| 1837 |
EXEC SQL END DECLARE SECTION; |
| 1838 |
|
| 1839 |
EXEC SQL INSERT INTO test_complex VALUES ('(1,1)', '(3,3)'); |
| 1840 |
|
| 1841 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT a, b FROM test_complex; |
| 1842 |
EXEC SQL OPEN cur1; |
| 1843 |
|
| 1844 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 1845 |
|
| 1846 |
while (1) |
| 1847 |
{ |
| 1848 |
EXEC SQL FETCH FROM cur1 INTO :a, :b; |
| 1849 |
printf("a=%s, b=%s\n", a.arr, b.arr); |
| 1850 |
} |
| 1851 |
|
| 1852 |
EXEC SQL CLOSE cur1; |
| 1853 |
</programlisting> |
| 1854 |
|
| 1855 |
This example shows following result: |
| 1856 |
<screen> |
| 1857 |
a=(1,1), b=(3,3) |
| 1858 |
</screen> |
| 1859 |
</para> |
| 1860 |
|
| 1861 |
<para> |
| 1862 |
Another workaround is avoiding the direct use of the user-defined |
| 1863 |
types in ECPG and instead create a function or cast that converts |
| 1864 |
between the user-defined type and a primitive type that ECPG can |
| 1865 |
handle. Note, however, that type casts, especially implicit |
| 1866 |
ones, should be introduced into the type system very carefully. |
| 1867 |
</para> |
| 1868 |
|
| 1869 |
<para> |
| 1870 |
For example, |
| 1871 |
<programlisting> |
| 1872 |
CREATE FUNCTION create_complex(r double, i double) RETURNS complex |
| 1873 |
LANGUAGE SQL |
| 1874 |
IMMUTABLE |
| 1875 |
AS $$ SELECT $1 * complex '(1,0')' + $2 * complex '(0,1)' $$; |
| 1876 |
</programlisting> |
| 1877 |
After this definition, the following |
| 1878 |
<programlisting> |
| 1879 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1880 |
double a, b, c, d; |
| 1881 |
EXEC SQL END DECLARE SECTION; |
| 1882 |
|
| 1883 |
a = 1; |
| 1884 |
b = 2; |
| 1885 |
c = 3; |
| 1886 |
d = 4; |
| 1887 |
|
| 1888 |
EXEC SQL INSERT INTO test_complex VALUES (create_complex(:a, :b), create_complex(:c, :d)); |
| 1889 |
</programlisting> |
| 1890 |
has the same effect as |
| 1891 |
<programlisting> |
| 1892 |
EXEC SQL INSERT INTO test_complex VALUES ('(1,2)', '(3,4)'); |
| 1893 |
</programlisting> |
| 1894 |
</para> |
| 1895 |
</sect3> |
| 1896 |
</sect2> |
| 1897 |
|
| 1898 |
<sect2 id="ecpg-indicators"> |
| 1899 |
<title>Indicators</title> |
| 1900 |
|
| 1901 |
<para> |
| 1902 |
The examples above do not handle null values. In fact, the |
| 1903 |
retrieval examples will raise an error if they fetch a null value |
| 1904 |
from the database. To be able to pass null values to the database |
| 1905 |
or retrieve null values from the database, you need to append a |
| 1906 |
second host variable specification to each host variable that |
| 1907 |
contains data. This second host variable is called the |
| 1908 |
<firstterm>indicator</firstterm> and contains a flag that tells |
| 1909 |
whether the datum is null, in which case the value of the real |
| 1910 |
host variable is ignored. Here is an example that handles the |
| 1911 |
retrieval of null values correctly: |
| 1912 |
<programlisting> |
| 1913 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1914 |
VARCHAR val; |
| 1915 |
int val_ind; |
| 1916 |
EXEC SQL END DECLARE SECTION: |
| 1917 |
|
| 1918 |
... |
| 1919 |
|
| 1920 |
EXEC SQL SELECT b INTO :val :val_ind FROM test1; |
| 1921 |
</programlisting> |
| 1922 |
The indicator variable <varname>val_ind</varname> will be zero if |
| 1923 |
the value was not null, and it will be negative if the value was |
| 1924 |
null. (See <xref linkend="ecpg-oracle-compat"/> to enable |
| 1925 |
Oracle-specific behavior.) |
| 1926 |
</para> |
| 1927 |
|
| 1928 |
<para> |
| 1929 |
The indicator has another function: if the indicator value is |
| 1930 |
positive, it means that the value is not null, but it was |
| 1931 |
truncated when it was stored in the host variable. |
| 1932 |
</para> |
| 1933 |
|
| 1934 |
<para> |
| 1935 |
If the argument <literal>-r no_indicator</literal> is passed to |
| 1936 |
the preprocessor <command>ecpg</command>, it works in |
| 1937 |
<quote>no-indicator</quote> mode. In no-indicator mode, if no |
| 1938 |
indicator variable is specified, null values are signaled (on |
| 1939 |
input and output) for character string types as empty string and |
| 1940 |
for integer types as the lowest possible value for type (for |
| 1941 |
example, <symbol>INT_MIN</symbol> for <type>int</type>). |
| 1942 |
</para> |
| 1943 |
</sect2> |
| 1944 |
</sect1> |
| 1945 |
|
| 1946 |
<sect1 id="ecpg-dynamic"> |
| 1947 |
<title>Dynamic SQL</title> |
| 1948 |
|
| 1949 |
<para> |
| 1950 |
In many cases, the particular SQL statements that an application |
| 1951 |
has to execute are known at the time the application is written. |
| 1952 |
In some cases, however, the SQL statements are composed at run time |
| 1953 |
or provided by an external source. In these cases you cannot embed |
| 1954 |
the SQL statements directly into the C source code, but there is a |
| 1955 |
facility that allows you to call arbitrary SQL statements that you |
| 1956 |
provide in a string variable. |
| 1957 |
</para> |
| 1958 |
|
| 1959 |
<sect2 id="ecpg-dynamic-without-result"> |
| 1960 |
<title>Executing Statements without a Result Set</title> |
| 1961 |
|
| 1962 |
<para> |
| 1963 |
The simplest way to execute an arbitrary SQL statement is to use |
| 1964 |
the command <command>EXECUTE IMMEDIATE</command>. For example: |
| 1965 |
<programlisting> |
| 1966 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1967 |
const char *stmt = "CREATE TABLE test1 (...);"; |
| 1968 |
EXEC SQL END DECLARE SECTION; |
| 1969 |
|
| 1970 |
EXEC SQL EXECUTE IMMEDIATE :stmt; |
| 1971 |
</programlisting> |
| 1972 |
<command>EXECUTE IMMEDIATE</command> can be used for SQL |
| 1973 |
statements that do not return a result set (e.g., |
| 1974 |
DDL, <command>INSERT</command>, <command>UPDATE</command>, |
| 1975 |
<command>DELETE</command>). You cannot execute statements that |
| 1976 |
retrieve data (e.g., <command>SELECT</command>) this way. The |
| 1977 |
next section describes how to do that. |
| 1978 |
</para> |
| 1979 |
</sect2> |
| 1980 |
|
| 1981 |
<sect2 id="ecpg-dynamic-input"> |
| 1982 |
<title>Executing a Statement with Input Parameters</title> |
| 1983 |
|
| 1984 |
<para> |
| 1985 |
A more powerful way to execute arbitrary SQL statements is to |
| 1986 |
prepare them once and execute the prepared statement as often as |
| 1987 |
you like. It is also possible to prepare a generalized version of |
| 1988 |
a statement and then execute specific versions of it by |
| 1989 |
substituting parameters. When preparing the statement, write |
| 1990 |
question marks where you want to substitute parameters later. For |
| 1991 |
example: |
| 1992 |
<programlisting> |
| 1993 |
EXEC SQL BEGIN DECLARE SECTION; |
| 1994 |
const char *stmt = "INSERT INTO test1 VALUES(?, ?);"; |
| 1995 |
EXEC SQL END DECLARE SECTION; |
| 1996 |
|
| 1997 |
EXEC SQL PREPARE mystmt FROM :stmt; |
| 1998 |
... |
| 1999 |
EXEC SQL EXECUTE mystmt USING 42, 'foobar'; |
| 2000 |
</programlisting> |
| 2001 |
</para> |
| 2002 |
|
| 2003 |
<para> |
| 2004 |
When you don't need the prepared statement anymore, you should |
| 2005 |
deallocate it: |
| 2006 |
<programlisting> |
| 2007 |
EXEC SQL DEALLOCATE PREPARE <replaceable>name</replaceable>; |
| 2008 |
</programlisting> |
| 2009 |
</para> |
| 2010 |
</sect2> |
| 2011 |
|
| 2012 |
<sect2 id="ecpg-dynamic-with-result"> |
| 2013 |
<title>Executing a Statement with a Result Set</title> |
| 2014 |
|
| 2015 |
<para> |
| 2016 |
To execute an SQL statement with a single result row, |
| 2017 |
<command>EXECUTE</command> can be used. To save the result, add |
| 2018 |
an <literal>INTO</literal> clause. |
| 2019 |
<programlisting><![CDATA[ |
| 2020 |
EXEC SQL BEGIN DECLARE SECTION; |
| 2021 |
const char *stmt = "SELECT a, b, c FROM test1 WHERE a > ?"; |
| 2022 |
int v1, v2; |
| 2023 |
VARCHAR v3[50]; |
| 2024 |
EXEC SQL END DECLARE SECTION; |
| 2025 |
|
| 2026 |
EXEC SQL PREPARE mystmt FROM :stmt; |
| 2027 |
... |
| 2028 |
EXEC SQL EXECUTE mystmt INTO :v1, :v2, :v3 USING 37; |
| 2029 |
]]> |
| 2030 |
</programlisting> |
| 2031 |
An <command>EXECUTE</command> command can have an |
| 2032 |
<literal>INTO</literal> clause, a <literal>USING</literal> clause, |
| 2033 |
both, or neither. |
| 2034 |
</para> |
| 2035 |
|
| 2036 |
<para> |
| 2037 |
If a query is expected to return more than one result row, a |
| 2038 |
cursor should be used, as in the following example. |
| 2039 |
(See <xref linkend="ecpg-cursors"/> for more details about the |
| 2040 |
cursor.) |
| 2041 |
<programlisting> |
| 2042 |
EXEC SQL BEGIN DECLARE SECTION; |
| 2043 |
char dbaname[128]; |
| 2044 |
char datname[128]; |
| 2045 |
char *stmt = "SELECT u.usename as dbaname, d.datname " |
| 2046 |
" FROM pg_database d, pg_user u " |
| 2047 |
" WHERE d.datdba = u.usesysid"; |
| 2048 |
EXEC SQL END DECLARE SECTION; |
| 2049 |
|
| 2050 |
EXEC SQL CONNECT TO testdb AS con1 USER testuser; |
| 2051 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 2052 |
|
| 2053 |
EXEC SQL PREPARE stmt1 FROM :stmt; |
| 2054 |
|
| 2055 |
EXEC SQL DECLARE cursor1 CURSOR FOR stmt1; |
| 2056 |
EXEC SQL OPEN cursor1; |
| 2057 |
|
| 2058 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 2059 |
|
| 2060 |
while (1) |
| 2061 |
{ |
| 2062 |
EXEC SQL FETCH cursor1 INTO :dbaname,:datname; |
| 2063 |
printf("dbaname=%s, datname=%s\n", dbaname, datname); |
| 2064 |
} |
| 2065 |
|
| 2066 |
EXEC SQL CLOSE cursor1; |
| 2067 |
|
| 2068 |
EXEC SQL COMMIT; |
| 2069 |
EXEC SQL DISCONNECT ALL; |
| 2070 |
</programlisting> |
| 2071 |
</para> |
| 2072 |
</sect2> |
| 2073 |
</sect1> |
| 2074 |
|
| 2075 |
<sect1 id="ecpg-pgtypes"> |
| 2076 |
<title>pgtypes Library</title> |
| 2077 |
|
| 2078 |
<para> |
| 2079 |
The pgtypes library maps <productname>PostgreSQL</productname> database |
| 2080 |
types to C equivalents that can be used in C programs. It also offers |
| 2081 |
functions to do basic calculations with those types within C, i.e., without |
| 2082 |
the help of the <productname>PostgreSQL</productname> server. See the |
| 2083 |
following example: |
| 2084 |
<programlisting><![CDATA[ |
| 2085 |
EXEC SQL BEGIN DECLARE SECTION; |
| 2086 |
date date1; |
| 2087 |
timestamp ts1, tsout; |
| 2088 |
interval iv1; |
| 2089 |
char *out; |
| 2090 |
EXEC SQL END DECLARE SECTION; |
| 2091 |
|
| 2092 |
PGTYPESdate_today(&date1); |
| 2093 |
EXEC SQL SELECT started, duration INTO :ts1, :iv1 FROM datetbl WHERE d=:date1; |
| 2094 |
PGTYPEStimestamp_add_interval(&ts1, &iv1, &tsout); |
| 2095 |
out = PGTYPEStimestamp_to_asc(&tsout); |
| 2096 |
printf("Started + duration: %s\n", out); |
| 2097 |
PGTYPESchar_free(out); |
| 2098 |
]]> |
| 2099 |
</programlisting> |
| 2100 |
</para> |
| 2101 |
|
| 2102 |
<sect2 id="ecpg-pgtypes-cstrings"> |
| 2103 |
<title>Character Strings</title> |
| 2104 |
<para> |
| 2105 |
Some functions such as <function>PGTYPESnumeric_to_asc</function> return |
| 2106 |
a pointer to a freshly allocated character string. These results should be |
| 2107 |
freed with <function>PGTYPESchar_free</function> instead of |
| 2108 |
<function>free</function>. (This is important only on Windows, where |
| 2109 |
memory allocation and release sometimes need to be done by the same |
| 2110 |
library.) |
| 2111 |
</para> |
| 2112 |
</sect2> |
| 2113 |
|
| 2114 |
<sect2 id="ecpg-pgtypes-numeric"> |
| 2115 |
<title>The numeric Type</title> |
| 2116 |
<para> |
| 2117 |
The numeric type offers to do calculations with arbitrary precision. See |
| 2118 |
<xref linkend="datatype-numeric"/> for the equivalent type in the |
| 2119 |
<productname>PostgreSQL</productname> server. Because of the arbitrary precision this |
| 2120 |
variable needs to be able to expand and shrink dynamically. That's why you |
| 2121 |
can only create numeric variables on the heap, by means of the |
| 2122 |
<function>PGTYPESnumeric_new</function> and <function>PGTYPESnumeric_free</function> |
| 2123 |
functions. The decimal type, which is similar but limited in precision, |
| 2124 |
can be created on the stack as well as on the heap. |
| 2125 |
</para> |
| 2126 |
<para> |
| 2127 |
The following functions can be used to work with the numeric type: |
| 2128 |
<variablelist> |
| 2129 |
<varlistentry id="ecpg-pgtypes-numeric-new"> |
| 2130 |
<term><function>PGTYPESnumeric_new</function></term> |
| 2131 |
<listitem> |
| 2132 |
<para> |
| 2133 |
Request a pointer to a newly allocated numeric variable. |
| 2134 |
<synopsis> |
| 2135 |
numeric *PGTYPESnumeric_new(void); |
| 2136 |
</synopsis> |
| 2137 |
</para> |
| 2138 |
</listitem> |
| 2139 |
</varlistentry> |
| 2140 |
|
| 2141 |
<varlistentry id="ecpg-pgtypes-numeric-free"> |
| 2142 |
<term><function>PGTYPESnumeric_free</function></term> |
| 2143 |
<listitem> |
| 2144 |
<para> |
| 2145 |
Free a numeric type, release all of its memory. |
| 2146 |
<synopsis> |
| 2147 |
void PGTYPESnumeric_free(numeric *var); |
| 2148 |
</synopsis> |
| 2149 |
</para> |
| 2150 |
</listitem> |
| 2151 |
</varlistentry> |
| 2152 |
|
| 2153 |
<varlistentry id="ecpg-pgtypes-numeric-from-asc"> |
| 2154 |
<term><function>PGTYPESnumeric_from_asc</function></term> |
| 2155 |
<listitem> |
| 2156 |
<para> |
| 2157 |
Parse a numeric type from its string notation. |
| 2158 |
<synopsis> |
| 2159 |
numeric *PGTYPESnumeric_from_asc(char *str, char **endptr); |
| 2160 |
</synopsis> |
| 2161 |
Valid formats are for example: |
| 2162 |
<literal>-2</literal>, |
| 2163 |
<literal>.794</literal>, |
| 2164 |
<literal>+3.44</literal>, |
| 2165 |
<literal>592.49E07</literal> or |
| 2166 |
<literal>-32.84e-4</literal>. |
| 2167 |
If the value could be parsed successfully, a valid pointer is returned, |
| 2168 |
else the NULL pointer. At the moment ECPG always parses the complete |
| 2169 |
string and so it currently does not support to store the address of the |
| 2170 |
first invalid character in <literal>*endptr</literal>. You can safely |
| 2171 |
set <literal>endptr</literal> to NULL. |
| 2172 |
</para> |
| 2173 |
</listitem> |
| 2174 |
</varlistentry> |
| 2175 |
|
| 2176 |
<varlistentry id="ecpg-pgtypes-numeric-to-asc"> |
| 2177 |
<term><function>PGTYPESnumeric_to_asc</function></term> |
| 2178 |
<listitem> |
| 2179 |
<para> |
| 2180 |
Returns a pointer to a string allocated by <function>malloc</function> that contains the string |
| 2181 |
representation of the numeric type <literal>num</literal>. |
| 2182 |
<synopsis> |
| 2183 |
char *PGTYPESnumeric_to_asc(numeric *num, int dscale); |
| 2184 |
</synopsis> |
| 2185 |
The numeric value will be printed with <literal>dscale</literal> decimal |
| 2186 |
digits, with rounding applied if necessary. |
| 2187 |
The result must be freed with <function>PGTYPESchar_free()</function>. |
| 2188 |
</para> |
| 2189 |
</listitem> |
| 2190 |
</varlistentry> |
| 2191 |
|
| 2192 |
<varlistentry id="ecpg-pgtypes-numeric-add"> |
| 2193 |
<term><function>PGTYPESnumeric_add</function></term> |
| 2194 |
<listitem> |
| 2195 |
<para> |
| 2196 |
Add two numeric variables into a third one. |
| 2197 |
<synopsis> |
| 2198 |
int PGTYPESnumeric_add(numeric *var1, numeric *var2, numeric *result); |
| 2199 |
</synopsis> |
| 2200 |
The function adds the variables <literal>var1</literal> and |
| 2201 |
<literal>var2</literal> into the result variable |
| 2202 |
<literal>result</literal>. |
| 2203 |
The function returns 0 on success and -1 in case of error. |
| 2204 |
</para> |
| 2205 |
</listitem> |
| 2206 |
</varlistentry> |
| 2207 |
|
| 2208 |
<varlistentry id="ecpg-pgtypes-numeric-sub"> |
| 2209 |
<term><function>PGTYPESnumeric_sub</function></term> |
| 2210 |
<listitem> |
| 2211 |
<para> |
| 2212 |
Subtract two numeric variables and return the result in a third one. |
| 2213 |
<synopsis> |
| 2214 |
int PGTYPESnumeric_sub(numeric *var1, numeric *var2, numeric *result); |
| 2215 |
</synopsis> |
| 2216 |
The function subtracts the variable <literal>var2</literal> from |
| 2217 |
the variable <literal>var1</literal>. The result of the operation is |
| 2218 |
stored in the variable <literal>result</literal>. |
| 2219 |
The function returns 0 on success and -1 in case of error. |
| 2220 |
</para> |
| 2221 |
</listitem> |
| 2222 |
</varlistentry> |
| 2223 |
|
| 2224 |
<varlistentry id="ecpg-pgtypes-numeric-mul"> |
| 2225 |
<term><function>PGTYPESnumeric_mul</function></term> |
| 2226 |
<listitem> |
| 2227 |
<para> |
| 2228 |
Multiply two numeric variables and return the result in a third one. |
| 2229 |
<synopsis> |
| 2230 |
int PGTYPESnumeric_mul(numeric *var1, numeric *var2, numeric *result); |
| 2231 |
</synopsis> |
| 2232 |
The function multiplies the variables <literal>var1</literal> and |
| 2233 |
<literal>var2</literal>. The result of the operation is stored in the |
| 2234 |
variable <literal>result</literal>. |
| 2235 |
The function returns 0 on success and -1 in case of error. |
| 2236 |
</para> |
| 2237 |
</listitem> |
| 2238 |
</varlistentry> |
| 2239 |
|
| 2240 |
<varlistentry id="ecpg-pgtypes-numeric-div"> |
| 2241 |
<term><function>PGTYPESnumeric_div</function></term> |
| 2242 |
<listitem> |
| 2243 |
<para> |
| 2244 |
Divide two numeric variables and return the result in a third one. |
| 2245 |
<synopsis> |
| 2246 |
int PGTYPESnumeric_div(numeric *var1, numeric *var2, numeric *result); |
| 2247 |
</synopsis> |
| 2248 |
The function divides the variables <literal>var1</literal> by |
| 2249 |
<literal>var2</literal>. The result of the operation is stored in the |
| 2250 |
variable <literal>result</literal>. |
| 2251 |
The function returns 0 on success and -1 in case of error. |
| 2252 |
</para> |
| 2253 |
</listitem> |
| 2254 |
</varlistentry> |
| 2255 |
|
| 2256 |
<varlistentry id="ecpg-pgtypes-numeric-cmp"> |
| 2257 |
<term><function>PGTYPESnumeric_cmp</function></term> |
| 2258 |
<listitem> |
| 2259 |
<para> |
| 2260 |
Compare two numeric variables. |
| 2261 |
<synopsis> |
| 2262 |
int PGTYPESnumeric_cmp(numeric *var1, numeric *var2) |
| 2263 |
</synopsis> |
| 2264 |
This function compares two numeric variables. In case of error, |
| 2265 |
<literal>INT_MAX</literal> is returned. On success, the function |
| 2266 |
returns one of three possible results: |
| 2267 |
<itemizedlist> |
| 2268 |
<listitem> |
| 2269 |
<para> |
| 2270 |
1, if <literal>var1</literal> is bigger than <literal>var2</literal> |
| 2271 |
</para> |
| 2272 |
</listitem> |
| 2273 |
<listitem> |
| 2274 |
<para> |
| 2275 |
-1, if <literal>var1</literal> is smaller than <literal>var2</literal> |
| 2276 |
</para> |
| 2277 |
</listitem> |
| 2278 |
<listitem> |
| 2279 |
<para> |
| 2280 |
0, if <literal>var1</literal> and <literal>var2</literal> are equal |
| 2281 |
</para> |
| 2282 |
</listitem> |
| 2283 |
</itemizedlist> |
| 2284 |
</para> |
| 2285 |
</listitem> |
| 2286 |
</varlistentry> |
| 2287 |
|
| 2288 |
<varlistentry id="ecpg-pgtypes-numeric-from-int"> |
| 2289 |
<term><function>PGTYPESnumeric_from_int</function></term> |
| 2290 |
<listitem> |
| 2291 |
<para> |
| 2292 |
Convert an int variable to a numeric variable. |
| 2293 |
<synopsis> |
| 2294 |
int PGTYPESnumeric_from_int(signed int int_val, numeric *var); |
| 2295 |
</synopsis> |
| 2296 |
This function accepts a variable of type signed int and stores it |
| 2297 |
in the numeric variable <literal>var</literal>. Upon success, 0 is returned and |
| 2298 |
-1 in case of a failure. |
| 2299 |
</para> |
| 2300 |
</listitem> |
| 2301 |
</varlistentry> |
| 2302 |
|
| 2303 |
<varlistentry id="ecpg-pgtypes-numeric-from-long"> |
| 2304 |
<term><function>PGTYPESnumeric_from_long</function></term> |
| 2305 |
<listitem> |
| 2306 |
<para> |
| 2307 |
Convert a long int variable to a numeric variable. |
| 2308 |
<synopsis> |
| 2309 |
int PGTYPESnumeric_from_long(signed long int long_val, numeric *var); |
| 2310 |
</synopsis> |
| 2311 |
This function accepts a variable of type signed long int and stores it |
| 2312 |
in the numeric variable <literal>var</literal>. Upon success, 0 is returned and |
| 2313 |
-1 in case of a failure. |
| 2314 |
</para> |
| 2315 |
</listitem> |
| 2316 |
</varlistentry> |
| 2317 |
|
| 2318 |
<varlistentry id="ecpg-pgtypes-numeric-copy"> |
| 2319 |
<term><function>PGTYPESnumeric_copy</function></term> |
| 2320 |
<listitem> |
| 2321 |
<para> |
| 2322 |
Copy over one numeric variable into another one. |
| 2323 |
<synopsis> |
| 2324 |
int PGTYPESnumeric_copy(numeric *src, numeric *dst); |
| 2325 |
</synopsis> |
| 2326 |
This function copies over the value of the variable that |
| 2327 |
<literal>src</literal> points to into the variable that <literal>dst</literal> |
| 2328 |
points to. It returns 0 on success and -1 if an error occurs. |
| 2329 |
</para> |
| 2330 |
</listitem> |
| 2331 |
</varlistentry> |
| 2332 |
|
| 2333 |
<varlistentry id="ecpg-pgtypes-numeric-from-double"> |
| 2334 |
<term><function>PGTYPESnumeric_from_double</function></term> |
| 2335 |
<listitem> |
| 2336 |
<para> |
| 2337 |
Convert a variable of type double to a numeric. |
| 2338 |
<synopsis> |
| 2339 |
int PGTYPESnumeric_from_double(double d, numeric *dst); |
| 2340 |
</synopsis> |
| 2341 |
This function accepts a variable of type double and stores the result |
| 2342 |
in the variable that <literal>dst</literal> points to. It returns 0 on success |
| 2343 |
and -1 if an error occurs. |
| 2344 |
</para> |
| 2345 |
</listitem> |
| 2346 |
</varlistentry> |
| 2347 |
|
| 2348 |
<varlistentry id="ecpg-pgtypes-numeric-to-double"> |
| 2349 |
<term><function>PGTYPESnumeric_to_double</function></term> |
| 2350 |
<listitem> |
| 2351 |
<para> |
| 2352 |
Convert a variable of type numeric to double. |
| 2353 |
<synopsis> |
| 2354 |
int PGTYPESnumeric_to_double(numeric *nv, double *dp) |
| 2355 |
</synopsis> |
| 2356 |
The function converts the numeric value from the variable that |
| 2357 |
<literal>nv</literal> points to into the double variable that <literal>dp</literal> points |
| 2358 |
to. It returns 0 on success and -1 if an error occurs, including |
| 2359 |
overflow. On overflow, the global variable <literal>errno</literal> will be set |
| 2360 |
to <literal>PGTYPES_NUM_OVERFLOW</literal> additionally. |
| 2361 |
</para> |
| 2362 |
</listitem> |
| 2363 |
</varlistentry> |
| 2364 |
|
| 2365 |
<varlistentry id="ecpg-pgtypes-numeric-to-int"> |
| 2366 |
<term><function>PGTYPESnumeric_to_int</function></term> |
| 2367 |
<listitem> |
| 2368 |
<para> |
| 2369 |
Convert a variable of type numeric to int. |
| 2370 |
<synopsis> |
| 2371 |
int PGTYPESnumeric_to_int(numeric *nv, int *ip); |
| 2372 |
</synopsis> |
| 2373 |
The function converts the numeric value from the variable that |
| 2374 |
<literal>nv</literal> points to into the integer variable that <literal>ip</literal> |
| 2375 |
points to. It returns 0 on success and -1 if an error occurs, including |
| 2376 |
overflow. On overflow, the global variable <literal>errno</literal> will be set |
| 2377 |
to <literal>PGTYPES_NUM_OVERFLOW</literal> additionally. |
| 2378 |
</para> |
| 2379 |
</listitem> |
| 2380 |
</varlistentry> |
| 2381 |
|
| 2382 |
<varlistentry id="ecpg-pgtypes-numeric-to-long"> |
| 2383 |
<term><function>PGTYPESnumeric_to_long</function></term> |
| 2384 |
<listitem> |
| 2385 |
<para> |
| 2386 |
Convert a variable of type numeric to long. |
| 2387 |
<synopsis> |
| 2388 |
int PGTYPESnumeric_to_long(numeric *nv, long *lp); |
| 2389 |
</synopsis> |
| 2390 |
The function converts the numeric value from the variable that |
| 2391 |
<literal>nv</literal> points to into the long integer variable that |
| 2392 |
<literal>lp</literal> points to. It returns 0 on success and -1 if an error |
| 2393 |
occurs, including overflow and underflow. On overflow, the global variable |
| 2394 |
<literal>errno</literal> will be set to <literal>PGTYPES_NUM_OVERFLOW</literal> |
| 2395 |
and on underflow <literal>errno</literal> will be set to |
| 2396 |
<literal>PGTYPES_NUM_UNDERFLOW</literal>. |
| 2397 |
</para> |
| 2398 |
</listitem> |
| 2399 |
</varlistentry> |
| 2400 |
|
| 2401 |
<varlistentry id="ecpg-pgtypes-numeric-to-decimal"> |
| 2402 |
<term><function>PGTYPESnumeric_to_decimal</function></term> |
| 2403 |
<listitem> |
| 2404 |
<para> |
| 2405 |
Convert a variable of type numeric to decimal. |
| 2406 |
<synopsis> |
| 2407 |
int PGTYPESnumeric_to_decimal(numeric *src, decimal *dst); |
| 2408 |
</synopsis> |
| 2409 |
The function converts the numeric value from the variable that |
| 2410 |
<literal>src</literal> points to into the decimal variable that |
| 2411 |
<literal>dst</literal> points to. It returns 0 on success and -1 if an error |
| 2412 |
occurs, including overflow. On overflow, the global variable |
| 2413 |
<literal>errno</literal> will be set to <literal>PGTYPES_NUM_OVERFLOW</literal> |
| 2414 |
additionally. |
| 2415 |
</para> |
| 2416 |
</listitem> |
| 2417 |
</varlistentry> |
| 2418 |
|
| 2419 |
<varlistentry id="ecpg-pgtypes-numeric-from-decimal"> |
| 2420 |
<term><function>PGTYPESnumeric_from_decimal</function></term> |
| 2421 |
<listitem> |
| 2422 |
<para> |
| 2423 |
Convert a variable of type decimal to numeric. |
| 2424 |
<synopsis> |
| 2425 |
int PGTYPESnumeric_from_decimal(decimal *src, numeric *dst); |
| 2426 |
</synopsis> |
| 2427 |
The function converts the decimal value from the variable that |
| 2428 |
<literal>src</literal> points to into the numeric variable that |
| 2429 |
<literal>dst</literal> points to. It returns 0 on success and -1 if an error |
| 2430 |
occurs. Since the decimal type is implemented as a limited version of |
| 2431 |
the numeric type, overflow cannot occur with this conversion. |
| 2432 |
</para> |
| 2433 |
</listitem> |
| 2434 |
</varlistentry> |
| 2435 |
</variablelist> |
| 2436 |
</para> |
| 2437 |
</sect2> |
| 2438 |
|
| 2439 |
<sect2 id="ecpg-pgtypes-date"> |
| 2440 |
<title>The date Type</title> |
| 2441 |
<para> |
| 2442 |
The date type in C enables your programs to deal with data of the SQL type |
| 2443 |
date. See <xref linkend="datatype-datetime"/> for the equivalent type in the |
| 2444 |
<productname>PostgreSQL</productname> server. |
| 2445 |
</para> |
| 2446 |
<para> |
| 2447 |
The following functions can be used to work with the date type: |
| 2448 |
<variablelist> |
| 2449 |
<varlistentry id="pgtypesdatefromtimestamp"> |
| 2450 |
<term><function>PGTYPESdate_from_timestamp</function></term> |
| 2451 |
<listitem> |
| 2452 |
<para> |
| 2453 |
Extract the date part from a timestamp. |
| 2454 |
<synopsis> |
| 2455 |
date PGTYPESdate_from_timestamp(timestamp dt); |
| 2456 |
</synopsis> |
| 2457 |
The function receives a timestamp as its only argument and returns the |
| 2458 |
extracted date part from this timestamp. |
| 2459 |
</para> |
| 2460 |
</listitem> |
| 2461 |
</varlistentry> |
| 2462 |
|
| 2463 |
<varlistentry id="pgtypesdatefromasc"> |
| 2464 |
<term><function>PGTYPESdate_from_asc</function></term> |
| 2465 |
<listitem> |
| 2466 |
<para> |
| 2467 |
Parse a date from its textual representation. |
| 2468 |
<synopsis> |
| 2469 |
date PGTYPESdate_from_asc(char *str, char **endptr); |
| 2470 |
</synopsis> |
| 2471 |
The function receives a C char* string <literal>str</literal> and a pointer to |
| 2472 |
a C char* string <literal>endptr</literal>. At the moment ECPG always parses |
| 2473 |
the complete string and so it currently does not support to store the |
| 2474 |
address of the first invalid character in <literal>*endptr</literal>. |
| 2475 |
You can safely set <literal>endptr</literal> to NULL. |
| 2476 |
</para> |
| 2477 |
<para> |
| 2478 |
Note that the function always assumes MDY-formatted dates and there is |
| 2479 |
currently no variable to change that within ECPG. |
| 2480 |
</para> |
| 2481 |
<para> |
| 2482 |
<xref linkend="ecpg-pgtypesdate-from-asc-table"/> shows the allowed input formats. |
| 2483 |
</para> |
| 2484 |
<table id="ecpg-pgtypesdate-from-asc-table"> |
| 2485 |
<title>Valid Input Formats for <function>PGTYPESdate_from_asc</function></title> |
| 2486 |
<tgroup cols="2"> |
| 2487 |
<thead> |
| 2488 |
<row> |
| 2489 |
<entry>Input</entry> |
| 2490 |
<entry>Result</entry> |
| 2491 |
</row> |
| 2492 |
</thead> |
| 2493 |
<tbody> |
| 2494 |
<row> |
| 2495 |
<entry><literal>January 8, 1999</literal></entry> |
| 2496 |
<entry><literal>January 8, 1999</literal></entry> |
| 2497 |
</row> |
| 2498 |
<row> |
| 2499 |
<entry><literal>1999-01-08</literal></entry> |
| 2500 |
<entry><literal>January 8, 1999</literal></entry> |
| 2501 |
</row> |
| 2502 |
<row> |
| 2503 |
<entry><literal>1/8/1999</literal></entry> |
| 2504 |
<entry><literal>January 8, 1999</literal></entry> |
| 2505 |
</row> |
| 2506 |
<row> |
| 2507 |
<entry><literal>1/18/1999</literal></entry> |
| 2508 |
<entry><literal>January 18, 1999</literal></entry> |
| 2509 |
</row> |
| 2510 |
<row> |
| 2511 |
<entry><literal>01/02/03</literal></entry> |
| 2512 |
<entry><literal>February 1, 2003</literal></entry> |
| 2513 |
</row> |
| 2514 |
<row> |
| 2515 |
<entry><literal>1999-Jan-08</literal></entry> |
| 2516 |
<entry><literal>January 8, 1999</literal></entry> |
| 2517 |
</row> |
| 2518 |
<row> |
| 2519 |
<entry><literal>Jan-08-1999</literal></entry> |
| 2520 |
<entry><literal>January 8, 1999</literal></entry> |
| 2521 |
</row> |
| 2522 |
<row> |
| 2523 |
<entry><literal>08-Jan-1999</literal></entry> |
| 2524 |
<entry><literal>January 8, 1999</literal></entry> |
| 2525 |
</row> |
| 2526 |
<row> |
| 2527 |
<entry><literal>99-Jan-08</literal></entry> |
| 2528 |
<entry><literal>January 8, 1999</literal></entry> |
| 2529 |
</row> |
| 2530 |
<row> |
| 2531 |
<entry><literal>08-Jan-99</literal></entry> |
| 2532 |
<entry><literal>January 8, 1999</literal></entry> |
| 2533 |
</row> |
| 2534 |
<row> |
| 2535 |
<entry><literal>08-Jan-06</literal></entry> |
| 2536 |
<entry><literal>January 8, 2006</literal></entry> |
| 2537 |
</row> |
| 2538 |
<row> |
| 2539 |
<entry><literal>Jan-08-99</literal></entry> |
| 2540 |
<entry><literal>January 8, 1999</literal></entry> |
| 2541 |
</row> |
| 2542 |
<row> |
| 2543 |
<entry><literal>19990108</literal></entry> |
| 2544 |
<entry><literal>ISO 8601; January 8, 1999</literal></entry> |
| 2545 |
</row> |
| 2546 |
<row> |
| 2547 |
<entry><literal>990108</literal></entry> |
| 2548 |
<entry><literal>ISO 8601; January 8, 1999</literal></entry> |
| 2549 |
</row> |
| 2550 |
<row> |
| 2551 |
<entry><literal>1999.008</literal></entry> |
| 2552 |
<entry><literal>year and day of year</literal></entry> |
| 2553 |
</row> |
| 2554 |
<row> |
| 2555 |
<entry><literal>J2451187</literal></entry> |
| 2556 |
<entry><literal>Julian day</literal></entry> |
| 2557 |
</row> |
| 2558 |
<row> |
| 2559 |
<entry><literal>January 8, 99 BC</literal></entry> |
| 2560 |
<entry><literal>year 99 before the Common Era</literal></entry> |
| 2561 |
</row> |
| 2562 |
</tbody> |
| 2563 |
</tgroup> |
| 2564 |
</table> |
| 2565 |
</listitem> |
| 2566 |
</varlistentry> |
| 2567 |
|
| 2568 |
<varlistentry id="pgtypesdatetoasc"> |
| 2569 |
<term><function>PGTYPESdate_to_asc</function></term> |
| 2570 |
<listitem> |
| 2571 |
<para> |
| 2572 |
Return the textual representation of a date variable. |
| 2573 |
<synopsis> |
| 2574 |
char *PGTYPESdate_to_asc(date dDate); |
| 2575 |
</synopsis> |
| 2576 |
The function receives the date <literal>dDate</literal> as its only parameter. |
| 2577 |
It will output the date in the form <literal>1999-01-18</literal>, i.e., in the |
| 2578 |
<literal>YYYY-MM-DD</literal> format. |
| 2579 |
The result must be freed with <function>PGTYPESchar_free()</function>. |
| 2580 |
</para> |
| 2581 |
</listitem> |
| 2582 |
</varlistentry> |
| 2583 |
|
| 2584 |
<varlistentry id="pgtypesdatejulmdy"> |
| 2585 |
<term><function>PGTYPESdate_julmdy</function></term> |
| 2586 |
<listitem> |
| 2587 |
<para> |
| 2588 |
Extract the values for the day, the month and the year from a variable |
| 2589 |
of type date. |
| 2590 |
<synopsis> |
| 2591 |
void PGTYPESdate_julmdy(date d, int *mdy); |
| 2592 |
</synopsis> |
| 2593 |
<!-- almost same description as for rjulmdy() --> |
| 2594 |
The function receives the date <literal>d</literal> and a pointer to an array |
| 2595 |
of 3 integer values <literal>mdy</literal>. The variable name indicates |
| 2596 |
the sequential order: <literal>mdy[0]</literal> will be set to contain the |
| 2597 |
number of the month, <literal>mdy[1]</literal> will be set to the value of the |
| 2598 |
day and <literal>mdy[2]</literal> will contain the year. |
| 2599 |
</para> |
| 2600 |
</listitem> |
| 2601 |
</varlistentry> |
| 2602 |
|
| 2603 |
<varlistentry id="pgtypesdatemdyjul"> |
| 2604 |
<term><function>PGTYPESdate_mdyjul</function></term> |
| 2605 |
<listitem> |
| 2606 |
<para> |
| 2607 |
Create a date value from an array of 3 integers that specify the |
| 2608 |
day, the month and the year of the date. |
| 2609 |
<synopsis> |
| 2610 |
void PGTYPESdate_mdyjul(int *mdy, date *jdate); |
| 2611 |
</synopsis> |
| 2612 |
The function receives the array of the 3 integers (<literal>mdy</literal>) as |
| 2613 |
its first argument and as its second argument a pointer to a variable |
| 2614 |
of type date that should hold the result of the operation. |
| 2615 |
</para> |
| 2616 |
</listitem> |
| 2617 |
</varlistentry> |
| 2618 |
|
| 2619 |
<varlistentry id="pgtypesdatedayofweek"> |
| 2620 |
<term><function>PGTYPESdate_dayofweek</function></term> |
| 2621 |
<listitem> |
| 2622 |
<para> |
| 2623 |
Return a number representing the day of the week for a date value. |
| 2624 |
<synopsis> |
| 2625 |
int PGTYPESdate_dayofweek(date d); |
| 2626 |
</synopsis> |
| 2627 |
The function receives the date variable <literal>d</literal> as its only |
| 2628 |
argument and returns an integer that indicates the day of the week for |
| 2629 |
this date. |
| 2630 |
<itemizedlist> |
| 2631 |
<listitem> |
| 2632 |
<para> |
| 2633 |
0 - Sunday |
| 2634 |
</para> |
| 2635 |
</listitem> |
| 2636 |
<listitem> |
| 2637 |
<para> |
| 2638 |
1 - Monday |
| 2639 |
</para> |
| 2640 |
</listitem> |
| 2641 |
<listitem> |
| 2642 |
<para> |
| 2643 |
2 - Tuesday |
| 2644 |
</para> |
| 2645 |
</listitem> |
| 2646 |
<listitem> |
| 2647 |
<para> |
| 2648 |
3 - Wednesday |
| 2649 |
</para> |
| 2650 |
</listitem> |
| 2651 |
<listitem> |
| 2652 |
<para> |
| 2653 |
4 - Thursday |
| 2654 |
</para> |
| 2655 |
</listitem> |
| 2656 |
<listitem> |
| 2657 |
<para> |
| 2658 |
5 - Friday |
| 2659 |
</para> |
| 2660 |
</listitem> |
| 2661 |
<listitem> |
| 2662 |
<para> |
| 2663 |
6 - Saturday |
| 2664 |
</para> |
| 2665 |
</listitem> |
| 2666 |
</itemizedlist> |
| 2667 |
</para> |
| 2668 |
</listitem> |
| 2669 |
</varlistentry> |
| 2670 |
|
| 2671 |
<varlistentry id="pgtypesdatetoday"> |
| 2672 |
<term><function>PGTYPESdate_today</function></term> |
| 2673 |
<listitem> |
| 2674 |
<para> |
| 2675 |
Get the current date. |
| 2676 |
<synopsis> |
| 2677 |
void PGTYPESdate_today(date *d); |
| 2678 |
</synopsis> |
| 2679 |
The function receives a pointer to a date variable (<literal>d</literal>) |
| 2680 |
that it sets to the current date. |
| 2681 |
</para> |
| 2682 |
</listitem> |
| 2683 |
</varlistentry> |
| 2684 |
|
| 2685 |
<varlistentry id="pgtypesdatefmtasc"> |
| 2686 |
<term><function>PGTYPESdate_fmt_asc</function></term> |
| 2687 |
<listitem> |
| 2688 |
<para> |
| 2689 |
Convert a variable of type date to its textual representation using a |
| 2690 |
format mask. |
| 2691 |
<synopsis> |
| 2692 |
int PGTYPESdate_fmt_asc(date dDate, char *fmtstring, char *outbuf); |
| 2693 |
</synopsis> |
| 2694 |
The function receives the date to convert (<literal>dDate</literal>), the |
| 2695 |
format mask (<literal>fmtstring</literal>) and the string that will hold the |
| 2696 |
textual representation of the date (<literal>outbuf</literal>). |
| 2697 |
</para> |
| 2698 |
<para> |
| 2699 |
On success, 0 is returned and a negative value if an error occurred. |
| 2700 |
</para> |
| 2701 |
<para> |
| 2702 |
The following literals are the field specifiers you can use: |
| 2703 |
<itemizedlist> |
| 2704 |
<listitem> |
| 2705 |
<para> |
| 2706 |
<literal>dd</literal> - The number of the day of the month. |
| 2707 |
</para> |
| 2708 |
</listitem> |
| 2709 |
<listitem> |
| 2710 |
<para> |
| 2711 |
<literal>mm</literal> - The number of the month of the year. |
| 2712 |
</para> |
| 2713 |
</listitem> |
| 2714 |
<listitem> |
| 2715 |
<para> |
| 2716 |
<literal>yy</literal> - The number of the year as a two digit number. |
| 2717 |
</para> |
| 2718 |
</listitem> |
| 2719 |
<listitem> |
| 2720 |
<para> |
| 2721 |
<literal>yyyy</literal> - The number of the year as a four digit number. |
| 2722 |
</para> |
| 2723 |
</listitem> |
| 2724 |
<listitem> |
| 2725 |
<para> |
| 2726 |
<literal>ddd</literal> - The name of the day (abbreviated). |
| 2727 |
</para> |
| 2728 |
</listitem> |
| 2729 |
<listitem> |
| 2730 |
<para> |
| 2731 |
<literal>mmm</literal> - The name of the month (abbreviated). |
| 2732 |
</para> |
| 2733 |
</listitem> |
| 2734 |
</itemizedlist> |
| 2735 |
All other characters are copied 1:1 to the output string. |
| 2736 |
</para> |
| 2737 |
<para> |
| 2738 |
<xref linkend="ecpg-pgtypesdate-fmt-asc-example-table"/> indicates a few possible formats. This will give |
| 2739 |
you an idea of how to use this function. All output lines are based on |
| 2740 |
the same date: November 23, 1959. |
| 2741 |
</para> |
| 2742 |
<table id="ecpg-pgtypesdate-fmt-asc-example-table"> |
| 2743 |
<title>Valid Input Formats for <function>PGTYPESdate_fmt_asc</function></title> |
| 2744 |
<tgroup cols="2"> |
| 2745 |
<thead> |
| 2746 |
<row> |
| 2747 |
<entry>Format</entry> |
| 2748 |
<entry>Result</entry> |
| 2749 |
</row> |
| 2750 |
</thead> |
| 2751 |
<tbody> |
| 2752 |
<row> |
| 2753 |
<entry><literal>mmddyy</literal></entry> |
| 2754 |
<entry><literal>112359</literal></entry> |
| 2755 |
</row> |
| 2756 |
<row> |
| 2757 |
<entry><literal>ddmmyy</literal></entry> |
| 2758 |
<entry><literal>231159</literal></entry> |
| 2759 |
</row> |
| 2760 |
<row> |
| 2761 |
<entry><literal>yymmdd</literal></entry> |
| 2762 |
<entry><literal>591123</literal></entry> |
| 2763 |
</row> |
| 2764 |
<row> |
| 2765 |
<entry><literal>yy/mm/dd</literal></entry> |
| 2766 |
<entry><literal>59/11/23</literal></entry> |
| 2767 |
</row> |
| 2768 |
<row> |
| 2769 |
<entry><literal>yy mm dd</literal></entry> |
| 2770 |
<entry><literal>59 11 23</literal></entry> |
| 2771 |
</row> |
| 2772 |
<row> |
| 2773 |
<entry><literal>yy.mm.dd</literal></entry> |
| 2774 |
<entry><literal>59.11.23</literal></entry> |
| 2775 |
</row> |
| 2776 |
<row> |
| 2777 |
<entry><literal>.mm.yyyy.dd.</literal></entry> |
| 2778 |
<entry><literal>.11.1959.23.</literal></entry> |
| 2779 |
</row> |
| 2780 |
<row> |
| 2781 |
<entry><literal>mmm. dd, yyyy</literal></entry> |
| 2782 |
<entry><literal>Nov. 23, 1959</literal></entry> |
| 2783 |
</row> |
| 2784 |
<row> |
| 2785 |
<entry><literal>mmm dd yyyy</literal></entry> |
| 2786 |
<entry><literal>Nov 23 1959</literal></entry> |
| 2787 |
</row> |
| 2788 |
<row> |
| 2789 |
<entry><literal>yyyy dd mm</literal></entry> |
| 2790 |
<entry><literal>1959 23 11</literal></entry> |
| 2791 |
</row> |
| 2792 |
<row> |
| 2793 |
<entry><literal>ddd, mmm. dd, yyyy</literal></entry> |
| 2794 |
<entry><literal>Mon, Nov. 23, 1959</literal></entry> |
| 2795 |
</row> |
| 2796 |
<row> |
| 2797 |
<entry><literal>(ddd) mmm. dd, yyyy</literal></entry> |
| 2798 |
<entry><literal>(Mon) Nov. 23, 1959</literal></entry> |
| 2799 |
</row> |
| 2800 |
</tbody> |
| 2801 |
</tgroup> |
| 2802 |
</table> |
| 2803 |
</listitem> |
| 2804 |
</varlistentry> |
| 2805 |
|
| 2806 |
<varlistentry id="pgtypesdatedefmtasc"> |
| 2807 |
<term><function>PGTYPESdate_defmt_asc</function></term> |
| 2808 |
<listitem> |
| 2809 |
<para> |
| 2810 |
Use a format mask to convert a C <type>char*</type> string to a value of type |
| 2811 |
date. |
| 2812 |
<synopsis> |
| 2813 |
int PGTYPESdate_defmt_asc(date *d, char *fmt, char *str); |
| 2814 |
</synopsis> |
| 2815 |
<!-- same description as rdefmtdate --> |
| 2816 |
The function receives a pointer to the date value that should hold the |
| 2817 |
result of the operation (<literal>d</literal>), the format mask to use for |
| 2818 |
parsing the date (<literal>fmt</literal>) and the C char* string containing |
| 2819 |
the textual representation of the date (<literal>str</literal>). The textual |
| 2820 |
representation is expected to match the format mask. However you do not |
| 2821 |
need to have a 1:1 mapping of the string to the format mask. The |
| 2822 |
function only analyzes the sequential order and looks for the literals |
| 2823 |
<literal>yy</literal> or <literal>yyyy</literal> that indicate the |
| 2824 |
position of the year, <literal>mm</literal> to indicate the position of |
| 2825 |
the month and <literal>dd</literal> to indicate the position of the |
| 2826 |
day. |
| 2827 |
</para> |
| 2828 |
<para> |
| 2829 |
<xref linkend="ecpg-rdefmtdate-example-table"/> indicates a few possible formats. This will give |
| 2830 |
you an idea of how to use this function. |
| 2831 |
</para> |
| 2832 |
<table id="ecpg-rdefmtdate-example-table"> |
| 2833 |
<title>Valid Input Formats for <function>rdefmtdate</function></title> |
| 2834 |
<tgroup cols="3"> |
| 2835 |
<thead> |
| 2836 |
<row> |
| 2837 |
<entry>Format</entry> |
| 2838 |
<entry>String</entry> |
| 2839 |
<entry>Result</entry> |
| 2840 |
</row> |
| 2841 |
</thead> |
| 2842 |
<tbody> |
| 2843 |
<row> |
| 2844 |
<entry><literal>ddmmyy</literal></entry> |
| 2845 |
<entry><literal>21-2-54</literal></entry> |
| 2846 |
<entry><literal>1954-02-21</literal></entry> |
| 2847 |
</row> |
| 2848 |
<row> |
| 2849 |
<entry><literal>ddmmyy</literal></entry> |
| 2850 |
<entry><literal>2-12-54</literal></entry> |
| 2851 |
<entry><literal>1954-12-02</literal></entry> |
| 2852 |
</row> |
| 2853 |
<row> |
| 2854 |
<entry><literal>ddmmyy</literal></entry> |
| 2855 |
<entry><literal>20111954</literal></entry> |
| 2856 |
<entry><literal>1954-11-20</literal></entry> |
| 2857 |
</row> |
| 2858 |
<row> |
| 2859 |
<entry><literal>ddmmyy</literal></entry> |
| 2860 |
<entry><literal>130464</literal></entry> |
| 2861 |
<entry><literal>1964-04-13</literal></entry> |
| 2862 |
</row> |
| 2863 |
<row> |
| 2864 |
<entry><literal>mmm.dd.yyyy</literal></entry> |
| 2865 |
<entry><literal>MAR-12-1967</literal></entry> |
| 2866 |
<entry><literal>1967-03-12</literal></entry> |
| 2867 |
</row> |
| 2868 |
<row> |
| 2869 |
<entry><literal>yy/mm/dd</literal></entry> |
| 2870 |
<entry><literal>1954, February 3rd</literal></entry> |
| 2871 |
<entry><literal>1954-02-03</literal></entry> |
| 2872 |
</row> |
| 2873 |
<row> |
| 2874 |
<entry><literal>mmm.dd.yyyy</literal></entry> |
| 2875 |
<entry><literal>041269</literal></entry> |
| 2876 |
<entry><literal>1969-04-12</literal></entry> |
| 2877 |
</row> |
| 2878 |
<row> |
| 2879 |
<entry><literal>yy/mm/dd</literal></entry> |
| 2880 |
<entry><literal>In the year 2525, in the month of July, mankind will be alive on the 28th day</literal></entry> |
| 2881 |
<entry><literal>2525-07-28</literal></entry> |
| 2882 |
</row> |
| 2883 |
<row> |
| 2884 |
<entry><literal>dd-mm-yy</literal></entry> |
| 2885 |
<entry><literal>I said on the 28th of July in the year 2525</literal></entry> |
| 2886 |
<entry><literal>2525-07-28</literal></entry> |
| 2887 |
</row> |
| 2888 |
<row> |
| 2889 |
<entry><literal>mmm.dd.yyyy</literal></entry> |
| 2890 |
<entry><literal>9/14/58</literal></entry> |
| 2891 |
<entry><literal>1958-09-14</literal></entry> |
| 2892 |
</row> |
| 2893 |
<row> |
| 2894 |
<entry><literal>yy/mm/dd</literal></entry> |
| 2895 |
<entry><literal>47/03/29</literal></entry> |
| 2896 |
<entry><literal>1947-03-29</literal></entry> |
| 2897 |
</row> |
| 2898 |
<row> |
| 2899 |
<entry><literal>mmm.dd.yyyy</literal></entry> |
| 2900 |
<entry><literal>oct 28 1975</literal></entry> |
| 2901 |
<entry><literal>1975-10-28</literal></entry> |
| 2902 |
</row> |
| 2903 |
<row> |
| 2904 |
<entry><literal>mmddyy</literal></entry> |
| 2905 |
<entry><literal>Nov 14th, 1985</literal></entry> |
| 2906 |
<entry><literal>1985-11-14</literal></entry> |
| 2907 |
</row> |
| 2908 |
</tbody> |
| 2909 |
</tgroup> |
| 2910 |
</table> |
| 2911 |
</listitem> |
| 2912 |
</varlistentry> |
| 2913 |
</variablelist> |
| 2914 |
</para> |
| 2915 |
</sect2> |
| 2916 |
|
| 2917 |
<sect2 id="ecpg-pgtypes-timestamp"> |
| 2918 |
<title>The timestamp Type</title> |
| 2919 |
<para> |
| 2920 |
The timestamp type in C enables your programs to deal with data of the SQL |
| 2921 |
type timestamp. See <xref linkend="datatype-datetime"/> for the equivalent |
| 2922 |
type in the <productname>PostgreSQL</productname> server. |
| 2923 |
</para> |
| 2924 |
<para> |
| 2925 |
The following functions can be used to work with the timestamp type: |
| 2926 |
<variablelist> |
| 2927 |
<varlistentry id="pgtypestimestampfromasc"> |
| 2928 |
<term><function>PGTYPEStimestamp_from_asc</function></term> |
| 2929 |
<listitem> |
| 2930 |
<para> |
| 2931 |
Parse a timestamp from its textual representation into a timestamp |
| 2932 |
variable. |
| 2933 |
<synopsis> |
| 2934 |
timestamp PGTYPEStimestamp_from_asc(char *str, char **endptr); |
| 2935 |
</synopsis> |
| 2936 |
The function receives the string to parse (<literal>str</literal>) and a |
| 2937 |
pointer to a C char* (<literal>endptr</literal>). |
| 2938 |
At the moment ECPG always parses |
| 2939 |
the complete string and so it currently does not support to store the |
| 2940 |
address of the first invalid character in <literal>*endptr</literal>. |
| 2941 |
You can safely set <literal>endptr</literal> to NULL. |
| 2942 |
</para> |
| 2943 |
<para> |
| 2944 |
The function returns the parsed timestamp on success. On error, |
| 2945 |
<literal>PGTYPESInvalidTimestamp</literal> is returned and <varname>errno</varname> is |
| 2946 |
set to <literal>PGTYPES_TS_BAD_TIMESTAMP</literal>. See <xref linkend="pgtypesinvalidtimestamp"/> for important notes on this value. |
| 2947 |
</para> |
| 2948 |
<para> |
| 2949 |
In general, the input string can contain any combination of an allowed |
| 2950 |
date specification, a whitespace character and an allowed time |
| 2951 |
specification. Note that time zones are not supported by ECPG. It can |
| 2952 |
parse them but does not apply any calculation as the |
| 2953 |
<productname>PostgreSQL</productname> server does for example. Timezone |
| 2954 |
specifiers are silently discarded. |
| 2955 |
</para> |
| 2956 |
<para> |
| 2957 |
<xref linkend="ecpg-pgtypestimestamp-from-asc-example-table"/> contains a few examples for input strings. |
| 2958 |
</para> |
| 2959 |
<table id="ecpg-pgtypestimestamp-from-asc-example-table"> |
| 2960 |
<title>Valid Input Formats for <function>PGTYPEStimestamp_from_asc</function></title> |
| 2961 |
<tgroup cols="2"> |
| 2962 |
<thead> |
| 2963 |
<row> |
| 2964 |
<entry>Input</entry> |
| 2965 |
<entry>Result</entry> |
| 2966 |
</row> |
| 2967 |
</thead> |
| 2968 |
<tbody> |
| 2969 |
<row> |
| 2970 |
<entry><literal>1999-01-08 04:05:06</literal></entry> |
| 2971 |
<entry><literal>1999-01-08 04:05:06</literal></entry> |
| 2972 |
</row> |
| 2973 |
<row> |
| 2974 |
<entry><literal>January 8 04:05:06 1999 PST</literal></entry> |
| 2975 |
<entry><literal>1999-01-08 04:05:06</literal></entry> |
| 2976 |
</row> |
| 2977 |
<row> |
| 2978 |
<entry><literal>1999-Jan-08 04:05:06.789-8</literal></entry> |
| 2979 |
<entry><literal>1999-01-08 04:05:06.789 (time zone specifier ignored)</literal></entry> |
| 2980 |
</row> |
| 2981 |
<row> |
| 2982 |
<entry><literal>J2451187 04:05-08:00</literal></entry> |
| 2983 |
<entry><literal>1999-01-08 04:05:00 (time zone specifier ignored)</literal></entry> |
| 2984 |
</row> |
| 2985 |
</tbody> |
| 2986 |
</tgroup> |
| 2987 |
</table> |
| 2988 |
</listitem> |
| 2989 |
</varlistentry> |
| 2990 |
|
| 2991 |
<varlistentry id="pgtypestimestamptoasc"> |
| 2992 |
<term><function>PGTYPEStimestamp_to_asc</function></term> |
| 2993 |
<listitem> |
| 2994 |
<para> |
| 2995 |
Converts a date to a C char* string. |
| 2996 |
<synopsis> |
| 2997 |
char *PGTYPEStimestamp_to_asc(timestamp tstamp); |
| 2998 |
</synopsis> |
| 2999 |
The function receives the timestamp <literal>tstamp</literal> as |
| 3000 |
its only argument and returns an allocated string that contains the |
| 3001 |
textual representation of the timestamp. |
| 3002 |
The result must be freed with <function>PGTYPESchar_free()</function>. |
| 3003 |
</para> |
| 3004 |
</listitem> |
| 3005 |
</varlistentry> |
| 3006 |
|
| 3007 |
<varlistentry id="pgtypestimestampcurrent"> |
| 3008 |
<term><function>PGTYPEStimestamp_current</function></term> |
| 3009 |
<listitem> |
| 3010 |
<para> |
| 3011 |
Retrieve the current timestamp. |
| 3012 |
<synopsis> |
| 3013 |
void PGTYPEStimestamp_current(timestamp *ts); |
| 3014 |
</synopsis> |
| 3015 |
The function retrieves the current timestamp and saves it into the |
| 3016 |
timestamp variable that <literal>ts</literal> points to. |
| 3017 |
</para> |
| 3018 |
</listitem> |
| 3019 |
</varlistentry> |
| 3020 |
|
| 3021 |
<varlistentry id="pgtypestimestampfmtasc"> |
| 3022 |
<term><function>PGTYPEStimestamp_fmt_asc</function></term> |
| 3023 |
<listitem> |
| 3024 |
<para> |
| 3025 |
Convert a timestamp variable to a C char* using a format mask. |
| 3026 |
<synopsis> |
| 3027 |
int PGTYPEStimestamp_fmt_asc(timestamp *ts, char *output, int str_len, char *fmtstr); |
| 3028 |
</synopsis> |
| 3029 |
The function receives a pointer to the timestamp to convert as its |
| 3030 |
first argument (<literal>ts</literal>), a pointer to the output buffer |
| 3031 |
(<literal>output</literal>), the maximal length that has been allocated for |
| 3032 |
the output buffer (<literal>str_len</literal>) and the format mask to |
| 3033 |
use for the conversion (<literal>fmtstr</literal>). |
| 3034 |
</para> |
| 3035 |
<para> |
| 3036 |
Upon success, the function returns 0 and a negative value if an |
| 3037 |
error occurred. |
| 3038 |
</para> |
| 3039 |
<para> |
| 3040 |
You can use the following format specifiers for the format mask. The |
| 3041 |
format specifiers are the same ones that are used in the |
| 3042 |
<function>strftime</function> function in <productname>libc</productname>. Any |
| 3043 |
non-format specifier will be copied into the output buffer. |
| 3044 |
<!-- This is from the FreeBSD man page: |
| 3045 |
http://www.freebsd.org/cgi/man.cgi?query=strftime&apropos=0&sektion=3&manpath=FreeBSD+7.0-current&format=html |
| 3046 |
--> |
| 3047 |
<itemizedlist> |
| 3048 |
<listitem> |
| 3049 |
<para> |
| 3050 |
<literal>%A</literal> - is replaced by national representation of |
| 3051 |
the full weekday name. |
| 3052 |
</para> |
| 3053 |
</listitem> |
| 3054 |
<listitem> |
| 3055 |
<para> |
| 3056 |
<literal>%a</literal> - is replaced by national representation of |
| 3057 |
the abbreviated weekday name. |
| 3058 |
</para> |
| 3059 |
</listitem> |
| 3060 |
<listitem> |
| 3061 |
<para> |
| 3062 |
<literal>%B</literal> - is replaced by national representation of |
| 3063 |
the full month name. |
| 3064 |
</para> |
| 3065 |
</listitem> |
| 3066 |
<listitem> |
| 3067 |
<para> |
| 3068 |
<literal>%b</literal> - is replaced by national representation of |
| 3069 |
the abbreviated month name. |
| 3070 |
</para> |
| 3071 |
</listitem> |
| 3072 |
<listitem> |
| 3073 |
<para> |
| 3074 |
<literal>%C</literal> - is replaced by (year / 100) as decimal |
| 3075 |
number; single digits are preceded by a zero. |
| 3076 |
</para> |
| 3077 |
</listitem> |
| 3078 |
<listitem> |
| 3079 |
<para> |
| 3080 |
<literal>%c</literal> - is replaced by national representation of |
| 3081 |
time and date. |
| 3082 |
</para> |
| 3083 |
</listitem> |
| 3084 |
<listitem> |
| 3085 |
<para> |
| 3086 |
<literal>%D</literal> - is equivalent to |
| 3087 |
<literal>%m/%d/%y</literal>. |
| 3088 |
</para> |
| 3089 |
</listitem> |
| 3090 |
<listitem> |
| 3091 |
<para> |
| 3092 |
<literal>%d</literal> - is replaced by the day of the month as a |
| 3093 |
decimal number (01–31). |
| 3094 |
</para> |
| 3095 |
</listitem> |
| 3096 |
<listitem> |
| 3097 |
<para> |
| 3098 |
<literal>%E*</literal> <literal>%O*</literal> - POSIX locale |
| 3099 |
extensions. The sequences |
| 3100 |
<literal>%Ec</literal> |
| 3101 |
<literal>%EC</literal> |
| 3102 |
<literal>%Ex</literal> |
| 3103 |
<literal>%EX</literal> |
| 3104 |
<literal>%Ey</literal> |
| 3105 |
<literal>%EY</literal> |
| 3106 |
<literal>%Od</literal> |
| 3107 |
<literal>%Oe</literal> |
| 3108 |
<literal>%OH</literal> |
| 3109 |
<literal>%OI</literal> |
| 3110 |
<literal>%Om</literal> |
| 3111 |
<literal>%OM</literal> |
| 3112 |
<literal>%OS</literal> |
| 3113 |
<literal>%Ou</literal> |
| 3114 |
<literal>%OU</literal> |
| 3115 |
<literal>%OV</literal> |
| 3116 |
<literal>%Ow</literal> |
| 3117 |
<literal>%OW</literal> |
| 3118 |
<literal>%Oy</literal> |
| 3119 |
are supposed to provide alternative representations. |
| 3120 |
</para> |
| 3121 |
<para> |
| 3122 |
Additionally <literal>%OB</literal> implemented to represent |
| 3123 |
alternative months names (used standalone, without day mentioned). |
| 3124 |
</para> |
| 3125 |
</listitem> |
| 3126 |
<listitem> |
| 3127 |
<para> |
| 3128 |
<literal>%e</literal> - is replaced by the day of month as a decimal |
| 3129 |
number (1–31); single digits are preceded by a blank. |
| 3130 |
</para> |
| 3131 |
</listitem> |
| 3132 |
<listitem> |
| 3133 |
<para> |
| 3134 |
<literal>%F</literal> - is equivalent to <literal>%Y-%m-%d</literal>. |
| 3135 |
</para> |
| 3136 |
</listitem> |
| 3137 |
<listitem> |
| 3138 |
<para> |
| 3139 |
<literal>%G</literal> - is replaced by a year as a decimal number |
| 3140 |
with century. This year is the one that contains the greater part of |
| 3141 |
the week (Monday as the first day of the week). |
| 3142 |
</para> |
| 3143 |
</listitem> |
| 3144 |
<listitem> |
| 3145 |
<para> |
| 3146 |
<literal>%g</literal> - is replaced by the same year as in |
| 3147 |
<literal>%G</literal>, but as a decimal number without century |
| 3148 |
(00–99). |
| 3149 |
</para> |
| 3150 |
</listitem> |
| 3151 |
<listitem> |
| 3152 |
<para> |
| 3153 |
<literal>%H</literal> - is replaced by the hour (24-hour clock) as a |
| 3154 |
decimal number (00–23). |
| 3155 |
</para> |
| 3156 |
</listitem> |
| 3157 |
<listitem> |
| 3158 |
<para> |
| 3159 |
<literal>%h</literal> - the same as <literal>%b</literal>. |
| 3160 |
</para> |
| 3161 |
</listitem> |
| 3162 |
<listitem> |
| 3163 |
<para> |
| 3164 |
<literal>%I</literal> - is replaced by the hour (12-hour clock) as a |
| 3165 |
decimal number (01–12). |
| 3166 |
</para> |
| 3167 |
</listitem> |
| 3168 |
<listitem> |
| 3169 |
<para> |
| 3170 |
<literal>%j</literal> - is replaced by the day of the year as a |
| 3171 |
decimal number (001–366). |
| 3172 |
</para> |
| 3173 |
</listitem> |
| 3174 |
<listitem> |
| 3175 |
<para> |
| 3176 |
<literal>%k</literal> - is replaced by the hour (24-hour clock) as a |
| 3177 |
decimal number (0–23); single digits are preceded by a blank. |
| 3178 |
</para> |
| 3179 |
</listitem> |
| 3180 |
<listitem> |
| 3181 |
<para> |
| 3182 |
<literal>%l</literal> - is replaced by the hour (12-hour clock) as a |
| 3183 |
decimal number (1–12); single digits are preceded by a blank. |
| 3184 |
</para> |
| 3185 |
</listitem> |
| 3186 |
<listitem> |
| 3187 |
<para> |
| 3188 |
<literal>%M</literal> - is replaced by the minute as a decimal |
| 3189 |
number (00–59). |
| 3190 |
</para> |
| 3191 |
</listitem> |
| 3192 |
<listitem> |
| 3193 |
<para> |
| 3194 |
<literal>%m</literal> - is replaced by the month as a decimal number |
| 3195 |
(01–12). |
| 3196 |
</para> |
| 3197 |
</listitem> |
| 3198 |
<listitem> |
| 3199 |
<para> |
| 3200 |
<literal>%n</literal> - is replaced by a newline. |
| 3201 |
</para> |
| 3202 |
</listitem> |
| 3203 |
<listitem> |
| 3204 |
<para> |
| 3205 |
<literal>%O*</literal> - the same as <literal>%E*</literal>. |
| 3206 |
</para> |
| 3207 |
</listitem> |
| 3208 |
<listitem> |
| 3209 |
<para> |
| 3210 |
<literal>%p</literal> - is replaced by national representation of |
| 3211 |
either <quote>ante meridiem</quote> or <quote>post meridiem</quote> as appropriate. |
| 3212 |
</para> |
| 3213 |
</listitem> |
| 3214 |
<listitem> |
| 3215 |
<para> |
| 3216 |
<literal>%R</literal> - is equivalent to <literal>%H:%M</literal>. |
| 3217 |
</para> |
| 3218 |
</listitem> |
| 3219 |
<listitem> |
| 3220 |
<para> |
| 3221 |
<literal>%r</literal> - is equivalent to <literal>%I:%M:%S |
| 3222 |
%p</literal>. |
| 3223 |
</para> |
| 3224 |
</listitem> |
| 3225 |
<listitem> |
| 3226 |
<para> |
| 3227 |
<literal>%S</literal> - is replaced by the second as a decimal |
| 3228 |
number (00–60). |
| 3229 |
</para> |
| 3230 |
</listitem> |
| 3231 |
<listitem> |
| 3232 |
<para> |
| 3233 |
<literal>%s</literal> - is replaced by the number of seconds since |
| 3234 |
the Epoch, UTC. |
| 3235 |
</para> |
| 3236 |
</listitem> |
| 3237 |
<listitem> |
| 3238 |
<para> |
| 3239 |
<literal>%T</literal> - is equivalent to <literal>%H:%M:%S</literal> |
| 3240 |
</para> |
| 3241 |
</listitem> |
| 3242 |
<listitem> |
| 3243 |
<para> |
| 3244 |
<literal>%t</literal> - is replaced by a tab. |
| 3245 |
</para> |
| 3246 |
</listitem> |
| 3247 |
<listitem> |
| 3248 |
<para> |
| 3249 |
<literal>%U</literal> - is replaced by the week number of the year |
| 3250 |
(Sunday as the first day of the week) as a decimal number (00–53). |
| 3251 |
</para> |
| 3252 |
</listitem> |
| 3253 |
<listitem> |
| 3254 |
<para> |
| 3255 |
<literal>%u</literal> - is replaced by the weekday (Monday as the |
| 3256 |
first day of the week) as a decimal number (1–7). |
| 3257 |
</para> |
| 3258 |
</listitem> |
| 3259 |
<listitem> |
| 3260 |
<para> |
| 3261 |
<literal>%V</literal> - is replaced by the week number of the year |
| 3262 |
(Monday as the first day of the week) as a decimal number (01–53). |
| 3263 |
If the week containing January 1 has four or more days in the new |
| 3264 |
year, then it is week 1; otherwise it is the last week of the |
| 3265 |
previous year, and the next week is week 1. |
| 3266 |
</para> |
| 3267 |
</listitem> |
| 3268 |
<listitem> |
| 3269 |
<para> |
| 3270 |
<literal>%v</literal> - is equivalent to |
| 3271 |
<literal>%e-%b-%Y</literal>. |
| 3272 |
</para> |
| 3273 |
</listitem> |
| 3274 |
<listitem> |
| 3275 |
<para> |
| 3276 |
<literal>%W</literal> - is replaced by the week number of the year |
| 3277 |
(Monday as the first day of the week) as a decimal number (00–53). |
| 3278 |
</para> |
| 3279 |
</listitem> |
| 3280 |
<listitem> |
| 3281 |
<para> |
| 3282 |
<literal>%w</literal> - is replaced by the weekday (Sunday as the |
| 3283 |
first day of the week) as a decimal number (0–6). |
| 3284 |
</para> |
| 3285 |
</listitem> |
| 3286 |
<listitem> |
| 3287 |
<para> |
| 3288 |
<literal>%X</literal> - is replaced by national representation of |
| 3289 |
the time. |
| 3290 |
</para> |
| 3291 |
</listitem> |
| 3292 |
<listitem> |
| 3293 |
<para> |
| 3294 |
<literal>%x</literal> - is replaced by national representation of |
| 3295 |
the date. |
| 3296 |
</para> |
| 3297 |
</listitem> |
| 3298 |
<listitem> |
| 3299 |
<para> |
| 3300 |
<literal>%Y</literal> - is replaced by the year with century as a |
| 3301 |
decimal number. |
| 3302 |
</para> |
| 3303 |
</listitem> |
| 3304 |
<listitem> |
| 3305 |
<para> |
| 3306 |
<literal>%y</literal> - is replaced by the year without century as a |
| 3307 |
decimal number (00–99). |
| 3308 |
</para> |
| 3309 |
</listitem> |
| 3310 |
<listitem> |
| 3311 |
<para> |
| 3312 |
<literal>%Z</literal> - is replaced by the time zone name. |
| 3313 |
</para> |
| 3314 |
</listitem> |
| 3315 |
<listitem> |
| 3316 |
<para> |
| 3317 |
<literal>%z</literal> - is replaced by the time zone offset from |
| 3318 |
UTC; a leading plus sign stands for east of UTC, a minus sign for |
| 3319 |
west of UTC, hours and minutes follow with two digits each and no |
| 3320 |
delimiter between them (common form for <ulink url="https://datatracker.ietf.org/doc/html/rfc822">RFC 822</ulink> date headers). |
| 3321 |
</para> |
| 3322 |
</listitem> |
| 3323 |
<listitem> |
| 3324 |
<para> |
| 3325 |
<literal>%+</literal> - is replaced by national representation of |
| 3326 |
the date and time. |
| 3327 |
</para> |
| 3328 |
</listitem> |
| 3329 |
<listitem> |
| 3330 |
<para> |
| 3331 |
<literal>%-*</literal> - GNU libc extension. Do not do any padding |
| 3332 |
when performing numerical outputs. |
| 3333 |
</para> |
| 3334 |
</listitem> |
| 3335 |
<listitem> |
| 3336 |
<para> |
| 3337 |
$_* - GNU libc extension. Explicitly specify space for padding. |
| 3338 |
</para> |
| 3339 |
</listitem> |
| 3340 |
<listitem> |
| 3341 |
<para> |
| 3342 |
<literal>%0*</literal> - GNU libc extension. Explicitly specify zero |
| 3343 |
for padding. |
| 3344 |
</para> |
| 3345 |
</listitem> |
| 3346 |
<listitem> |
| 3347 |
<para> |
| 3348 |
<literal>%%</literal> - is replaced by <literal>%</literal>. |
| 3349 |
</para> |
| 3350 |
</listitem> |
| 3351 |
</itemizedlist> |
| 3352 |
</para> |
| 3353 |
</listitem> |
| 3354 |
</varlistentry> |
| 3355 |
|
| 3356 |
<varlistentry id="pgtypestimestampsub"> |
| 3357 |
<term><function>PGTYPEStimestamp_sub</function></term> |
| 3358 |
<listitem> |
| 3359 |
<para> |
| 3360 |
Subtract one timestamp from another one and save the result in a |
| 3361 |
variable of type interval. |
| 3362 |
<synopsis> |
| 3363 |
int PGTYPEStimestamp_sub(timestamp *ts1, timestamp *ts2, interval *iv); |
| 3364 |
</synopsis> |
| 3365 |
The function will subtract the timestamp variable that <literal>ts2</literal> |
| 3366 |
points to from the timestamp variable that <literal>ts1</literal> points to |
| 3367 |
and will store the result in the interval variable that <literal>iv</literal> |
| 3368 |
points to. |
| 3369 |
</para> |
| 3370 |
<para> |
| 3371 |
Upon success, the function returns 0 and a negative value if an |
| 3372 |
error occurred. |
| 3373 |
</para> |
| 3374 |
</listitem> |
| 3375 |
</varlistentry> |
| 3376 |
|
| 3377 |
<varlistentry id="pgtypestimestampdefmtasc"> |
| 3378 |
<term><function>PGTYPEStimestamp_defmt_asc</function></term> |
| 3379 |
<listitem> |
| 3380 |
<para> |
| 3381 |
Parse a timestamp value from its textual representation using a |
| 3382 |
formatting mask. |
| 3383 |
<synopsis> |
| 3384 |
int PGTYPEStimestamp_defmt_asc(char *str, char *fmt, timestamp *d); |
| 3385 |
</synopsis> |
| 3386 |
The function receives the textual representation of a timestamp in the |
| 3387 |
variable <literal>str</literal> as well as the formatting mask to use in the |
| 3388 |
variable <literal>fmt</literal>. The result will be stored in the variable |
| 3389 |
that <literal>d</literal> points to. |
| 3390 |
</para> |
| 3391 |
<para> |
| 3392 |
If the formatting mask <literal>fmt</literal> is NULL, the function will fall |
| 3393 |
back to the default formatting mask which is <literal>%Y-%m-%d |
| 3394 |
%H:%M:%S</literal>. |
| 3395 |
</para> |
| 3396 |
<para> |
| 3397 |
This is the reverse function to <xref |
| 3398 |
linkend="pgtypestimestampfmtasc"/>. See the documentation there in |
| 3399 |
order to find out about the possible formatting mask entries. |
| 3400 |
</para> |
| 3401 |
</listitem> |
| 3402 |
</varlistentry> |
| 3403 |
|
| 3404 |
<varlistentry id="pgtypestimestampaddinterval"> |
| 3405 |
<term><function>PGTYPEStimestamp_add_interval</function></term> |
| 3406 |
<listitem> |
| 3407 |
<para> |
| 3408 |
Add an interval variable to a timestamp variable. |
| 3409 |
<synopsis> |
| 3410 |
int PGTYPEStimestamp_add_interval(timestamp *tin, interval *span, timestamp *tout); |
| 3411 |
</synopsis> |
| 3412 |
The function receives a pointer to a timestamp variable <literal>tin</literal> |
| 3413 |
and a pointer to an interval variable <literal>span</literal>. It adds the |
| 3414 |
interval to the timestamp and saves the resulting timestamp in the |
| 3415 |
variable that <literal>tout</literal> points to. |
| 3416 |
</para> |
| 3417 |
<para> |
| 3418 |
Upon success, the function returns 0 and a negative value if an |
| 3419 |
error occurred. |
| 3420 |
</para> |
| 3421 |
</listitem> |
| 3422 |
</varlistentry> |
| 3423 |
|
| 3424 |
<varlistentry id="pgtypestimestampsubinterval"> |
| 3425 |
<term><function>PGTYPEStimestamp_sub_interval</function></term> |
| 3426 |
<listitem> |
| 3427 |
<para> |
| 3428 |
Subtract an interval variable from a timestamp variable. |
| 3429 |
<synopsis> |
| 3430 |
int PGTYPEStimestamp_sub_interval(timestamp *tin, interval *span, timestamp *tout); |
| 3431 |
</synopsis> |
| 3432 |
The function subtracts the interval variable that <literal>span</literal> |
| 3433 |
points to from the timestamp variable that <literal>tin</literal> points to |
| 3434 |
and saves the result into the variable that <literal>tout</literal> points |
| 3435 |
to. |
| 3436 |
</para> |
| 3437 |
<para> |
| 3438 |
Upon success, the function returns 0 and a negative value if an |
| 3439 |
error occurred. |
| 3440 |
</para> |
| 3441 |
</listitem> |
| 3442 |
</varlistentry> |
| 3443 |
</variablelist> |
| 3444 |
</para> |
| 3445 |
</sect2> |
| 3446 |
|
| 3447 |
<sect2 id="ecpg-pgtypes-interval"> |
| 3448 |
<title>The interval Type</title> |
| 3449 |
<para> |
| 3450 |
The interval type in C enables your programs to deal with data of the SQL |
| 3451 |
type interval. See <xref linkend="datatype-datetime"/> for the equivalent |
| 3452 |
type in the <productname>PostgreSQL</productname> server. |
| 3453 |
</para> |
| 3454 |
<para> |
| 3455 |
The following functions can be used to work with the interval type: |
| 3456 |
<variablelist> |
| 3457 |
|
| 3458 |
<varlistentry id="pgtypesintervalnew"> |
| 3459 |
<term><function>PGTYPESinterval_new</function></term> |
| 3460 |
<listitem> |
| 3461 |
<para> |
| 3462 |
Return a pointer to a newly allocated interval variable. |
| 3463 |
<synopsis> |
| 3464 |
interval *PGTYPESinterval_new(void); |
| 3465 |
</synopsis> |
| 3466 |
</para> |
| 3467 |
</listitem> |
| 3468 |
</varlistentry> |
| 3469 |
|
| 3470 |
<varlistentry id="pgtypesintervalfree"> |
| 3471 |
<term><function>PGTYPESinterval_free</function></term> |
| 3472 |
<listitem> |
| 3473 |
<para> |
| 3474 |
Release the memory of a previously allocated interval variable. |
| 3475 |
<synopsis> |
| 3476 |
void PGTYPESinterval_free(interval *intvl); |
| 3477 |
</synopsis> |
| 3478 |
</para> |
| 3479 |
</listitem> |
| 3480 |
</varlistentry> |
| 3481 |
|
| 3482 |
<varlistentry id="pgtypesintervalfromasc"> |
| 3483 |
<term><function>PGTYPESinterval_from_asc</function></term> |
| 3484 |
<listitem> |
| 3485 |
<para> |
| 3486 |
Parse an interval from its textual representation. |
| 3487 |
<synopsis> |
| 3488 |
interval *PGTYPESinterval_from_asc(char *str, char **endptr); |
| 3489 |
</synopsis> |
| 3490 |
The function parses the input string <literal>str</literal> and returns a |
| 3491 |
pointer to an allocated interval variable. |
| 3492 |
At the moment ECPG always parses |
| 3493 |
the complete string and so it currently does not support to store the |
| 3494 |
address of the first invalid character in <literal>*endptr</literal>. |
| 3495 |
You can safely set <literal>endptr</literal> to NULL. |
| 3496 |
</para> |
| 3497 |
</listitem> |
| 3498 |
</varlistentry> |
| 3499 |
|
| 3500 |
<varlistentry id="pgtypesintervaltoasc"> |
| 3501 |
<term><function>PGTYPESinterval_to_asc</function></term> |
| 3502 |
<listitem> |
| 3503 |
<para> |
| 3504 |
Convert a variable of type interval to its textual representation. |
| 3505 |
<synopsis> |
| 3506 |
char *PGTYPESinterval_to_asc(interval *span); |
| 3507 |
</synopsis> |
| 3508 |
The function converts the interval variable that <literal>span</literal> |
| 3509 |
points to into a C char*. The output looks like this example: |
| 3510 |
<literal>@ 1 day 12 hours 59 mins 10 secs</literal>. |
| 3511 |
The result must be freed with <function>PGTYPESchar_free()</function>. |
| 3512 |
</para> |
| 3513 |
</listitem> |
| 3514 |
</varlistentry> |
| 3515 |
|
| 3516 |
<varlistentry id="pgtypesintervalcopy"> |
| 3517 |
<term><function>PGTYPESinterval_copy</function></term> |
| 3518 |
<listitem> |
| 3519 |
<para> |
| 3520 |
Copy a variable of type interval. |
| 3521 |
<synopsis> |
| 3522 |
int PGTYPESinterval_copy(interval *intvlsrc, interval *intvldest); |
| 3523 |
</synopsis> |
| 3524 |
The function copies the interval variable that <literal>intvlsrc</literal> |
| 3525 |
points to into the variable that <literal>intvldest</literal> points to. Note |
| 3526 |
that you need to allocate the memory for the destination variable |
| 3527 |
before. |
| 3528 |
</para> |
| 3529 |
</listitem> |
| 3530 |
</varlistentry> |
| 3531 |
</variablelist> |
| 3532 |
</para> |
| 3533 |
</sect2> |
| 3534 |
|
| 3535 |
<sect2 id="ecpg-pgtypes-decimal"> |
| 3536 |
<title>The decimal Type</title> |
| 3537 |
<para> |
| 3538 |
The decimal type is similar to the numeric type. However it is limited to |
| 3539 |
a maximum precision of 30 significant digits. In contrast to the numeric |
| 3540 |
type which can be created on the heap only, the decimal type can be |
| 3541 |
created either on the stack or on the heap (by means of the functions |
| 3542 |
<function>PGTYPESdecimal_new</function> and |
| 3543 |
<function>PGTYPESdecimal_free</function>). |
| 3544 |
There are a lot of other functions that deal with the decimal type in the |
| 3545 |
<productname>Informix</productname> compatibility mode described in <xref |
| 3546 |
linkend="ecpg-informix-compat"/>. |
| 3547 |
</para> |
| 3548 |
<para> |
| 3549 |
The following functions can be used to work with the decimal type and are |
| 3550 |
not only contained in the <literal>libcompat</literal> library. |
| 3551 |
<variablelist> |
| 3552 |
<varlistentry id="ecpg-pgtypes-decimal-new"> |
| 3553 |
<term><function>PGTYPESdecimal_new</function></term> |
| 3554 |
<listitem> |
| 3555 |
<para> |
| 3556 |
Request a pointer to a newly allocated decimal variable. |
| 3557 |
<synopsis> |
| 3558 |
decimal *PGTYPESdecimal_new(void); |
| 3559 |
</synopsis> |
| 3560 |
</para> |
| 3561 |
</listitem> |
| 3562 |
</varlistentry> |
| 3563 |
|
| 3564 |
<varlistentry id="ecpg-pgtypes-decimal-free"> |
| 3565 |
<term><function>PGTYPESdecimal_free</function></term> |
| 3566 |
<listitem> |
| 3567 |
<para> |
| 3568 |
Free a decimal type, release all of its memory. |
| 3569 |
<synopsis> |
| 3570 |
void PGTYPESdecimal_free(decimal *var); |
| 3571 |
</synopsis> |
| 3572 |
</para> |
| 3573 |
</listitem> |
| 3574 |
</varlistentry> |
| 3575 |
</variablelist> |
| 3576 |
</para> |
| 3577 |
</sect2> |
| 3578 |
|
| 3579 |
<sect2 id="ecpg-pgtypes-errno"> |
| 3580 |
<title>errno Values of pgtypeslib</title> |
| 3581 |
<para> |
| 3582 |
<variablelist> |
| 3583 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-num-bad-numeric"> |
| 3584 |
<term><literal>PGTYPES_NUM_BAD_NUMERIC</literal></term> |
| 3585 |
<listitem> |
| 3586 |
<para> |
| 3587 |
An argument should contain a numeric variable (or point to a numeric |
| 3588 |
variable) but in fact its in-memory representation was invalid. |
| 3589 |
</para> |
| 3590 |
</listitem> |
| 3591 |
</varlistentry> |
| 3592 |
|
| 3593 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-num-overflow"> |
| 3594 |
<term><literal>PGTYPES_NUM_OVERFLOW</literal></term> |
| 3595 |
<listitem> |
| 3596 |
<para> |
| 3597 |
An overflow occurred. Since the numeric type can deal with almost |
| 3598 |
arbitrary precision, converting a numeric variable into other types |
| 3599 |
might cause overflow. |
| 3600 |
</para> |
| 3601 |
</listitem> |
| 3602 |
</varlistentry> |
| 3603 |
|
| 3604 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-num-underflow"> |
| 3605 |
<term><literal>PGTYPES_NUM_UNDERFLOW</literal></term> |
| 3606 |
<listitem> |
| 3607 |
<para> |
| 3608 |
An underflow occurred. Since the numeric type can deal with almost |
| 3609 |
arbitrary precision, converting a numeric variable into other types |
| 3610 |
might cause underflow. |
| 3611 |
</para> |
| 3612 |
</listitem> |
| 3613 |
</varlistentry> |
| 3614 |
|
| 3615 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-num-divide-zero"> |
| 3616 |
<term><literal>PGTYPES_NUM_DIVIDE_ZERO</literal></term> |
| 3617 |
<listitem> |
| 3618 |
<para> |
| 3619 |
A division by zero has been attempted. |
| 3620 |
</para> |
| 3621 |
</listitem> |
| 3622 |
</varlistentry> |
| 3623 |
|
| 3624 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-date-bad-date"> |
| 3625 |
<term><literal>PGTYPES_DATE_BAD_DATE</literal></term> |
| 3626 |
<listitem> |
| 3627 |
<para> |
| 3628 |
An invalid date string was passed to |
| 3629 |
the <function>PGTYPESdate_from_asc</function> function. |
| 3630 |
</para> |
| 3631 |
</listitem> |
| 3632 |
</varlistentry> |
| 3633 |
|
| 3634 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-date-err-eargs"> |
| 3635 |
<term><literal>PGTYPES_DATE_ERR_EARGS</literal></term> |
| 3636 |
<listitem> |
| 3637 |
<para> |
| 3638 |
Invalid arguments were passed to the |
| 3639 |
<function>PGTYPESdate_defmt_asc</function> function. |
| 3640 |
</para> |
| 3641 |
</listitem> |
| 3642 |
</varlistentry> |
| 3643 |
|
| 3644 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-date-err-enoshortdate"> |
| 3645 |
<term><literal>PGTYPES_DATE_ERR_ENOSHORTDATE</literal></term> |
| 3646 |
<listitem> |
| 3647 |
<para> |
| 3648 |
An invalid token in the input string was found by the |
| 3649 |
<function>PGTYPESdate_defmt_asc</function> function. |
| 3650 |
</para> |
| 3651 |
</listitem> |
| 3652 |
</varlistentry> |
| 3653 |
|
| 3654 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-intvl-bad-interval"> |
| 3655 |
<term><literal>PGTYPES_INTVL_BAD_INTERVAL</literal></term> |
| 3656 |
<listitem> |
| 3657 |
<para> |
| 3658 |
An invalid interval string was passed to the |
| 3659 |
<function>PGTYPESinterval_from_asc</function> function, or an |
| 3660 |
invalid interval value was passed to the |
| 3661 |
<function>PGTYPESinterval_to_asc</function> function. |
| 3662 |
</para> |
| 3663 |
</listitem> |
| 3664 |
</varlistentry> |
| 3665 |
|
| 3666 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-date-err-enotdmy"> |
| 3667 |
<term><literal>PGTYPES_DATE_ERR_ENOTDMY</literal></term> |
| 3668 |
<listitem> |
| 3669 |
<para> |
| 3670 |
There was a mismatch in the day/month/year assignment in the |
| 3671 |
<function>PGTYPESdate_defmt_asc</function> function. |
| 3672 |
</para> |
| 3673 |
</listitem> |
| 3674 |
</varlistentry> |
| 3675 |
|
| 3676 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-date-bad-day"> |
| 3677 |
<term><literal>PGTYPES_DATE_BAD_DAY</literal></term> |
| 3678 |
<listitem> |
| 3679 |
<para> |
| 3680 |
An invalid day of the month value was found by |
| 3681 |
the <function>PGTYPESdate_defmt_asc</function> function. |
| 3682 |
</para> |
| 3683 |
</listitem> |
| 3684 |
</varlistentry> |
| 3685 |
|
| 3686 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-date-bad-month"> |
| 3687 |
<term><literal>PGTYPES_DATE_BAD_MONTH</literal></term> |
| 3688 |
<listitem> |
| 3689 |
<para> |
| 3690 |
An invalid month value was found by |
| 3691 |
the <function>PGTYPESdate_defmt_asc</function> function. |
| 3692 |
</para> |
| 3693 |
</listitem> |
| 3694 |
</varlistentry> |
| 3695 |
|
| 3696 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-ts-bad-timestamp"> |
| 3697 |
<term><literal>PGTYPES_TS_BAD_TIMESTAMP</literal></term> |
| 3698 |
<listitem> |
| 3699 |
<para> |
| 3700 |
An invalid timestamp string pass passed to |
| 3701 |
the <function>PGTYPEStimestamp_from_asc</function> function, |
| 3702 |
or an invalid timestamp value was passed to |
| 3703 |
the <function>PGTYPEStimestamp_to_asc</function> function. |
| 3704 |
</para> |
| 3705 |
</listitem> |
| 3706 |
</varlistentry> |
| 3707 |
|
| 3708 |
<varlistentry id="ecpg-pgtypes-errno-pgtypes-ts-err-einftime"> |
| 3709 |
<term><literal>PGTYPES_TS_ERR_EINFTIME</literal></term> |
| 3710 |
<listitem> |
| 3711 |
<para> |
| 3712 |
An infinite timestamp value was encountered in a context that |
| 3713 |
cannot handle it. |
| 3714 |
</para> |
| 3715 |
</listitem> |
| 3716 |
</varlistentry> |
| 3717 |
</variablelist> |
| 3718 |
</para> |
| 3719 |
</sect2> |
| 3720 |
|
| 3721 |
<sect2 id="ecpg-pgtypes-constants"> |
| 3722 |
<title>Special Constants of pgtypeslib</title> |
| 3723 |
<para> |
| 3724 |
<variablelist> |
| 3725 |
<varlistentry id="pgtypesinvalidtimestamp"> |
| 3726 |
<term><literal>PGTYPESInvalidTimestamp</literal></term> |
| 3727 |
<listitem> |
| 3728 |
<para> |
| 3729 |
A value of type timestamp representing an invalid time stamp. This is |
| 3730 |
returned by the function <function>PGTYPEStimestamp_from_asc</function> on |
| 3731 |
parse error. |
| 3732 |
Note that due to the internal representation of the <type>timestamp</type> data type, |
| 3733 |
<literal>PGTYPESInvalidTimestamp</literal> is also a valid timestamp at |
| 3734 |
the same time. It is set to <literal>1899-12-31 23:59:59</literal>. In order |
| 3735 |
to detect errors, make sure that your application does not only test |
| 3736 |
for <literal>PGTYPESInvalidTimestamp</literal> but also for |
| 3737 |
<literal>errno != 0</literal> after each call to |
| 3738 |
<function>PGTYPEStimestamp_from_asc</function>. |
| 3739 |
</para> |
| 3740 |
</listitem> |
| 3741 |
</varlistentry> |
| 3742 |
</variablelist> |
| 3743 |
</para> |
| 3744 |
</sect2> |
| 3745 |
</sect1> |
| 3746 |
|
| 3747 |
<sect1 id="ecpg-descriptors"> |
| 3748 |
<title>Using Descriptor Areas</title> |
| 3749 |
|
| 3750 |
<para> |
| 3751 |
An SQL descriptor area is a more sophisticated method for processing |
| 3752 |
the result of a <command>SELECT</command>, <command>FETCH</command> or |
| 3753 |
a <command>DESCRIBE</command> statement. An SQL descriptor area groups |
| 3754 |
the data of one row of data together with metadata items into one |
| 3755 |
data structure. The metadata is particularly useful when executing |
| 3756 |
dynamic SQL statements, where the nature of the result columns might |
| 3757 |
not be known ahead of time. PostgreSQL provides two ways to use |
| 3758 |
Descriptor Areas: the named SQL Descriptor Areas and the C-structure |
| 3759 |
SQLDAs. |
| 3760 |
</para> |
| 3761 |
|
| 3762 |
<sect2 id="ecpg-named-descriptors"> |
| 3763 |
<title>Named SQL Descriptor Areas</title> |
| 3764 |
|
| 3765 |
<para> |
| 3766 |
A named SQL descriptor area consists of a header, which contains |
| 3767 |
information concerning the entire descriptor, and one or more item |
| 3768 |
descriptor areas, which basically each describe one column in the |
| 3769 |
result row. |
| 3770 |
</para> |
| 3771 |
|
| 3772 |
<para> |
| 3773 |
Before you can use an SQL descriptor area, you need to allocate one: |
| 3774 |
<programlisting> |
| 3775 |
EXEC SQL ALLOCATE DESCRIPTOR <replaceable>identifier</replaceable>; |
| 3776 |
</programlisting> |
| 3777 |
The identifier serves as the <quote>variable name</quote> of the |
| 3778 |
descriptor area. <!-- The scope of the allocated descriptor is WHAT?. --> |
| 3779 |
When you don't need the descriptor anymore, you should deallocate |
| 3780 |
it: |
| 3781 |
<programlisting> |
| 3782 |
EXEC SQL DEALLOCATE DESCRIPTOR <replaceable>identifier</replaceable>; |
| 3783 |
</programlisting> |
| 3784 |
</para> |
| 3785 |
|
| 3786 |
<para> |
| 3787 |
To use a descriptor area, specify it as the storage target in an |
| 3788 |
<literal>INTO</literal> clause, instead of listing host variables: |
| 3789 |
<programlisting> |
| 3790 |
EXEC SQL FETCH NEXT FROM mycursor INTO SQL DESCRIPTOR mydesc; |
| 3791 |
</programlisting> |
| 3792 |
If the result set is empty, the Descriptor Area will still contain |
| 3793 |
the metadata from the query, i.e., the field names. |
| 3794 |
</para> |
| 3795 |
|
| 3796 |
<para> |
| 3797 |
For not yet executed prepared queries, the <command>DESCRIBE</command> |
| 3798 |
statement can be used to get the metadata of the result set: |
| 3799 |
<programlisting> |
| 3800 |
EXEC SQL BEGIN DECLARE SECTION; |
| 3801 |
char *sql_stmt = "SELECT * FROM table1"; |
| 3802 |
EXEC SQL END DECLARE SECTION; |
| 3803 |
|
| 3804 |
EXEC SQL PREPARE stmt1 FROM :sql_stmt; |
| 3805 |
EXEC SQL DESCRIBE stmt1 INTO SQL DESCRIPTOR mydesc; |
| 3806 |
</programlisting> |
| 3807 |
</para> |
| 3808 |
|
| 3809 |
<para> |
| 3810 |
Before PostgreSQL 9.0, the <literal>SQL</literal> keyword was optional, |
| 3811 |
so using <literal>DESCRIPTOR</literal> and <literal>SQL DESCRIPTOR</literal> |
| 3812 |
produced named SQL Descriptor Areas. Now it is mandatory, omitting |
| 3813 |
the <literal>SQL</literal> keyword produces SQLDA Descriptor Areas, |
| 3814 |
see <xref linkend="ecpg-sqlda-descriptors"/>. |
| 3815 |
</para> |
| 3816 |
|
| 3817 |
<para> |
| 3818 |
In <command>DESCRIBE</command> and <command>FETCH</command> statements, |
| 3819 |
the <literal>INTO</literal> and <literal>USING</literal> keywords can be |
| 3820 |
used to similarly: they produce the result set and the metadata in a |
| 3821 |
Descriptor Area. |
| 3822 |
</para> |
| 3823 |
|
| 3824 |
<para> |
| 3825 |
Now how do you get the data out of the descriptor area? You can |
| 3826 |
think of the descriptor area as a structure with named fields. To |
| 3827 |
retrieve the value of a field from the header and store it into a |
| 3828 |
host variable, use the following command: |
| 3829 |
<programlisting> |
| 3830 |
EXEC SQL GET DESCRIPTOR <replaceable>name</replaceable> :<replaceable>hostvar</replaceable> = <replaceable>field</replaceable>; |
| 3831 |
</programlisting> |
| 3832 |
Currently, there is only one header field defined: |
| 3833 |
<replaceable>COUNT</replaceable>, which tells how many item |
| 3834 |
descriptor areas exist (that is, how many columns are contained in |
| 3835 |
the result). The host variable needs to be of an integer type. To |
| 3836 |
get a field from the item descriptor area, use the following |
| 3837 |
command: |
| 3838 |
<programlisting> |
| 3839 |
EXEC SQL GET DESCRIPTOR <replaceable>name</replaceable> VALUE <replaceable>num</replaceable> :<replaceable>hostvar</replaceable> = <replaceable>field</replaceable>; |
| 3840 |
</programlisting> |
| 3841 |
<replaceable>num</replaceable> can be a literal integer or a host |
| 3842 |
variable containing an integer. Possible fields are: |
| 3843 |
|
| 3844 |
<variablelist> |
| 3845 |
<varlistentry id="ecpg-named-descriptors-cardinality"> |
| 3846 |
<term><literal>CARDINALITY</literal> (integer)</term> |
| 3847 |
<listitem> |
| 3848 |
<para> |
| 3849 |
number of rows in the result set |
| 3850 |
</para> |
| 3851 |
</listitem> |
| 3852 |
</varlistentry> |
| 3853 |
|
| 3854 |
<varlistentry id="ecpg-named-descriptors-data"> |
| 3855 |
<term><literal>DATA</literal></term> |
| 3856 |
<listitem> |
| 3857 |
<para> |
| 3858 |
actual data item (therefore, the data type of this field |
| 3859 |
depends on the query) |
| 3860 |
</para> |
| 3861 |
</listitem> |
| 3862 |
</varlistentry> |
| 3863 |
|
| 3864 |
<varlistentry id="ecpg-named-descriptors-datetime-interval-code"> |
| 3865 |
<term><literal>DATETIME_INTERVAL_CODE</literal> (integer)</term> |
| 3866 |
<listitem> |
| 3867 |
<para> |
| 3868 |
When <literal>TYPE</literal> is <literal>9</literal>, |
| 3869 |
<literal>DATETIME_INTERVAL_CODE</literal> will have a value of |
| 3870 |
<literal>1</literal> for <literal>DATE</literal>, |
| 3871 |
<literal>2</literal> for <literal>TIME</literal>, |
| 3872 |
<literal>3</literal> for <literal>TIMESTAMP</literal>, |
| 3873 |
<literal>4</literal> for <literal>TIME WITH TIME ZONE</literal>, or |
| 3874 |
<literal>5</literal> for <literal>TIMESTAMP WITH TIME ZONE</literal>. |
| 3875 |
</para> |
| 3876 |
</listitem> |
| 3877 |
</varlistentry> |
| 3878 |
|
| 3879 |
<varlistentry id="ecpg-named-descriptors-datetime-interval-precision"> |
| 3880 |
<term><literal>DATETIME_INTERVAL_PRECISION</literal> (integer)</term> |
| 3881 |
<listitem> |
| 3882 |
<para> |
| 3883 |
not implemented |
| 3884 |
</para> |
| 3885 |
</listitem> |
| 3886 |
</varlistentry> |
| 3887 |
|
| 3888 |
<varlistentry id="ecpg-named-descriptors-indicator"> |
| 3889 |
<term><literal>INDICATOR</literal> (integer)</term> |
| 3890 |
<listitem> |
| 3891 |
<para> |
| 3892 |
the indicator (indicating a null value or a value truncation) |
| 3893 |
</para> |
| 3894 |
</listitem> |
| 3895 |
</varlistentry> |
| 3896 |
|
| 3897 |
<varlistentry id="ecpg-named-descriptors-key-member"> |
| 3898 |
<term><literal>KEY_MEMBER</literal> (integer)</term> |
| 3899 |
<listitem> |
| 3900 |
<para> |
| 3901 |
not implemented |
| 3902 |
</para> |
| 3903 |
</listitem> |
| 3904 |
</varlistentry> |
| 3905 |
|
| 3906 |
<varlistentry id="ecpg-named-descriptors-length"> |
| 3907 |
<term><literal>LENGTH</literal> (integer)</term> |
| 3908 |
<listitem> |
| 3909 |
<para> |
| 3910 |
length of the datum in characters |
| 3911 |
</para> |
| 3912 |
</listitem> |
| 3913 |
</varlistentry> |
| 3914 |
|
| 3915 |
<varlistentry id="ecpg-named-descriptors-name"> |
| 3916 |
<term><literal>NAME</literal> (string)</term> |
| 3917 |
<listitem> |
| 3918 |
<para> |
| 3919 |
name of the column |
| 3920 |
</para> |
| 3921 |
</listitem> |
| 3922 |
</varlistentry> |
| 3923 |
|
| 3924 |
<varlistentry id="ecpg-named-descriptors-nullable"> |
| 3925 |
<term><literal>NULLABLE</literal> (integer)</term> |
| 3926 |
<listitem> |
| 3927 |
<para> |
| 3928 |
not implemented |
| 3929 |
</para> |
| 3930 |
</listitem> |
| 3931 |
</varlistentry> |
| 3932 |
|
| 3933 |
<varlistentry id="ecpg-named-descriptors-octet-length"> |
| 3934 |
<term><literal>OCTET_LENGTH</literal> (integer)</term> |
| 3935 |
<listitem> |
| 3936 |
<para> |
| 3937 |
length of the character representation of the datum in bytes |
| 3938 |
</para> |
| 3939 |
</listitem> |
| 3940 |
</varlistentry> |
| 3941 |
|
| 3942 |
<varlistentry id="ecpg-named-descriptors-precision"> |
| 3943 |
<term><literal>PRECISION</literal> (integer)</term> |
| 3944 |
<listitem> |
| 3945 |
<para> |
| 3946 |
precision (for type <type>numeric</type>) |
| 3947 |
</para> |
| 3948 |
</listitem> |
| 3949 |
</varlistentry> |
| 3950 |
|
| 3951 |
<varlistentry id="ecpg-named-descriptors-returned-length"> |
| 3952 |
<term><literal>RETURNED_LENGTH</literal> (integer)</term> |
| 3953 |
<listitem> |
| 3954 |
<para> |
| 3955 |
length of the datum in characters |
| 3956 |
</para> |
| 3957 |
</listitem> |
| 3958 |
</varlistentry> |
| 3959 |
|
| 3960 |
<varlistentry id="ecpg-named-descriptors-returned-octet-length"> |
| 3961 |
<term><literal>RETURNED_OCTET_LENGTH</literal> (integer)</term> |
| 3962 |
<listitem> |
| 3963 |
<para> |
| 3964 |
length of the character representation of the datum in bytes |
| 3965 |
</para> |
| 3966 |
</listitem> |
| 3967 |
</varlistentry> |
| 3968 |
|
| 3969 |
<varlistentry id="ecpg-named-descriptors-scale"> |
| 3970 |
<term><literal>SCALE</literal> (integer)</term> |
| 3971 |
<listitem> |
| 3972 |
<para> |
| 3973 |
scale (for type <type>numeric</type>) |
| 3974 |
</para> |
| 3975 |
</listitem> |
| 3976 |
</varlistentry> |
| 3977 |
|
| 3978 |
<varlistentry id="ecpg-named-descriptors-type"> |
| 3979 |
<term><literal>TYPE</literal> (integer)</term> |
| 3980 |
<listitem> |
| 3981 |
<para> |
| 3982 |
numeric code of the data type of the column |
| 3983 |
</para> |
| 3984 |
</listitem> |
| 3985 |
</varlistentry> |
| 3986 |
</variablelist> |
| 3987 |
</para> |
| 3988 |
|
| 3989 |
<para> |
| 3990 |
In <command>EXECUTE</command>, <command>DECLARE</command> and <command>OPEN</command> |
| 3991 |
statements, the effect of the <literal>INTO</literal> and <literal>USING</literal> |
| 3992 |
keywords are different. A Descriptor Area can also be manually built to |
| 3993 |
provide the input parameters for a query or a cursor and |
| 3994 |
<literal>USING SQL DESCRIPTOR <replaceable>name</replaceable></literal> |
| 3995 |
is the way to pass the input parameters into a parameterized query. The statement |
| 3996 |
to build a named SQL Descriptor Area is below: |
| 3997 |
<programlisting> |
| 3998 |
EXEC SQL SET DESCRIPTOR <replaceable>name</replaceable> VALUE <replaceable>num</replaceable> <replaceable>field</replaceable> = :<replaceable>hostvar</replaceable>; |
| 3999 |
</programlisting> |
| 4000 |
</para> |
| 4001 |
|
| 4002 |
<para> |
| 4003 |
PostgreSQL supports retrieving more that one record in one <command>FETCH</command> |
| 4004 |
statement and storing the data in host variables in this case assumes that the |
| 4005 |
variable is an array. E.g.: |
| 4006 |
<programlisting> |
| 4007 |
EXEC SQL BEGIN DECLARE SECTION; |
| 4008 |
int id[5]; |
| 4009 |
EXEC SQL END DECLARE SECTION; |
| 4010 |
|
| 4011 |
EXEC SQL FETCH 5 FROM mycursor INTO SQL DESCRIPTOR mydesc; |
| 4012 |
|
| 4013 |
EXEC SQL GET DESCRIPTOR mydesc VALUE 1 :id = DATA; |
| 4014 |
</programlisting> |
| 4015 |
|
| 4016 |
</para> |
| 4017 |
|
| 4018 |
</sect2> |
| 4019 |
|
| 4020 |
<sect2 id="ecpg-sqlda-descriptors"> |
| 4021 |
<title>SQLDA Descriptor Areas</title> |
| 4022 |
|
| 4023 |
<para> |
| 4024 |
An SQLDA Descriptor Area is a C language structure which can be also used |
| 4025 |
to get the result set and the metadata of a query. One structure stores one |
| 4026 |
record from the result set. |
| 4027 |
<programlisting> |
| 4028 |
EXEC SQL include sqlda.h; |
| 4029 |
sqlda_t *mysqlda; |
| 4030 |
|
| 4031 |
EXEC SQL FETCH 3 FROM mycursor INTO DESCRIPTOR mysqlda; |
| 4032 |
</programlisting> |
| 4033 |
Note that the <literal>SQL</literal> keyword is omitted. The paragraphs about |
| 4034 |
the use cases of the <literal>INTO</literal> and <literal>USING</literal> |
| 4035 |
keywords in <xref linkend="ecpg-named-descriptors"/> also apply here with an addition. |
| 4036 |
In a <command>DESCRIBE</command> statement the <literal>DESCRIPTOR</literal> |
| 4037 |
keyword can be completely omitted if the <literal>INTO</literal> keyword is used: |
| 4038 |
<programlisting> |
| 4039 |
EXEC SQL DESCRIBE prepared_statement INTO mysqlda; |
| 4040 |
</programlisting> |
| 4041 |
</para> |
| 4042 |
|
| 4043 |
<procedure> |
| 4044 |
<para> |
| 4045 |
The general flow of a program that uses SQLDA is: |
| 4046 |
</para> |
| 4047 |
<step><simpara>Prepare a query, and declare a cursor for it.</simpara></step> |
| 4048 |
<step><simpara>Declare an SQLDA for the result rows.</simpara></step> |
| 4049 |
<step><simpara>Declare an SQLDA for the input parameters, and initialize them (memory allocation, parameter settings).</simpara></step> |
| 4050 |
<step><simpara>Open a cursor with the input SQLDA.</simpara></step> |
| 4051 |
<step><simpara>Fetch rows from the cursor, and store them into an output SQLDA.</simpara></step> |
| 4052 |
<step><simpara>Read values from the output SQLDA into the host variables (with conversion if necessary).</simpara></step> |
| 4053 |
<step><simpara>Close the cursor.</simpara></step> |
| 4054 |
<step><simpara>Free the memory area allocated for the input SQLDA.</simpara></step> |
| 4055 |
</procedure> |
| 4056 |
|
| 4057 |
<sect3 id="ecpg-sqlda-descriptors-sqlda"> |
| 4058 |
<title>SQLDA Data Structure</title> |
| 4059 |
|
| 4060 |
<para> |
| 4061 |
SQLDA uses three data structure |
| 4062 |
types: <type>sqlda_t</type>, <type>sqlvar_t</type>, |
| 4063 |
and <type>struct sqlname</type>. |
| 4064 |
</para> |
| 4065 |
|
| 4066 |
<tip> |
| 4067 |
<para> |
| 4068 |
PostgreSQL's SQLDA has a similar data structure to the one in |
| 4069 |
IBM DB2 Universal Database, so some technical information on |
| 4070 |
DB2's SQLDA could help understanding PostgreSQL's one better. |
| 4071 |
</para> |
| 4072 |
</tip> |
| 4073 |
|
| 4074 |
<sect4 id="ecpg-sqlda-sqlda"> |
| 4075 |
<title>sqlda_t Structure</title> |
| 4076 |
|
| 4077 |
<para> |
| 4078 |
The structure type <type>sqlda_t</type> is the type of the |
| 4079 |
actual SQLDA. It holds one record. And two or |
| 4080 |
more <type>sqlda_t</type> structures can be connected in a |
| 4081 |
linked list with the pointer in |
| 4082 |
the <structfield>desc_next</structfield> field, thus |
| 4083 |
representing an ordered collection of rows. So, when two or |
| 4084 |
more rows are fetched, the application can read them by |
| 4085 |
following the <structfield>desc_next</structfield> pointer in |
| 4086 |
each <type>sqlda_t</type> node. |
| 4087 |
</para> |
| 4088 |
|
| 4089 |
<para> |
| 4090 |
The definition of <type>sqlda_t</type> is: |
| 4091 |
<programlisting> |
| 4092 |
struct sqlda_struct |
| 4093 |
{ |
| 4094 |
char sqldaid[8]; |
| 4095 |
long sqldabc; |
| 4096 |
short sqln; |
| 4097 |
short sqld; |
| 4098 |
struct sqlda_struct *desc_next; |
| 4099 |
struct sqlvar_struct sqlvar[1]; |
| 4100 |
}; |
| 4101 |
|
| 4102 |
typedef struct sqlda_struct sqlda_t; |
| 4103 |
</programlisting> |
| 4104 |
|
| 4105 |
The meaning of the fields is: |
| 4106 |
|
| 4107 |
<variablelist> |
| 4108 |
<varlistentry id="ecpg-sqlda-sqlda-sqldaid"> |
| 4109 |
<term><literal>sqldaid</literal></term> |
| 4110 |
<listitem> |
| 4111 |
<para> |
| 4112 |
It contains the literal string <literal>"SQLDA "</literal>. |
| 4113 |
</para> |
| 4114 |
</listitem> |
| 4115 |
</varlistentry> |
| 4116 |
|
| 4117 |
<varlistentry id="ecpg-sqlda-sqlda-sqldabc"> |
| 4118 |
<term><literal>sqldabc</literal></term> |
| 4119 |
<listitem> |
| 4120 |
<para> |
| 4121 |
It contains the size of the allocated space in bytes. |
| 4122 |
</para> |
| 4123 |
</listitem> |
| 4124 |
</varlistentry> |
| 4125 |
|
| 4126 |
<varlistentry id="ecpg-sqlda-sqlda-sqln"> |
| 4127 |
<term><literal>sqln</literal></term> |
| 4128 |
<listitem> |
| 4129 |
<para> |
| 4130 |
It contains the number of input parameters for a parameterized query in |
| 4131 |
case it's passed into <command>OPEN</command>, <command>DECLARE</command> or |
| 4132 |
<command>EXECUTE</command> statements using the <literal>USING</literal> |
| 4133 |
keyword. In case it's used as output of <command>SELECT</command>, |
| 4134 |
<command>EXECUTE</command> or <command>FETCH</command> statements, |
| 4135 |
its value is the same as <literal>sqld</literal> |
| 4136 |
statement |
| 4137 |
</para> |
| 4138 |
</listitem> |
| 4139 |
</varlistentry> |
| 4140 |
|
| 4141 |
<varlistentry id="ecpg-sqlda-sqlda-sqld"> |
| 4142 |
<term><literal>sqld</literal></term> |
| 4143 |
<listitem> |
| 4144 |
<para> |
| 4145 |
It contains the number of fields in a result set. |
| 4146 |
</para> |
| 4147 |
</listitem> |
| 4148 |
</varlistentry> |
| 4149 |
|
| 4150 |
<varlistentry id="ecpg-sqlda-sqlda-desc-next"> |
| 4151 |
<term><literal>desc_next</literal></term> |
| 4152 |
<listitem> |
| 4153 |
<para> |
| 4154 |
If the query returns more than one record, multiple linked |
| 4155 |
SQLDA structures are returned, and <literal>desc_next</literal> holds |
| 4156 |
a pointer to the next entry in the list. |
| 4157 |
</para> |
| 4158 |
</listitem> |
| 4159 |
</varlistentry> |
| 4160 |
<varlistentry id="ecpg-sqlda-sqlda-sqlvar"> |
| 4161 |
<term><literal>sqlvar</literal></term> |
| 4162 |
<listitem> |
| 4163 |
<para> |
| 4164 |
This is the array of the columns in the result set. |
| 4165 |
</para> |
| 4166 |
</listitem> |
| 4167 |
</varlistentry> |
| 4168 |
</variablelist> |
| 4169 |
</para> |
| 4170 |
</sect4> |
| 4171 |
|
| 4172 |
<sect4 id="ecpg-sqlda-sqlvar"> |
| 4173 |
<title>sqlvar_t Structure</title> |
| 4174 |
|
| 4175 |
<para> |
| 4176 |
The structure type <type>sqlvar_t</type> holds a column value |
| 4177 |
and metadata such as type and length. The definition of the type |
| 4178 |
is: |
| 4179 |
|
| 4180 |
<programlisting> |
| 4181 |
struct sqlvar_struct |
| 4182 |
{ |
| 4183 |
short sqltype; |
| 4184 |
short sqllen; |
| 4185 |
char *sqldata; |
| 4186 |
short *sqlind; |
| 4187 |
struct sqlname sqlname; |
| 4188 |
}; |
| 4189 |
|
| 4190 |
typedef struct sqlvar_struct sqlvar_t; |
| 4191 |
</programlisting> |
| 4192 |
|
| 4193 |
The meaning of the fields is: |
| 4194 |
|
| 4195 |
<variablelist> |
| 4196 |
<varlistentry id="ecpg-sqlda-sqlvar-sqltype"> |
| 4197 |
<term><literal>sqltype</literal></term> |
| 4198 |
<listitem> |
| 4199 |
<para> |
| 4200 |
Contains the type identifier of the field. For values, |
| 4201 |
see <literal>enum ECPGttype</literal> in <literal>ecpgtype.h</literal>. |
| 4202 |
</para> |
| 4203 |
</listitem> |
| 4204 |
</varlistentry> |
| 4205 |
|
| 4206 |
<varlistentry id="ecpg-sqlda-sqlvar-sqllen"> |
| 4207 |
<term><literal>sqllen</literal></term> |
| 4208 |
<listitem> |
| 4209 |
<para> |
| 4210 |
Contains the binary length of the field. e.g., 4 bytes for <type>ECPGt_int</type>. |
| 4211 |
</para> |
| 4212 |
</listitem> |
| 4213 |
</varlistentry> |
| 4214 |
|
| 4215 |
<varlistentry id="ecpg-sqlda-sqlvar-sqldata"> |
| 4216 |
<term><literal>sqldata</literal></term> |
| 4217 |
<listitem> |
| 4218 |
<para> |
| 4219 |
Points to the data. The format of the data is described |
| 4220 |
in <xref linkend="ecpg-variables-type-mapping"/>. |
| 4221 |
</para> |
| 4222 |
</listitem> |
| 4223 |
</varlistentry> |
| 4224 |
|
| 4225 |
<varlistentry id="ecpg-sqlda-sqlvar-sqlind"> |
| 4226 |
<term><literal>sqlind</literal></term> |
| 4227 |
<listitem> |
| 4228 |
<para> |
| 4229 |
Points to the null indicator. 0 means not null, -1 means |
| 4230 |
null. |
| 4231 |
</para> |
| 4232 |
</listitem> |
| 4233 |
</varlistentry> |
| 4234 |
|
| 4235 |
<varlistentry id="ecpg-sqlda-sqlvar-sqlname"> |
| 4236 |
<term><literal>sqlname</literal></term> |
| 4237 |
<listitem> |
| 4238 |
<para> |
| 4239 |
The name of the field. |
| 4240 |
</para> |
| 4241 |
</listitem> |
| 4242 |
</varlistentry> |
| 4243 |
</variablelist> |
| 4244 |
</para> |
| 4245 |
</sect4> |
| 4246 |
|
| 4247 |
<sect4 id="ecpg-sqlda-sqlname"> |
| 4248 |
<title>struct sqlname Structure</title> |
| 4249 |
|
| 4250 |
<para> |
| 4251 |
A <type>struct sqlname</type> structure holds a column name. It |
| 4252 |
is used as a member of the <type>sqlvar_t</type> structure. The |
| 4253 |
definition of the structure is: |
| 4254 |
<programlisting> |
| 4255 |
#define NAMEDATALEN 64 |
| 4256 |
|
| 4257 |
struct sqlname |
| 4258 |
{ |
| 4259 |
short length; |
| 4260 |
char data[NAMEDATALEN]; |
| 4261 |
}; |
| 4262 |
</programlisting> |
| 4263 |
The meaning of the fields is: |
| 4264 |
<variablelist> |
| 4265 |
<varlistentry id="ecpg-sqlda-sqlname-length"> |
| 4266 |
<term><literal>length</literal></term> |
| 4267 |
<listitem> |
| 4268 |
<para> |
| 4269 |
Contains the length of the field name. |
| 4270 |
</para> |
| 4271 |
</listitem> |
| 4272 |
</varlistentry> |
| 4273 |
<varlistentry id="ecpg-sqlda-sqlname-data"> |
| 4274 |
<term><literal>data</literal></term> |
| 4275 |
<listitem> |
| 4276 |
<para> |
| 4277 |
Contains the actual field name. |
| 4278 |
</para> |
| 4279 |
</listitem> |
| 4280 |
</varlistentry> |
| 4281 |
</variablelist> |
| 4282 |
</para> |
| 4283 |
</sect4> |
| 4284 |
</sect3> |
| 4285 |
|
| 4286 |
<sect3 id="ecpg-sqlda-output"> |
| 4287 |
<title>Retrieving a Result Set Using an SQLDA</title> |
| 4288 |
|
| 4289 |
<procedure> |
| 4290 |
<para> |
| 4291 |
The general steps to retrieve a query result set through an |
| 4292 |
SQLDA are: |
| 4293 |
</para> |
| 4294 |
<step><simpara>Declare an <type>sqlda_t</type> structure to receive the result set.</simpara></step> |
| 4295 |
<step><simpara>Execute <command>FETCH</command>/<command>EXECUTE</command>/<command>DESCRIBE</command> commands to process a query specifying the declared SQLDA.</simpara></step> |
| 4296 |
<step><simpara>Check the number of records in the result set by looking at <structfield>sqln</structfield>, a member of the <type>sqlda_t</type> structure.</simpara></step> |
| 4297 |
<step><simpara>Get the values of each column from <literal>sqlvar[0]</literal>, <literal>sqlvar[1]</literal>, etc., members of the <type>sqlda_t</type> structure.</simpara></step> |
| 4298 |
<step><simpara>Go to next row (<type>sqlda_t</type> structure) by following the <structfield>desc_next</structfield> pointer, a member of the <type>sqlda_t</type> structure.</simpara></step> |
| 4299 |
<step><simpara>Repeat above as you need.</simpara></step> |
| 4300 |
</procedure> |
| 4301 |
|
| 4302 |
<para> |
| 4303 |
Here is an example retrieving a result set through an SQLDA. |
| 4304 |
</para> |
| 4305 |
|
| 4306 |
<para> |
| 4307 |
First, declare a <type>sqlda_t</type> structure to receive the result set. |
| 4308 |
<programlisting> |
| 4309 |
sqlda_t *sqlda1; |
| 4310 |
</programlisting> |
| 4311 |
</para> |
| 4312 |
|
| 4313 |
<para> |
| 4314 |
Next, specify the SQLDA in a command. This is |
| 4315 |
a <command>FETCH</command> command example. |
| 4316 |
<programlisting> |
| 4317 |
EXEC SQL FETCH NEXT FROM cur1 INTO DESCRIPTOR sqlda1; |
| 4318 |
</programlisting> |
| 4319 |
</para> |
| 4320 |
|
| 4321 |
<para> |
| 4322 |
Run a loop following the linked list to retrieve the rows. |
| 4323 |
<programlisting> |
| 4324 |
sqlda_t *cur_sqlda; |
| 4325 |
|
| 4326 |
for (cur_sqlda = sqlda1; |
| 4327 |
cur_sqlda != NULL; |
| 4328 |
cur_sqlda = cur_sqlda->desc_next) |
| 4329 |
{ |
| 4330 |
... |
| 4331 |
} |
| 4332 |
</programlisting> |
| 4333 |
</para> |
| 4334 |
|
| 4335 |
<para> |
| 4336 |
Inside the loop, run another loop to retrieve each column data |
| 4337 |
(<type>sqlvar_t</type> structure) of the row. |
| 4338 |
<programlisting> |
| 4339 |
for (i = 0; i < cur_sqlda->sqld; i++) |
| 4340 |
{ |
| 4341 |
sqlvar_t v = cur_sqlda->sqlvar[i]; |
| 4342 |
char *sqldata = v.sqldata; |
| 4343 |
short sqllen = v.sqllen; |
| 4344 |
... |
| 4345 |
} |
| 4346 |
</programlisting> |
| 4347 |
</para> |
| 4348 |
|
| 4349 |
<para> |
| 4350 |
To get a column value, check the <structfield>sqltype</structfield> value, |
| 4351 |
a member of the <type>sqlvar_t</type> structure. Then, switch |
| 4352 |
to an appropriate way, depending on the column type, to copy |
| 4353 |
data from the <structfield>sqlvar</structfield> field to a host variable. |
| 4354 |
<programlisting> |
| 4355 |
char var_buf[1024]; |
| 4356 |
|
| 4357 |
switch (v.sqltype) |
| 4358 |
{ |
| 4359 |
case ECPGt_char: |
| 4360 |
memset(&var_buf, 0, sizeof(var_buf)); |
| 4361 |
memcpy(&var_buf, sqldata, (sizeof(var_buf) <= sqllen ? sizeof(var_buf) - 1 : sqllen)); |
| 4362 |
break; |
| 4363 |
|
| 4364 |
case ECPGt_int: /* integer */ |
| 4365 |
memcpy(&intval, sqldata, sqllen); |
| 4366 |
snprintf(var_buf, sizeof(var_buf), "%d", intval); |
| 4367 |
break; |
| 4368 |
|
| 4369 |
... |
| 4370 |
} |
| 4371 |
</programlisting> |
| 4372 |
</para> |
| 4373 |
</sect3> |
| 4374 |
|
| 4375 |
<sect3 id="ecpg-sqlda-input"> |
| 4376 |
<title>Passing Query Parameters Using an SQLDA</title> |
| 4377 |
|
| 4378 |
<procedure> |
| 4379 |
<para> |
| 4380 |
The general steps to use an SQLDA to pass input |
| 4381 |
parameters to a prepared query are: |
| 4382 |
</para> |
| 4383 |
<step><simpara>Create a prepared query (prepared statement)</simpara></step> |
| 4384 |
<step><simpara>Declare an sqlda_t structure as an input SQLDA.</simpara></step> |
| 4385 |
<step><simpara>Allocate memory area (as sqlda_t structure) for the input SQLDA.</simpara></step> |
| 4386 |
<step><simpara>Set (copy) input values in the allocated memory.</simpara></step> |
| 4387 |
<step><simpara>Open a cursor with specifying the input SQLDA.</simpara></step> |
| 4388 |
</procedure> |
| 4389 |
|
| 4390 |
<para> |
| 4391 |
Here is an example. |
| 4392 |
</para> |
| 4393 |
|
| 4394 |
<para> |
| 4395 |
First, create a prepared statement. |
| 4396 |
<programlisting> |
| 4397 |
EXEC SQL BEGIN DECLARE SECTION; |
| 4398 |
char query[1024] = "SELECT d.oid, * FROM pg_database d, pg_stat_database s WHERE d.oid = s.datid AND (d.datname = ? OR d.oid = ?)"; |
| 4399 |
EXEC SQL END DECLARE SECTION; |
| 4400 |
|
| 4401 |
EXEC SQL PREPARE stmt1 FROM :query; |
| 4402 |
</programlisting> |
| 4403 |
</para> |
| 4404 |
|
| 4405 |
<para> |
| 4406 |
Next, allocate memory for an SQLDA, and set the number of input |
| 4407 |
parameters in <structfield>sqln</structfield>, a member variable of |
| 4408 |
the <type>sqlda_t</type> structure. When two or more input |
| 4409 |
parameters are required for the prepared query, the application |
| 4410 |
has to allocate additional memory space which is calculated by |
| 4411 |
(nr. of params - 1) * sizeof(sqlvar_t). The example shown here |
| 4412 |
allocates memory space for two input parameters. |
| 4413 |
<programlisting> |
| 4414 |
sqlda_t *sqlda2; |
| 4415 |
|
| 4416 |
sqlda2 = (sqlda_t *) malloc(sizeof(sqlda_t) + sizeof(sqlvar_t)); |
| 4417 |
memset(sqlda2, 0, sizeof(sqlda_t) + sizeof(sqlvar_t)); |
| 4418 |
|
| 4419 |
sqlda2->sqln = 2; /* number of input variables */ |
| 4420 |
</programlisting> |
| 4421 |
</para> |
| 4422 |
|
| 4423 |
<para> |
| 4424 |
After memory allocation, store the parameter values into the |
| 4425 |
<literal>sqlvar[]</literal> array. (This is same array used for |
| 4426 |
retrieving column values when the SQLDA is receiving a result |
| 4427 |
set.) In this example, the input parameters |
| 4428 |
are <literal>"postgres"</literal>, having a string type, |
| 4429 |
and <literal>1</literal>, having an integer type. |
| 4430 |
<programlisting> |
| 4431 |
sqlda2->sqlvar[0].sqltype = ECPGt_char; |
| 4432 |
sqlda2->sqlvar[0].sqldata = "postgres"; |
| 4433 |
sqlda2->sqlvar[0].sqllen = 8; |
| 4434 |
|
| 4435 |
int intval = 1; |
| 4436 |
sqlda2->sqlvar[1].sqltype = ECPGt_int; |
| 4437 |
sqlda2->sqlvar[1].sqldata = (char *) &intval; |
| 4438 |
sqlda2->sqlvar[1].sqllen = sizeof(intval); |
| 4439 |
</programlisting> |
| 4440 |
</para> |
| 4441 |
|
| 4442 |
<para> |
| 4443 |
By opening a cursor and specifying the SQLDA that was set up |
| 4444 |
beforehand, the input parameters are passed to the prepared |
| 4445 |
statement. |
| 4446 |
<programlisting> |
| 4447 |
EXEC SQL OPEN cur1 USING DESCRIPTOR sqlda2; |
| 4448 |
</programlisting> |
| 4449 |
</para> |
| 4450 |
|
| 4451 |
<para> |
| 4452 |
Finally, after using input SQLDAs, the allocated memory space |
| 4453 |
must be freed explicitly, unlike SQLDAs used for receiving query |
| 4454 |
results. |
| 4455 |
<programlisting> |
| 4456 |
free(sqlda2); |
| 4457 |
</programlisting> |
| 4458 |
</para> |
| 4459 |
</sect3> |
| 4460 |
|
| 4461 |
<sect3 id="ecpg-sqlda-example"> |
| 4462 |
<title>A Sample Application Using SQLDA</title> |
| 4463 |
|
| 4464 |
<para> |
| 4465 |
Here is an example program, which describes how to fetch access |
| 4466 |
statistics of the databases, specified by the input parameters, |
| 4467 |
from the system catalogs. |
| 4468 |
</para> |
| 4469 |
|
| 4470 |
<para> |
| 4471 |
This application joins two system tables, pg_database and |
| 4472 |
pg_stat_database on the database OID, and also fetches and shows |
| 4473 |
the database statistics which are retrieved by two input |
| 4474 |
parameters (a database <literal>postgres</literal>, and OID <literal>1</literal>). |
| 4475 |
</para> |
| 4476 |
|
| 4477 |
<para> |
| 4478 |
First, declare an SQLDA for input and an SQLDA for output. |
| 4479 |
<programlisting> |
| 4480 |
EXEC SQL include sqlda.h; |
| 4481 |
|
| 4482 |
sqlda_t *sqlda1; /* an output descriptor */ |
| 4483 |
sqlda_t *sqlda2; /* an input descriptor */ |
| 4484 |
</programlisting> |
| 4485 |
</para> |
| 4486 |
|
| 4487 |
<para> |
| 4488 |
Next, connect to the database, prepare a statement, and declare a |
| 4489 |
cursor for the prepared statement. |
| 4490 |
<programlisting> |
| 4491 |
int |
| 4492 |
main(void) |
| 4493 |
{ |
| 4494 |
EXEC SQL BEGIN DECLARE SECTION; |
| 4495 |
char query[1024] = "SELECT d.oid,* FROM pg_database d, pg_stat_database s WHERE d.oid=s.datid AND ( d.datname=? OR d.oid=? )"; |
| 4496 |
EXEC SQL END DECLARE SECTION; |
| 4497 |
|
| 4498 |
EXEC SQL CONNECT TO testdb AS con1 USER testuser; |
| 4499 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 4500 |
|
| 4501 |
EXEC SQL PREPARE stmt1 FROM :query; |
| 4502 |
EXEC SQL DECLARE cur1 CURSOR FOR stmt1; |
| 4503 |
</programlisting> |
| 4504 |
</para> |
| 4505 |
|
| 4506 |
<para> |
| 4507 |
Next, put some values in the input SQLDA for the input |
| 4508 |
parameters. Allocate memory for the input SQLDA, and set the |
| 4509 |
number of input parameters to <literal>sqln</literal>. Store |
| 4510 |
type, value, and value length into <literal>sqltype</literal>, |
| 4511 |
<literal>sqldata</literal>, and <literal>sqllen</literal> in the |
| 4512 |
<literal>sqlvar</literal> structure. |
| 4513 |
|
| 4514 |
<programlisting> |
| 4515 |
/* Create SQLDA structure for input parameters. */ |
| 4516 |
sqlda2 = (sqlda_t *) malloc(sizeof(sqlda_t) + sizeof(sqlvar_t)); |
| 4517 |
memset(sqlda2, 0, sizeof(sqlda_t) + sizeof(sqlvar_t)); |
| 4518 |
sqlda2->sqln = 2; /* number of input variables */ |
| 4519 |
|
| 4520 |
sqlda2->sqlvar[0].sqltype = ECPGt_char; |
| 4521 |
sqlda2->sqlvar[0].sqldata = "postgres"; |
| 4522 |
sqlda2->sqlvar[0].sqllen = 8; |
| 4523 |
|
| 4524 |
intval = 1; |
| 4525 |
sqlda2->sqlvar[1].sqltype = ECPGt_int; |
| 4526 |
sqlda2->sqlvar[1].sqldata = (char *)&intval; |
| 4527 |
sqlda2->sqlvar[1].sqllen = sizeof(intval); |
| 4528 |
</programlisting> |
| 4529 |
</para> |
| 4530 |
|
| 4531 |
<para> |
| 4532 |
After setting up the input SQLDA, open a cursor with the input |
| 4533 |
SQLDA. |
| 4534 |
|
| 4535 |
<programlisting> |
| 4536 |
/* Open a cursor with input parameters. */ |
| 4537 |
EXEC SQL OPEN cur1 USING DESCRIPTOR sqlda2; |
| 4538 |
</programlisting> |
| 4539 |
</para> |
| 4540 |
|
| 4541 |
<para> |
| 4542 |
Fetch rows into the output SQLDA from the opened cursor. |
| 4543 |
(Generally, you have to call <command>FETCH</command> repeatedly |
| 4544 |
in the loop, to fetch all rows in the result set.) |
| 4545 |
<programlisting> |
| 4546 |
while (1) |
| 4547 |
{ |
| 4548 |
sqlda_t *cur_sqlda; |
| 4549 |
|
| 4550 |
/* Assign descriptor to the cursor */ |
| 4551 |
EXEC SQL FETCH NEXT FROM cur1 INTO DESCRIPTOR sqlda1; |
| 4552 |
</programlisting> |
| 4553 |
</para> |
| 4554 |
|
| 4555 |
<para> |
| 4556 |
Next, retrieve the fetched records from the SQLDA, by following |
| 4557 |
the linked list of the <type>sqlda_t</type> structure. |
| 4558 |
<programlisting> |
| 4559 |
for (cur_sqlda = sqlda1 ; |
| 4560 |
cur_sqlda != NULL ; |
| 4561 |
cur_sqlda = cur_sqlda->desc_next) |
| 4562 |
{ |
| 4563 |
... |
| 4564 |
</programlisting> |
| 4565 |
</para> |
| 4566 |
|
| 4567 |
<para> |
| 4568 |
Read each columns in the first record. The number of columns is |
| 4569 |
stored in <structfield>sqld</structfield>, the actual data of the first |
| 4570 |
column is stored in <literal>sqlvar[0]</literal>, both members of |
| 4571 |
the <type>sqlda_t</type> structure. |
| 4572 |
|
| 4573 |
<programlisting> |
| 4574 |
/* Print every column in a row. */ |
| 4575 |
for (i = 0; i < sqlda1->sqld; i++) |
| 4576 |
{ |
| 4577 |
sqlvar_t v = sqlda1->sqlvar[i]; |
| 4578 |
char *sqldata = v.sqldata; |
| 4579 |
short sqllen = v.sqllen; |
| 4580 |
|
| 4581 |
strncpy(name_buf, v.sqlname.data, v.sqlname.length); |
| 4582 |
name_buf[v.sqlname.length] = '\0'; |
| 4583 |
</programlisting> |
| 4584 |
</para> |
| 4585 |
|
| 4586 |
<para> |
| 4587 |
Now, the column data is stored in the variable <varname>v</varname>. |
| 4588 |
Copy every datum into host variables, looking |
| 4589 |
at <literal>v.sqltype</literal> for the type of the column. |
| 4590 |
<programlisting> |
| 4591 |
switch (v.sqltype) { |
| 4592 |
int intval; |
| 4593 |
double doubleval; |
| 4594 |
unsigned long long int longlongval; |
| 4595 |
|
| 4596 |
case ECPGt_char: |
| 4597 |
memset(&var_buf, 0, sizeof(var_buf)); |
| 4598 |
memcpy(&var_buf, sqldata, (sizeof(var_buf) <= sqllen ? sizeof(var_buf)-1 : sqllen)); |
| 4599 |
break; |
| 4600 |
|
| 4601 |
case ECPGt_int: /* integer */ |
| 4602 |
memcpy(&intval, sqldata, sqllen); |
| 4603 |
snprintf(var_buf, sizeof(var_buf), "%d", intval); |
| 4604 |
break; |
| 4605 |
|
| 4606 |
... |
| 4607 |
|
| 4608 |
default: |
| 4609 |
... |
| 4610 |
} |
| 4611 |
|
| 4612 |
printf("%s = %s (type: %d)\n", name_buf, var_buf, v.sqltype); |
| 4613 |
} |
| 4614 |
</programlisting> |
| 4615 |
</para> |
| 4616 |
|
| 4617 |
<para> |
| 4618 |
Close the cursor after processing all of records, and disconnect |
| 4619 |
from the database. |
| 4620 |
<programlisting> |
| 4621 |
EXEC SQL CLOSE cur1; |
| 4622 |
EXEC SQL COMMIT; |
| 4623 |
|
| 4624 |
EXEC SQL DISCONNECT ALL; |
| 4625 |
</programlisting> |
| 4626 |
</para> |
| 4627 |
|
| 4628 |
<para> |
| 4629 |
The whole program is shown |
| 4630 |
in <xref linkend="ecpg-sqlda-example-example"/>. |
| 4631 |
</para> |
| 4632 |
|
| 4633 |
<example id="ecpg-sqlda-example-example"> |
| 4634 |
<title>Example SQLDA Program</title> |
| 4635 |
<programlisting> |
| 4636 |
#include <stdlib.h> |
| 4637 |
#include <string.h> |
| 4638 |
#include <stdlib.h> |
| 4639 |
#include <stdio.h> |
| 4640 |
#include <unistd.h> |
| 4641 |
|
| 4642 |
EXEC SQL include sqlda.h; |
| 4643 |
|
| 4644 |
sqlda_t *sqlda1; /* descriptor for output */ |
| 4645 |
sqlda_t *sqlda2; /* descriptor for input */ |
| 4646 |
|
| 4647 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 4648 |
EXEC SQL WHENEVER SQLERROR STOP; |
| 4649 |
|
| 4650 |
int |
| 4651 |
main(void) |
| 4652 |
{ |
| 4653 |
EXEC SQL BEGIN DECLARE SECTION; |
| 4654 |
char query[1024] = "SELECT d.oid,* FROM pg_database d, pg_stat_database s WHERE d.oid=s.datid AND ( d.datname=? OR d.oid=? )"; |
| 4655 |
|
| 4656 |
int intval; |
| 4657 |
unsigned long long int longlongval; |
| 4658 |
EXEC SQL END DECLARE SECTION; |
| 4659 |
|
| 4660 |
EXEC SQL CONNECT TO uptimedb AS con1 USER uptime; |
| 4661 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 4662 |
|
| 4663 |
EXEC SQL PREPARE stmt1 FROM :query; |
| 4664 |
EXEC SQL DECLARE cur1 CURSOR FOR stmt1; |
| 4665 |
|
| 4666 |
/* Create an SQLDA structure for an input parameter */ |
| 4667 |
sqlda2 = (sqlda_t *)malloc(sizeof(sqlda_t) + sizeof(sqlvar_t)); |
| 4668 |
memset(sqlda2, 0, sizeof(sqlda_t) + sizeof(sqlvar_t)); |
| 4669 |
sqlda2->sqln = 2; /* a number of input variables */ |
| 4670 |
|
| 4671 |
sqlda2->sqlvar[0].sqltype = ECPGt_char; |
| 4672 |
sqlda2->sqlvar[0].sqldata = "postgres"; |
| 4673 |
sqlda2->sqlvar[0].sqllen = 8; |
| 4674 |
|
| 4675 |
intval = 1; |
| 4676 |
sqlda2->sqlvar[1].sqltype = ECPGt_int; |
| 4677 |
sqlda2->sqlvar[1].sqldata = (char *) &intval; |
| 4678 |
sqlda2->sqlvar[1].sqllen = sizeof(intval); |
| 4679 |
|
| 4680 |
/* Open a cursor with input parameters. */ |
| 4681 |
EXEC SQL OPEN cur1 USING DESCRIPTOR sqlda2; |
| 4682 |
|
| 4683 |
while (1) |
| 4684 |
{ |
| 4685 |
sqlda_t *cur_sqlda; |
| 4686 |
|
| 4687 |
/* Assign descriptor to the cursor */ |
| 4688 |
EXEC SQL FETCH NEXT FROM cur1 INTO DESCRIPTOR sqlda1; |
| 4689 |
|
| 4690 |
for (cur_sqlda = sqlda1 ; |
| 4691 |
cur_sqlda != NULL ; |
| 4692 |
cur_sqlda = cur_sqlda->desc_next) |
| 4693 |
{ |
| 4694 |
int i; |
| 4695 |
char name_buf[1024]; |
| 4696 |
char var_buf[1024]; |
| 4697 |
|
| 4698 |
/* Print every column in a row. */ |
| 4699 |
for (i=0 ; i<cur_sqlda->sqld ; i++) |
| 4700 |
{ |
| 4701 |
sqlvar_t v = cur_sqlda->sqlvar[i]; |
| 4702 |
char *sqldata = v.sqldata; |
| 4703 |
short sqllen = v.sqllen; |
| 4704 |
|
| 4705 |
strncpy(name_buf, v.sqlname.data, v.sqlname.length); |
| 4706 |
name_buf[v.sqlname.length] = '\0'; |
| 4707 |
|
| 4708 |
switch (v.sqltype) |
| 4709 |
{ |
| 4710 |
case ECPGt_char: |
| 4711 |
memset(&var_buf, 0, sizeof(var_buf)); |
| 4712 |
memcpy(&var_buf, sqldata, (sizeof(var_buf)<=sqllen ? sizeof(var_buf)-1 : sqllen) ); |
| 4713 |
break; |
| 4714 |
|
| 4715 |
case ECPGt_int: /* integer */ |
| 4716 |
memcpy(&intval, sqldata, sqllen); |
| 4717 |
snprintf(var_buf, sizeof(var_buf), "%d", intval); |
| 4718 |
break; |
| 4719 |
|
| 4720 |
case ECPGt_long_long: /* bigint */ |
| 4721 |
memcpy(&longlongval, sqldata, sqllen); |
| 4722 |
snprintf(var_buf, sizeof(var_buf), "%lld", longlongval); |
| 4723 |
break; |
| 4724 |
|
| 4725 |
default: |
| 4726 |
{ |
| 4727 |
int i; |
| 4728 |
memset(var_buf, 0, sizeof(var_buf)); |
| 4729 |
for (i = 0; i < sqllen; i++) |
| 4730 |
{ |
| 4731 |
char tmpbuf[16]; |
| 4732 |
snprintf(tmpbuf, sizeof(tmpbuf), "%02x ", (unsigned char) sqldata[i]); |
| 4733 |
strncat(var_buf, tmpbuf, sizeof(var_buf)); |
| 4734 |
} |
| 4735 |
} |
| 4736 |
break; |
| 4737 |
} |
| 4738 |
|
| 4739 |
printf("%s = %s (type: %d)\n", name_buf, var_buf, v.sqltype); |
| 4740 |
} |
| 4741 |
|
| 4742 |
printf("\n"); |
| 4743 |
} |
| 4744 |
} |
| 4745 |
|
| 4746 |
EXEC SQL CLOSE cur1; |
| 4747 |
EXEC SQL COMMIT; |
| 4748 |
|
| 4749 |
EXEC SQL DISCONNECT ALL; |
| 4750 |
|
| 4751 |
return 0; |
| 4752 |
} |
| 4753 |
</programlisting> |
| 4754 |
|
| 4755 |
<para> |
| 4756 |
The output of this example should look something like the |
| 4757 |
following (some numbers will vary). |
| 4758 |
</para> |
| 4759 |
|
| 4760 |
<screen> |
| 4761 |
oid = 1 (type: 1) |
| 4762 |
datname = template1 (type: 1) |
| 4763 |
datdba = 10 (type: 1) |
| 4764 |
encoding = 0 (type: 5) |
| 4765 |
datistemplate = t (type: 1) |
| 4766 |
datallowconn = t (type: 1) |
| 4767 |
dathasloginevt = f (type: 1) |
| 4768 |
datconnlimit = -1 (type: 5) |
| 4769 |
datfrozenxid = 379 (type: 1) |
| 4770 |
dattablespace = 1663 (type: 1) |
| 4771 |
datconfig = (type: 1) |
| 4772 |
datacl = {=c/uptime,uptime=CTc/uptime} (type: 1) |
| 4773 |
datid = 1 (type: 1) |
| 4774 |
datname = template1 (type: 1) |
| 4775 |
numbackends = 0 (type: 5) |
| 4776 |
xact_commit = 113606 (type: 9) |
| 4777 |
xact_rollback = 0 (type: 9) |
| 4778 |
blks_read = 130 (type: 9) |
| 4779 |
blks_hit = 7341714 (type: 9) |
| 4780 |
tup_returned = 38262679 (type: 9) |
| 4781 |
tup_fetched = 1836281 (type: 9) |
| 4782 |
tup_inserted = 0 (type: 9) |
| 4783 |
tup_updated = 0 (type: 9) |
| 4784 |
tup_deleted = 0 (type: 9) |
| 4785 |
|
| 4786 |
oid = 11511 (type: 1) |
| 4787 |
datname = postgres (type: 1) |
| 4788 |
datdba = 10 (type: 1) |
| 4789 |
encoding = 0 (type: 5) |
| 4790 |
datistemplate = f (type: 1) |
| 4791 |
datallowconn = t (type: 1) |
| 4792 |
dathasloginevt = f (type: 1) |
| 4793 |
datconnlimit = -1 (type: 5) |
| 4794 |
datfrozenxid = 379 (type: 1) |
| 4795 |
dattablespace = 1663 (type: 1) |
| 4796 |
datconfig = (type: 1) |
| 4797 |
datacl = (type: 1) |
| 4798 |
datid = 11511 (type: 1) |
| 4799 |
datname = postgres (type: 1) |
| 4800 |
numbackends = 0 (type: 5) |
| 4801 |
xact_commit = 221069 (type: 9) |
| 4802 |
xact_rollback = 18 (type: 9) |
| 4803 |
blks_read = 1176 (type: 9) |
| 4804 |
blks_hit = 13943750 (type: 9) |
| 4805 |
tup_returned = 77410091 (type: 9) |
| 4806 |
tup_fetched = 3253694 (type: 9) |
| 4807 |
tup_inserted = 0 (type: 9) |
| 4808 |
tup_updated = 0 (type: 9) |
| 4809 |
tup_deleted = 0 (type: 9) |
| 4810 |
</screen> |
| 4811 |
</example> |
| 4812 |
</sect3> |
| 4813 |
</sect2> |
| 4814 |
</sect1> |
| 4815 |
|
| 4816 |
<sect1 id="ecpg-errors"> |
| 4817 |
<title>Error Handling</title> |
| 4818 |
|
| 4819 |
<para> |
| 4820 |
This section describes how you can handle exceptional conditions |
| 4821 |
and warnings in an embedded SQL program. There are two |
| 4822 |
nonexclusive facilities for this. |
| 4823 |
|
| 4824 |
<itemizedlist> |
| 4825 |
<listitem> |
| 4826 |
<simpara> |
| 4827 |
Callbacks can be configured to handle warning and error |
| 4828 |
conditions using the <literal>WHENEVER</literal> command. |
| 4829 |
</simpara> |
| 4830 |
</listitem> |
| 4831 |
|
| 4832 |
<listitem> |
| 4833 |
<simpara> |
| 4834 |
Detailed information about the error or warning can be obtained |
| 4835 |
from the <varname>sqlca</varname> variable. |
| 4836 |
</simpara> |
| 4837 |
</listitem> |
| 4838 |
</itemizedlist> |
| 4839 |
</para> |
| 4840 |
|
| 4841 |
<sect2 id="ecpg-whenever"> |
| 4842 |
<title>Setting Callbacks</title> |
| 4843 |
|
| 4844 |
<para> |
| 4845 |
One simple method to catch errors and warnings is to set a |
| 4846 |
specific action to be executed whenever a particular condition |
| 4847 |
occurs. In general: |
| 4848 |
<programlisting> |
| 4849 |
EXEC SQL WHENEVER <replaceable>condition</replaceable> <replaceable>action</replaceable>; |
| 4850 |
</programlisting> |
| 4851 |
</para> |
| 4852 |
|
| 4853 |
<para> |
| 4854 |
<replaceable>condition</replaceable> can be one of the following: |
| 4855 |
|
| 4856 |
<variablelist> |
| 4857 |
<varlistentry id="ecpg-whenever-sqlerror"> |
| 4858 |
<term><literal>SQLERROR</literal></term> |
| 4859 |
<listitem> |
| 4860 |
<para> |
| 4861 |
The specified action is called whenever an error occurs during |
| 4862 |
the execution of an SQL statement. |
| 4863 |
</para> |
| 4864 |
</listitem> |
| 4865 |
</varlistentry> |
| 4866 |
|
| 4867 |
<varlistentry id="ecpg-whenever-sqlwarning"> |
| 4868 |
<term><literal>SQLWARNING</literal></term> |
| 4869 |
<listitem> |
| 4870 |
<para> |
| 4871 |
The specified action is called whenever a warning occurs |
| 4872 |
during the execution of an SQL statement. |
| 4873 |
</para> |
| 4874 |
</listitem> |
| 4875 |
</varlistentry> |
| 4876 |
|
| 4877 |
<varlistentry id="ecpg-whenever-not-found"> |
| 4878 |
<term><literal>NOT FOUND</literal></term> |
| 4879 |
<listitem> |
| 4880 |
<para> |
| 4881 |
The specified action is called whenever an SQL statement |
| 4882 |
retrieves or affects zero rows. (This condition is not an |
| 4883 |
error, but you might be interested in handling it specially.) |
| 4884 |
</para> |
| 4885 |
</listitem> |
| 4886 |
</varlistentry> |
| 4887 |
</variablelist> |
| 4888 |
</para> |
| 4889 |
|
| 4890 |
<para> |
| 4891 |
<replaceable>action</replaceable> can be one of the following: |
| 4892 |
|
| 4893 |
<variablelist> |
| 4894 |
<varlistentry id="ecpg-whenever-continue"> |
| 4895 |
<term><literal>CONTINUE</literal></term> |
| 4896 |
<listitem> |
| 4897 |
<para> |
| 4898 |
This effectively means that the condition is ignored. This is |
| 4899 |
the default. |
| 4900 |
</para> |
| 4901 |
</listitem> |
| 4902 |
</varlistentry> |
| 4903 |
|
| 4904 |
<varlistentry id="ecpg-whenever-goto"> |
| 4905 |
<term><literal>GOTO <replaceable>label</replaceable></literal></term> |
| 4906 |
<term><literal>GO TO <replaceable>label</replaceable></literal></term> |
| 4907 |
<listitem> |
| 4908 |
<para> |
| 4909 |
Jump to the specified label (using a C <literal>goto</literal> |
| 4910 |
statement). |
| 4911 |
</para> |
| 4912 |
</listitem> |
| 4913 |
</varlistentry> |
| 4914 |
|
| 4915 |
<varlistentry id="ecpg-whenever-sqlprint"> |
| 4916 |
<term><literal>SQLPRINT</literal></term> |
| 4917 |
<listitem> |
| 4918 |
<para> |
| 4919 |
Print a message to standard error. This is useful for simple |
| 4920 |
programs or during prototyping. The details of the message |
| 4921 |
cannot be configured. |
| 4922 |
</para> |
| 4923 |
</listitem> |
| 4924 |
</varlistentry> |
| 4925 |
|
| 4926 |
<varlistentry id="ecpg-whenever-stop"> |
| 4927 |
<term><literal>STOP</literal></term> |
| 4928 |
<listitem> |
| 4929 |
<para> |
| 4930 |
Call <literal>exit(1)</literal>, which will terminate the |
| 4931 |
program. |
| 4932 |
</para> |
| 4933 |
</listitem> |
| 4934 |
</varlistentry> |
| 4935 |
|
| 4936 |
<varlistentry id="ecpg-whenever-do-break"> |
| 4937 |
<term><literal>DO BREAK</literal></term> |
| 4938 |
<listitem> |
| 4939 |
<para> |
| 4940 |
Execute the C statement <literal>break</literal>. This should |
| 4941 |
only be used in loops or <literal>switch</literal> statements. |
| 4942 |
</para> |
| 4943 |
</listitem> |
| 4944 |
</varlistentry> |
| 4945 |
|
| 4946 |
<varlistentry id="ecpg-whenever-do-continue"> |
| 4947 |
<term><literal>DO CONTINUE</literal></term> |
| 4948 |
<listitem> |
| 4949 |
<para> |
| 4950 |
Execute the C statement <literal>continue</literal>. This should |
| 4951 |
only be used in loops statements. if executed, will cause the flow |
| 4952 |
of control to return to the top of the loop. |
| 4953 |
</para> |
| 4954 |
</listitem> |
| 4955 |
</varlistentry> |
| 4956 |
|
| 4957 |
<varlistentry id="ecpg-whenever-call"> |
| 4958 |
<term><literal>CALL <replaceable>name</replaceable> (<replaceable>args</replaceable>)</literal></term> |
| 4959 |
<term><literal>DO <replaceable>name</replaceable> (<replaceable>args</replaceable>)</literal></term> |
| 4960 |
<listitem> |
| 4961 |
<para> |
| 4962 |
Call the specified C functions with the specified arguments. (This |
| 4963 |
use is different from the meaning of <literal>CALL</literal> |
| 4964 |
and <literal>DO</literal> in the normal PostgreSQL grammar.) |
| 4965 |
</para> |
| 4966 |
</listitem> |
| 4967 |
</varlistentry> |
| 4968 |
</variablelist> |
| 4969 |
|
| 4970 |
The SQL standard only provides for the actions |
| 4971 |
<literal>CONTINUE</literal> and <literal>GOTO</literal> (and |
| 4972 |
<literal>GO TO</literal>). |
| 4973 |
</para> |
| 4974 |
|
| 4975 |
<para> |
| 4976 |
Here is an example that you might want to use in a simple program. |
| 4977 |
It prints a simple message when a warning occurs and aborts the |
| 4978 |
program when an error happens: |
| 4979 |
<programlisting> |
| 4980 |
EXEC SQL WHENEVER SQLWARNING SQLPRINT; |
| 4981 |
EXEC SQL WHENEVER SQLERROR STOP; |
| 4982 |
</programlisting> |
| 4983 |
</para> |
| 4984 |
|
| 4985 |
<para> |
| 4986 |
The statement <literal>EXEC SQL WHENEVER</literal> is a directive |
| 4987 |
of the SQL preprocessor, not a C statement. The error or warning |
| 4988 |
actions that it sets apply to all embedded SQL statements that |
| 4989 |
appear below the point where the handler is set, unless a |
| 4990 |
different action was set for the same condition between the first |
| 4991 |
<literal>EXEC SQL WHENEVER</literal> and the SQL statement causing |
| 4992 |
the condition, regardless of the flow of control in the C program. |
| 4993 |
So neither of the two following C program excerpts will have the |
| 4994 |
desired effect: |
| 4995 |
<programlisting> |
| 4996 |
/* |
| 4997 |
* WRONG |
| 4998 |
*/ |
| 4999 |
int main(int argc, char *argv[]) |
| 5000 |
{ |
| 5001 |
... |
| 5002 |
if (verbose) { |
| 5003 |
EXEC SQL WHENEVER SQLWARNING SQLPRINT; |
| 5004 |
} |
| 5005 |
... |
| 5006 |
EXEC SQL SELECT ...; |
| 5007 |
... |
| 5008 |
} |
| 5009 |
</programlisting> |
| 5010 |
|
| 5011 |
<programlisting> |
| 5012 |
/* |
| 5013 |
* WRONG |
| 5014 |
*/ |
| 5015 |
int main(int argc, char *argv[]) |
| 5016 |
{ |
| 5017 |
... |
| 5018 |
set_error_handler(); |
| 5019 |
... |
| 5020 |
EXEC SQL SELECT ...; |
| 5021 |
... |
| 5022 |
} |
| 5023 |
|
| 5024 |
static void set_error_handler(void) |
| 5025 |
{ |
| 5026 |
EXEC SQL WHENEVER SQLERROR STOP; |
| 5027 |
} |
| 5028 |
</programlisting> |
| 5029 |
</para> |
| 5030 |
</sect2> |
| 5031 |
|
| 5032 |
<sect2 id="ecpg-sqlca"> |
| 5033 |
<title>sqlca</title> |
| 5034 |
|
| 5035 |
<para> |
| 5036 |
For more powerful error handling, the embedded SQL interface |
| 5037 |
provides a global variable with the name <varname>sqlca</varname> |
| 5038 |
(SQL communication area) |
| 5039 |
that has the following structure: |
| 5040 |
<programlisting> |
| 5041 |
struct |
| 5042 |
{ |
| 5043 |
char sqlcaid[8]; |
| 5044 |
long sqlabc; |
| 5045 |
long sqlcode; |
| 5046 |
struct |
| 5047 |
{ |
| 5048 |
int sqlerrml; |
| 5049 |
char sqlerrmc[SQLERRMC_LEN]; |
| 5050 |
} sqlerrm; |
| 5051 |
char sqlerrp[8]; |
| 5052 |
long sqlerrd[6]; |
| 5053 |
char sqlwarn[8]; |
| 5054 |
char sqlstate[5]; |
| 5055 |
} sqlca; |
| 5056 |
</programlisting> |
| 5057 |
(In a multithreaded program, every thread automatically gets its |
| 5058 |
own copy of <varname>sqlca</varname>. This works similarly to the |
| 5059 |
handling of the standard C global variable |
| 5060 |
<varname>errno</varname>.) |
| 5061 |
</para> |
| 5062 |
|
| 5063 |
<para> |
| 5064 |
<varname>sqlca</varname> covers both warnings and errors. If |
| 5065 |
multiple warnings or errors occur during the execution of a |
| 5066 |
statement, then <varname>sqlca</varname> will only contain |
| 5067 |
information about the last one. |
| 5068 |
</para> |
| 5069 |
|
| 5070 |
<para> |
| 5071 |
If no error occurred in the last <acronym>SQL</acronym> statement, |
| 5072 |
<literal>sqlca.sqlcode</literal> will be 0 and |
| 5073 |
<literal>sqlca.sqlstate</literal> will be |
| 5074 |
<literal>"00000"</literal>. If a warning or error occurred, then |
| 5075 |
<literal>sqlca.sqlcode</literal> will be negative and |
| 5076 |
<literal>sqlca.sqlstate</literal> will be different from |
| 5077 |
<literal>"00000"</literal>. A positive |
| 5078 |
<literal>sqlca.sqlcode</literal> indicates a harmless condition, |
| 5079 |
such as that the last query returned zero rows. |
| 5080 |
<literal>sqlcode</literal> and <literal>sqlstate</literal> are two |
| 5081 |
different error code schemes; details appear below. |
| 5082 |
</para> |
| 5083 |
|
| 5084 |
<para> |
| 5085 |
If the last SQL statement was successful, then |
| 5086 |
<literal>sqlca.sqlerrd[1]</literal> contains the OID of the |
| 5087 |
processed row, if applicable, and |
| 5088 |
<literal>sqlca.sqlerrd[2]</literal> contains the number of |
| 5089 |
processed or returned rows, if applicable to the command. |
| 5090 |
</para> |
| 5091 |
|
| 5092 |
<para> |
| 5093 |
In case of an error or warning, |
| 5094 |
<literal>sqlca.sqlerrm.sqlerrmc</literal> will contain a string |
| 5095 |
that describes the error. The field |
| 5096 |
<literal>sqlca.sqlerrm.sqlerrml</literal> contains the length of |
| 5097 |
the error message that is stored in |
| 5098 |
<literal>sqlca.sqlerrm.sqlerrmc</literal> (the result of |
| 5099 |
<function>strlen()</function>, not really interesting for a C |
| 5100 |
programmer). Note that some messages are too long to fit in the |
| 5101 |
fixed-size <literal>sqlerrmc</literal> array; they will be truncated. |
| 5102 |
</para> |
| 5103 |
|
| 5104 |
<para> |
| 5105 |
In case of a warning, <literal>sqlca.sqlwarn[2]</literal> is set |
| 5106 |
to <literal>W</literal>. (In all other cases, it is set to |
| 5107 |
something different from <literal>W</literal>.) If |
| 5108 |
<literal>sqlca.sqlwarn[1]</literal> is set to |
| 5109 |
<literal>W</literal>, then a value was truncated when it was |
| 5110 |
stored in a host variable. <literal>sqlca.sqlwarn[0]</literal> is |
| 5111 |
set to <literal>W</literal> if any of the other elements are set |
| 5112 |
to indicate a warning. |
| 5113 |
</para> |
| 5114 |
|
| 5115 |
<para> |
| 5116 |
The fields <structfield>sqlcaid</structfield>, |
| 5117 |
<structfield>sqlabc</structfield>, |
| 5118 |
<structfield>sqlerrp</structfield>, and the remaining elements of |
| 5119 |
<structfield>sqlerrd</structfield> and |
| 5120 |
<structfield>sqlwarn</structfield> currently contain no useful |
| 5121 |
information. |
| 5122 |
</para> |
| 5123 |
|
| 5124 |
<para> |
| 5125 |
The structure <varname>sqlca</varname> is not defined in the SQL |
| 5126 |
standard, but is implemented in several other SQL database |
| 5127 |
systems. The definitions are similar at the core, but if you want |
| 5128 |
to write portable applications, then you should investigate the |
| 5129 |
different implementations carefully. |
| 5130 |
</para> |
| 5131 |
|
| 5132 |
<para> |
| 5133 |
Here is one example that combines the use of <literal>WHENEVER</literal> |
| 5134 |
and <varname>sqlca</varname>, printing out the contents |
| 5135 |
of <varname>sqlca</varname> when an error occurs. This is perhaps |
| 5136 |
useful for debugging or prototyping applications, before |
| 5137 |
installing a more <quote>user-friendly</quote> error handler. |
| 5138 |
|
| 5139 |
<programlisting> |
| 5140 |
EXEC SQL WHENEVER SQLERROR CALL print_sqlca(); |
| 5141 |
|
| 5142 |
void |
| 5143 |
print_sqlca() |
| 5144 |
{ |
| 5145 |
fprintf(stderr, "==== sqlca ====\n"); |
| 5146 |
fprintf(stderr, "sqlcode: %ld\n", sqlca.sqlcode); |
| 5147 |
fprintf(stderr, "sqlerrm.sqlerrml: %d\n", sqlca.sqlerrm.sqlerrml); |
| 5148 |
fprintf(stderr, "sqlerrm.sqlerrmc: %s\n", sqlca.sqlerrm.sqlerrmc); |
| 5149 |
fprintf(stderr, "sqlerrd: %ld %ld %ld %ld %ld %ld\n", sqlca.sqlerrd[0],sqlca.sqlerrd[1],sqlca.sqlerrd[2], |
| 5150 |
sqlca.sqlerrd[3],sqlca.sqlerrd[4],sqlca.sqlerrd[5]); |
| 5151 |
fprintf(stderr, "sqlwarn: %d %d %d %d %d %d %d %d\n", sqlca.sqlwarn[0], sqlca.sqlwarn[1], sqlca.sqlwarn[2], |
| 5152 |
sqlca.sqlwarn[3], sqlca.sqlwarn[4], sqlca.sqlwarn[5], |
| 5153 |
sqlca.sqlwarn[6], sqlca.sqlwarn[7]); |
| 5154 |
fprintf(stderr, "sqlstate: %5s\n", sqlca.sqlstate); |
| 5155 |
fprintf(stderr, "===============\n"); |
| 5156 |
} |
| 5157 |
</programlisting> |
| 5158 |
|
| 5159 |
The result could look as follows (here an error due to a |
| 5160 |
misspelled table name): |
| 5161 |
|
| 5162 |
<screen> |
| 5163 |
==== sqlca ==== |
| 5164 |
sqlcode: -400 |
| 5165 |
sqlerrm.sqlerrml: 49 |
| 5166 |
sqlerrm.sqlerrmc: relation "pg_databasep" does not exist on line 38 |
| 5167 |
sqlerrd: 0 0 0 0 0 0 |
| 5168 |
sqlwarn: 0 0 0 0 0 0 0 0 |
| 5169 |
sqlstate: 42P01 |
| 5170 |
=============== |
| 5171 |
</screen> |
| 5172 |
</para> |
| 5173 |
</sect2> |
| 5174 |
|
| 5175 |
<sect2 id="ecpg-sqlstate-sqlcode"> |
| 5176 |
<title><literal>SQLSTATE</literal> vs. <literal>SQLCODE</literal></title> |
| 5177 |
|
| 5178 |
<para> |
| 5179 |
The fields <literal>sqlca.sqlstate</literal> and |
| 5180 |
<literal>sqlca.sqlcode</literal> are two different schemes that |
| 5181 |
provide error codes. Both are derived from the SQL standard, but |
| 5182 |
<literal>SQLCODE</literal> has been marked deprecated in the SQL-92 |
| 5183 |
edition of the standard and has been dropped in later editions. |
| 5184 |
Therefore, new applications are strongly encouraged to use |
| 5185 |
<literal>SQLSTATE</literal>. |
| 5186 |
</para> |
| 5187 |
|
| 5188 |
<para> |
| 5189 |
<literal>SQLSTATE</literal> is a five-character array. The five |
| 5190 |
characters contain digits or upper-case letters that represent |
| 5191 |
codes of various error and warning conditions. |
| 5192 |
<literal>SQLSTATE</literal> has a hierarchical scheme: the first |
| 5193 |
two characters indicate the general class of the condition, the |
| 5194 |
last three characters indicate a subclass of the general |
| 5195 |
condition. A successful state is indicated by the code |
| 5196 |
<literal>00000</literal>. The <literal>SQLSTATE</literal> codes are for |
| 5197 |
the most part defined in the SQL standard. The |
| 5198 |
<productname>PostgreSQL</productname> server natively supports |
| 5199 |
<literal>SQLSTATE</literal> error codes; therefore a high degree |
| 5200 |
of consistency can be achieved by using this error code scheme |
| 5201 |
throughout all applications. For further information see |
| 5202 |
<xref linkend="errcodes-appendix"/>. |
| 5203 |
</para> |
| 5204 |
|
| 5205 |
<para> |
| 5206 |
<literal>SQLCODE</literal>, the deprecated error code scheme, is a |
| 5207 |
simple integer. A value of 0 indicates success, a positive value |
| 5208 |
indicates success with additional information, a negative value |
| 5209 |
indicates an error. The SQL standard only defines the positive |
| 5210 |
value +100, which indicates that the last command returned or |
| 5211 |
affected zero rows, and no specific negative values. Therefore, |
| 5212 |
this scheme can only achieve poor portability and does not have a |
| 5213 |
hierarchical code assignment. Historically, the embedded SQL |
| 5214 |
processor for <productname>PostgreSQL</productname> has assigned |
| 5215 |
some specific <literal>SQLCODE</literal> values for its use, which |
| 5216 |
are listed below with their numeric value and their symbolic name. |
| 5217 |
Remember that these are not portable to other SQL implementations. |
| 5218 |
To simplify the porting of applications to the |
| 5219 |
<literal>SQLSTATE</literal> scheme, the corresponding |
| 5220 |
<literal>SQLSTATE</literal> is also listed. There is, however, no |
| 5221 |
one-to-one or one-to-many mapping between the two schemes (indeed |
| 5222 |
it is many-to-many), so you should consult the global |
| 5223 |
<literal>SQLSTATE</literal> listing in <xref linkend="errcodes-appendix"/> |
| 5224 |
in each case. |
| 5225 |
</para> |
| 5226 |
|
| 5227 |
<para> |
| 5228 |
These are the assigned <literal>SQLCODE</literal> values: |
| 5229 |
|
| 5230 |
<variablelist> |
| 5231 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-no-error"> |
| 5232 |
<term>0 (<symbol>ECPG_NO_ERROR</symbol>)</term> |
| 5233 |
<listitem> |
| 5234 |
<para> |
| 5235 |
Indicates no error. (SQLSTATE 00000) |
| 5236 |
</para> |
| 5237 |
</listitem> |
| 5238 |
</varlistentry> |
| 5239 |
|
| 5240 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-not-found"> |
| 5241 |
<term>100 (<symbol>ECPG_NOT_FOUND</symbol>)</term> |
| 5242 |
<listitem> |
| 5243 |
<para> |
| 5244 |
This is a harmless condition indicating that the last command |
| 5245 |
retrieved or processed zero rows, or that you are at the end of |
| 5246 |
the cursor. (SQLSTATE 02000) |
| 5247 |
</para> |
| 5248 |
|
| 5249 |
<para> |
| 5250 |
When processing a cursor in a loop, you could use this code as |
| 5251 |
a way to detect when to abort the loop, like this: |
| 5252 |
<programlisting> |
| 5253 |
while (1) |
| 5254 |
{ |
| 5255 |
EXEC SQL FETCH ... ; |
| 5256 |
if (sqlca.sqlcode == ECPG_NOT_FOUND) |
| 5257 |
break; |
| 5258 |
} |
| 5259 |
</programlisting> |
| 5260 |
But <literal>WHENEVER NOT FOUND DO BREAK</literal> effectively |
| 5261 |
does this internally, so there is usually no advantage in |
| 5262 |
writing this out explicitly. |
| 5263 |
</para> |
| 5264 |
</listitem> |
| 5265 |
</varlistentry> |
| 5266 |
|
| 5267 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-out-of-memory"> |
| 5268 |
<term>-12 (<symbol>ECPG_OUT_OF_MEMORY</symbol>)</term> |
| 5269 |
<listitem> |
| 5270 |
<para> |
| 5271 |
Indicates that your virtual memory is exhausted. The numeric |
| 5272 |
value is defined as <literal>-ENOMEM</literal>. (SQLSTATE |
| 5273 |
YE001) |
| 5274 |
</para> |
| 5275 |
</listitem> |
| 5276 |
</varlistentry> |
| 5277 |
|
| 5278 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-unsupported"> |
| 5279 |
<term>-200 (<symbol>ECPG_UNSUPPORTED</symbol>)</term> |
| 5280 |
<listitem> |
| 5281 |
<para> |
| 5282 |
Indicates the preprocessor has generated something that the |
| 5283 |
library does not know about. Perhaps you are running |
| 5284 |
incompatible versions of the preprocessor and the |
| 5285 |
library. (SQLSTATE YE002) |
| 5286 |
</para> |
| 5287 |
</listitem> |
| 5288 |
</varlistentry> |
| 5289 |
|
| 5290 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-too-many-arguments"> |
| 5291 |
<term>-201 (<symbol>ECPG_TOO_MANY_ARGUMENTS</symbol>)</term> |
| 5292 |
<listitem> |
| 5293 |
<para> |
| 5294 |
This means that the command specified more host variables than |
| 5295 |
the command expected. (SQLSTATE 07001 or 07002) |
| 5296 |
</para> |
| 5297 |
</listitem> |
| 5298 |
</varlistentry> |
| 5299 |
|
| 5300 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-too-few-arguments"> |
| 5301 |
<term>-202 (<symbol>ECPG_TOO_FEW_ARGUMENTS</symbol>)</term> |
| 5302 |
<listitem> |
| 5303 |
<para> |
| 5304 |
This means that the command specified fewer host variables than |
| 5305 |
the command expected. (SQLSTATE 07001 or 07002) |
| 5306 |
</para> |
| 5307 |
</listitem> |
| 5308 |
</varlistentry> |
| 5309 |
|
| 5310 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-too-many-matches"> |
| 5311 |
<term>-203 (<symbol>ECPG_TOO_MANY_MATCHES</symbol>)</term> |
| 5312 |
<listitem> |
| 5313 |
<para> |
| 5314 |
This means a query has returned multiple rows but the statement |
| 5315 |
was only prepared to store one result row (for example, because |
| 5316 |
the specified variables are not arrays). (SQLSTATE 21000) |
| 5317 |
</para> |
| 5318 |
</listitem> |
| 5319 |
</varlistentry> |
| 5320 |
|
| 5321 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-int-format"> |
| 5322 |
<term>-204 (<symbol>ECPG_INT_FORMAT</symbol>)</term> |
| 5323 |
<listitem> |
| 5324 |
<para> |
| 5325 |
The host variable is of type <type>int</type> and the datum in |
| 5326 |
the database is of a different type and contains a value that |
| 5327 |
cannot be interpreted as an <type>int</type>. The library uses |
| 5328 |
<function>strtol()</function> for this conversion. (SQLSTATE |
| 5329 |
42804) |
| 5330 |
</para> |
| 5331 |
</listitem> |
| 5332 |
</varlistentry> |
| 5333 |
|
| 5334 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-uint-format"> |
| 5335 |
<term>-205 (<symbol>ECPG_UINT_FORMAT</symbol>)</term> |
| 5336 |
<listitem> |
| 5337 |
<para> |
| 5338 |
The host variable is of type <type>unsigned int</type> and the |
| 5339 |
datum in the database is of a different type and contains a |
| 5340 |
value that cannot be interpreted as an <type>unsigned |
| 5341 |
int</type>. The library uses <function>strtoul()</function> |
| 5342 |
for this conversion. (SQLSTATE 42804) |
| 5343 |
</para> |
| 5344 |
</listitem> |
| 5345 |
</varlistentry> |
| 5346 |
|
| 5347 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-float-format"> |
| 5348 |
<term>-206 (<symbol>ECPG_FLOAT_FORMAT</symbol>)</term> |
| 5349 |
<listitem> |
| 5350 |
<para> |
| 5351 |
The host variable is of type <type>float</type> and the datum |
| 5352 |
in the database is of another type and contains a value that |
| 5353 |
cannot be interpreted as a <type>float</type>. The library |
| 5354 |
uses <function>strtod()</function> for this conversion. |
| 5355 |
(SQLSTATE 42804) |
| 5356 |
</para> |
| 5357 |
</listitem> |
| 5358 |
</varlistentry> |
| 5359 |
|
| 5360 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-numeric-format"> |
| 5361 |
<term>-207 (<symbol>ECPG_NUMERIC_FORMAT</symbol>)</term> |
| 5362 |
<listitem> |
| 5363 |
<para> |
| 5364 |
The host variable is of type <type>numeric</type> and the datum |
| 5365 |
in the database is of another type and contains a value that |
| 5366 |
cannot be interpreted as a <type>numeric</type> value. |
| 5367 |
(SQLSTATE 42804) |
| 5368 |
</para> |
| 5369 |
</listitem> |
| 5370 |
</varlistentry> |
| 5371 |
|
| 5372 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-interval-format"> |
| 5373 |
<term>-208 (<symbol>ECPG_INTERVAL_FORMAT</symbol>)</term> |
| 5374 |
<listitem> |
| 5375 |
<para> |
| 5376 |
The host variable is of type <type>interval</type> and the datum |
| 5377 |
in the database is of another type and contains a value that |
| 5378 |
cannot be interpreted as an <type>interval</type> value. |
| 5379 |
(SQLSTATE 42804) |
| 5380 |
</para> |
| 5381 |
</listitem> |
| 5382 |
</varlistentry> |
| 5383 |
|
| 5384 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-date-format"> |
| 5385 |
<term>-209 (<symbol>ECPG_DATE_FORMAT</symbol>)</term> |
| 5386 |
<listitem> |
| 5387 |
<para> |
| 5388 |
The host variable is of type <type>date</type> and the datum in |
| 5389 |
the database is of another type and contains a value that |
| 5390 |
cannot be interpreted as a <type>date</type> value. |
| 5391 |
(SQLSTATE 42804) |
| 5392 |
</para> |
| 5393 |
</listitem> |
| 5394 |
</varlistentry> |
| 5395 |
|
| 5396 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-timestamp-format"> |
| 5397 |
<term>-210 (<symbol>ECPG_TIMESTAMP_FORMAT</symbol>)</term> |
| 5398 |
<listitem> |
| 5399 |
<para> |
| 5400 |
The host variable is of type <type>timestamp</type> and the |
| 5401 |
datum in the database is of another type and contains a value |
| 5402 |
that cannot be interpreted as a <type>timestamp</type> value. |
| 5403 |
(SQLSTATE 42804) |
| 5404 |
</para> |
| 5405 |
</listitem> |
| 5406 |
</varlistentry> |
| 5407 |
|
| 5408 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-convert-bool"> |
| 5409 |
<term>-211 (<symbol>ECPG_CONVERT_BOOL</symbol>)</term> |
| 5410 |
<listitem> |
| 5411 |
<para> |
| 5412 |
This means the host variable is of type <type>bool</type> and |
| 5413 |
the datum in the database is neither <literal>'t'</literal> nor |
| 5414 |
<literal>'f'</literal>. (SQLSTATE 42804) |
| 5415 |
</para> |
| 5416 |
</listitem> |
| 5417 |
</varlistentry> |
| 5418 |
|
| 5419 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-empty"> |
| 5420 |
<term>-212 (<symbol>ECPG_EMPTY</symbol>)</term> |
| 5421 |
<listitem> |
| 5422 |
<para> |
| 5423 |
The statement sent to the <productname>PostgreSQL</productname> |
| 5424 |
server was empty. (This cannot normally happen in an embedded |
| 5425 |
SQL program, so it might point to an internal error.) (SQLSTATE |
| 5426 |
YE002) |
| 5427 |
</para> |
| 5428 |
</listitem> |
| 5429 |
</varlistentry> |
| 5430 |
|
| 5431 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-missing-indicator"> |
| 5432 |
<term>-213 (<symbol>ECPG_MISSING_INDICATOR</symbol>)</term> |
| 5433 |
<listitem> |
| 5434 |
<para> |
| 5435 |
A null value was returned and no null indicator variable was |
| 5436 |
supplied. (SQLSTATE 22002) |
| 5437 |
</para> |
| 5438 |
</listitem> |
| 5439 |
</varlistentry> |
| 5440 |
|
| 5441 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-no-array"> |
| 5442 |
<term>-214 (<symbol>ECPG_NO_ARRAY</symbol>)</term> |
| 5443 |
<listitem> |
| 5444 |
<para> |
| 5445 |
An ordinary variable was used in a place that requires an |
| 5446 |
array. (SQLSTATE 42804) |
| 5447 |
</para> |
| 5448 |
</listitem> |
| 5449 |
</varlistentry> |
| 5450 |
|
| 5451 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-data-not-array"> |
| 5452 |
<term>-215 (<symbol>ECPG_DATA_NOT_ARRAY</symbol>)</term> |
| 5453 |
<listitem> |
| 5454 |
<para> |
| 5455 |
The database returned an ordinary variable in a place that |
| 5456 |
requires array value. (SQLSTATE 42804) |
| 5457 |
</para> |
| 5458 |
</listitem> |
| 5459 |
</varlistentry> |
| 5460 |
|
| 5461 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-array-insert"> |
| 5462 |
<term>-216 (<symbol>ECPG_ARRAY_INSERT</symbol>)</term> |
| 5463 |
<listitem> |
| 5464 |
<para> |
| 5465 |
The value could not be inserted into the array. (SQLSTATE |
| 5466 |
42804) |
| 5467 |
</para> |
| 5468 |
</listitem> |
| 5469 |
</varlistentry> |
| 5470 |
|
| 5471 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-no-conn"> |
| 5472 |
<term>-220 (<symbol>ECPG_NO_CONN</symbol>)</term> |
| 5473 |
<listitem> |
| 5474 |
<para> |
| 5475 |
The program tried to access a connection that does not exist. |
| 5476 |
(SQLSTATE 08003) |
| 5477 |
</para> |
| 5478 |
</listitem> |
| 5479 |
</varlistentry> |
| 5480 |
|
| 5481 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-not-conn"> |
| 5482 |
<term>-221 (<symbol>ECPG_NOT_CONN</symbol>)</term> |
| 5483 |
<listitem> |
| 5484 |
<para> |
| 5485 |
The program tried to access a connection that does exist but is |
| 5486 |
not open. (This is an internal error.) (SQLSTATE YE002) |
| 5487 |
</para> |
| 5488 |
</listitem> |
| 5489 |
</varlistentry> |
| 5490 |
|
| 5491 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-invalid-stmt"> |
| 5492 |
<term>-230 (<symbol>ECPG_INVALID_STMT</symbol>)</term> |
| 5493 |
<listitem> |
| 5494 |
<para> |
| 5495 |
The statement you are trying to use has not been prepared. |
| 5496 |
(SQLSTATE 26000) |
| 5497 |
</para> |
| 5498 |
</listitem> |
| 5499 |
</varlistentry> |
| 5500 |
|
| 5501 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-informix-duplicate-key"> |
| 5502 |
<term>-239 (<symbol>ECPG_INFORMIX_DUPLICATE_KEY</symbol>)</term> |
| 5503 |
<listitem> |
| 5504 |
<para> |
| 5505 |
Duplicate key error, violation of unique constraint (Informix |
| 5506 |
compatibility mode). (SQLSTATE 23505) |
| 5507 |
</para> |
| 5508 |
</listitem> |
| 5509 |
</varlistentry> |
| 5510 |
|
| 5511 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-unknown-descriptor"> |
| 5512 |
<term>-240 (<symbol>ECPG_UNKNOWN_DESCRIPTOR</symbol>)</term> |
| 5513 |
<listitem> |
| 5514 |
<para> |
| 5515 |
The descriptor specified was not found. The statement you are |
| 5516 |
trying to use has not been prepared. (SQLSTATE 33000) |
| 5517 |
</para> |
| 5518 |
</listitem> |
| 5519 |
</varlistentry> |
| 5520 |
|
| 5521 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-invalid-descriptor-index"> |
| 5522 |
<term>-241 (<symbol>ECPG_INVALID_DESCRIPTOR_INDEX</symbol>)</term> |
| 5523 |
<listitem> |
| 5524 |
<para> |
| 5525 |
The descriptor index specified was out of range. (SQLSTATE |
| 5526 |
07009) |
| 5527 |
</para> |
| 5528 |
</listitem> |
| 5529 |
</varlistentry> |
| 5530 |
|
| 5531 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-unknown-descriptor-item"> |
| 5532 |
<term>-242 (<symbol>ECPG_UNKNOWN_DESCRIPTOR_ITEM</symbol>)</term> |
| 5533 |
<listitem> |
| 5534 |
<para> |
| 5535 |
An invalid descriptor item was requested. (This is an internal |
| 5536 |
error.) (SQLSTATE YE002) |
| 5537 |
</para> |
| 5538 |
</listitem> |
| 5539 |
</varlistentry> |
| 5540 |
|
| 5541 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-var-not-numeric"> |
| 5542 |
<term>-243 (<symbol>ECPG_VAR_NOT_NUMERIC</symbol>)</term> |
| 5543 |
<listitem> |
| 5544 |
<para> |
| 5545 |
During the execution of a dynamic statement, the database |
| 5546 |
returned a numeric value and the host variable was not numeric. |
| 5547 |
(SQLSTATE 07006) |
| 5548 |
</para> |
| 5549 |
</listitem> |
| 5550 |
</varlistentry> |
| 5551 |
|
| 5552 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-var-not-char"> |
| 5553 |
<term>-244 (<symbol>ECPG_VAR_NOT_CHAR</symbol>)</term> |
| 5554 |
<listitem> |
| 5555 |
<para> |
| 5556 |
During the execution of a dynamic statement, the database |
| 5557 |
returned a non-numeric value and the host variable was numeric. |
| 5558 |
(SQLSTATE 07006) |
| 5559 |
</para> |
| 5560 |
</listitem> |
| 5561 |
</varlistentry> |
| 5562 |
|
| 5563 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-informix-subselect-not-one"> |
| 5564 |
<term>-284 (<symbol>ECPG_INFORMIX_SUBSELECT_NOT_ONE</symbol>)</term> |
| 5565 |
<listitem> |
| 5566 |
<para> |
| 5567 |
A result of the subquery is not single row (Informix |
| 5568 |
compatibility mode). (SQLSTATE 21000) |
| 5569 |
</para> |
| 5570 |
</listitem> |
| 5571 |
</varlistentry> |
| 5572 |
|
| 5573 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-pgsql"> |
| 5574 |
<term>-400 (<symbol>ECPG_PGSQL</symbol>)</term> |
| 5575 |
<listitem> |
| 5576 |
<para> |
| 5577 |
Some error caused by the <productname>PostgreSQL</productname> |
| 5578 |
server. The message contains the error message from the |
| 5579 |
<productname>PostgreSQL</productname> server. |
| 5580 |
</para> |
| 5581 |
</listitem> |
| 5582 |
</varlistentry> |
| 5583 |
|
| 5584 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-trans"> |
| 5585 |
<term>-401 (<symbol>ECPG_TRANS</symbol>)</term> |
| 5586 |
<listitem> |
| 5587 |
<para> |
| 5588 |
The <productname>PostgreSQL</productname> server signaled that |
| 5589 |
we cannot start, commit, or rollback the transaction. |
| 5590 |
(SQLSTATE 08007) |
| 5591 |
</para> |
| 5592 |
</listitem> |
| 5593 |
</varlistentry> |
| 5594 |
|
| 5595 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-connect"> |
| 5596 |
<term>-402 (<symbol>ECPG_CONNECT</symbol>)</term> |
| 5597 |
<listitem> |
| 5598 |
<para> |
| 5599 |
The connection attempt to the database did not succeed. |
| 5600 |
(SQLSTATE 08001) |
| 5601 |
</para> |
| 5602 |
</listitem> |
| 5603 |
</varlistentry> |
| 5604 |
|
| 5605 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-duplicate-key"> |
| 5606 |
<term>-403 (<symbol>ECPG_DUPLICATE_KEY</symbol>)</term> |
| 5607 |
<listitem> |
| 5608 |
<para> |
| 5609 |
Duplicate key error, violation of unique constraint. (SQLSTATE |
| 5610 |
23505) |
| 5611 |
</para> |
| 5612 |
</listitem> |
| 5613 |
</varlistentry> |
| 5614 |
|
| 5615 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-subselect-not-one"> |
| 5616 |
<term>-404 (<symbol>ECPG_SUBSELECT_NOT_ONE</symbol>)</term> |
| 5617 |
<listitem> |
| 5618 |
<para> |
| 5619 |
A result for the subquery is not single row. (SQLSTATE 21000) |
| 5620 |
</para> |
| 5621 |
</listitem> |
| 5622 |
</varlistentry> |
| 5623 |
|
| 5624 |
<!-- currently not used by the code --> |
| 5625 |
<!-- |
| 5626 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-warning-unrecognized"> |
| 5627 |
<term>-600 (<symbol>ECPG_WARNING_UNRECOGNIZED</symbol>)</term> |
| 5628 |
<listitem> |
| 5629 |
<para> |
| 5630 |
An unrecognized warning was received from the server. |
| 5631 |
</para> |
| 5632 |
</listitem> |
| 5633 |
</varlistentry> |
| 5634 |
|
| 5635 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-warning-query-ignored"> |
| 5636 |
<term>-601 (<symbol>ECPG_WARNING_QUERY_IGNORED</symbol>)</term> |
| 5637 |
<listitem> |
| 5638 |
<para> |
| 5639 |
Current transaction is aborted. Queries are ignored until the |
| 5640 |
end of the transaction block. |
| 5641 |
</para> |
| 5642 |
</listitem> |
| 5643 |
</varlistentry> |
| 5644 |
--> |
| 5645 |
|
| 5646 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-warning-unknown-portal"> |
| 5647 |
<term>-602 (<symbol>ECPG_WARNING_UNKNOWN_PORTAL</symbol>)</term> |
| 5648 |
<listitem> |
| 5649 |
<para> |
| 5650 |
An invalid cursor name was specified. (SQLSTATE 34000) |
| 5651 |
</para> |
| 5652 |
</listitem> |
| 5653 |
</varlistentry> |
| 5654 |
|
| 5655 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-warning-in-transaction"> |
| 5656 |
<term>-603 (<symbol>ECPG_WARNING_IN_TRANSACTION</symbol>)</term> |
| 5657 |
<listitem> |
| 5658 |
<para> |
| 5659 |
Transaction is in progress. (SQLSTATE 25001) |
| 5660 |
</para> |
| 5661 |
</listitem> |
| 5662 |
</varlistentry> |
| 5663 |
|
| 5664 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-warning-no-transaction"> |
| 5665 |
<term>-604 (<symbol>ECPG_WARNING_NO_TRANSACTION</symbol>)</term> |
| 5666 |
<listitem> |
| 5667 |
<para> |
| 5668 |
There is no active (in-progress) transaction. (SQLSTATE 25P01) |
| 5669 |
</para> |
| 5670 |
</listitem> |
| 5671 |
</varlistentry> |
| 5672 |
|
| 5673 |
<varlistentry id="ecpg-sqlstate-sqlcode-ecpg-warning-portal-exists"> |
| 5674 |
<term>-605 (<symbol>ECPG_WARNING_PORTAL_EXISTS</symbol>)</term> |
| 5675 |
<listitem> |
| 5676 |
<para> |
| 5677 |
An existing cursor name was specified. (SQLSTATE 42P03) |
| 5678 |
</para> |
| 5679 |
</listitem> |
| 5680 |
</varlistentry> |
| 5681 |
|
| 5682 |
</variablelist> |
| 5683 |
</para> |
| 5684 |
</sect2> |
| 5685 |
</sect1> |
| 5686 |
|
| 5687 |
<sect1 id="ecpg-preproc"> |
| 5688 |
<title>Preprocessor Directives</title> |
| 5689 |
|
| 5690 |
<para> |
| 5691 |
Several preprocessor directives are available that modify how |
| 5692 |
the <command>ecpg</command> preprocessor parses and processes a |
| 5693 |
file. |
| 5694 |
</para> |
| 5695 |
|
| 5696 |
<sect2 id="ecpg-include"> |
| 5697 |
<title>Including Files</title> |
| 5698 |
|
| 5699 |
<para> |
| 5700 |
To include an external file into your embedded SQL program, use: |
| 5701 |
<programlisting> |
| 5702 |
EXEC SQL INCLUDE <replaceable>filename</replaceable>; |
| 5703 |
EXEC SQL INCLUDE <<replaceable>filename</replaceable>>; |
| 5704 |
EXEC SQL INCLUDE "<replaceable>filename</replaceable>"; |
| 5705 |
</programlisting> |
| 5706 |
The embedded SQL preprocessor will look for a file named |
| 5707 |
<literal><replaceable>filename</replaceable>.h</literal>, |
| 5708 |
preprocess it, and include it in the resulting C output. Thus, |
| 5709 |
embedded SQL statements in the included file are handled correctly. |
| 5710 |
</para> |
| 5711 |
|
| 5712 |
<para> |
| 5713 |
The <command>ecpg</command> preprocessor will search a file at |
| 5714 |
several directories in following order: |
| 5715 |
|
| 5716 |
<itemizedlist> |
| 5717 |
<listitem><simpara>current directory</simpara></listitem> |
| 5718 |
<listitem><simpara><filename>/usr/local/include</filename></simpara></listitem> |
| 5719 |
<listitem><simpara>PostgreSQL include directory, defined at build time (e.g., <filename>/usr/local/pgsql/include</filename>)</simpara></listitem> |
| 5720 |
<listitem><simpara><filename>/usr/include</filename></simpara></listitem> |
| 5721 |
</itemizedlist> |
| 5722 |
|
| 5723 |
But when <literal>EXEC SQL INCLUDE |
| 5724 |
"<replaceable>filename</replaceable>"</literal> is used, only the |
| 5725 |
current directory is searched. |
| 5726 |
</para> |
| 5727 |
|
| 5728 |
<para> |
| 5729 |
In each directory, the preprocessor will first look for the file |
| 5730 |
name as given, and if not found will append <literal>.h</literal> |
| 5731 |
to the file name and try again (unless the specified file name |
| 5732 |
already has that suffix). |
| 5733 |
</para> |
| 5734 |
|
| 5735 |
<para> |
| 5736 |
Note that <command>EXEC SQL INCLUDE</command> is <emphasis>not</emphasis> the same as: |
| 5737 |
<programlisting> |
| 5738 |
#include <<replaceable>filename</replaceable>.h> |
| 5739 |
</programlisting> |
| 5740 |
because this file would not be subject to SQL command preprocessing. |
| 5741 |
Naturally, you can continue to use the C |
| 5742 |
<literal>#include</literal> directive to include other header |
| 5743 |
files. |
| 5744 |
</para> |
| 5745 |
|
| 5746 |
<note> |
| 5747 |
<para> |
| 5748 |
The include file name is case-sensitive, even though the rest of |
| 5749 |
the <literal>EXEC SQL INCLUDE</literal> command follows the normal |
| 5750 |
SQL case-sensitivity rules. |
| 5751 |
</para> |
| 5752 |
</note> |
| 5753 |
</sect2> |
| 5754 |
|
| 5755 |
<sect2 id="ecpg-define"> |
| 5756 |
<title>The define and undef Directives</title> |
| 5757 |
<para> |
| 5758 |
Similar to the directive <literal>#define</literal> that is known from C, |
| 5759 |
embedded SQL has a similar concept: |
| 5760 |
<programlisting> |
| 5761 |
EXEC SQL DEFINE <replaceable>name</replaceable>; |
| 5762 |
EXEC SQL DEFINE <replaceable>name</replaceable> <replaceable>value</replaceable>; |
| 5763 |
</programlisting> |
| 5764 |
So you can define a name: |
| 5765 |
<programlisting> |
| 5766 |
EXEC SQL DEFINE HAVE_FEATURE; |
| 5767 |
</programlisting> |
| 5768 |
And you can also define constants: |
| 5769 |
<programlisting> |
| 5770 |
EXEC SQL DEFINE MYNUMBER 12; |
| 5771 |
EXEC SQL DEFINE MYSTRING 'abc'; |
| 5772 |
</programlisting> |
| 5773 |
Use <literal>undef</literal> to remove a previous definition: |
| 5774 |
<programlisting> |
| 5775 |
EXEC SQL UNDEF MYNUMBER; |
| 5776 |
</programlisting> |
| 5777 |
</para> |
| 5778 |
|
| 5779 |
<para> |
| 5780 |
Of course you can continue to use the C versions <literal>#define</literal> |
| 5781 |
and <literal>#undef</literal> in your embedded SQL program. The difference |
| 5782 |
is where your defined values get evaluated. If you use <literal>EXEC SQL |
| 5783 |
DEFINE</literal> then the <command>ecpg</command> preprocessor evaluates the defines and substitutes |
| 5784 |
the values. For example if you write: |
| 5785 |
<programlisting> |
| 5786 |
EXEC SQL DEFINE MYNUMBER 12; |
| 5787 |
... |
| 5788 |
EXEC SQL UPDATE Tbl SET col = MYNUMBER; |
| 5789 |
</programlisting> |
| 5790 |
then <command>ecpg</command> will already do the substitution and your C compiler will never |
| 5791 |
see any name or identifier <literal>MYNUMBER</literal>. Note that you cannot use |
| 5792 |
<literal>#define</literal> for a constant that you are going to use in an |
| 5793 |
embedded SQL query because in this case the embedded SQL precompiler is not |
| 5794 |
able to see this declaration. |
| 5795 |
</para> |
| 5796 |
|
| 5797 |
<para> |
| 5798 |
If multiple input files are named on the <command>ecpg</command> |
| 5799 |
preprocessor's command line, the effects of <literal>EXEC SQL |
| 5800 |
DEFINE</literal> and <literal>EXEC SQL UNDEF</literal> do not carry |
| 5801 |
across files: each file starts with only the symbols defined |
| 5802 |
by <option>-D</option> switches on the command line. |
| 5803 |
</para> |
| 5804 |
</sect2> |
| 5805 |
|
| 5806 |
<sect2 id="ecpg-ifdef"> |
| 5807 |
<title>ifdef, ifndef, elif, else, and endif Directives</title> |
| 5808 |
<para> |
| 5809 |
You can use the following directives to compile code sections conditionally: |
| 5810 |
|
| 5811 |
<variablelist> |
| 5812 |
<varlistentry id="ecpg-ifdef-ifdef"> |
| 5813 |
<term><literal>EXEC SQL ifdef <replaceable>name</replaceable>;</literal></term> |
| 5814 |
<listitem> |
| 5815 |
<para> |
| 5816 |
Checks a <replaceable>name</replaceable> and processes subsequent lines if |
| 5817 |
<replaceable>name</replaceable> has been defined via <literal>EXEC SQL define |
| 5818 |
<replaceable>name</replaceable></literal>. |
| 5819 |
</para> |
| 5820 |
</listitem> |
| 5821 |
</varlistentry> |
| 5822 |
|
| 5823 |
<varlistentry id="ecpg-ifdef-ifndef"> |
| 5824 |
<term><literal>EXEC SQL ifndef <replaceable>name</replaceable>;</literal></term> |
| 5825 |
<listitem> |
| 5826 |
<para> |
| 5827 |
Checks a <replaceable>name</replaceable> and processes subsequent lines if |
| 5828 |
<replaceable>name</replaceable> has <emphasis>not</emphasis> been defined via |
| 5829 |
<literal>EXEC SQL define <replaceable>name</replaceable></literal>. |
| 5830 |
</para> |
| 5831 |
</listitem> |
| 5832 |
</varlistentry> |
| 5833 |
|
| 5834 |
<varlistentry id="ecpg-ifdef-elif"> |
| 5835 |
<term><literal>EXEC SQL elif <replaceable>name</replaceable>;</literal></term> |
| 5836 |
<listitem> |
| 5837 |
<para> |
| 5838 |
Begins an optional alternative section after an |
| 5839 |
<literal>EXEC SQL ifdef <replaceable>name</replaceable></literal> or |
| 5840 |
<literal>EXEC SQL ifndef <replaceable>name</replaceable></literal> |
| 5841 |
directive. Any number of <literal>elif</literal> sections can appear. |
| 5842 |
Lines following an <literal>elif</literal> will be processed |
| 5843 |
if <replaceable>name</replaceable> has been |
| 5844 |
defined <emphasis>and</emphasis> no previous section of the same |
| 5845 |
<literal>ifdef</literal>/<literal>ifndef</literal>...<literal>endif</literal> |
| 5846 |
construct has been processed. |
| 5847 |
</para> |
| 5848 |
</listitem> |
| 5849 |
</varlistentry> |
| 5850 |
|
| 5851 |
<varlistentry id="ecpg-ifdef-else"> |
| 5852 |
<term><literal>EXEC SQL else;</literal></term> |
| 5853 |
<listitem> |
| 5854 |
<para> |
| 5855 |
Begins an optional, final alternative section after an |
| 5856 |
<literal>EXEC SQL ifdef <replaceable>name</replaceable></literal> or |
| 5857 |
<literal>EXEC SQL ifndef <replaceable>name</replaceable></literal> |
| 5858 |
directive. Subsequent lines will be processed if no previous section |
| 5859 |
of the same |
| 5860 |
<literal>ifdef</literal>/<literal>ifndef</literal>...<literal>endif</literal> |
| 5861 |
construct has been processed. |
| 5862 |
</para> |
| 5863 |
</listitem> |
| 5864 |
</varlistentry> |
| 5865 |
|
| 5866 |
<varlistentry id="ecpg-ifdef-endif"> |
| 5867 |
<term><literal>EXEC SQL endif;</literal></term> |
| 5868 |
<listitem> |
| 5869 |
<para> |
| 5870 |
Ends an |
| 5871 |
<literal>ifdef</literal>/<literal>ifndef</literal>...<literal>endif</literal> |
| 5872 |
construct. Subsequent lines are processed normally. |
| 5873 |
</para> |
| 5874 |
</listitem> |
| 5875 |
</varlistentry> |
| 5876 |
</variablelist> |
| 5877 |
</para> |
| 5878 |
|
| 5879 |
<para> |
| 5880 |
<literal>ifdef</literal>/<literal>ifndef</literal>...<literal>endif</literal> |
| 5881 |
constructs can be nested, up to 127 levels deep. |
| 5882 |
</para> |
| 5883 |
|
| 5884 |
<para> |
| 5885 |
This example will compile exactly one of the three <literal>SET |
| 5886 |
TIMEZONE</literal> commands: |
| 5887 |
<programlisting> |
| 5888 |
EXEC SQL ifdef TZVAR; |
| 5889 |
EXEC SQL SET TIMEZONE TO TZVAR; |
| 5890 |
EXEC SQL elif TZNAME; |
| 5891 |
EXEC SQL SET TIMEZONE TO TZNAME; |
| 5892 |
EXEC SQL else; |
| 5893 |
EXEC SQL SET TIMEZONE TO 'GMT'; |
| 5894 |
EXEC SQL endif; |
| 5895 |
</programlisting> |
| 5896 |
</para> |
| 5897 |
|
| 5898 |
</sect2> |
| 5899 |
</sect1> |
| 5900 |
|
| 5901 |
<sect1 id="ecpg-process"> |
| 5902 |
<title>Processing Embedded SQL Programs</title> |
| 5903 |
|
| 5904 |
<para> |
| 5905 |
Now that you have an idea how to form embedded SQL C programs, you |
| 5906 |
probably want to know how to compile them. Before compiling you |
| 5907 |
run the file through the embedded <acronym>SQL</acronym> |
| 5908 |
<acronym>C</acronym> preprocessor, which converts the |
| 5909 |
<acronym>SQL</acronym> statements you used to special function |
| 5910 |
calls. After compiling, you must link with a special library that |
| 5911 |
contains the needed functions. These functions fetch information |
| 5912 |
from the arguments, perform the <acronym>SQL</acronym> command using |
| 5913 |
the <application>libpq</application> interface, and put the result |
| 5914 |
in the arguments specified for output. |
| 5915 |
</para> |
| 5916 |
|
| 5917 |
<para> |
| 5918 |
The preprocessor program is called <filename>ecpg</filename> and is |
| 5919 |
included in a normal <productname>PostgreSQL</productname> installation. |
| 5920 |
Embedded SQL programs are typically named with an extension |
| 5921 |
<filename>.pgc</filename>. If you have a program file called |
| 5922 |
<filename>prog1.pgc</filename>, you can preprocess it by simply |
| 5923 |
calling: |
| 5924 |
<programlisting> |
| 5925 |
ecpg prog1.pgc |
| 5926 |
</programlisting> |
| 5927 |
This will create a file called <filename>prog1.c</filename>. If |
| 5928 |
your input files do not follow the suggested naming pattern, you |
| 5929 |
can specify the output file explicitly using the |
| 5930 |
<option>-o</option> option. |
| 5931 |
</para> |
| 5932 |
|
| 5933 |
<para> |
| 5934 |
The preprocessed file can be compiled normally, for example: |
| 5935 |
<programlisting> |
| 5936 |
cc -c prog1.c |
| 5937 |
</programlisting> |
| 5938 |
The generated C source files include header files from the |
| 5939 |
<productname>PostgreSQL</productname> installation, so if you installed |
| 5940 |
<productname>PostgreSQL</productname> in a location that is not searched by |
| 5941 |
default, you have to add an option such as |
| 5942 |
<literal>-I/usr/local/pgsql/include</literal> to the compilation |
| 5943 |
command line. |
| 5944 |
</para> |
| 5945 |
|
| 5946 |
<para> |
| 5947 |
To link an embedded SQL program, you need to include the |
| 5948 |
<filename>libecpg</filename> library, like so: |
| 5949 |
<programlisting> |
| 5950 |
cc -o myprog prog1.o prog2.o ... -lecpg |
| 5951 |
</programlisting> |
| 5952 |
Again, you might have to add an option like |
| 5953 |
<literal>-L/usr/local/pgsql/lib</literal> to that command line. |
| 5954 |
</para> |
| 5955 |
|
| 5956 |
<para> |
| 5957 |
You can |
| 5958 |
use <command>pg_config</command><indexterm><primary>pg_config</primary><secondary sortas="ecpg">with |
| 5959 |
ecpg</secondary></indexterm> |
| 5960 |
or <command>pkg-config</command><indexterm><primary>pkg-config</primary><secondary sortas="ecpg">with |
| 5961 |
ecpg</secondary></indexterm> with package name <literal>libecpg</literal> to |
| 5962 |
get the paths for your installation. |
| 5963 |
</para> |
| 5964 |
|
| 5965 |
<para> |
| 5966 |
If you manage the build process of a larger project using |
| 5967 |
<application>make</application>, it might be convenient to include |
| 5968 |
the following implicit rule to your makefiles: |
| 5969 |
<programlisting> |
| 5970 |
ECPG = ecpg |
| 5971 |
|
| 5972 |
%.c: %.pgc |
| 5973 |
$(ECPG) $< |
| 5974 |
</programlisting> |
| 5975 |
</para> |
| 5976 |
|
| 5977 |
<para> |
| 5978 |
The complete syntax of the <command>ecpg</command> command is |
| 5979 |
detailed in <xref linkend="app-ecpg"/>. |
| 5980 |
</para> |
| 5981 |
|
| 5982 |
<para> |
| 5983 |
The <application>ecpg</application> library is thread-safe by |
| 5984 |
default. However, you might need to use some threading |
| 5985 |
command-line options to compile your client code. |
| 5986 |
</para> |
| 5987 |
</sect1> |
| 5988 |
|
| 5989 |
<sect1 id="ecpg-library"> |
| 5990 |
<title>Library Functions</title> |
| 5991 |
|
| 5992 |
<para> |
| 5993 |
The <filename>libecpg</filename> library primarily contains |
| 5994 |
<quote>hidden</quote> functions that are used to implement the |
| 5995 |
functionality expressed by the embedded SQL commands. But there |
| 5996 |
are some functions that can usefully be called directly. Note that |
| 5997 |
this makes your code unportable. |
| 5998 |
</para> |
| 5999 |
|
| 6000 |
<itemizedlist> |
| 6001 |
<listitem> |
| 6002 |
<para> |
| 6003 |
<function>ECPGdebug(int <replaceable>on</replaceable>, FILE |
| 6004 |
*<replaceable>stream</replaceable>)</function> turns on debug |
| 6005 |
logging if called with the first argument non-zero. Debug logging |
| 6006 |
is done on <replaceable>stream</replaceable>. The log contains |
| 6007 |
all <acronym>SQL</acronym> statements with all the input |
| 6008 |
variables inserted, and the results from the |
| 6009 |
<productname>PostgreSQL</productname> server. This can be very |
| 6010 |
useful when searching for errors in your <acronym>SQL</acronym> |
| 6011 |
statements. |
| 6012 |
</para> |
| 6013 |
<note> |
| 6014 |
<para> |
| 6015 |
On Windows, if the <application>ecpg</application> libraries and an application are |
| 6016 |
compiled with different flags, this function call will crash the |
| 6017 |
application because the internal representation of the |
| 6018 |
<literal>FILE</literal> pointers differ. Specifically, |
| 6019 |
multithreaded/single-threaded, release/debug, and static/dynamic |
| 6020 |
flags should be the same for the library and all applications using |
| 6021 |
that library. |
| 6022 |
</para> |
| 6023 |
</note> |
| 6024 |
</listitem> |
| 6025 |
|
| 6026 |
<listitem> |
| 6027 |
<para> |
| 6028 |
<function>ECPGget_PGconn(const char *<replaceable>connection_name</replaceable>) |
| 6029 |
</function> returns the library database connection handle identified by the given name. |
| 6030 |
If <replaceable>connection_name</replaceable> is set to <literal>NULL</literal>, the current |
| 6031 |
connection handle is returned. If no connection handle can be identified, the function returns |
| 6032 |
<literal>NULL</literal>. The returned connection handle can be used to call any other functions |
| 6033 |
from <application>libpq</application>, if necessary. |
| 6034 |
</para> |
| 6035 |
<note> |
| 6036 |
<para> |
| 6037 |
It is a bad idea to manipulate database connection handles made from <application>ecpg</application> directly |
| 6038 |
with <application>libpq</application> routines. |
| 6039 |
</para> |
| 6040 |
</note> |
| 6041 |
</listitem> |
| 6042 |
|
| 6043 |
<listitem> |
| 6044 |
<para> |
| 6045 |
<function>ECPGtransactionStatus(const char *<replaceable>connection_name</replaceable>)</function> |
| 6046 |
returns the current transaction status of the given connection identified by <replaceable>connection_name</replaceable>. |
| 6047 |
See <xref linkend="libpq-status"/> and libpq's <xref linkend="libpq-PQtransactionStatus"/> for details about the returned status codes. |
| 6048 |
</para> |
| 6049 |
</listitem> |
| 6050 |
|
| 6051 |
<listitem> |
| 6052 |
<para> |
| 6053 |
<function>ECPGstatus(int <replaceable>lineno</replaceable>, |
| 6054 |
const char* <replaceable>connection_name</replaceable>)</function> |
| 6055 |
returns true if you are connected to a database and false if not. |
| 6056 |
<replaceable>connection_name</replaceable> can be <literal>NULL</literal> |
| 6057 |
if a single connection is being used. |
| 6058 |
</para> |
| 6059 |
</listitem> |
| 6060 |
</itemizedlist> |
| 6061 |
</sect1> |
| 6062 |
|
| 6063 |
<sect1 id="ecpg-lo"> |
| 6064 |
<title>Large Objects</title> |
| 6065 |
|
| 6066 |
<para> |
| 6067 |
Large objects are not directly supported by ECPG, but ECPG |
| 6068 |
application can manipulate large objects through the libpq large |
| 6069 |
object functions, obtaining the necessary <type>PGconn</type> |
| 6070 |
object by calling the <function>ECPGget_PGconn()</function> |
| 6071 |
function. (However, use of |
| 6072 |
the <function>ECPGget_PGconn()</function> function and touching |
| 6073 |
<type>PGconn</type> objects directly should be done very carefully |
| 6074 |
and ideally not mixed with other ECPG database access calls.) |
| 6075 |
</para> |
| 6076 |
|
| 6077 |
<para> |
| 6078 |
For more details about the <function>ECPGget_PGconn()</function>, see |
| 6079 |
<xref linkend="ecpg-library"/>. For information about the large |
| 6080 |
object function interface, see <xref linkend="largeobjects"/>. |
| 6081 |
</para> |
| 6082 |
|
| 6083 |
<para> |
| 6084 |
Large object functions have to be called in a transaction block, so |
| 6085 |
when autocommit is off, <command>BEGIN</command> commands have to |
| 6086 |
be issued explicitly. |
| 6087 |
</para> |
| 6088 |
|
| 6089 |
<para> |
| 6090 |
<xref linkend="ecpg-lo-example"/> shows an example program that |
| 6091 |
illustrates how to create, write, and read a large object in an |
| 6092 |
ECPG application. |
| 6093 |
</para> |
| 6094 |
|
| 6095 |
<example id="ecpg-lo-example"> |
| 6096 |
<title>ECPG Program Accessing Large Objects</title> |
| 6097 |
<programlisting><![CDATA[ |
| 6098 |
#include <stdio.h> |
| 6099 |
#include <stdlib.h> |
| 6100 |
#include <libpq-fe.h> |
| 6101 |
#include <libpq/libpq-fs.h> |
| 6102 |
|
| 6103 |
EXEC SQL WHENEVER SQLERROR STOP; |
| 6104 |
|
| 6105 |
int |
| 6106 |
main(void) |
| 6107 |
{ |
| 6108 |
PGconn *conn; |
| 6109 |
Oid loid; |
| 6110 |
int fd; |
| 6111 |
char buf[256]; |
| 6112 |
int buflen = 256; |
| 6113 |
char buf2[256]; |
| 6114 |
int rc; |
| 6115 |
|
| 6116 |
memset(buf, 1, buflen); |
| 6117 |
|
| 6118 |
EXEC SQL CONNECT TO testdb AS con1; |
| 6119 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 6120 |
|
| 6121 |
conn = ECPGget_PGconn("con1"); |
| 6122 |
printf("conn = %p\n", conn); |
| 6123 |
|
| 6124 |
/* create */ |
| 6125 |
loid = lo_create(conn, 0); |
| 6126 |
if (loid < 0) |
| 6127 |
printf("lo_create() failed: %s", PQerrorMessage(conn)); |
| 6128 |
|
| 6129 |
printf("loid = %d\n", loid); |
| 6130 |
|
| 6131 |
/* write test */ |
| 6132 |
fd = lo_open(conn, loid, INV_READ|INV_WRITE); |
| 6133 |
if (fd < 0) |
| 6134 |
printf("lo_open() failed: %s", PQerrorMessage(conn)); |
| 6135 |
|
| 6136 |
printf("fd = %d\n", fd); |
| 6137 |
|
| 6138 |
rc = lo_write(conn, fd, buf, buflen); |
| 6139 |
if (rc < 0) |
| 6140 |
printf("lo_write() failed\n"); |
| 6141 |
|
| 6142 |
rc = lo_close(conn, fd); |
| 6143 |
if (rc < 0) |
| 6144 |
printf("lo_close() failed: %s", PQerrorMessage(conn)); |
| 6145 |
|
| 6146 |
/* read test */ |
| 6147 |
fd = lo_open(conn, loid, INV_READ); |
| 6148 |
if (fd < 0) |
| 6149 |
printf("lo_open() failed: %s", PQerrorMessage(conn)); |
| 6150 |
|
| 6151 |
printf("fd = %d\n", fd); |
| 6152 |
|
| 6153 |
rc = lo_read(conn, fd, buf2, buflen); |
| 6154 |
if (rc < 0) |
| 6155 |
printf("lo_read() failed\n"); |
| 6156 |
|
| 6157 |
rc = lo_close(conn, fd); |
| 6158 |
if (rc < 0) |
| 6159 |
printf("lo_close() failed: %s", PQerrorMessage(conn)); |
| 6160 |
|
| 6161 |
/* check */ |
| 6162 |
rc = memcmp(buf, buf2, buflen); |
| 6163 |
printf("memcmp() = %d\n", rc); |
| 6164 |
|
| 6165 |
/* cleanup */ |
| 6166 |
rc = lo_unlink(conn, loid); |
| 6167 |
if (rc < 0) |
| 6168 |
printf("lo_unlink() failed: %s", PQerrorMessage(conn)); |
| 6169 |
|
| 6170 |
EXEC SQL COMMIT; |
| 6171 |
EXEC SQL DISCONNECT ALL; |
| 6172 |
return 0; |
| 6173 |
} |
| 6174 |
]]></programlisting> |
| 6175 |
</example> |
| 6176 |
</sect1> |
| 6177 |
|
| 6178 |
<sect1 id="ecpg-cpp"> |
| 6179 |
<title><acronym>C++</acronym> Applications</title> |
| 6180 |
|
| 6181 |
<para> |
| 6182 |
ECPG has some limited support for C++ applications. This section |
| 6183 |
describes some caveats. |
| 6184 |
</para> |
| 6185 |
|
| 6186 |
<para> |
| 6187 |
The <command>ecpg</command> preprocessor takes an input file |
| 6188 |
written in C (or something like C) and embedded SQL commands, |
| 6189 |
converts the embedded SQL commands into C language chunks, and |
| 6190 |
finally generates a <filename>.c</filename> file. The header file |
| 6191 |
declarations of the library functions used by the C language chunks |
| 6192 |
that <command>ecpg</command> generates are wrapped |
| 6193 |
in <literal>extern "C" { ... }</literal> blocks when used under |
| 6194 |
C++, so they should work seamlessly in C++. |
| 6195 |
</para> |
| 6196 |
|
| 6197 |
<para> |
| 6198 |
In general, however, the <command>ecpg</command> preprocessor only |
| 6199 |
understands C; it does not handle the special syntax and reserved |
| 6200 |
words of the C++ language. So, some embedded SQL code written in |
| 6201 |
C++ application code that uses complicated features specific to C++ |
| 6202 |
might fail to be preprocessed correctly or might not work as |
| 6203 |
expected. |
| 6204 |
</para> |
| 6205 |
|
| 6206 |
<para> |
| 6207 |
A safe way to use the embedded SQL code in a C++ application is |
| 6208 |
hiding the ECPG calls in a C module, which the C++ application code |
| 6209 |
calls into to access the database, and linking that together with |
| 6210 |
the rest of the C++ code. See <xref linkend="ecpg-cpp-and-c"/> |
| 6211 |
about that. |
| 6212 |
</para> |
| 6213 |
|
| 6214 |
<sect2 id="ecpg-cpp-scope"> |
| 6215 |
<title>Scope for Host Variables</title> |
| 6216 |
|
| 6217 |
<para> |
| 6218 |
The <command>ecpg</command> preprocessor understands the scope of |
| 6219 |
variables in C. In the C language, this is rather simple because |
| 6220 |
the scopes of variables is based on their code blocks. In C++, |
| 6221 |
however, the class member variables are referenced in a different |
| 6222 |
code block from the declared position, so |
| 6223 |
the <command>ecpg</command> preprocessor will not understand the |
| 6224 |
scope of the class member variables. |
| 6225 |
</para> |
| 6226 |
|
| 6227 |
<para> |
| 6228 |
For example, in the following case, the <command>ecpg</command> |
| 6229 |
preprocessor cannot find any declaration for the |
| 6230 |
variable <literal>dbname</literal> in the <literal>test</literal> |
| 6231 |
method, so an error will occur. |
| 6232 |
|
| 6233 |
<programlisting> |
| 6234 |
class TestCpp |
| 6235 |
{ |
| 6236 |
EXEC SQL BEGIN DECLARE SECTION; |
| 6237 |
char dbname[1024]; |
| 6238 |
EXEC SQL END DECLARE SECTION; |
| 6239 |
|
| 6240 |
public: |
| 6241 |
TestCpp(); |
| 6242 |
void test(); |
| 6243 |
~TestCpp(); |
| 6244 |
}; |
| 6245 |
|
| 6246 |
TestCpp::TestCpp() |
| 6247 |
{ |
| 6248 |
EXEC SQL CONNECT TO testdb1; |
| 6249 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 6250 |
} |
| 6251 |
|
| 6252 |
void Test::test() |
| 6253 |
{ |
| 6254 |
EXEC SQL SELECT current_database() INTO :dbname; |
| 6255 |
printf("current_database = %s\n", dbname); |
| 6256 |
} |
| 6257 |
|
| 6258 |
TestCpp::~TestCpp() |
| 6259 |
{ |
| 6260 |
EXEC SQL DISCONNECT ALL; |
| 6261 |
} |
| 6262 |
</programlisting> |
| 6263 |
|
| 6264 |
This code will result in an error like this: |
| 6265 |
<screen> |
| 6266 |
<userinput>ecpg test_cpp.pgc</userinput> |
| 6267 |
test_cpp.pgc:28: ERROR: variable "dbname" is not declared |
| 6268 |
</screen> |
| 6269 |
</para> |
| 6270 |
|
| 6271 |
<para> |
| 6272 |
To avoid this scope issue, the <literal>test</literal> method |
| 6273 |
could be modified to use a local variable as intermediate storage. |
| 6274 |
But this approach is only a poor workaround, because it uglifies |
| 6275 |
the code and reduces performance. |
| 6276 |
|
| 6277 |
<programlisting> |
| 6278 |
void TestCpp::test() |
| 6279 |
{ |
| 6280 |
EXEC SQL BEGIN DECLARE SECTION; |
| 6281 |
char tmp[1024]; |
| 6282 |
EXEC SQL END DECLARE SECTION; |
| 6283 |
|
| 6284 |
EXEC SQL SELECT current_database() INTO :tmp; |
| 6285 |
strlcpy(dbname, tmp, sizeof(tmp)); |
| 6286 |
|
| 6287 |
printf("current_database = %s\n", dbname); |
| 6288 |
} |
| 6289 |
</programlisting> |
| 6290 |
</para> |
| 6291 |
</sect2> |
| 6292 |
|
| 6293 |
<sect2 id="ecpg-cpp-and-c"> |
| 6294 |
<title>C++ Application Development with External C Module</title> |
| 6295 |
|
| 6296 |
<para> |
| 6297 |
If you understand these technical limitations of |
| 6298 |
the <command>ecpg</command> preprocessor in C++, you might come to |
| 6299 |
the conclusion that linking C objects and C++ objects at the link |
| 6300 |
stage to enable C++ applications to use ECPG features could be |
| 6301 |
better than writing some embedded SQL commands in C++ code |
| 6302 |
directly. This section describes a way to separate some embedded |
| 6303 |
SQL commands from C++ application code with a simple example. In |
| 6304 |
this example, the application is implemented in C++, while C and |
| 6305 |
ECPG is used to connect to the PostgreSQL server. |
| 6306 |
</para> |
| 6307 |
|
| 6308 |
<para> |
| 6309 |
Three kinds of files have to be created: a C file |
| 6310 |
(<filename>*.pgc</filename>), a header file, and a C++ file: |
| 6311 |
|
| 6312 |
<variablelist> |
| 6313 |
<varlistentry id="ecpg-cpp-and-c-test-mod-pgc"> |
| 6314 |
<term><filename>test_mod.pgc</filename></term> |
| 6315 |
<listitem> |
| 6316 |
<para> |
| 6317 |
A sub-routine module to execute SQL commands embedded in C. |
| 6318 |
It is going to be converted |
| 6319 |
into <filename>test_mod.c</filename> by the preprocessor. |
| 6320 |
|
| 6321 |
<programlisting> |
| 6322 |
#include "test_mod.h" |
| 6323 |
#include <stdio.h> |
| 6324 |
|
| 6325 |
void |
| 6326 |
db_connect() |
| 6327 |
{ |
| 6328 |
EXEC SQL CONNECT TO testdb1; |
| 6329 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 6330 |
} |
| 6331 |
|
| 6332 |
void |
| 6333 |
db_test() |
| 6334 |
{ |
| 6335 |
EXEC SQL BEGIN DECLARE SECTION; |
| 6336 |
char dbname[1024]; |
| 6337 |
EXEC SQL END DECLARE SECTION; |
| 6338 |
|
| 6339 |
EXEC SQL SELECT current_database() INTO :dbname; |
| 6340 |
printf("current_database = %s\n", dbname); |
| 6341 |
} |
| 6342 |
|
| 6343 |
void |
| 6344 |
db_disconnect() |
| 6345 |
{ |
| 6346 |
EXEC SQL DISCONNECT ALL; |
| 6347 |
} |
| 6348 |
</programlisting> |
| 6349 |
</para> |
| 6350 |
</listitem> |
| 6351 |
</varlistentry> |
| 6352 |
|
| 6353 |
<varlistentry id="ecpg-cpp-and-c-test-mod-h"> |
| 6354 |
<term><filename>test_mod.h</filename></term> |
| 6355 |
<listitem> |
| 6356 |
<para> |
| 6357 |
A header file with declarations of the functions in the C |
| 6358 |
module (<filename>test_mod.pgc</filename>). It is included by |
| 6359 |
<filename>test_cpp.cpp</filename>. This file has to have an |
| 6360 |
<literal>extern "C"</literal> block around the declarations, |
| 6361 |
because it will be linked from the C++ module. |
| 6362 |
|
| 6363 |
<programlisting> |
| 6364 |
#ifdef __cplusplus |
| 6365 |
extern "C" { |
| 6366 |
#endif |
| 6367 |
|
| 6368 |
void db_connect(); |
| 6369 |
void db_test(); |
| 6370 |
void db_disconnect(); |
| 6371 |
|
| 6372 |
#ifdef __cplusplus |
| 6373 |
} |
| 6374 |
#endif |
| 6375 |
</programlisting> |
| 6376 |
</para> |
| 6377 |
</listitem> |
| 6378 |
</varlistentry> |
| 6379 |
|
| 6380 |
<varlistentry id="ecpg-cpp-and-c-test-cpp-cpp"> |
| 6381 |
<term><filename>test_cpp.cpp</filename></term> |
| 6382 |
<listitem> |
| 6383 |
<para> |
| 6384 |
The main code for the application, including |
| 6385 |
the <function>main</function> routine, and in this example a |
| 6386 |
C++ class. |
| 6387 |
|
| 6388 |
<programlisting> |
| 6389 |
#include "test_mod.h" |
| 6390 |
|
| 6391 |
class TestCpp |
| 6392 |
{ |
| 6393 |
public: |
| 6394 |
TestCpp(); |
| 6395 |
void test(); |
| 6396 |
~TestCpp(); |
| 6397 |
}; |
| 6398 |
|
| 6399 |
TestCpp::TestCpp() |
| 6400 |
{ |
| 6401 |
db_connect(); |
| 6402 |
} |
| 6403 |
|
| 6404 |
void |
| 6405 |
TestCpp::test() |
| 6406 |
{ |
| 6407 |
db_test(); |
| 6408 |
} |
| 6409 |
|
| 6410 |
TestCpp::~TestCpp() |
| 6411 |
{ |
| 6412 |
db_disconnect(); |
| 6413 |
} |
| 6414 |
|
| 6415 |
int |
| 6416 |
main(void) |
| 6417 |
{ |
| 6418 |
TestCpp *t = new TestCpp(); |
| 6419 |
|
| 6420 |
t->test(); |
| 6421 |
return 0; |
| 6422 |
} |
| 6423 |
</programlisting> |
| 6424 |
</para> |
| 6425 |
</listitem> |
| 6426 |
</varlistentry> |
| 6427 |
</variablelist> |
| 6428 |
</para> |
| 6429 |
|
| 6430 |
<para> |
| 6431 |
To build the application, proceed as follows. Convert |
| 6432 |
<filename>test_mod.pgc</filename> into <filename>test_mod.c</filename> by |
| 6433 |
running <command>ecpg</command>, and generate |
| 6434 |
<filename>test_mod.o</filename> by compiling |
| 6435 |
<filename>test_mod.c</filename> with the C compiler: |
| 6436 |
<programlisting> |
| 6437 |
ecpg -o test_mod.c test_mod.pgc |
| 6438 |
cc -c test_mod.c -o test_mod.o |
| 6439 |
</programlisting> |
| 6440 |
</para> |
| 6441 |
|
| 6442 |
<para> |
| 6443 |
Next, generate <filename>test_cpp.o</filename> by compiling |
| 6444 |
<filename>test_cpp.cpp</filename> with the C++ compiler: |
| 6445 |
<programlisting> |
| 6446 |
c++ -c test_cpp.cpp -o test_cpp.o |
| 6447 |
</programlisting> |
| 6448 |
</para> |
| 6449 |
|
| 6450 |
<para> |
| 6451 |
Finally, link these object files, <filename>test_cpp.o</filename> |
| 6452 |
and <filename>test_mod.o</filename>, into one executable, using the C++ |
| 6453 |
compiler driver: |
| 6454 |
<programlisting> |
| 6455 |
c++ test_cpp.o test_mod.o -lecpg -o test_cpp |
| 6456 |
</programlisting> |
| 6457 |
</para> |
| 6458 |
</sect2> |
| 6459 |
</sect1> |
| 6460 |
|
| 6461 |
<sect1 id="ecpg-sql-commands"> |
| 6462 |
<title>Embedded SQL Commands</title> |
| 6463 |
|
| 6464 |
<para> |
| 6465 |
This section describes all SQL commands that are specific to |
| 6466 |
embedded SQL. Also refer to the SQL commands listed |
| 6467 |
in <xref linkend="sql-commands"/>, which can also be used in |
| 6468 |
embedded SQL, unless stated otherwise. |
| 6469 |
</para> |
| 6470 |
|
| 6471 |
<refentry id="ecpg-sql-allocate-descriptor"> |
| 6472 |
<refnamediv> |
| 6473 |
<refname>ALLOCATE DESCRIPTOR</refname> |
| 6474 |
<refpurpose>allocate an SQL descriptor area</refpurpose> |
| 6475 |
</refnamediv> |
| 6476 |
|
| 6477 |
<refsynopsisdiv> |
| 6478 |
<synopsis> |
| 6479 |
ALLOCATE DESCRIPTOR <replaceable class="parameter">name</replaceable> |
| 6480 |
</synopsis> |
| 6481 |
</refsynopsisdiv> |
| 6482 |
|
| 6483 |
<refsect1> |
| 6484 |
<title>Description</title> |
| 6485 |
|
| 6486 |
<para> |
| 6487 |
<command>ALLOCATE DESCRIPTOR</command> allocates a new named SQL |
| 6488 |
descriptor area, which can be used to exchange data between the |
| 6489 |
PostgreSQL server and the host program. |
| 6490 |
</para> |
| 6491 |
|
| 6492 |
<para> |
| 6493 |
Descriptor areas should be freed after use using |
| 6494 |
the <command>DEALLOCATE DESCRIPTOR</command> command. |
| 6495 |
</para> |
| 6496 |
</refsect1> |
| 6497 |
|
| 6498 |
<refsect1> |
| 6499 |
<title>Parameters</title> |
| 6500 |
|
| 6501 |
<variablelist> |
| 6502 |
<varlistentry id="ecpg-sql-allocate-descriptor-name"> |
| 6503 |
<term><replaceable class="parameter">name</replaceable></term> |
| 6504 |
<listitem> |
| 6505 |
<para> |
| 6506 |
A name of SQL descriptor, case sensitive. This can be an SQL |
| 6507 |
identifier or a host variable. |
| 6508 |
</para> |
| 6509 |
</listitem> |
| 6510 |
</varlistentry> |
| 6511 |
</variablelist> |
| 6512 |
</refsect1> |
| 6513 |
|
| 6514 |
<refsect1> |
| 6515 |
<title>Examples</title> |
| 6516 |
|
| 6517 |
<programlisting> |
| 6518 |
EXEC SQL ALLOCATE DESCRIPTOR mydesc; |
| 6519 |
</programlisting> |
| 6520 |
</refsect1> |
| 6521 |
|
| 6522 |
<refsect1> |
| 6523 |
<title>Compatibility</title> |
| 6524 |
|
| 6525 |
<para> |
| 6526 |
<command>ALLOCATE DESCRIPTOR</command> is specified in the SQL |
| 6527 |
standard. |
| 6528 |
</para> |
| 6529 |
</refsect1> |
| 6530 |
|
| 6531 |
<refsect1> |
| 6532 |
<title>See Also</title> |
| 6533 |
|
| 6534 |
<simplelist type="inline"> |
| 6535 |
<member><xref linkend="ecpg-sql-deallocate-descriptor"/></member> |
| 6536 |
<member><xref linkend="ecpg-sql-get-descriptor"/></member> |
| 6537 |
<member><xref linkend="ecpg-sql-set-descriptor"/></member> |
| 6538 |
</simplelist> |
| 6539 |
</refsect1> |
| 6540 |
</refentry> |
| 6541 |
|
| 6542 |
<refentry id="ecpg-sql-connect"> |
| 6543 |
<refnamediv> |
| 6544 |
<refname>CONNECT</refname> |
| 6545 |
<refpurpose>establish a database connection</refpurpose> |
| 6546 |
</refnamediv> |
| 6547 |
|
| 6548 |
<refsynopsisdiv> |
| 6549 |
<synopsis> |
| 6550 |
CONNECT TO <replaceable>connection_target</replaceable> [ AS <replaceable>connection_name</replaceable> ] [ USER <replaceable>connection_user</replaceable> ] |
| 6551 |
CONNECT TO DEFAULT |
| 6552 |
CONNECT <replaceable>connection_user</replaceable> |
| 6553 |
DATABASE <replaceable>connection_target</replaceable> |
| 6554 |
</synopsis> |
| 6555 |
</refsynopsisdiv> |
| 6556 |
|
| 6557 |
<refsect1> |
| 6558 |
<title>Description</title> |
| 6559 |
|
| 6560 |
<para> |
| 6561 |
The <command>CONNECT</command> command establishes a connection |
| 6562 |
between the client and the PostgreSQL server. |
| 6563 |
</para> |
| 6564 |
</refsect1> |
| 6565 |
|
| 6566 |
<refsect1> |
| 6567 |
<title>Parameters</title> |
| 6568 |
|
| 6569 |
<variablelist> |
| 6570 |
<varlistentry id="ecpg-sql-connect-connection-target"> |
| 6571 |
<term><replaceable class="parameter">connection_target</replaceable></term> |
| 6572 |
<listitem> |
| 6573 |
<para> |
| 6574 |
<replaceable class="parameter">connection_target</replaceable> |
| 6575 |
specifies the target server of the connection on one of |
| 6576 |
several forms. |
| 6577 |
|
| 6578 |
<variablelist> |
| 6579 |
<varlistentry id="ecpg-sql-connect-connection-target-database-name"> |
| 6580 |
<term>[ <replaceable>database_name</replaceable> ] [ <literal>@</literal><replaceable>host</replaceable> ] [ <literal>:</literal><replaceable>port</replaceable> ]</term> |
| 6581 |
<listitem> |
| 6582 |
<para> |
| 6583 |
Connect over TCP/IP |
| 6584 |
</para> |
| 6585 |
</listitem> |
| 6586 |
</varlistentry> |
| 6587 |
|
| 6588 |
<varlistentry id="ecpg-sql-connect-connection-target-unix-domain-sockets"> |
| 6589 |
<term><literal>unix:postgresql://</literal><replaceable>host</replaceable> [ <literal>:</literal><replaceable>port</replaceable> ] <literal>/</literal> [ <replaceable>database_name</replaceable> ] [ <literal>?</literal><replaceable>connection_option</replaceable> ]</term> |
| 6590 |
<listitem> |
| 6591 |
<para> |
| 6592 |
Connect over Unix-domain sockets |
| 6593 |
</para> |
| 6594 |
</listitem> |
| 6595 |
</varlistentry> |
| 6596 |
|
| 6597 |
<varlistentry id="ecpg-sql-connect-connection-target-tcp-ip"> |
| 6598 |
<term><literal>tcp:postgresql://</literal><replaceable>host</replaceable> [ <literal>:</literal><replaceable>port</replaceable> ] <literal>/</literal> [ <replaceable>database_name</replaceable> ] [ <literal>?</literal><replaceable>connection_option</replaceable> ]</term> |
| 6599 |
<listitem> |
| 6600 |
<para> |
| 6601 |
Connect over TCP/IP |
| 6602 |
</para> |
| 6603 |
</listitem> |
| 6604 |
</varlistentry> |
| 6605 |
|
| 6606 |
<varlistentry id="ecpg-sql-connect-connection-target-constant"> |
| 6607 |
<term>SQL string constant</term> |
| 6608 |
<listitem> |
| 6609 |
<para> |
| 6610 |
containing a value in one of the above forms |
| 6611 |
</para> |
| 6612 |
</listitem> |
| 6613 |
</varlistentry> |
| 6614 |
|
| 6615 |
<varlistentry id="ecpg-sql-connect-connection-target-host-variable"> |
| 6616 |
<term>host variable</term> |
| 6617 |
<listitem> |
| 6618 |
<para> |
| 6619 |
host variable of type <type>char[]</type> |
| 6620 |
or <type>VARCHAR[]</type> containing a value in one of the |
| 6621 |
above forms |
| 6622 |
</para> |
| 6623 |
</listitem> |
| 6624 |
</varlistentry> |
| 6625 |
</variablelist> |
| 6626 |
</para> |
| 6627 |
</listitem> |
| 6628 |
</varlistentry> |
| 6629 |
|
| 6630 |
<varlistentry id="ecpg-sql-connect-connection-name"> |
| 6631 |
<term><replaceable class="parameter">connection_name</replaceable></term> |
| 6632 |
<listitem> |
| 6633 |
<para> |
| 6634 |
An optional identifier for the connection, so that it can be |
| 6635 |
referred to in other commands. This can be an SQL identifier |
| 6636 |
or a host variable. |
| 6637 |
</para> |
| 6638 |
</listitem> |
| 6639 |
</varlistentry> |
| 6640 |
|
| 6641 |
<varlistentry id="ecpg-sql-connect-connection-user"> |
| 6642 |
<term><replaceable class="parameter">connection_user</replaceable></term> |
| 6643 |
<listitem> |
| 6644 |
<para> |
| 6645 |
The user name for the database connection. |
| 6646 |
</para> |
| 6647 |
|
| 6648 |
<para> |
| 6649 |
This parameter can also specify user name and password, using one the forms |
| 6650 |
<literal><replaceable>user_name</replaceable>/<replaceable>password</replaceable></literal>, |
| 6651 |
<literal><replaceable>user_name</replaceable> IDENTIFIED BY <replaceable>password</replaceable></literal>, or |
| 6652 |
<literal><replaceable>user_name</replaceable> USING <replaceable>password</replaceable></literal>. |
| 6653 |
</para> |
| 6654 |
|
| 6655 |
<para> |
| 6656 |
User name and password can be SQL identifiers, string |
| 6657 |
constants, or host variables. |
| 6658 |
</para> |
| 6659 |
</listitem> |
| 6660 |
</varlistentry> |
| 6661 |
|
| 6662 |
<varlistentry id="ecpg-sql-connect-default"> |
| 6663 |
<term><literal>DEFAULT</literal></term> |
| 6664 |
<listitem> |
| 6665 |
<para> |
| 6666 |
Use all default connection parameters, as defined by libpq. |
| 6667 |
</para> |
| 6668 |
</listitem> |
| 6669 |
</varlistentry> |
| 6670 |
</variablelist> |
| 6671 |
</refsect1> |
| 6672 |
|
| 6673 |
<refsect1> |
| 6674 |
<title>Examples</title> |
| 6675 |
|
| 6676 |
<para> |
| 6677 |
Here a several variants for specifying connection parameters: |
| 6678 |
<programlisting> |
| 6679 |
EXEC SQL CONNECT TO "connectdb" AS main; |
| 6680 |
EXEC SQL CONNECT TO "connectdb" AS second; |
| 6681 |
EXEC SQL CONNECT TO "unix:postgresql://200.46.204.71/connectdb" AS main USER connectuser; |
| 6682 |
EXEC SQL CONNECT TO "unix:postgresql://localhost/connectdb" AS main USER connectuser; |
| 6683 |
EXEC SQL CONNECT TO 'connectdb' AS main; |
| 6684 |
EXEC SQL CONNECT TO 'unix:postgresql://localhost/connectdb' AS main USER :user; |
| 6685 |
EXEC SQL CONNECT TO :db AS :id; |
| 6686 |
EXEC SQL CONNECT TO :db USER connectuser USING :pw; |
| 6687 |
EXEC SQL CONNECT TO @localhost AS main USER connectdb; |
| 6688 |
EXEC SQL CONNECT TO REGRESSDB1 as main; |
| 6689 |
EXEC SQL CONNECT TO AS main USER connectdb; |
| 6690 |
EXEC SQL CONNECT TO connectdb AS :id; |
| 6691 |
EXEC SQL CONNECT TO connectdb AS main USER connectuser/connectdb; |
| 6692 |
EXEC SQL CONNECT TO connectdb AS main; |
| 6693 |
EXEC SQL CONNECT TO connectdb@localhost AS main; |
| 6694 |
EXEC SQL CONNECT TO tcp:postgresql://localhost/ USER connectdb; |
| 6695 |
EXEC SQL CONNECT TO tcp:postgresql://localhost/connectdb USER connectuser IDENTIFIED BY connectpw; |
| 6696 |
EXEC SQL CONNECT TO tcp:postgresql://localhost:20/connectdb USER connectuser IDENTIFIED BY connectpw; |
| 6697 |
EXEC SQL CONNECT TO unix:postgresql://localhost/ AS main USER connectdb; |
| 6698 |
EXEC SQL CONNECT TO unix:postgresql://localhost/connectdb AS main USER connectuser; |
| 6699 |
EXEC SQL CONNECT TO unix:postgresql://localhost/connectdb USER connectuser IDENTIFIED BY "connectpw"; |
| 6700 |
EXEC SQL CONNECT TO unix:postgresql://localhost/connectdb USER connectuser USING "connectpw"; |
| 6701 |
EXEC SQL CONNECT TO unix:postgresql://localhost/connectdb?connect_timeout=14 USER connectuser; |
| 6702 |
</programlisting> |
| 6703 |
</para> |
| 6704 |
|
| 6705 |
<para> |
| 6706 |
Here is an example program that illustrates the use of host |
| 6707 |
variables to specify connection parameters: |
| 6708 |
<programlisting> |
| 6709 |
int |
| 6710 |
main(void) |
| 6711 |
{ |
| 6712 |
EXEC SQL BEGIN DECLARE SECTION; |
| 6713 |
char *dbname = "testdb"; /* database name */ |
| 6714 |
char *user = "testuser"; /* connection user name */ |
| 6715 |
char *connection = "tcp:postgresql://localhost:5432/testdb"; |
| 6716 |
/* connection string */ |
| 6717 |
char ver[256]; /* buffer to store the version string */ |
| 6718 |
EXEC SQL END DECLARE SECTION; |
| 6719 |
|
| 6720 |
ECPGdebug(1, stderr); |
| 6721 |
|
| 6722 |
EXEC SQL CONNECT TO :dbname USER :user; |
| 6723 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 6724 |
EXEC SQL SELECT version() INTO :ver; |
| 6725 |
EXEC SQL DISCONNECT; |
| 6726 |
|
| 6727 |
printf("version: %s\n", ver); |
| 6728 |
|
| 6729 |
EXEC SQL CONNECT TO :connection USER :user; |
| 6730 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 6731 |
EXEC SQL SELECT version() INTO :ver; |
| 6732 |
EXEC SQL DISCONNECT; |
| 6733 |
|
| 6734 |
printf("version: %s\n", ver); |
| 6735 |
|
| 6736 |
return 0; |
| 6737 |
} |
| 6738 |
</programlisting> |
| 6739 |
</para> |
| 6740 |
</refsect1> |
| 6741 |
|
| 6742 |
<refsect1> |
| 6743 |
<title>Compatibility</title> |
| 6744 |
|
| 6745 |
<para> |
| 6746 |
<command>CONNECT</command> is specified in the SQL standard, but |
| 6747 |
the format of the connection parameters is |
| 6748 |
implementation-specific. |
| 6749 |
</para> |
| 6750 |
</refsect1> |
| 6751 |
|
| 6752 |
<refsect1> |
| 6753 |
<title>See Also</title> |
| 6754 |
|
| 6755 |
<simplelist type="inline"> |
| 6756 |
<member><xref linkend="ecpg-sql-disconnect"/></member> |
| 6757 |
<member><xref linkend="ecpg-sql-set-connection"/></member> |
| 6758 |
</simplelist> |
| 6759 |
</refsect1> |
| 6760 |
</refentry> |
| 6761 |
|
| 6762 |
<refentry id="ecpg-sql-deallocate-descriptor"> |
| 6763 |
<refnamediv> |
| 6764 |
<refname>DEALLOCATE DESCRIPTOR</refname> |
| 6765 |
<refpurpose>deallocate an SQL descriptor area</refpurpose> |
| 6766 |
</refnamediv> |
| 6767 |
|
| 6768 |
<refsynopsisdiv> |
| 6769 |
<synopsis> |
| 6770 |
DEALLOCATE DESCRIPTOR <replaceable class="parameter">name</replaceable> |
| 6771 |
</synopsis> |
| 6772 |
</refsynopsisdiv> |
| 6773 |
|
| 6774 |
<refsect1> |
| 6775 |
<title>Description</title> |
| 6776 |
|
| 6777 |
<para> |
| 6778 |
<command>DEALLOCATE DESCRIPTOR</command> deallocates a named SQL |
| 6779 |
descriptor area. |
| 6780 |
</para> |
| 6781 |
</refsect1> |
| 6782 |
|
| 6783 |
<refsect1> |
| 6784 |
<title>Parameters</title> |
| 6785 |
|
| 6786 |
<variablelist> |
| 6787 |
<varlistentry id="ecpg-sql-deallocate-descriptor-name"> |
| 6788 |
<term><replaceable class="parameter">name</replaceable></term> |
| 6789 |
<listitem> |
| 6790 |
<para> |
| 6791 |
The name of the descriptor which is going to be deallocated. |
| 6792 |
It is case sensitive. This can be an SQL identifier or a host |
| 6793 |
variable. |
| 6794 |
</para> |
| 6795 |
</listitem> |
| 6796 |
</varlistentry> |
| 6797 |
</variablelist> |
| 6798 |
</refsect1> |
| 6799 |
|
| 6800 |
<refsect1> |
| 6801 |
<title>Examples</title> |
| 6802 |
|
| 6803 |
<programlisting> |
| 6804 |
EXEC SQL DEALLOCATE DESCRIPTOR mydesc; |
| 6805 |
</programlisting> |
| 6806 |
</refsect1> |
| 6807 |
|
| 6808 |
<refsect1> |
| 6809 |
<title>Compatibility</title> |
| 6810 |
|
| 6811 |
<para> |
| 6812 |
<command>DEALLOCATE DESCRIPTOR</command> is specified in the SQL |
| 6813 |
standard. |
| 6814 |
</para> |
| 6815 |
</refsect1> |
| 6816 |
|
| 6817 |
<refsect1> |
| 6818 |
<title>See Also</title> |
| 6819 |
|
| 6820 |
<simplelist type="inline"> |
| 6821 |
<member><xref linkend="ecpg-sql-allocate-descriptor"/></member> |
| 6822 |
<member><xref linkend="ecpg-sql-get-descriptor"/></member> |
| 6823 |
<member><xref linkend="ecpg-sql-set-descriptor"/></member> |
| 6824 |
</simplelist> |
| 6825 |
</refsect1> |
| 6826 |
</refentry> |
| 6827 |
|
| 6828 |
<refentry id="ecpg-sql-declare"> |
| 6829 |
<refnamediv> |
| 6830 |
<refname>DECLARE</refname> |
| 6831 |
<refpurpose>define a cursor</refpurpose> |
| 6832 |
</refnamediv> |
| 6833 |
|
| 6834 |
<refsynopsisdiv> |
| 6835 |
<synopsis> |
| 6836 |
DECLARE <replaceable class="parameter">cursor_name</replaceable> [ BINARY ] [ ASENSITIVE | INSENSITIVE ] [ [ NO ] SCROLL ] CURSOR [ { WITH | WITHOUT } HOLD ] FOR <replaceable class="parameter">prepared_name</replaceable> |
| 6837 |
DECLARE <replaceable class="parameter">cursor_name</replaceable> [ BINARY ] [ ASENSITIVE | INSENSITIVE ] [ [ NO ] SCROLL ] CURSOR [ { WITH | WITHOUT } HOLD ] FOR <replaceable class="parameter">query</replaceable> |
| 6838 |
</synopsis> |
| 6839 |
</refsynopsisdiv> |
| 6840 |
|
| 6841 |
<refsect1> |
| 6842 |
<title>Description</title> |
| 6843 |
|
| 6844 |
<para> |
| 6845 |
<command>DECLARE</command> declares a cursor for iterating over |
| 6846 |
the result set of a prepared statement. This command has |
| 6847 |
slightly different semantics from the direct SQL |
| 6848 |
command <command>DECLARE</command>: Whereas the latter executes a |
| 6849 |
query and prepares the result set for retrieval, this embedded |
| 6850 |
SQL command merely declares a name as a <quote>loop |
| 6851 |
variable</quote> for iterating over the result set of a query; |
| 6852 |
the actual execution happens when the cursor is opened with |
| 6853 |
the <command>OPEN</command> command. |
| 6854 |
</para> |
| 6855 |
</refsect1> |
| 6856 |
|
| 6857 |
<refsect1> |
| 6858 |
<title>Parameters</title> |
| 6859 |
<variablelist> |
| 6860 |
|
| 6861 |
<varlistentry id="ecpg-sql-declare-cursor-name"> |
| 6862 |
<term><replaceable class="parameter">cursor_name</replaceable></term> |
| 6863 |
<listitem> |
| 6864 |
<para> |
| 6865 |
A cursor name, case sensitive. This can be an SQL identifier |
| 6866 |
or a host variable. |
| 6867 |
</para> |
| 6868 |
</listitem> |
| 6869 |
</varlistentry> |
| 6870 |
|
| 6871 |
<varlistentry id="ecpg-sql-declare-prepared-name"> |
| 6872 |
<term><replaceable class="parameter">prepared_name</replaceable></term> |
| 6873 |
<listitem> |
| 6874 |
<para> |
| 6875 |
The name of a prepared query, either as an SQL identifier or a |
| 6876 |
host variable. |
| 6877 |
</para> |
| 6878 |
</listitem> |
| 6879 |
</varlistentry> |
| 6880 |
|
| 6881 |
<varlistentry id="ecpg-sql-declare-query"> |
| 6882 |
<term><replaceable class="parameter">query</replaceable></term> |
| 6883 |
<listitem> |
| 6884 |
<para> |
| 6885 |
A <xref linkend="sql-select"/> or |
| 6886 |
<xref linkend="sql-values"/> command which will provide the |
| 6887 |
rows to be returned by the cursor. |
| 6888 |
</para> |
| 6889 |
</listitem> |
| 6890 |
</varlistentry> |
| 6891 |
</variablelist> |
| 6892 |
|
| 6893 |
<para> |
| 6894 |
For the meaning of the cursor options, |
| 6895 |
see <xref linkend="sql-declare"/>. |
| 6896 |
</para> |
| 6897 |
</refsect1> |
| 6898 |
|
| 6899 |
<refsect1> |
| 6900 |
<title>Examples</title> |
| 6901 |
|
| 6902 |
<para> |
| 6903 |
Examples declaring a cursor for a query: |
| 6904 |
<programlisting> |
| 6905 |
EXEC SQL DECLARE C CURSOR FOR SELECT * FROM My_Table; |
| 6906 |
EXEC SQL DECLARE C CURSOR FOR SELECT Item1 FROM T; |
| 6907 |
EXEC SQL DECLARE cur1 CURSOR FOR SELECT version(); |
| 6908 |
</programlisting> |
| 6909 |
</para> |
| 6910 |
|
| 6911 |
<para> |
| 6912 |
An example declaring a cursor for a prepared statement: |
| 6913 |
<programlisting> |
| 6914 |
EXEC SQL PREPARE stmt1 AS SELECT version(); |
| 6915 |
EXEC SQL DECLARE cur1 CURSOR FOR stmt1; |
| 6916 |
</programlisting> |
| 6917 |
</para> |
| 6918 |
</refsect1> |
| 6919 |
|
| 6920 |
<refsect1> |
| 6921 |
<title>Compatibility</title> |
| 6922 |
|
| 6923 |
<para> |
| 6924 |
<command>DECLARE</command> is specified in the SQL standard. |
| 6925 |
</para> |
| 6926 |
</refsect1> |
| 6927 |
|
| 6928 |
<refsect1> |
| 6929 |
<title>See Also</title> |
| 6930 |
|
| 6931 |
<simplelist type="inline"> |
| 6932 |
<member><xref linkend="ecpg-sql-open"/></member> |
| 6933 |
<member><xref linkend="sql-close"/></member> |
| 6934 |
<member><xref linkend="sql-declare"/></member> |
| 6935 |
</simplelist> |
| 6936 |
</refsect1> |
| 6937 |
</refentry> |
| 6938 |
|
| 6939 |
<refentry id="ecpg-sql-declare-statement"> |
| 6940 |
<refnamediv> |
| 6941 |
<refname>DECLARE STATEMENT</refname> |
| 6942 |
<refpurpose>declare SQL statement identifier</refpurpose> |
| 6943 |
</refnamediv> |
| 6944 |
|
| 6945 |
<refsynopsisdiv> |
| 6946 |
<synopsis> |
| 6947 |
EXEC SQL [ AT <replaceable class="parameter">connection_name</replaceable> ] DECLARE <replaceable class="parameter">statement_name</replaceable> STATEMENT |
| 6948 |
</synopsis> |
| 6949 |
</refsynopsisdiv> |
| 6950 |
|
| 6951 |
<refsect1> |
| 6952 |
<title>Description</title> |
| 6953 |
|
| 6954 |
<para> |
| 6955 |
<command>DECLARE STATEMENT</command> declares an SQL statement identifier. |
| 6956 |
SQL statement identifier can be associated with the connection. |
| 6957 |
When the identifier is used by dynamic SQL statements, the statements |
| 6958 |
are executed using the associated connection. |
| 6959 |
The namespace of the declaration is the precompile unit, and multiple |
| 6960 |
declarations to the same SQL statement identifier are not allowed. |
| 6961 |
Note that if the precompiler runs in Informix compatibility mode and |
| 6962 |
some SQL statement is declared, "database" can not be used as a cursor |
| 6963 |
name. |
| 6964 |
</para> |
| 6965 |
</refsect1> |
| 6966 |
|
| 6967 |
<refsect1> |
| 6968 |
<title>Parameters</title> |
| 6969 |
|
| 6970 |
<variablelist> |
| 6971 |
<varlistentry id="ecpg-sql-declare-statement-connection-name"> |
| 6972 |
<term><replaceable class="parameter">connection_name</replaceable></term> |
| 6973 |
<listitem> |
| 6974 |
<para> |
| 6975 |
A database connection name established by the <command>CONNECT</command> command. |
| 6976 |
</para> |
| 6977 |
<para> |
| 6978 |
AT clause can be omitted, but such statement has no meaning. |
| 6979 |
</para> |
| 6980 |
</listitem> |
| 6981 |
</varlistentry> |
| 6982 |
</variablelist> |
| 6983 |
|
| 6984 |
<variablelist> |
| 6985 |
<varlistentry id="ecpg-sql-declare-statement-statement-name"> |
| 6986 |
<term><replaceable class="parameter">statement_name</replaceable></term> |
| 6987 |
<listitem> |
| 6988 |
<para> |
| 6989 |
The name of an SQL statement identifier, either as an SQL identifier or a host variable. |
| 6990 |
</para> |
| 6991 |
</listitem> |
| 6992 |
</varlistentry> |
| 6993 |
</variablelist> |
| 6994 |
</refsect1> |
| 6995 |
|
| 6996 |
<refsect1> |
| 6997 |
<title>Notes</title> |
| 6998 |
<para> |
| 6999 |
This association is valid only if the declaration is physically placed on top of a dynamic statement. |
| 7000 |
</para> |
| 7001 |
</refsect1> |
| 7002 |
|
| 7003 |
<refsect1> |
| 7004 |
<title>Examples</title> |
| 7005 |
|
| 7006 |
<programlisting> |
| 7007 |
EXEC SQL CONNECT TO postgres AS con1; |
| 7008 |
EXEC SQL AT con1 DECLARE sql_stmt STATEMENT; |
| 7009 |
EXEC SQL DECLARE cursor_name CURSOR FOR sql_stmt; |
| 7010 |
EXEC SQL PREPARE sql_stmt FROM :dyn_string; |
| 7011 |
EXEC SQL OPEN cursor_name; |
| 7012 |
EXEC SQL FETCH cursor_name INTO :column1; |
| 7013 |
EXEC SQL CLOSE cursor_name; |
| 7014 |
</programlisting> |
| 7015 |
</refsect1> |
| 7016 |
|
| 7017 |
<refsect1> |
| 7018 |
<title>Compatibility</title> |
| 7019 |
|
| 7020 |
<para> |
| 7021 |
<command>DECLARE STATEMENT</command> is an extension of the SQL standard, |
| 7022 |
but can be used in famous DBMSs. |
| 7023 |
</para> |
| 7024 |
</refsect1> |
| 7025 |
|
| 7026 |
<refsect1> |
| 7027 |
<title>See Also</title> |
| 7028 |
|
| 7029 |
<simplelist type="inline"> |
| 7030 |
<member><xref linkend="ecpg-sql-connect"/></member> |
| 7031 |
<member><xref linkend="ecpg-sql-declare"/></member> |
| 7032 |
<member><xref linkend="ecpg-sql-open"/></member> |
| 7033 |
</simplelist> |
| 7034 |
</refsect1> |
| 7035 |
</refentry> |
| 7036 |
|
| 7037 |
<refentry id="ecpg-sql-describe"> |
| 7038 |
<refnamediv> |
| 7039 |
<refname>DESCRIBE</refname> |
| 7040 |
<refpurpose>obtain information about a prepared statement or result set</refpurpose> |
| 7041 |
</refnamediv> |
| 7042 |
|
| 7043 |
<refsynopsisdiv> |
| 7044 |
<synopsis> |
| 7045 |
DESCRIBE [ OUTPUT ] <replaceable class="parameter">prepared_name</replaceable> USING [ SQL ] DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> |
| 7046 |
DESCRIBE [ OUTPUT ] <replaceable class="parameter">prepared_name</replaceable> INTO [ SQL ] DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> |
| 7047 |
DESCRIBE [ OUTPUT ] <replaceable class="parameter">prepared_name</replaceable> INTO <replaceable class="parameter">sqlda_name</replaceable> |
| 7048 |
</synopsis> |
| 7049 |
</refsynopsisdiv> |
| 7050 |
|
| 7051 |
<refsect1> |
| 7052 |
<title>Description</title> |
| 7053 |
|
| 7054 |
<para> |
| 7055 |
<command>DESCRIBE</command> retrieves metadata information about |
| 7056 |
the result columns contained in a prepared statement, without |
| 7057 |
actually fetching a row. |
| 7058 |
</para> |
| 7059 |
</refsect1> |
| 7060 |
|
| 7061 |
<refsect1> |
| 7062 |
<title>Parameters</title> |
| 7063 |
|
| 7064 |
<variablelist> |
| 7065 |
<varlistentry id="ecpg-sql-describe-prepared-name"> |
| 7066 |
<term><replaceable class="parameter">prepared_name</replaceable></term> |
| 7067 |
<listitem> |
| 7068 |
<para> |
| 7069 |
The name of a prepared statement. This can be an SQL |
| 7070 |
identifier or a host variable. |
| 7071 |
</para> |
| 7072 |
</listitem> |
| 7073 |
</varlistentry> |
| 7074 |
|
| 7075 |
<varlistentry id="ecpg-sql-describe-descriptor-name"> |
| 7076 |
<term><replaceable class="parameter">descriptor_name</replaceable></term> |
| 7077 |
<listitem> |
| 7078 |
<para> |
| 7079 |
A descriptor name. It is case sensitive. It can be an SQL |
| 7080 |
identifier or a host variable. |
| 7081 |
</para> |
| 7082 |
</listitem> |
| 7083 |
</varlistentry> |
| 7084 |
|
| 7085 |
<varlistentry id="ecpg-sql-describe-sqlda-name"> |
| 7086 |
<term><replaceable class="parameter">sqlda_name</replaceable></term> |
| 7087 |
<listitem> |
| 7088 |
<para> |
| 7089 |
The name of an SQLDA variable. |
| 7090 |
</para> |
| 7091 |
</listitem> |
| 7092 |
</varlistentry> |
| 7093 |
</variablelist> |
| 7094 |
</refsect1> |
| 7095 |
|
| 7096 |
<refsect1> |
| 7097 |
<title>Examples</title> |
| 7098 |
|
| 7099 |
<programlisting> |
| 7100 |
EXEC SQL ALLOCATE DESCRIPTOR mydesc; |
| 7101 |
EXEC SQL PREPARE stmt1 FROM :sql_stmt; |
| 7102 |
EXEC SQL DESCRIBE stmt1 INTO SQL DESCRIPTOR mydesc; |
| 7103 |
EXEC SQL GET DESCRIPTOR mydesc VALUE 1 :charvar = NAME; |
| 7104 |
EXEC SQL DEALLOCATE DESCRIPTOR mydesc; |
| 7105 |
</programlisting> |
| 7106 |
</refsect1> |
| 7107 |
|
| 7108 |
<refsect1> |
| 7109 |
<title>Compatibility</title> |
| 7110 |
|
| 7111 |
<para> |
| 7112 |
<command>DESCRIBE</command> is specified in the SQL standard. |
| 7113 |
</para> |
| 7114 |
</refsect1> |
| 7115 |
|
| 7116 |
<refsect1> |
| 7117 |
<title>See Also</title> |
| 7118 |
|
| 7119 |
<simplelist type="inline"> |
| 7120 |
<member><xref linkend="ecpg-sql-allocate-descriptor"/></member> |
| 7121 |
<member><xref linkend="ecpg-sql-get-descriptor"/></member> |
| 7122 |
</simplelist> |
| 7123 |
</refsect1> |
| 7124 |
</refentry> |
| 7125 |
|
| 7126 |
<refentry id="ecpg-sql-disconnect"> |
| 7127 |
<refnamediv> |
| 7128 |
<refname>DISCONNECT</refname> |
| 7129 |
<refpurpose>terminate a database connection</refpurpose> |
| 7130 |
</refnamediv> |
| 7131 |
|
| 7132 |
<refsynopsisdiv> |
| 7133 |
<synopsis> |
| 7134 |
DISCONNECT <replaceable class="parameter">connection_name</replaceable> |
| 7135 |
DISCONNECT [ CURRENT ] |
| 7136 |
DISCONNECT ALL |
| 7137 |
</synopsis> |
| 7138 |
</refsynopsisdiv> |
| 7139 |
|
| 7140 |
<refsect1> |
| 7141 |
<title>Description</title> |
| 7142 |
|
| 7143 |
<para> |
| 7144 |
<command>DISCONNECT</command> closes a connection (or all |
| 7145 |
connections) to the database. |
| 7146 |
</para> |
| 7147 |
</refsect1> |
| 7148 |
|
| 7149 |
<refsect1> |
| 7150 |
<title>Parameters</title> |
| 7151 |
|
| 7152 |
<variablelist> |
| 7153 |
<varlistentry id="ecpg-sql-disconnect-connection-name"> |
| 7154 |
<term><replaceable class="parameter">connection_name</replaceable></term> |
| 7155 |
<listitem> |
| 7156 |
<para> |
| 7157 |
A database connection name established by |
| 7158 |
the <command>CONNECT</command> command. |
| 7159 |
</para> |
| 7160 |
</listitem> |
| 7161 |
</varlistentry> |
| 7162 |
|
| 7163 |
<varlistentry id="ecpg-sql-disconnect-current"> |
| 7164 |
<term><literal>CURRENT</literal></term> |
| 7165 |
<listitem> |
| 7166 |
<para> |
| 7167 |
Close the <quote>current</quote> connection, which is either |
| 7168 |
the most recently opened connection, or the connection set by |
| 7169 |
the <command>SET CONNECTION</command> command. This is also |
| 7170 |
the default if no argument is given to |
| 7171 |
the <command>DISCONNECT</command> command. |
| 7172 |
</para> |
| 7173 |
</listitem> |
| 7174 |
</varlistentry> |
| 7175 |
|
| 7176 |
<varlistentry id="ecpg-sql-disconnect-all"> |
| 7177 |
<term><literal>ALL</literal></term> |
| 7178 |
<listitem> |
| 7179 |
<para> |
| 7180 |
Close all open connections. |
| 7181 |
</para> |
| 7182 |
</listitem> |
| 7183 |
</varlistentry> |
| 7184 |
</variablelist> |
| 7185 |
</refsect1> |
| 7186 |
|
| 7187 |
<refsect1> |
| 7188 |
<title>Examples</title> |
| 7189 |
|
| 7190 |
<programlisting> |
| 7191 |
int |
| 7192 |
main(void) |
| 7193 |
{ |
| 7194 |
EXEC SQL CONNECT TO testdb AS con1 USER testuser; |
| 7195 |
EXEC SQL CONNECT TO testdb AS con2 USER testuser; |
| 7196 |
EXEC SQL CONNECT TO testdb AS con3 USER testuser; |
| 7197 |
|
| 7198 |
EXEC SQL DISCONNECT CURRENT; /* close con3 */ |
| 7199 |
EXEC SQL DISCONNECT ALL; /* close con2 and con1 */ |
| 7200 |
|
| 7201 |
return 0; |
| 7202 |
} |
| 7203 |
</programlisting> |
| 7204 |
</refsect1> |
| 7205 |
|
| 7206 |
<refsect1> |
| 7207 |
<title>Compatibility</title> |
| 7208 |
|
| 7209 |
<para> |
| 7210 |
<command>DISCONNECT</command> is specified in the SQL standard. |
| 7211 |
</para> |
| 7212 |
</refsect1> |
| 7213 |
|
| 7214 |
<refsect1> |
| 7215 |
<title>See Also</title> |
| 7216 |
|
| 7217 |
<simplelist type="inline"> |
| 7218 |
<member><xref linkend="ecpg-sql-connect"/></member> |
| 7219 |
<member><xref linkend="ecpg-sql-set-connection"/></member> |
| 7220 |
</simplelist> |
| 7221 |
</refsect1> |
| 7222 |
</refentry> |
| 7223 |
|
| 7224 |
<refentry id="ecpg-sql-execute-immediate"> |
| 7225 |
<refnamediv> |
| 7226 |
<refname>EXECUTE IMMEDIATE</refname> |
| 7227 |
<refpurpose>dynamically prepare and execute a statement</refpurpose> |
| 7228 |
</refnamediv> |
| 7229 |
|
| 7230 |
<refsynopsisdiv> |
| 7231 |
<synopsis> |
| 7232 |
EXECUTE IMMEDIATE <replaceable class="parameter">string</replaceable> |
| 7233 |
</synopsis> |
| 7234 |
</refsynopsisdiv> |
| 7235 |
|
| 7236 |
<refsect1> |
| 7237 |
<title>Description</title> |
| 7238 |
|
| 7239 |
<para> |
| 7240 |
<command>EXECUTE IMMEDIATE</command> immediately prepares and |
| 7241 |
executes a dynamically specified SQL statement, without |
| 7242 |
retrieving result rows. |
| 7243 |
</para> |
| 7244 |
</refsect1> |
| 7245 |
|
| 7246 |
<refsect1> |
| 7247 |
<title>Parameters</title> |
| 7248 |
|
| 7249 |
<variablelist> |
| 7250 |
<varlistentry id="ecpg-sql-execute-immediate-string"> |
| 7251 |
<term><replaceable class="parameter">string</replaceable></term> |
| 7252 |
<listitem> |
| 7253 |
<para> |
| 7254 |
A literal string or a host variable containing the SQL |
| 7255 |
statement to be executed. |
| 7256 |
</para> |
| 7257 |
</listitem> |
| 7258 |
</varlistentry> |
| 7259 |
</variablelist> |
| 7260 |
</refsect1> |
| 7261 |
|
| 7262 |
<refsect1> |
| 7263 |
<title>Notes</title> |
| 7264 |
|
| 7265 |
<para> |
| 7266 |
In typical usage, the <replaceable>string</replaceable> is a host |
| 7267 |
variable reference to a string containing a dynamically-constructed |
| 7268 |
SQL statement. The case of a literal string is not very useful; |
| 7269 |
you might as well just write the SQL statement directly, without |
| 7270 |
the extra typing of <command>EXECUTE IMMEDIATE</command>. |
| 7271 |
</para> |
| 7272 |
|
| 7273 |
<para> |
| 7274 |
If you do use a literal string, keep in mind that any double quotes |
| 7275 |
you might wish to include in the SQL statement must be written as |
| 7276 |
octal escapes (<literal>\042</literal>) not the usual C |
| 7277 |
idiom <literal>\"</literal>. This is because the string is inside |
| 7278 |
an <literal>EXEC SQL</literal> section, so the ECPG lexer parses it |
| 7279 |
according to SQL rules not C rules. Any embedded backslashes will |
| 7280 |
later be handled according to C rules; but <literal>\"</literal> |
| 7281 |
causes an immediate syntax error because it is seen as ending the |
| 7282 |
literal. |
| 7283 |
</para> |
| 7284 |
</refsect1> |
| 7285 |
|
| 7286 |
<refsect1> |
| 7287 |
<title>Examples</title> |
| 7288 |
|
| 7289 |
<para> |
| 7290 |
Here is an example that executes an <command>INSERT</command> |
| 7291 |
statement using <command>EXECUTE IMMEDIATE</command> and a host |
| 7292 |
variable named <varname>command</varname>: |
| 7293 |
<programlisting> |
| 7294 |
sprintf(command, "INSERT INTO test (name, amount, letter) VALUES ('db: ''r1''', 1, 'f')"); |
| 7295 |
EXEC SQL EXECUTE IMMEDIATE :command; |
| 7296 |
</programlisting> |
| 7297 |
</para> |
| 7298 |
</refsect1> |
| 7299 |
|
| 7300 |
<refsect1> |
| 7301 |
<title>Compatibility</title> |
| 7302 |
|
| 7303 |
<para> |
| 7304 |
<command>EXECUTE IMMEDIATE</command> is specified in the SQL standard. |
| 7305 |
</para> |
| 7306 |
</refsect1> |
| 7307 |
</refentry> |
| 7308 |
|
| 7309 |
<refentry id="ecpg-sql-get-descriptor"> |
| 7310 |
<refnamediv> |
| 7311 |
<refname>GET DESCRIPTOR</refname> |
| 7312 |
<refpurpose>get information from an SQL descriptor area</refpurpose> |
| 7313 |
</refnamediv> |
| 7314 |
|
| 7315 |
<refsynopsisdiv> |
| 7316 |
<synopsis> |
| 7317 |
GET DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> <replaceable class="parameter">:cvariable</replaceable> = <replaceable class="parameter">descriptor_header_item</replaceable> |
| 7318 |
GET DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> VALUE <replaceable class="parameter">column_number</replaceable> <replaceable class="parameter">:cvariable</replaceable> = <replaceable class="parameter">descriptor_item</replaceable> [, ... ] |
| 7319 |
</synopsis> |
| 7320 |
</refsynopsisdiv> |
| 7321 |
|
| 7322 |
<refsect1> |
| 7323 |
<title>Description</title> |
| 7324 |
|
| 7325 |
<para> |
| 7326 |
<command>GET DESCRIPTOR</command> retrieves information about a |
| 7327 |
query result set from an SQL descriptor area and stores it into |
| 7328 |
host variables. A descriptor area is typically populated |
| 7329 |
using <command>FETCH</command> or <command>SELECT</command> |
| 7330 |
before using this command to transfer the information into host |
| 7331 |
language variables. |
| 7332 |
</para> |
| 7333 |
|
| 7334 |
<para> |
| 7335 |
This command has two forms: The first form retrieves |
| 7336 |
descriptor <quote>header</quote> item, which applies to the result |
| 7337 |
set in its entirety. One example is the row count. The second |
| 7338 |
form, which requires the column number as additional parameter, |
| 7339 |
retrieves information about a particular column. Examples are |
| 7340 |
the column name and the actual column value. |
| 7341 |
</para> |
| 7342 |
</refsect1> |
| 7343 |
|
| 7344 |
<refsect1> |
| 7345 |
<title>Parameters</title> |
| 7346 |
|
| 7347 |
<variablelist> |
| 7348 |
<varlistentry id="ecpg-sql-get-descriptor-descriptor-name"> |
| 7349 |
<term><replaceable class="parameter">descriptor_name</replaceable></term> |
| 7350 |
<listitem> |
| 7351 |
<para> |
| 7352 |
A descriptor name. |
| 7353 |
</para> |
| 7354 |
</listitem> |
| 7355 |
</varlistentry> |
| 7356 |
|
| 7357 |
<varlistentry id="ecpg-sql-get-descriptor-descriptor-header-item"> |
| 7358 |
<term><replaceable class="parameter">descriptor_header_item</replaceable></term> |
| 7359 |
<listitem> |
| 7360 |
<para> |
| 7361 |
A token identifying which header information item to retrieve. |
| 7362 |
Only <literal>COUNT</literal>, to get the number of columns in the |
| 7363 |
result set, is currently supported. |
| 7364 |
</para> |
| 7365 |
</listitem> |
| 7366 |
</varlistentry> |
| 7367 |
|
| 7368 |
<varlistentry id="ecpg-sql-get-descriptor-column-number"> |
| 7369 |
<term><replaceable class="parameter">column_number</replaceable></term> |
| 7370 |
<listitem> |
| 7371 |
<para> |
| 7372 |
The number of the column about which information is to be |
| 7373 |
retrieved. The count starts at 1. |
| 7374 |
</para> |
| 7375 |
</listitem> |
| 7376 |
</varlistentry> |
| 7377 |
|
| 7378 |
<varlistentry id="ecpg-sql-get-descriptor-descriptor-item"> |
| 7379 |
<term><replaceable class="parameter">descriptor_item</replaceable></term> |
| 7380 |
<listitem> |
| 7381 |
<para> |
| 7382 |
A token identifying which item of information about a column |
| 7383 |
to retrieve. See <xref linkend="ecpg-named-descriptors"/> for |
| 7384 |
a list of supported items. |
| 7385 |
</para> |
| 7386 |
</listitem> |
| 7387 |
</varlistentry> |
| 7388 |
|
| 7389 |
<varlistentry id="ecpg-sql-get-descriptor-cvariable"> |
| 7390 |
<term><replaceable class="parameter">cvariable</replaceable></term> |
| 7391 |
<listitem> |
| 7392 |
<para> |
| 7393 |
A host variable that will receive the data retrieved from the |
| 7394 |
descriptor area. |
| 7395 |
</para> |
| 7396 |
</listitem> |
| 7397 |
</varlistentry> |
| 7398 |
</variablelist> |
| 7399 |
</refsect1> |
| 7400 |
|
| 7401 |
<refsect1> |
| 7402 |
<title>Examples</title> |
| 7403 |
|
| 7404 |
<para> |
| 7405 |
An example to retrieve the number of columns in a result set: |
| 7406 |
<programlisting> |
| 7407 |
EXEC SQL GET DESCRIPTOR d :d_count = COUNT; |
| 7408 |
</programlisting> |
| 7409 |
</para> |
| 7410 |
|
| 7411 |
<para> |
| 7412 |
An example to retrieve a data length in the first column: |
| 7413 |
<programlisting> |
| 7414 |
EXEC SQL GET DESCRIPTOR d VALUE 1 :d_returned_octet_length = RETURNED_OCTET_LENGTH; |
| 7415 |
</programlisting> |
| 7416 |
</para> |
| 7417 |
|
| 7418 |
<para> |
| 7419 |
An example to retrieve the data body of the second column as a |
| 7420 |
string: |
| 7421 |
<programlisting> |
| 7422 |
EXEC SQL GET DESCRIPTOR d VALUE 2 :d_data = DATA; |
| 7423 |
</programlisting> |
| 7424 |
</para> |
| 7425 |
|
| 7426 |
<para> |
| 7427 |
Here is an example for a whole procedure of |
| 7428 |
executing <literal>SELECT current_database();</literal> and showing the number of |
| 7429 |
columns, the column data length, and the column data: |
| 7430 |
<programlisting> |
| 7431 |
int |
| 7432 |
main(void) |
| 7433 |
{ |
| 7434 |
EXEC SQL BEGIN DECLARE SECTION; |
| 7435 |
int d_count; |
| 7436 |
char d_data[1024]; |
| 7437 |
int d_returned_octet_length; |
| 7438 |
EXEC SQL END DECLARE SECTION; |
| 7439 |
|
| 7440 |
EXEC SQL CONNECT TO testdb AS con1 USER testuser; |
| 7441 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 7442 |
EXEC SQL ALLOCATE DESCRIPTOR d; |
| 7443 |
|
| 7444 |
/* Declare, open a cursor, and assign a descriptor to the cursor */ |
| 7445 |
EXEC SQL DECLARE cur CURSOR FOR SELECT current_database(); |
| 7446 |
EXEC SQL OPEN cur; |
| 7447 |
EXEC SQL FETCH NEXT FROM cur INTO SQL DESCRIPTOR d; |
| 7448 |
|
| 7449 |
/* Get a number of total columns */ |
| 7450 |
EXEC SQL GET DESCRIPTOR d :d_count = COUNT; |
| 7451 |
printf("d_count = %d\n", d_count); |
| 7452 |
|
| 7453 |
/* Get length of a returned column */ |
| 7454 |
EXEC SQL GET DESCRIPTOR d VALUE 1 :d_returned_octet_length = RETURNED_OCTET_LENGTH; |
| 7455 |
printf("d_returned_octet_length = %d\n", d_returned_octet_length); |
| 7456 |
|
| 7457 |
/* Fetch the returned column as a string */ |
| 7458 |
EXEC SQL GET DESCRIPTOR d VALUE 1 :d_data = DATA; |
| 7459 |
printf("d_data = %s\n", d_data); |
| 7460 |
|
| 7461 |
/* Closing */ |
| 7462 |
EXEC SQL CLOSE cur; |
| 7463 |
EXEC SQL COMMIT; |
| 7464 |
|
| 7465 |
EXEC SQL DEALLOCATE DESCRIPTOR d; |
| 7466 |
EXEC SQL DISCONNECT ALL; |
| 7467 |
|
| 7468 |
return 0; |
| 7469 |
} |
| 7470 |
</programlisting> |
| 7471 |
When the example is executed, the result will look like this: |
| 7472 |
<screen> |
| 7473 |
d_count = 1 |
| 7474 |
d_returned_octet_length = 6 |
| 7475 |
d_data = testdb |
| 7476 |
</screen> |
| 7477 |
</para> |
| 7478 |
</refsect1> |
| 7479 |
|
| 7480 |
<refsect1> |
| 7481 |
<title>Compatibility</title> |
| 7482 |
|
| 7483 |
<para> |
| 7484 |
<command>GET DESCRIPTOR</command> is specified in the SQL standard. |
| 7485 |
</para> |
| 7486 |
</refsect1> |
| 7487 |
|
| 7488 |
<refsect1> |
| 7489 |
<title>See Also</title> |
| 7490 |
|
| 7491 |
<simplelist type="inline"> |
| 7492 |
<member><xref linkend="ecpg-sql-allocate-descriptor"/></member> |
| 7493 |
<member><xref linkend="ecpg-sql-set-descriptor"/></member> |
| 7494 |
</simplelist> |
| 7495 |
</refsect1> |
| 7496 |
</refentry> |
| 7497 |
|
| 7498 |
<refentry id="ecpg-sql-open"> |
| 7499 |
<refnamediv> |
| 7500 |
<refname>OPEN</refname> |
| 7501 |
<refpurpose>open a dynamic cursor</refpurpose> |
| 7502 |
</refnamediv> |
| 7503 |
|
| 7504 |
<refsynopsisdiv> |
| 7505 |
<synopsis> |
| 7506 |
OPEN <replaceable class="parameter">cursor_name</replaceable> |
| 7507 |
OPEN <replaceable class="parameter">cursor_name</replaceable> USING <replaceable class="parameter">value</replaceable> [, ... ] |
| 7508 |
OPEN <replaceable class="parameter">cursor_name</replaceable> USING SQL DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> |
| 7509 |
</synopsis> |
| 7510 |
</refsynopsisdiv> |
| 7511 |
|
| 7512 |
<refsect1> |
| 7513 |
<title>Description</title> |
| 7514 |
|
| 7515 |
<para> |
| 7516 |
<command>OPEN</command> opens a cursor and optionally binds |
| 7517 |
actual values to the placeholders in the cursor's declaration. |
| 7518 |
The cursor must previously have been declared with |
| 7519 |
the <command>DECLARE</command> command. The execution |
| 7520 |
of <command>OPEN</command> causes the query to start executing on |
| 7521 |
the server. |
| 7522 |
</para> |
| 7523 |
</refsect1> |
| 7524 |
|
| 7525 |
<refsect1> |
| 7526 |
<title>Parameters</title> |
| 7527 |
|
| 7528 |
<variablelist> |
| 7529 |
<varlistentry id="ecpg-sql-open-cursor-name"> |
| 7530 |
<term><replaceable class="parameter">cursor_name</replaceable></term> |
| 7531 |
<listitem> |
| 7532 |
<para> |
| 7533 |
The name of the cursor to be opened. This can be an SQL |
| 7534 |
identifier or a host variable. |
| 7535 |
</para> |
| 7536 |
</listitem> |
| 7537 |
</varlistentry> |
| 7538 |
|
| 7539 |
<varlistentry id="ecpg-sql-open-value"> |
| 7540 |
<term><replaceable class="parameter">value</replaceable></term> |
| 7541 |
<listitem> |
| 7542 |
<para> |
| 7543 |
A value to be bound to a placeholder in the cursor. This can |
| 7544 |
be an SQL constant, a host variable, or a host variable with |
| 7545 |
indicator. |
| 7546 |
</para> |
| 7547 |
</listitem> |
| 7548 |
</varlistentry> |
| 7549 |
|
| 7550 |
<varlistentry id="ecpg-sql-open-descriptor-name"> |
| 7551 |
<term><replaceable class="parameter">descriptor_name</replaceable></term> |
| 7552 |
<listitem> |
| 7553 |
<para> |
| 7554 |
The name of a descriptor containing values to be bound to the |
| 7555 |
placeholders in the cursor. This can be an SQL identifier or |
| 7556 |
a host variable. |
| 7557 |
</para> |
| 7558 |
</listitem> |
| 7559 |
</varlistentry> |
| 7560 |
</variablelist> |
| 7561 |
</refsect1> |
| 7562 |
|
| 7563 |
<refsect1> |
| 7564 |
<title>Examples</title> |
| 7565 |
|
| 7566 |
<programlisting> |
| 7567 |
EXEC SQL OPEN a; |
| 7568 |
EXEC SQL OPEN d USING 1, 'test'; |
| 7569 |
EXEC SQL OPEN c1 USING SQL DESCRIPTOR mydesc; |
| 7570 |
EXEC SQL OPEN :curname1; |
| 7571 |
</programlisting> |
| 7572 |
</refsect1> |
| 7573 |
|
| 7574 |
<refsect1> |
| 7575 |
<title>Compatibility</title> |
| 7576 |
|
| 7577 |
<para> |
| 7578 |
<command>OPEN</command> is specified in the SQL standard. |
| 7579 |
</para> |
| 7580 |
</refsect1> |
| 7581 |
|
| 7582 |
<refsect1> |
| 7583 |
<title>See Also</title> |
| 7584 |
|
| 7585 |
<simplelist type="inline"> |
| 7586 |
<member><xref linkend="ecpg-sql-declare"/></member> |
| 7587 |
<member><xref linkend="sql-close"/></member> |
| 7588 |
</simplelist> |
| 7589 |
</refsect1> |
| 7590 |
</refentry> |
| 7591 |
|
| 7592 |
<refentry id="ecpg-sql-prepare"> |
| 7593 |
<refnamediv> |
| 7594 |
<refname>PREPARE</refname> |
| 7595 |
<refpurpose>prepare a statement for execution</refpurpose> |
| 7596 |
</refnamediv> |
| 7597 |
|
| 7598 |
<refsynopsisdiv> |
| 7599 |
<synopsis> |
| 7600 |
PREPARE <replaceable class="parameter">prepared_name</replaceable> FROM <replaceable class="parameter">string</replaceable> |
| 7601 |
</synopsis> |
| 7602 |
</refsynopsisdiv> |
| 7603 |
|
| 7604 |
<refsect1> |
| 7605 |
<title>Description</title> |
| 7606 |
|
| 7607 |
<para> |
| 7608 |
<command>PREPARE</command> prepares a statement dynamically |
| 7609 |
specified as a string for execution. This is different from the |
| 7610 |
direct SQL statement <xref linkend="sql-prepare"/>, which can also |
| 7611 |
be used in embedded programs. The <xref linkend="sql-execute"/> |
| 7612 |
command is used to execute either kind of prepared statement. |
| 7613 |
</para> |
| 7614 |
</refsect1> |
| 7615 |
|
| 7616 |
<refsect1> |
| 7617 |
<title>Parameters</title> |
| 7618 |
|
| 7619 |
<variablelist> |
| 7620 |
<varlistentry id="ecpg-sql-prepare-prepared-name"> |
| 7621 |
<term><replaceable class="parameter">prepared_name</replaceable></term> |
| 7622 |
<listitem> |
| 7623 |
<para> |
| 7624 |
An identifier for the prepared query. |
| 7625 |
</para> |
| 7626 |
</listitem> |
| 7627 |
</varlistentry> |
| 7628 |
|
| 7629 |
<varlistentry id="ecpg-sql-prepare-string"> |
| 7630 |
<term><replaceable class="parameter">string</replaceable></term> |
| 7631 |
<listitem> |
| 7632 |
<para> |
| 7633 |
A literal string or a host variable containing a preparable |
| 7634 |
SQL statement, one of SELECT, INSERT, UPDATE, or DELETE. |
| 7635 |
Use question marks (<literal>?</literal>) for parameter values |
| 7636 |
to be supplied at execution. |
| 7637 |
</para> |
| 7638 |
</listitem> |
| 7639 |
</varlistentry> |
| 7640 |
</variablelist> |
| 7641 |
</refsect1> |
| 7642 |
|
| 7643 |
<refsect1> |
| 7644 |
<title>Notes</title> |
| 7645 |
|
| 7646 |
<para> |
| 7647 |
In typical usage, the <replaceable>string</replaceable> is a host |
| 7648 |
variable reference to a string containing a dynamically-constructed |
| 7649 |
SQL statement. The case of a literal string is not very useful; |
| 7650 |
you might as well just write a direct SQL <command>PREPARE</command> |
| 7651 |
statement. |
| 7652 |
</para> |
| 7653 |
|
| 7654 |
<para> |
| 7655 |
If you do use a literal string, keep in mind that any double quotes |
| 7656 |
you might wish to include in the SQL statement must be written as |
| 7657 |
octal escapes (<literal>\042</literal>) not the usual C |
| 7658 |
idiom <literal>\"</literal>. This is because the string is inside |
| 7659 |
an <literal>EXEC SQL</literal> section, so the ECPG lexer parses it |
| 7660 |
according to SQL rules not C rules. Any embedded backslashes will |
| 7661 |
later be handled according to C rules; but <literal>\"</literal> |
| 7662 |
causes an immediate syntax error because it is seen as ending the |
| 7663 |
literal. |
| 7664 |
</para> |
| 7665 |
</refsect1> |
| 7666 |
|
| 7667 |
<refsect1> |
| 7668 |
<title>Examples</title> |
| 7669 |
<programlisting> |
| 7670 |
char *stmt = "SELECT * FROM test1 WHERE a = ? AND b = ?"; |
| 7671 |
|
| 7672 |
EXEC SQL ALLOCATE DESCRIPTOR outdesc; |
| 7673 |
EXEC SQL PREPARE foo FROM :stmt; |
| 7674 |
|
| 7675 |
EXEC SQL EXECUTE foo USING SQL DESCRIPTOR indesc INTO SQL DESCRIPTOR outdesc; |
| 7676 |
</programlisting> |
| 7677 |
</refsect1> |
| 7678 |
|
| 7679 |
<refsect1> |
| 7680 |
<title>Compatibility</title> |
| 7681 |
|
| 7682 |
<para> |
| 7683 |
<command>PREPARE</command> is specified in the SQL standard. |
| 7684 |
</para> |
| 7685 |
</refsect1> |
| 7686 |
|
| 7687 |
<refsect1> |
| 7688 |
<title>See Also</title> |
| 7689 |
|
| 7690 |
<simplelist type="inline"> |
| 7691 |
<member><xref linkend="sql-execute"/></member> |
| 7692 |
</simplelist> |
| 7693 |
</refsect1> |
| 7694 |
</refentry> |
| 7695 |
|
| 7696 |
<refentry id="ecpg-sql-set-autocommit"> |
| 7697 |
<refnamediv> |
| 7698 |
<refname>SET AUTOCOMMIT</refname> |
| 7699 |
<refpurpose>set the autocommit behavior of the current session</refpurpose> |
| 7700 |
</refnamediv> |
| 7701 |
|
| 7702 |
<refsynopsisdiv> |
| 7703 |
<synopsis> |
| 7704 |
SET AUTOCOMMIT { = | TO } { ON | OFF } |
| 7705 |
</synopsis> |
| 7706 |
</refsynopsisdiv> |
| 7707 |
|
| 7708 |
<refsect1> |
| 7709 |
<title>Description</title> |
| 7710 |
|
| 7711 |
<para> |
| 7712 |
<command>SET AUTOCOMMIT</command> sets the autocommit behavior of |
| 7713 |
the current database session. By default, embedded SQL programs |
| 7714 |
are <emphasis>not</emphasis> in autocommit mode, |
| 7715 |
so <command>COMMIT</command> needs to be issued explicitly when |
| 7716 |
desired. This command can change the session to autocommit mode, |
| 7717 |
where each individual statement is committed implicitly. |
| 7718 |
</para> |
| 7719 |
</refsect1> |
| 7720 |
|
| 7721 |
<refsect1> |
| 7722 |
<title>Compatibility</title> |
| 7723 |
|
| 7724 |
<para> |
| 7725 |
<command>SET AUTOCOMMIT</command> is an extension of PostgreSQL ECPG. |
| 7726 |
</para> |
| 7727 |
</refsect1> |
| 7728 |
</refentry> |
| 7729 |
|
| 7730 |
<refentry id="ecpg-sql-set-connection"> |
| 7731 |
<refnamediv> |
| 7732 |
<refname>SET CONNECTION</refname> |
| 7733 |
<refpurpose>select a database connection</refpurpose> |
| 7734 |
</refnamediv> |
| 7735 |
|
| 7736 |
<refsynopsisdiv> |
| 7737 |
<synopsis> |
| 7738 |
SET CONNECTION [ TO | = ] <replaceable class="parameter">connection_name</replaceable> |
| 7739 |
</synopsis> |
| 7740 |
</refsynopsisdiv> |
| 7741 |
|
| 7742 |
<refsect1> |
| 7743 |
<title>Description</title> |
| 7744 |
|
| 7745 |
<para> |
| 7746 |
<command>SET CONNECTION</command> sets the <quote>current</quote> |
| 7747 |
database connection, which is the one that all commands use |
| 7748 |
unless overridden. |
| 7749 |
</para> |
| 7750 |
</refsect1> |
| 7751 |
|
| 7752 |
<refsect1> |
| 7753 |
<title>Parameters</title> |
| 7754 |
|
| 7755 |
<variablelist> |
| 7756 |
<varlistentry id="ecpg-sql-set-connection-connection-name"> |
| 7757 |
<term><replaceable class="parameter">connection_name</replaceable></term> |
| 7758 |
<listitem> |
| 7759 |
<para> |
| 7760 |
A database connection name established by |
| 7761 |
the <command>CONNECT</command> command. |
| 7762 |
</para> |
| 7763 |
</listitem> |
| 7764 |
</varlistentry> |
| 7765 |
|
| 7766 |
<varlistentry id="ecpg-sql-set-connection-current"> |
| 7767 |
<term><literal>CURRENT</literal></term> |
| 7768 |
<listitem> |
| 7769 |
<para> |
| 7770 |
Set the connection to the current connection (thus, nothing happens). |
| 7771 |
</para> |
| 7772 |
</listitem> |
| 7773 |
</varlistentry> |
| 7774 |
</variablelist> |
| 7775 |
</refsect1> |
| 7776 |
|
| 7777 |
<refsect1> |
| 7778 |
<title>Examples</title> |
| 7779 |
|
| 7780 |
<programlisting> |
| 7781 |
EXEC SQL SET CONNECTION TO con2; |
| 7782 |
EXEC SQL SET CONNECTION = con1; |
| 7783 |
</programlisting> |
| 7784 |
</refsect1> |
| 7785 |
|
| 7786 |
<refsect1> |
| 7787 |
<title>Compatibility</title> |
| 7788 |
|
| 7789 |
<para> |
| 7790 |
<command>SET CONNECTION</command> is specified in the SQL standard. |
| 7791 |
</para> |
| 7792 |
</refsect1> |
| 7793 |
|
| 7794 |
<refsect1> |
| 7795 |
<title>See Also</title> |
| 7796 |
|
| 7797 |
<simplelist type="inline"> |
| 7798 |
<member><xref linkend="ecpg-sql-connect"/></member> |
| 7799 |
<member><xref linkend="ecpg-sql-disconnect"/></member> |
| 7800 |
</simplelist> |
| 7801 |
</refsect1> |
| 7802 |
</refentry> |
| 7803 |
|
| 7804 |
<refentry id="ecpg-sql-set-descriptor"> |
| 7805 |
<refnamediv> |
| 7806 |
<refname>SET DESCRIPTOR</refname> |
| 7807 |
<refpurpose>set information in an SQL descriptor area</refpurpose> |
| 7808 |
</refnamediv> |
| 7809 |
|
| 7810 |
<refsynopsisdiv> |
| 7811 |
<synopsis> |
| 7812 |
SET DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> <replaceable class="parameter">descriptor_header_item</replaceable> = <replaceable>value</replaceable> |
| 7813 |
SET DESCRIPTOR <replaceable class="parameter">descriptor_name</replaceable> VALUE <replaceable class="parameter">number</replaceable> <replaceable class="parameter">descriptor_item</replaceable> = <replaceable>value</replaceable> [, ...] |
| 7814 |
</synopsis> |
| 7815 |
</refsynopsisdiv> |
| 7816 |
|
| 7817 |
<refsect1> |
| 7818 |
<title>Description</title> |
| 7819 |
|
| 7820 |
<para> |
| 7821 |
<command>SET DESCRIPTOR</command> populates an SQL descriptor |
| 7822 |
area with values. The descriptor area is then typically used to |
| 7823 |
bind parameters in a prepared query execution. |
| 7824 |
</para> |
| 7825 |
|
| 7826 |
<para> |
| 7827 |
This command has two forms: The first form applies to the |
| 7828 |
descriptor <quote>header</quote>, which is independent of a |
| 7829 |
particular datum. The second form assigns values to particular |
| 7830 |
datums, identified by number. |
| 7831 |
</para> |
| 7832 |
</refsect1> |
| 7833 |
|
| 7834 |
<refsect1> |
| 7835 |
<title>Parameters</title> |
| 7836 |
|
| 7837 |
<variablelist> |
| 7838 |
<varlistentry id="ecpg-sql-set-descriptor-descriptor-name"> |
| 7839 |
<term><replaceable class="parameter">descriptor_name</replaceable></term> |
| 7840 |
<listitem> |
| 7841 |
<para> |
| 7842 |
A descriptor name. |
| 7843 |
</para> |
| 7844 |
</listitem> |
| 7845 |
</varlistentry> |
| 7846 |
|
| 7847 |
<varlistentry id="ecpg-sql-set-descriptor-descriptor-header-item"> |
| 7848 |
<term><replaceable class="parameter">descriptor_header_item</replaceable></term> |
| 7849 |
<listitem> |
| 7850 |
<para> |
| 7851 |
A token identifying which header information item to set. |
| 7852 |
Only <literal>COUNT</literal>, to set the number of descriptor |
| 7853 |
items, is currently supported. |
| 7854 |
</para> |
| 7855 |
</listitem> |
| 7856 |
</varlistentry> |
| 7857 |
|
| 7858 |
<varlistentry id="ecpg-sql-set-descriptor-number"> |
| 7859 |
<term><replaceable class="parameter">number</replaceable></term> |
| 7860 |
<listitem> |
| 7861 |
<para> |
| 7862 |
The number of the descriptor item to set. The count starts at |
| 7863 |
1. |
| 7864 |
</para> |
| 7865 |
</listitem> |
| 7866 |
</varlistentry> |
| 7867 |
|
| 7868 |
<varlistentry id="ecpg-sql-set-descriptor-descriptor-item"> |
| 7869 |
<term><replaceable class="parameter">descriptor_item</replaceable></term> |
| 7870 |
<listitem> |
| 7871 |
<para> |
| 7872 |
A token identifying which item of information to set in the |
| 7873 |
descriptor. See <xref linkend="ecpg-named-descriptors"/> for a |
| 7874 |
list of supported items. |
| 7875 |
</para> |
| 7876 |
</listitem> |
| 7877 |
</varlistentry> |
| 7878 |
|
| 7879 |
<varlistentry id="ecpg-sql-set-descriptor-value"> |
| 7880 |
<term><replaceable class="parameter">value</replaceable></term> |
| 7881 |
<listitem> |
| 7882 |
<para> |
| 7883 |
A value to store into the descriptor item. This can be an SQL |
| 7884 |
constant or a host variable. |
| 7885 |
</para> |
| 7886 |
</listitem> |
| 7887 |
</varlistentry> |
| 7888 |
</variablelist> |
| 7889 |
</refsect1> |
| 7890 |
|
| 7891 |
<refsect1> |
| 7892 |
<title>Examples</title> |
| 7893 |
<programlisting> |
| 7894 |
EXEC SQL SET DESCRIPTOR indesc COUNT = 1; |
| 7895 |
EXEC SQL SET DESCRIPTOR indesc VALUE 1 DATA = 2; |
| 7896 |
EXEC SQL SET DESCRIPTOR indesc VALUE 1 DATA = :val1; |
| 7897 |
EXEC SQL SET DESCRIPTOR indesc VALUE 2 INDICATOR = :val1, DATA = 'some string'; |
| 7898 |
EXEC SQL SET DESCRIPTOR indesc VALUE 2 INDICATOR = :val2null, DATA = :val2; |
| 7899 |
</programlisting> |
| 7900 |
</refsect1> |
| 7901 |
|
| 7902 |
<refsect1> |
| 7903 |
<title>Compatibility</title> |
| 7904 |
|
| 7905 |
<para> |
| 7906 |
<command>SET DESCRIPTOR</command> is specified in the SQL standard. |
| 7907 |
</para> |
| 7908 |
</refsect1> |
| 7909 |
|
| 7910 |
<refsect1> |
| 7911 |
<title>See Also</title> |
| 7912 |
|
| 7913 |
<simplelist type="inline"> |
| 7914 |
<member><xref linkend="ecpg-sql-allocate-descriptor"/></member> |
| 7915 |
<member><xref linkend="ecpg-sql-get-descriptor"/></member> |
| 7916 |
</simplelist> |
| 7917 |
</refsect1> |
| 7918 |
</refentry> |
| 7919 |
|
| 7920 |
<refentry id="ecpg-sql-type"> |
| 7921 |
<refnamediv> |
| 7922 |
<refname>TYPE</refname> |
| 7923 |
<refpurpose>define a new data type</refpurpose> |
| 7924 |
</refnamediv> |
| 7925 |
|
| 7926 |
<refsynopsisdiv> |
| 7927 |
<synopsis> |
| 7928 |
TYPE <replaceable class="parameter">type_name</replaceable> IS <replaceable class="parameter">ctype</replaceable> |
| 7929 |
</synopsis> |
| 7930 |
</refsynopsisdiv> |
| 7931 |
|
| 7932 |
<refsect1> |
| 7933 |
<title>Description</title> |
| 7934 |
|
| 7935 |
<para> |
| 7936 |
The <command>TYPE</command> command defines a new C type. It is |
| 7937 |
equivalent to putting a <literal>typedef</literal> into a declare |
| 7938 |
section. |
| 7939 |
</para> |
| 7940 |
|
| 7941 |
<para> |
| 7942 |
This command is only recognized when <command>ecpg</command> is |
| 7943 |
run with the <option>-c</option> option. |
| 7944 |
</para> |
| 7945 |
</refsect1> |
| 7946 |
|
| 7947 |
<refsect1> |
| 7948 |
<title>Parameters</title> |
| 7949 |
|
| 7950 |
<variablelist> |
| 7951 |
<varlistentry id="ecpg-sql-type-type-name"> |
| 7952 |
<term><replaceable class="parameter">type_name</replaceable></term> |
| 7953 |
<listitem> |
| 7954 |
<para> |
| 7955 |
The name for the new type. It must be a valid C type name. |
| 7956 |
</para> |
| 7957 |
</listitem> |
| 7958 |
</varlistentry> |
| 7959 |
|
| 7960 |
<varlistentry id="ecpg-sql-type-ctype"> |
| 7961 |
<term><replaceable class="parameter">ctype</replaceable></term> |
| 7962 |
<listitem> |
| 7963 |
<para> |
| 7964 |
A C type specification. |
| 7965 |
</para> |
| 7966 |
</listitem> |
| 7967 |
</varlistentry> |
| 7968 |
</variablelist> |
| 7969 |
</refsect1> |
| 7970 |
|
| 7971 |
<refsect1> |
| 7972 |
<title>Examples</title> |
| 7973 |
|
| 7974 |
<programlisting> |
| 7975 |
EXEC SQL TYPE customer IS |
| 7976 |
struct |
| 7977 |
{ |
| 7978 |
varchar name[50]; |
| 7979 |
int phone; |
| 7980 |
}; |
| 7981 |
|
| 7982 |
EXEC SQL TYPE cust_ind IS |
| 7983 |
struct ind |
| 7984 |
{ |
| 7985 |
short name_ind; |
| 7986 |
short phone_ind; |
| 7987 |
}; |
| 7988 |
|
| 7989 |
EXEC SQL TYPE c IS char reference; |
| 7990 |
EXEC SQL TYPE ind IS union { int integer; short smallint; }; |
| 7991 |
EXEC SQL TYPE intarray IS int[AMOUNT]; |
| 7992 |
EXEC SQL TYPE str IS varchar[BUFFERSIZ]; |
| 7993 |
EXEC SQL TYPE string IS char[11]; |
| 7994 |
</programlisting> |
| 7995 |
|
| 7996 |
<para> |
| 7997 |
Here is an example program that uses <command>EXEC SQL |
| 7998 |
TYPE</command>: |
| 7999 |
<programlisting> |
| 8000 |
EXEC SQL WHENEVER SQLERROR SQLPRINT; |
| 8001 |
|
| 8002 |
EXEC SQL TYPE tt IS |
| 8003 |
struct |
| 8004 |
{ |
| 8005 |
varchar v[256]; |
| 8006 |
int i; |
| 8007 |
}; |
| 8008 |
|
| 8009 |
EXEC SQL TYPE tt_ind IS |
| 8010 |
struct ind { |
| 8011 |
short v_ind; |
| 8012 |
short i_ind; |
| 8013 |
}; |
| 8014 |
|
| 8015 |
int |
| 8016 |
main(void) |
| 8017 |
{ |
| 8018 |
EXEC SQL BEGIN DECLARE SECTION; |
| 8019 |
tt t; |
| 8020 |
tt_ind t_ind; |
| 8021 |
EXEC SQL END DECLARE SECTION; |
| 8022 |
|
| 8023 |
EXEC SQL CONNECT TO testdb AS con1; |
| 8024 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 8025 |
|
| 8026 |
EXEC SQL SELECT current_database(), 256 INTO :t:t_ind LIMIT 1; |
| 8027 |
|
| 8028 |
printf("t.v = %s\n", t.v.arr); |
| 8029 |
printf("t.i = %d\n", t.i); |
| 8030 |
|
| 8031 |
printf("t_ind.v_ind = %d\n", t_ind.v_ind); |
| 8032 |
printf("t_ind.i_ind = %d\n", t_ind.i_ind); |
| 8033 |
|
| 8034 |
EXEC SQL DISCONNECT con1; |
| 8035 |
|
| 8036 |
return 0; |
| 8037 |
} |
| 8038 |
</programlisting> |
| 8039 |
|
| 8040 |
The output from this program looks like this: |
| 8041 |
<screen> |
| 8042 |
t.v = testdb |
| 8043 |
t.i = 256 |
| 8044 |
t_ind.v_ind = 0 |
| 8045 |
t_ind.i_ind = 0 |
| 8046 |
</screen> |
| 8047 |
</para> |
| 8048 |
</refsect1> |
| 8049 |
|
| 8050 |
<refsect1> |
| 8051 |
<title>Compatibility</title> |
| 8052 |
|
| 8053 |
<para> |
| 8054 |
The <command>TYPE</command> command is a PostgreSQL extension. |
| 8055 |
</para> |
| 8056 |
</refsect1> |
| 8057 |
</refentry> |
| 8058 |
|
| 8059 |
<refentry id="ecpg-sql-var"> |
| 8060 |
<refnamediv> |
| 8061 |
<refname>VAR</refname> |
| 8062 |
<refpurpose>define a variable</refpurpose> |
| 8063 |
</refnamediv> |
| 8064 |
|
| 8065 |
<refsynopsisdiv> |
| 8066 |
<synopsis> |
| 8067 |
VAR <replaceable>varname</replaceable> IS <replaceable>ctype</replaceable> |
| 8068 |
</synopsis> |
| 8069 |
</refsynopsisdiv> |
| 8070 |
|
| 8071 |
<refsect1> |
| 8072 |
<title>Description</title> |
| 8073 |
|
| 8074 |
<para> |
| 8075 |
The <command>VAR</command> command assigns a new C data type |
| 8076 |
to a host variable. The host variable must be previously |
| 8077 |
declared in a declare section. |
| 8078 |
</para> |
| 8079 |
</refsect1> |
| 8080 |
|
| 8081 |
<refsect1> |
| 8082 |
<title>Parameters</title> |
| 8083 |
|
| 8084 |
<variablelist> |
| 8085 |
<varlistentry id="ecpg-sql-var-varname"> |
| 8086 |
<term><replaceable class="parameter">varname</replaceable></term> |
| 8087 |
<listitem> |
| 8088 |
<para> |
| 8089 |
A C variable name. |
| 8090 |
</para> |
| 8091 |
</listitem> |
| 8092 |
</varlistentry> |
| 8093 |
|
| 8094 |
<varlistentry id="ecpg-sql-var-ctype"> |
| 8095 |
<term><replaceable class="parameter">ctype</replaceable></term> |
| 8096 |
<listitem> |
| 8097 |
<para> |
| 8098 |
A C type specification. |
| 8099 |
</para> |
| 8100 |
</listitem> |
| 8101 |
</varlistentry> |
| 8102 |
</variablelist> |
| 8103 |
</refsect1> |
| 8104 |
|
| 8105 |
<refsect1> |
| 8106 |
<title>Examples</title> |
| 8107 |
|
| 8108 |
<programlisting> |
| 8109 |
Exec sql begin declare section; |
| 8110 |
short a; |
| 8111 |
exec sql end declare section; |
| 8112 |
EXEC SQL VAR a IS int; |
| 8113 |
</programlisting> |
| 8114 |
</refsect1> |
| 8115 |
|
| 8116 |
<refsect1> |
| 8117 |
<title>Compatibility</title> |
| 8118 |
|
| 8119 |
<para> |
| 8120 |
The <command>VAR</command> command is a PostgreSQL extension. |
| 8121 |
</para> |
| 8122 |
</refsect1> |
| 8123 |
</refentry> |
| 8124 |
|
| 8125 |
<refentry id="ecpg-sql-whenever"> |
| 8126 |
<refnamediv> |
| 8127 |
<refname>WHENEVER</refname> |
| 8128 |
<refpurpose>specify the action to be taken when an SQL statement causes a specific class condition to be raised</refpurpose> |
| 8129 |
</refnamediv> |
| 8130 |
|
| 8131 |
<refsynopsisdiv> |
| 8132 |
<synopsis> |
| 8133 |
WHENEVER { NOT FOUND | SQLERROR | SQLWARNING } <replaceable class="parameter">action</replaceable> |
| 8134 |
</synopsis> |
| 8135 |
</refsynopsisdiv> |
| 8136 |
|
| 8137 |
<refsect1> |
| 8138 |
<title>Description</title> |
| 8139 |
|
| 8140 |
<para> |
| 8141 |
Define a behavior which is called on the special cases (Rows not |
| 8142 |
found, SQL warnings or errors) in the result of SQL execution. |
| 8143 |
</para> |
| 8144 |
</refsect1> |
| 8145 |
|
| 8146 |
<refsect1> |
| 8147 |
<title>Parameters</title> |
| 8148 |
|
| 8149 |
<para> |
| 8150 |
See <xref linkend="ecpg-whenever"/> for a description of the |
| 8151 |
parameters. |
| 8152 |
</para> |
| 8153 |
</refsect1> |
| 8154 |
|
| 8155 |
<refsect1> |
| 8156 |
<title>Examples</title> |
| 8157 |
|
| 8158 |
<programlisting> |
| 8159 |
EXEC SQL WHENEVER NOT FOUND CONTINUE; |
| 8160 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 8161 |
EXEC SQL WHENEVER NOT FOUND DO CONTINUE; |
| 8162 |
EXEC SQL WHENEVER SQLWARNING SQLPRINT; |
| 8163 |
EXEC SQL WHENEVER SQLWARNING DO warn(); |
| 8164 |
EXEC SQL WHENEVER SQLERROR sqlprint; |
| 8165 |
EXEC SQL WHENEVER SQLERROR CALL print2(); |
| 8166 |
EXEC SQL WHENEVER SQLERROR DO handle_error("select"); |
| 8167 |
EXEC SQL WHENEVER SQLERROR DO sqlnotice(NULL, NONO); |
| 8168 |
EXEC SQL WHENEVER SQLERROR DO sqlprint(); |
| 8169 |
EXEC SQL WHENEVER SQLERROR GOTO error_label; |
| 8170 |
EXEC SQL WHENEVER SQLERROR STOP; |
| 8171 |
</programlisting> |
| 8172 |
|
| 8173 |
<para> |
| 8174 |
A typical application is the use of <literal>WHENEVER NOT FOUND |
| 8175 |
BREAK</literal> to handle looping through result sets: |
| 8176 |
<programlisting> |
| 8177 |
int |
| 8178 |
main(void) |
| 8179 |
{ |
| 8180 |
EXEC SQL CONNECT TO testdb AS con1; |
| 8181 |
EXEC SQL SELECT pg_catalog.set_config('search_path', '', false); EXEC SQL COMMIT; |
| 8182 |
EXEC SQL ALLOCATE DESCRIPTOR d; |
| 8183 |
EXEC SQL DECLARE cur CURSOR FOR SELECT current_database(), 'hoge', 256; |
| 8184 |
EXEC SQL OPEN cur; |
| 8185 |
|
| 8186 |
/* when end of result set reached, break out of while loop */ |
| 8187 |
EXEC SQL WHENEVER NOT FOUND DO BREAK; |
| 8188 |
|
| 8189 |
while (1) |
| 8190 |
{ |
| 8191 |
EXEC SQL FETCH NEXT FROM cur INTO SQL DESCRIPTOR d; |
| 8192 |
... |
| 8193 |
} |
| 8194 |
|
| 8195 |
EXEC SQL CLOSE cur; |
| 8196 |
EXEC SQL COMMIT; |
| 8197 |
|
| 8198 |
EXEC SQL DEALLOCATE DESCRIPTOR d; |
| 8199 |
EXEC SQL DISCONNECT ALL; |
| 8200 |
|
| 8201 |
return 0; |
| 8202 |
} |
| 8203 |
</programlisting> |
| 8204 |
</para> |
| 8205 |
</refsect1> |
| 8206 |
|
| 8207 |
<refsect1> |
| 8208 |
<title>Compatibility</title> |
| 8209 |
|
| 8210 |
<para> |
| 8211 |
<command>WHENEVER</command> is specified in the SQL standard, but |
| 8212 |
most of the actions are PostgreSQL extensions. |
| 8213 |
</para> |
| 8214 |
</refsect1> |
| 8215 |
</refentry> |
| 8216 |
</sect1> |
| 8217 |
|
| 8218 |
<sect1 id="ecpg-informix-compat"> |
| 8219 |
<title><productname>Informix</productname> Compatibility Mode</title> |
| 8220 |
<para> |
| 8221 |
<command>ecpg</command> can be run in a so-called <firstterm>Informix compatibility mode</firstterm>. If |
| 8222 |
this mode is active, it tries to behave as if it were the <productname>Informix</productname> |
| 8223 |
precompiler for <productname>Informix</productname> E/SQL. Generally spoken this will allow you to use |
| 8224 |
the dollar sign instead of the <literal>EXEC SQL</literal> primitive to introduce |
| 8225 |
embedded SQL commands: |
| 8226 |
<programlisting> |
| 8227 |
$int j = 3; |
| 8228 |
$CONNECT TO :dbname; |
| 8229 |
$CREATE TABLE test(i INT PRIMARY KEY, j INT); |
| 8230 |
$INSERT INTO test(i, j) VALUES (7, :j); |
| 8231 |
$COMMIT; |
| 8232 |
</programlisting> |
| 8233 |
</para> |
| 8234 |
|
| 8235 |
<note> |
| 8236 |
<para> |
| 8237 |
There must not be any white space between the <literal>$</literal> |
| 8238 |
and a following preprocessor directive, that is, |
| 8239 |
<literal>include</literal>, <literal>define</literal>, <literal>ifdef</literal>, |
| 8240 |
etc. Otherwise, the preprocessor will parse the token as a host |
| 8241 |
variable. |
| 8242 |
</para> |
| 8243 |
</note> |
| 8244 |
|
| 8245 |
<para> |
| 8246 |
There are two compatibility modes: <literal>INFORMIX</literal>, <literal>INFORMIX_SE</literal> |
| 8247 |
</para> |
| 8248 |
<para> |
| 8249 |
When linking programs that use this compatibility mode, remember to link |
| 8250 |
against <literal>libcompat</literal> that is shipped with ECPG. |
| 8251 |
</para> |
| 8252 |
<para> |
| 8253 |
Besides the previously explained syntactic sugar, the <productname>Informix</productname> compatibility |
| 8254 |
mode ports some functions for input, output and transformation of data as |
| 8255 |
well as embedded SQL statements known from E/SQL to ECPG. |
| 8256 |
</para> |
| 8257 |
<para> |
| 8258 |
<productname>Informix</productname> compatibility mode is closely connected to the pgtypeslib library |
| 8259 |
of ECPG. pgtypeslib maps SQL data types to data types within the C host |
| 8260 |
program and most of the additional functions of the <productname>Informix</productname> compatibility |
| 8261 |
mode allow you to operate on those C host program types. Note however that |
| 8262 |
the extent of the compatibility is limited. It does not try to copy <productname>Informix</productname> |
| 8263 |
behavior; it allows you to do more or less the same operations and gives |
| 8264 |
you functions that have the same name and the same basic behavior but it is |
| 8265 |
no drop-in replacement if you are using <productname>Informix</productname> at the moment. Moreover, |
| 8266 |
some of the data types are different. For example, |
| 8267 |
<productname>PostgreSQL</productname>'s datetime and interval types do not |
| 8268 |
know about ranges like for example <literal>YEAR TO MINUTE</literal> so you won't |
| 8269 |
find support in ECPG for that either. |
| 8270 |
</para> |
| 8271 |
|
| 8272 |
<sect2 id="ecpg-informix-types"> |
| 8273 |
<title>Additional Types</title> |
| 8274 |
<para> |
| 8275 |
The Informix-special "string" pseudo-type for storing right-trimmed character string data is now |
| 8276 |
supported in Informix-mode without using <literal>typedef</literal>. In fact, in Informix-mode, |
| 8277 |
ECPG refuses to process source files that contain <literal>typedef sometype string;</literal> |
| 8278 |
<programlisting> |
| 8279 |
EXEC SQL BEGIN DECLARE SECTION; |
| 8280 |
string userid; /* this variable will contain trimmed data */ |
| 8281 |
EXEC SQL END DECLARE SECTION; |
| 8282 |
|
| 8283 |
EXEC SQL FETCH MYCUR INTO :userid; |
| 8284 |
</programlisting> |
| 8285 |
</para> |
| 8286 |
</sect2> |
| 8287 |
|
| 8288 |
<sect2 id="ecpg-informix-statements"> |
| 8289 |
<title>Additional/Missing Embedded SQL Statements</title> |
| 8290 |
<para> |
| 8291 |
<variablelist> |
| 8292 |
<varlistentry id="ecpg-informix-statements-close-database"> |
| 8293 |
<term><literal>CLOSE DATABASE</literal></term> |
| 8294 |
<listitem> |
| 8295 |
<para> |
| 8296 |
This statement closes the current connection. In fact, this is a |
| 8297 |
synonym for ECPG's <literal>DISCONNECT CURRENT</literal>: |
| 8298 |
<programlisting> |
| 8299 |
$CLOSE DATABASE; /* close the current connection */ |
| 8300 |
EXEC SQL CLOSE DATABASE; |
| 8301 |
</programlisting> |
| 8302 |
</para> |
| 8303 |
</listitem> |
| 8304 |
</varlistentry> |
| 8305 |
<varlistentry id="ecpg-informix-statements-free-cursor-name"> |
| 8306 |
<term><literal>FREE cursor_name</literal></term> |
| 8307 |
<listitem> |
| 8308 |
<para> |
| 8309 |
Due to differences in how ECPG works compared to Informix's ESQL/C (namely, which steps |
| 8310 |
are purely grammar transformations and which steps rely on the underlying run-time library) |
| 8311 |
there is no <literal>FREE cursor_name</literal> statement in ECPG. This is because in ECPG, |
| 8312 |
<literal>DECLARE CURSOR</literal> doesn't translate to a function call into |
| 8313 |
the run-time library that uses to the cursor name. This means that there's no run-time |
| 8314 |
bookkeeping of SQL cursors in the ECPG run-time library, only in the PostgreSQL server. |
| 8315 |
</para> |
| 8316 |
</listitem> |
| 8317 |
</varlistentry> |
| 8318 |
<varlistentry id="ecpg-informix-statements-free-statement-name"> |
| 8319 |
<term><literal>FREE statement_name</literal></term> |
| 8320 |
<listitem> |
| 8321 |
<para> |
| 8322 |
<literal>FREE statement_name</literal> is a synonym for <literal>DEALLOCATE PREPARE statement_name</literal>. |
| 8323 |
</para> |
| 8324 |
</listitem> |
| 8325 |
</varlistentry> |
| 8326 |
</variablelist> |
| 8327 |
</para> |
| 8328 |
</sect2> |
| 8329 |
|
| 8330 |
<sect2 id="ecpg-informix-sqlda"> |
| 8331 |
<title>Informix-compatible SQLDA Descriptor Areas</title> |
| 8332 |
<para> |
| 8333 |
Informix-compatible mode supports a different structure than the one described in |
| 8334 |
<xref linkend="ecpg-sqlda-descriptors"/>. See below: |
| 8335 |
<programlisting> |
| 8336 |
struct sqlvar_compat |
| 8337 |
{ |
| 8338 |
short sqltype; |
| 8339 |
int sqllen; |
| 8340 |
char *sqldata; |
| 8341 |
short *sqlind; |
| 8342 |
char *sqlname; |
| 8343 |
char *sqlformat; |
| 8344 |
short sqlitype; |
| 8345 |
short sqlilen; |
| 8346 |
char *sqlidata; |
| 8347 |
int sqlxid; |
| 8348 |
char *sqltypename; |
| 8349 |
short sqltypelen; |
| 8350 |
short sqlownerlen; |
| 8351 |
short sqlsourcetype; |
| 8352 |
char *sqlownername; |
| 8353 |
int sqlsourceid; |
| 8354 |
char *sqlilongdata; |
| 8355 |
int sqlflags; |
| 8356 |
void *sqlreserved; |
| 8357 |
}; |
| 8358 |
|
| 8359 |
struct sqlda_compat |
| 8360 |
{ |
| 8361 |
short sqld; |
| 8362 |
struct sqlvar_compat *sqlvar; |
| 8363 |
char desc_name[19]; |
| 8364 |
short desc_occ; |
| 8365 |
struct sqlda_compat *desc_next; |
| 8366 |
void *reserved; |
| 8367 |
}; |
| 8368 |
|
| 8369 |
typedef struct sqlvar_compat sqlvar_t; |
| 8370 |
typedef struct sqlda_compat sqlda_t; |
| 8371 |
</programlisting> |
| 8372 |
</para> |
| 8373 |
|
| 8374 |
<para> |
| 8375 |
The global properties are: |
| 8376 |
<variablelist> |
| 8377 |
|
| 8378 |
<varlistentry id="ecpg-informix-sqlda-sqld"> |
| 8379 |
<term><literal>sqld</literal></term> |
| 8380 |
<listitem> |
| 8381 |
<para> |
| 8382 |
The number of fields in the <literal>SQLDA</literal> descriptor. |
| 8383 |
</para> |
| 8384 |
</listitem> |
| 8385 |
</varlistentry> |
| 8386 |
|
| 8387 |
<varlistentry id="ecpg-informix-sqlda-sqlvar"> |
| 8388 |
<term><literal>sqlvar</literal></term> |
| 8389 |
<listitem> |
| 8390 |
<para> |
| 8391 |
Pointer to the per-field properties. |
| 8392 |
</para> |
| 8393 |
</listitem> |
| 8394 |
</varlistentry> |
| 8395 |
|
| 8396 |
<varlistentry id="ecpg-informix-sqlda-desc-name"> |
| 8397 |
<term><literal>desc_name</literal></term> |
| 8398 |
<listitem> |
| 8399 |
<para> |
| 8400 |
Unused, filled with zero-bytes. |
| 8401 |
</para> |
| 8402 |
</listitem> |
| 8403 |
</varlistentry> |
| 8404 |
|
| 8405 |
<varlistentry id="ecpg-informix-sqlda-desc-occ"> |
| 8406 |
<term><literal>desc_occ</literal></term> |
| 8407 |
<listitem> |
| 8408 |
<para> |
| 8409 |
Size of the allocated structure. |
| 8410 |
</para> |
| 8411 |
</listitem> |
| 8412 |
</varlistentry> |
| 8413 |
|
| 8414 |
<varlistentry id="ecpg-informix-sqlda-desc-next"> |
| 8415 |
<term><literal>desc_next</literal></term> |
| 8416 |
<listitem> |
| 8417 |
<para> |
| 8418 |
Pointer to the next SQLDA structure if the result set contains more than one record. |
| 8419 |
</para> |
| 8420 |
</listitem> |
| 8421 |
</varlistentry> |
| 8422 |
|
| 8423 |
<varlistentry id="ecpg-informix-sqlda-reserved"> |
| 8424 |
<term><literal>reserved</literal></term> |
| 8425 |
<listitem> |
| 8426 |
<para> |
| 8427 |
Unused pointer, contains NULL. Kept for Informix-compatibility. |
| 8428 |
</para> |
| 8429 |
</listitem> |
| 8430 |
</varlistentry> |
| 8431 |
|
| 8432 |
</variablelist> |
| 8433 |
|
| 8434 |
The per-field properties are below, they are stored in the <literal>sqlvar</literal> array: |
| 8435 |
|
| 8436 |
<variablelist> |
| 8437 |
|
| 8438 |
<varlistentry id="ecpg-informix-sqlda-sqltype"> |
| 8439 |
<term><literal>sqltype</literal></term> |
| 8440 |
<listitem> |
| 8441 |
<para> |
| 8442 |
Type of the field. Constants are in <literal>sqltypes.h</literal> |
| 8443 |
</para> |
| 8444 |
</listitem> |
| 8445 |
</varlistentry> |
| 8446 |
|
| 8447 |
<varlistentry id="ecpg-informix-sqlda-sqllen"> |
| 8448 |
<term><literal>sqllen</literal></term> |
| 8449 |
<listitem> |
| 8450 |
<para> |
| 8451 |
Length of the field data. |
| 8452 |
</para> |
| 8453 |
</listitem> |
| 8454 |
</varlistentry> |
| 8455 |
|
| 8456 |
<varlistentry id="ecpg-informix-sqlda-sqldata"> |
| 8457 |
<term><literal>sqldata</literal></term> |
| 8458 |
<listitem> |
| 8459 |
<para> |
| 8460 |
Pointer to the field data. The pointer is of <literal>char *</literal> type, |
| 8461 |
the data pointed by it is in a binary format. Example: |
| 8462 |
<programlisting> |
| 8463 |
int intval; |
| 8464 |
|
| 8465 |
switch (sqldata->sqlvar[i].sqltype) |
| 8466 |
{ |
| 8467 |
case SQLINTEGER: |
| 8468 |
intval = *(int *)sqldata->sqlvar[i].sqldata; |
| 8469 |
break; |
| 8470 |
... |
| 8471 |
} |
| 8472 |
</programlisting> |
| 8473 |
</para> |
| 8474 |
</listitem> |
| 8475 |
</varlistentry> |
| 8476 |
|
| 8477 |
<varlistentry id="ecpg-informix-sqlda-sqlind"> |
| 8478 |
<term><literal>sqlind</literal></term> |
| 8479 |
<listitem> |
| 8480 |
<para> |
| 8481 |
Pointer to the NULL indicator. If returned by DESCRIBE or FETCH then it's always a valid pointer. |
| 8482 |
If used as input for <literal>EXECUTE ... USING sqlda;</literal> then NULL-pointer value means |
| 8483 |
that the value for this field is non-NULL. Otherwise a valid pointer and <literal>sqlitype</literal> |
| 8484 |
has to be properly set. Example: |
| 8485 |
<programlisting> |
| 8486 |
if (*(int2 *)sqldata->sqlvar[i].sqlind != 0) |
| 8487 |
printf("value is NULL\n"); |
| 8488 |
</programlisting> |
| 8489 |
</para> |
| 8490 |
</listitem> |
| 8491 |
</varlistentry> |
| 8492 |
|
| 8493 |
<varlistentry id="ecpg-informix-sqlda-sqlname"> |
| 8494 |
<term><literal>sqlname</literal></term> |
| 8495 |
<listitem> |
| 8496 |
<para> |
| 8497 |
Name of the field. 0-terminated string. |
| 8498 |
</para> |
| 8499 |
</listitem> |
| 8500 |
</varlistentry> |
| 8501 |
|
| 8502 |
<varlistentry id="ecpg-informix-sqlda-sqlformat"> |
| 8503 |
<term><literal>sqlformat</literal></term> |
| 8504 |
<listitem> |
| 8505 |
<para> |
| 8506 |
Reserved in Informix, value of <xref linkend="libpq-PQfformat"/> for the field. |
| 8507 |
</para> |
| 8508 |
</listitem> |
| 8509 |
</varlistentry> |
| 8510 |
|
| 8511 |
<varlistentry id="ecpg-informix-sqlda-sqlitype"> |
| 8512 |
<term><literal>sqlitype</literal></term> |
| 8513 |
<listitem> |
| 8514 |
<para> |
| 8515 |
Type of the NULL indicator data. It's always SQLSMINT when returning data from the server. |
| 8516 |
When the <literal>SQLDA</literal> is used for a parameterized query, the data is treated |
| 8517 |
according to the set type. |
| 8518 |
</para> |
| 8519 |
</listitem> |
| 8520 |
</varlistentry> |
| 8521 |
|
| 8522 |
<varlistentry id="ecpg-informix-sqlda-sqlilen"> |
| 8523 |
<term><literal>sqlilen</literal></term> |
| 8524 |
<listitem> |
| 8525 |
<para> |
| 8526 |
Length of the NULL indicator data. |
| 8527 |
</para> |
| 8528 |
</listitem> |
| 8529 |
</varlistentry> |
| 8530 |
|
| 8531 |
<varlistentry id="ecpg-informix-sqlda-sqlxid"> |
| 8532 |
<term><literal>sqlxid</literal></term> |
| 8533 |
<listitem> |
| 8534 |
<para> |
| 8535 |
Extended type of the field, result of <xref linkend="libpq-PQftype"/>. |
| 8536 |
</para> |
| 8537 |
</listitem> |
| 8538 |
</varlistentry> |
| 8539 |
|
| 8540 |
<varlistentry id="ecpg-informix-sqlda-sqltypename"> |
| 8541 |
<term><literal>sqltypename</literal></term> |
| 8542 |
<term><literal>sqltypelen</literal></term> |
| 8543 |
<term><literal>sqlownerlen</literal></term> |
| 8544 |
<term><literal>sqlsourcetype</literal></term> |
| 8545 |
<term><literal>sqlownername</literal></term> |
| 8546 |
<term><literal>sqlsourceid</literal></term> |
| 8547 |
<term><literal>sqlflags</literal></term> |
| 8548 |
<term><literal>sqlreserved</literal></term> |
| 8549 |
<listitem> |
| 8550 |
<para> |
| 8551 |
Unused. |
| 8552 |
</para> |
| 8553 |
</listitem> |
| 8554 |
</varlistentry> |
| 8555 |
|
| 8556 |
<varlistentry id="ecpg-informix-sqlda-sqlilongdata"> |
| 8557 |
<term><literal>sqlilongdata</literal></term> |
| 8558 |
<listitem> |
| 8559 |
<para> |
| 8560 |
It equals to <literal>sqldata</literal> if <literal>sqllen</literal> is larger than 32kB. |
| 8561 |
</para> |
| 8562 |
</listitem> |
| 8563 |
</varlistentry> |
| 8564 |
|
| 8565 |
</variablelist> |
| 8566 |
|
| 8567 |
Example: |
| 8568 |
<programlisting> |
| 8569 |
EXEC SQL INCLUDE sqlda.h; |
| 8570 |
|
| 8571 |
sqlda_t *sqlda; /* This doesn't need to be under embedded DECLARE SECTION */ |
| 8572 |
|
| 8573 |
EXEC SQL BEGIN DECLARE SECTION; |
| 8574 |
char *prep_stmt = "select * from table1"; |
| 8575 |
int i; |
| 8576 |
EXEC SQL END DECLARE SECTION; |
| 8577 |
|
| 8578 |
... |
| 8579 |
|
| 8580 |
EXEC SQL PREPARE mystmt FROM :prep_stmt; |
| 8581 |
|
| 8582 |
EXEC SQL DESCRIBE mystmt INTO sqlda; |
| 8583 |
|
| 8584 |
printf("# of fields: %d\n", sqlda->sqld); |
| 8585 |
for (i = 0; i < sqlda->sqld; i++) |
| 8586 |
printf("field %d: \"%s\"\n", sqlda->sqlvar[i]->sqlname); |
| 8587 |
|
| 8588 |
EXEC SQL DECLARE mycursor CURSOR FOR mystmt; |
| 8589 |
EXEC SQL OPEN mycursor; |
| 8590 |
EXEC SQL WHENEVER NOT FOUND GOTO out; |
| 8591 |
|
| 8592 |
while (1) |
| 8593 |
{ |
| 8594 |
EXEC SQL FETCH mycursor USING sqlda; |
| 8595 |
} |
| 8596 |
|
| 8597 |
EXEC SQL CLOSE mycursor; |
| 8598 |
|
| 8599 |
free(sqlda); /* The main structure is all to be free(), |
| 8600 |
* sqlda and sqlda->sqlvar is in one allocated area */ |
| 8601 |
</programlisting> |
| 8602 |
For more information, see the <literal>sqlda.h</literal> header and the |
| 8603 |
<literal>src/interfaces/ecpg/test/compat_informix/sqlda.pgc</literal> regression test. |
| 8604 |
</para> |
| 8605 |
</sect2> |
| 8606 |
|
| 8607 |
<sect2 id="ecpg-informix-functions"> |
| 8608 |
<title>Additional Functions</title> |
| 8609 |
<para> |
| 8610 |
<variablelist> |
| 8611 |
<varlistentry id="ecpg-informix-functions-decadd"> |
| 8612 |
<term><function>decadd</function></term> |
| 8613 |
<listitem> |
| 8614 |
<para> |
| 8615 |
Add two decimal type values. |
| 8616 |
<synopsis> |
| 8617 |
int decadd(decimal *arg1, decimal *arg2, decimal *sum); |
| 8618 |
</synopsis> |
| 8619 |
The function receives a pointer to the first operand of type decimal |
| 8620 |
(<literal>arg1</literal>), a pointer to the second operand of type decimal |
| 8621 |
(<literal>arg2</literal>) and a pointer to a value of type decimal that will |
| 8622 |
contain the sum (<literal>sum</literal>). On success, the function returns 0. |
| 8623 |
<symbol>ECPG_INFORMIX_NUM_OVERFLOW</symbol> is returned in case of overflow and |
| 8624 |
<symbol>ECPG_INFORMIX_NUM_UNDERFLOW</symbol> in case of underflow. -1 is returned for |
| 8625 |
other failures and <varname>errno</varname> is set to the respective <varname>errno</varname> number of the |
| 8626 |
pgtypeslib. |
| 8627 |
</para> |
| 8628 |
</listitem> |
| 8629 |
</varlistentry> |
| 8630 |
|
| 8631 |
<varlistentry id="ecpg-informix-functions-deccmp"> |
| 8632 |
<term><function>deccmp</function></term> |
| 8633 |
<listitem> |
| 8634 |
<para> |
| 8635 |
Compare two variables of type decimal. |
| 8636 |
<synopsis> |
| 8637 |
int deccmp(decimal *arg1, decimal *arg2); |
| 8638 |
</synopsis> |
| 8639 |
The function receives a pointer to the first decimal value |
| 8640 |
(<literal>arg1</literal>), a pointer to the second decimal value |
| 8641 |
(<literal>arg2</literal>) and returns an integer value that indicates which is |
| 8642 |
the bigger value. |
| 8643 |
<itemizedlist> |
| 8644 |
<listitem> |
| 8645 |
<para> |
| 8646 |
1, if the value that <literal>arg1</literal> points to is bigger than the |
| 8647 |
value that <literal>var2</literal> points to |
| 8648 |
</para> |
| 8649 |
</listitem> |
| 8650 |
<listitem> |
| 8651 |
<para> |
| 8652 |
-1, if the value that <literal>arg1</literal> points to is smaller than the |
| 8653 |
value that <literal>arg2</literal> points to </para> |
| 8654 |
</listitem> |
| 8655 |
<listitem> |
| 8656 |
<para> |
| 8657 |
0, if the value that <literal>arg1</literal> points to and the value that |
| 8658 |
<literal>arg2</literal> points to are equal |
| 8659 |
</para> |
| 8660 |
</listitem> |
| 8661 |
</itemizedlist> |
| 8662 |
</para> |
| 8663 |
</listitem> |
| 8664 |
</varlistentry> |
| 8665 |
|
| 8666 |
<varlistentry id="ecpg-informix-functions-deccopy"> |
| 8667 |
<term><function>deccopy</function></term> |
| 8668 |
<listitem> |
| 8669 |
<para> |
| 8670 |
Copy a decimal value. |
| 8671 |
<synopsis> |
| 8672 |
void deccopy(decimal *src, decimal *target); |
| 8673 |
</synopsis> |
| 8674 |
The function receives a pointer to the decimal value that should be |
| 8675 |
copied as the first argument (<literal>src</literal>) and a pointer to the |
| 8676 |
target structure of type decimal (<literal>target</literal>) as the second |
| 8677 |
argument. |
| 8678 |
</para> |
| 8679 |
</listitem> |
| 8680 |
</varlistentry> |
| 8681 |
|
| 8682 |
<varlistentry id="ecpg-informix-functions-deccvasc"> |
| 8683 |
<term><function>deccvasc</function></term> |
| 8684 |
<listitem> |
| 8685 |
<para> |
| 8686 |
Convert a value from its ASCII representation into a decimal type. |
| 8687 |
<synopsis> |
| 8688 |
int deccvasc(char *cp, int len, decimal *np); |
| 8689 |
</synopsis> |
| 8690 |
The function receives a pointer to string that contains the string |
| 8691 |
representation of the number to be converted (<literal>cp</literal>) as well |
| 8692 |
as its length <literal>len</literal>. <literal>np</literal> is a pointer to the |
| 8693 |
decimal value that saves the result of the operation. |
| 8694 |
</para> |
| 8695 |
<para> |
| 8696 |
Valid formats are for example: |
| 8697 |
<literal>-2</literal>, |
| 8698 |
<literal>.794</literal>, |
| 8699 |
<literal>+3.44</literal>, |
| 8700 |
<literal>592.49E07</literal> or |
| 8701 |
<literal>-32.84e-4</literal>. |
| 8702 |
</para> |
| 8703 |
<para> |
| 8704 |
The function returns 0 on success. If overflow or underflow occurred, |
| 8705 |
<literal>ECPG_INFORMIX_NUM_OVERFLOW</literal> or |
| 8706 |
<literal>ECPG_INFORMIX_NUM_UNDERFLOW</literal> is returned. If the ASCII |
| 8707 |
representation could not be parsed, |
| 8708 |
<literal>ECPG_INFORMIX_BAD_NUMERIC</literal> is returned or |
| 8709 |
<literal>ECPG_INFORMIX_BAD_EXPONENT</literal> if this problem occurred while |
| 8710 |
parsing the exponent. |
| 8711 |
</para> |
| 8712 |
</listitem> |
| 8713 |
</varlistentry> |
| 8714 |
|
| 8715 |
<varlistentry id="ecpg-informix-functions-deccvdbl"> |
| 8716 |
<term><function>deccvdbl</function></term> |
| 8717 |
<listitem> |
| 8718 |
<para> |
| 8719 |
Convert a value of type double to a value of type decimal. |
| 8720 |
<synopsis> |
| 8721 |
int deccvdbl(double dbl, decimal *np); |
| 8722 |
</synopsis> |
| 8723 |
The function receives the variable of type double that should be |
| 8724 |
converted as its first argument (<literal>dbl</literal>). As the second |
| 8725 |
argument (<literal>np</literal>), the function receives a pointer to the |
| 8726 |
decimal variable that should hold the result of the operation. |
| 8727 |
</para> |
| 8728 |
<para> |
| 8729 |
The function returns 0 on success and a negative value if the |
| 8730 |
conversion failed. |
| 8731 |
</para> |
| 8732 |
</listitem> |
| 8733 |
</varlistentry> |
| 8734 |
|
| 8735 |
<varlistentry id="ecpg-informix-functions-deccvint"> |
| 8736 |
<term><function>deccvint</function></term> |
| 8737 |
<listitem> |
| 8738 |
<para> |
| 8739 |
Convert a value of type int to a value of type decimal. |
| 8740 |
<synopsis> |
| 8741 |
int deccvint(int in, decimal *np); |
| 8742 |
</synopsis> |
| 8743 |
The function receives the variable of type int that should be |
| 8744 |
converted as its first argument (<literal>in</literal>). As the second |
| 8745 |
argument (<literal>np</literal>), the function receives a pointer to the |
| 8746 |
decimal variable that should hold the result of the operation. |
| 8747 |
</para> |
| 8748 |
<para> |
| 8749 |
The function returns 0 on success and a negative value if the |
| 8750 |
conversion failed. |
| 8751 |
</para> |
| 8752 |
</listitem> |
| 8753 |
</varlistentry> |
| 8754 |
|
| 8755 |
<varlistentry id="ecpg-informix-functions-deccvlong"> |
| 8756 |
<term><function>deccvlong</function></term> |
| 8757 |
<listitem> |
| 8758 |
<para> |
| 8759 |
Convert a value of type long to a value of type decimal. |
| 8760 |
<synopsis> |
| 8761 |
int deccvlong(long lng, decimal *np); |
| 8762 |
</synopsis> |
| 8763 |
The function receives the variable of type long that should be |
| 8764 |
converted as its first argument (<literal>lng</literal>). As the second |
| 8765 |
argument (<literal>np</literal>), the function receives a pointer to the |
| 8766 |
decimal variable that should hold the result of the operation. |
| 8767 |
</para> |
| 8768 |
<para> |
| 8769 |
The function returns 0 on success and a negative value if the |
| 8770 |
conversion failed. |
| 8771 |
</para> |
| 8772 |
</listitem> |
| 8773 |
</varlistentry> |
| 8774 |
|
| 8775 |
<varlistentry id="ecpg-informix-functions-decdiv"> |
| 8776 |
<term><function>decdiv</function></term> |
| 8777 |
<listitem> |
| 8778 |
<para> |
| 8779 |
Divide two variables of type decimal. |
| 8780 |
<synopsis> |
| 8781 |
int decdiv(decimal *n1, decimal *n2, decimal *result); |
| 8782 |
</synopsis> |
| 8783 |
The function receives pointers to the variables that are the first |
| 8784 |
(<literal>n1</literal>) and the second (<literal>n2</literal>) operands and |
| 8785 |
calculates <literal>n1</literal>/<literal>n2</literal>. <literal>result</literal> is a |
| 8786 |
pointer to the variable that should hold the result of the operation. |
| 8787 |
</para> |
| 8788 |
<para> |
| 8789 |
On success, 0 is returned and a negative value if the division fails. |
| 8790 |
If overflow or underflow occurred, the function returns |
| 8791 |
<literal>ECPG_INFORMIX_NUM_OVERFLOW</literal> or |
| 8792 |
<literal>ECPG_INFORMIX_NUM_UNDERFLOW</literal> respectively. If an attempt to |
| 8793 |
divide by zero is observed, the function returns |
| 8794 |
<literal>ECPG_INFORMIX_DIVIDE_ZERO</literal>. |
| 8795 |
</para> |
| 8796 |
</listitem> |
| 8797 |
</varlistentry> |
| 8798 |
|
| 8799 |
<varlistentry id="ecpg-informix-functions-decmul"> |
| 8800 |
<term><function>decmul</function></term> |
| 8801 |
<listitem> |
| 8802 |
<para> |
| 8803 |
Multiply two decimal values. |
| 8804 |
<synopsis> |
| 8805 |
int decmul(decimal *n1, decimal *n2, decimal *result); |
| 8806 |
</synopsis> |
| 8807 |
The function receives pointers to the variables that are the first |
| 8808 |
(<literal>n1</literal>) and the second (<literal>n2</literal>) operands and |
| 8809 |
calculates <literal>n1</literal>*<literal>n2</literal>. <literal>result</literal> is a |
| 8810 |
pointer to the variable that should hold the result of the operation. |
| 8811 |
</para> |
| 8812 |
<para> |
| 8813 |
On success, 0 is returned and a negative value if the multiplication |
| 8814 |
fails. If overflow or underflow occurred, the function returns |
| 8815 |
<literal>ECPG_INFORMIX_NUM_OVERFLOW</literal> or |
| 8816 |
<literal>ECPG_INFORMIX_NUM_UNDERFLOW</literal> respectively. |
| 8817 |
</para> |
| 8818 |
</listitem> |
| 8819 |
</varlistentry> |
| 8820 |
|
| 8821 |
<varlistentry id="ecpg-informix-functions-decsub"> |
| 8822 |
<term><function>decsub</function></term> |
| 8823 |
<listitem> |
| 8824 |
<para> |
| 8825 |
Subtract one decimal value from another. |
| 8826 |
<synopsis> |
| 8827 |
int decsub(decimal *n1, decimal *n2, decimal *result); |
| 8828 |
</synopsis> |
| 8829 |
The function receives pointers to the variables that are the first |
| 8830 |
(<literal>n1</literal>) and the second (<literal>n2</literal>) operands and |
| 8831 |
calculates <literal>n1</literal>-<literal>n2</literal>. <literal>result</literal> is a |
| 8832 |
pointer to the variable that should hold the result of the operation. |
| 8833 |
</para> |
| 8834 |
<para> |
| 8835 |
On success, 0 is returned and a negative value if the subtraction |
| 8836 |
fails. If overflow or underflow occurred, the function returns |
| 8837 |
<literal>ECPG_INFORMIX_NUM_OVERFLOW</literal> or |
| 8838 |
<literal>ECPG_INFORMIX_NUM_UNDERFLOW</literal> respectively. |
| 8839 |
</para> |
| 8840 |
</listitem> |
| 8841 |
</varlistentry> |
| 8842 |
|
| 8843 |
<varlistentry id="ecpg-informix-functions-dectoasc"> |
| 8844 |
<term><function>dectoasc</function></term> |
| 8845 |
<listitem> |
| 8846 |
<para> |
| 8847 |
Convert a variable of type decimal to its ASCII representation in a C |
| 8848 |
char* string. |
| 8849 |
<synopsis> |
| 8850 |
int dectoasc(decimal *np, char *cp, int len, int right) |
| 8851 |
</synopsis> |
| 8852 |
The function receives a pointer to a variable of type decimal |
| 8853 |
(<literal>np</literal>) that it converts to its textual representation. |
| 8854 |
<literal>cp</literal> is the buffer that should hold the result of the |
| 8855 |
operation. The parameter <literal>right</literal> specifies, how many digits |
| 8856 |
right of the decimal point should be included in the output. The result |
| 8857 |
will be rounded to this number of decimal digits. Setting |
| 8858 |
<literal>right</literal> to -1 indicates that all available decimal digits |
| 8859 |
should be included in the output. If the length of the output buffer, |
| 8860 |
which is indicated by <literal>len</literal> is not sufficient to hold the |
| 8861 |
textual representation including the trailing zero byte, only a |
| 8862 |
single <literal>*</literal> character is stored in the result and -1 is |
| 8863 |
returned. |
| 8864 |
</para> |
| 8865 |
<para> |
| 8866 |
The function returns either -1 if the buffer <literal>cp</literal> was too |
| 8867 |
small or <literal>ECPG_INFORMIX_OUT_OF_MEMORY</literal> if memory was |
| 8868 |
exhausted. |
| 8869 |
</para> |
| 8870 |
</listitem> |
| 8871 |
</varlistentry> |
| 8872 |
|
| 8873 |
<varlistentry id="ecpg-informix-functions-dectodbl"> |
| 8874 |
<term><function>dectodbl</function></term> |
| 8875 |
<listitem> |
| 8876 |
<para> |
| 8877 |
Convert a variable of type decimal to a double. |
| 8878 |
<synopsis> |
| 8879 |
int dectodbl(decimal *np, double *dblp); |
| 8880 |
</synopsis> |
| 8881 |
The function receives a pointer to the decimal value to convert |
| 8882 |
(<literal>np</literal>) and a pointer to the double variable that |
| 8883 |
should hold the result of the operation (<literal>dblp</literal>). |
| 8884 |
</para> |
| 8885 |
<para> |
| 8886 |
On success, 0 is returned and a negative value if the conversion |
| 8887 |
failed. |
| 8888 |
</para> |
| 8889 |
</listitem> |
| 8890 |
</varlistentry> |
| 8891 |
|
| 8892 |
<varlistentry id="ecpg-informix-functions-dectoint"> |
| 8893 |
<term><function>dectoint</function></term> |
| 8894 |
<listitem> |
| 8895 |
<para> |
| 8896 |
Convert a variable of type decimal to an integer. |
| 8897 |
<synopsis> |
| 8898 |
int dectoint(decimal *np, int *ip); |
| 8899 |
</synopsis> |
| 8900 |
The function receives a pointer to the decimal value to convert |
| 8901 |
(<literal>np</literal>) and a pointer to the integer variable that |
| 8902 |
should hold the result of the operation (<literal>ip</literal>). |
| 8903 |
</para> |
| 8904 |
<para> |
| 8905 |
On success, 0 is returned and a negative value if the conversion |
| 8906 |
failed. If an overflow occurred, <literal>ECPG_INFORMIX_NUM_OVERFLOW</literal> |
| 8907 |
is returned. |
| 8908 |
</para> |
| 8909 |
<para> |
| 8910 |
Note that the ECPG implementation differs from the <productname>Informix</productname> |
| 8911 |
implementation. <productname>Informix</productname> limits an integer to the range from -32767 to |
| 8912 |
32767, while the limits in the ECPG implementation depend on the |
| 8913 |
architecture (<literal>INT_MIN .. INT_MAX</literal>). |
| 8914 |
</para> |
| 8915 |
</listitem> |
| 8916 |
</varlistentry> |
| 8917 |
|
| 8918 |
<varlistentry id="ecpg-informix-functions-dectolong"> |
| 8919 |
<term><function>dectolong</function></term> |
| 8920 |
<listitem> |
| 8921 |
<para> |
| 8922 |
Convert a variable of type decimal to a long integer. |
| 8923 |
<synopsis> |
| 8924 |
int dectolong(decimal *np, long *lngp); |
| 8925 |
</synopsis> |
| 8926 |
The function receives a pointer to the decimal value to convert |
| 8927 |
(<literal>np</literal>) and a pointer to the long variable that |
| 8928 |
should hold the result of the operation (<literal>lngp</literal>). |
| 8929 |
</para> |
| 8930 |
<para> |
| 8931 |
On success, 0 is returned and a negative value if the conversion |
| 8932 |
failed. If an overflow occurred, <literal>ECPG_INFORMIX_NUM_OVERFLOW</literal> |
| 8933 |
is returned. |
| 8934 |
</para> |
| 8935 |
<para> |
| 8936 |
Note that the ECPG implementation differs from the <productname>Informix</productname> |
| 8937 |
implementation. <productname>Informix</productname> limits a long integer to the range from |
| 8938 |
-2,147,483,647 to 2,147,483,647, while the limits in the ECPG |
| 8939 |
implementation depend on the architecture (<literal>-LONG_MAX .. |
| 8940 |
LONG_MAX</literal>). |
| 8941 |
</para> |
| 8942 |
</listitem> |
| 8943 |
</varlistentry> |
| 8944 |
|
| 8945 |
<varlistentry id="ecpg-informix-functions-rdatestr"> |
| 8946 |
<term><function>rdatestr</function></term> |
| 8947 |
<listitem> |
| 8948 |
<para> |
| 8949 |
Converts a date to a C char* string. |
| 8950 |
<synopsis> |
| 8951 |
int rdatestr(date d, char *str); |
| 8952 |
</synopsis> |
| 8953 |
The function receives two arguments, the first one is the date to |
| 8954 |
convert (<literal>d</literal>) and the second one is a pointer to the target |
| 8955 |
string. The output format is always <literal>yyyy-mm-dd</literal>, so you need |
| 8956 |
to allocate at least 11 bytes (including the zero-byte terminator) for the |
| 8957 |
string. |
| 8958 |
</para> |
| 8959 |
<para> |
| 8960 |
The function returns 0 on success and a negative value in case of |
| 8961 |
error. |
| 8962 |
</para> |
| 8963 |
<para> |
| 8964 |
Note that ECPG's implementation differs from the <productname>Informix</productname> |
| 8965 |
implementation. In <productname>Informix</productname> the format can be influenced by setting |
| 8966 |
environment variables. In ECPG however, you cannot change the output |
| 8967 |
format. |
| 8968 |
</para> |
| 8969 |
</listitem> |
| 8970 |
</varlistentry> |
| 8971 |
|
| 8972 |
<varlistentry id="ecpg-informix-functions-rstrdate"> |
| 8973 |
<term><function>rstrdate</function></term> |
| 8974 |
<listitem> |
| 8975 |
<para> |
| 8976 |
Parse the textual representation of a date. |
| 8977 |
<synopsis> |
| 8978 |
int rstrdate(char *str, date *d); |
| 8979 |
</synopsis> |
| 8980 |
The function receives the textual representation of the date to convert |
| 8981 |
(<literal>str</literal>) and a pointer to a variable of type date |
| 8982 |
(<literal>d</literal>). This function does not allow you to specify a format |
| 8983 |
mask. It uses the default format mask of <productname>Informix</productname> which is |
| 8984 |
<literal>mm/dd/yyyy</literal>. Internally, this function is implemented by |
| 8985 |
means of <function>rdefmtdate</function>. Therefore, <function>rstrdate</function> is |
| 8986 |
not faster and if you have the choice you should opt for |
| 8987 |
<function>rdefmtdate</function> which allows you to specify the format mask |
| 8988 |
explicitly. |
| 8989 |
</para> |
| 8990 |
<para> |
| 8991 |
The function returns the same values as <function>rdefmtdate</function>. |
| 8992 |
</para> |
| 8993 |
</listitem> |
| 8994 |
</varlistentry> |
| 8995 |
|
| 8996 |
<varlistentry id="ecpg-informix-functions-rtoday"> |
| 8997 |
<term><function>rtoday</function></term> |
| 8998 |
<listitem> |
| 8999 |
<para> |
| 9000 |
Get the current date. |
| 9001 |
<synopsis> |
| 9002 |
void rtoday(date *d); |
| 9003 |
</synopsis> |
| 9004 |
The function receives a pointer to a date variable (<literal>d</literal>) |
| 9005 |
that it sets to the current date. |
| 9006 |
</para> |
| 9007 |
<para> |
| 9008 |
Internally this function uses the <xref linkend="pgtypesdatetoday"/> |
| 9009 |
function. |
| 9010 |
</para> |
| 9011 |
</listitem> |
| 9012 |
</varlistentry> |
| 9013 |
|
| 9014 |
<varlistentry id="ecpg-informix-functions-rjulmdy"> |
| 9015 |
<term><function>rjulmdy</function></term> |
| 9016 |
<listitem> |
| 9017 |
<para> |
| 9018 |
Extract the values for the day, the month and the year from a variable |
| 9019 |
of type date. |
| 9020 |
<synopsis> |
| 9021 |
int rjulmdy(date d, short mdy[3]); |
| 9022 |
</synopsis> |
| 9023 |
The function receives the date <literal>d</literal> and a pointer to an array |
| 9024 |
of 3 short integer values <literal>mdy</literal>. The variable name indicates |
| 9025 |
the sequential order: <literal>mdy[0]</literal> will be set to contain the |
| 9026 |
number of the month, <literal>mdy[1]</literal> will be set to the value of the |
| 9027 |
day and <literal>mdy[2]</literal> will contain the year. |
| 9028 |
</para> |
| 9029 |
<para> |
| 9030 |
The function always returns 0 at the moment. |
| 9031 |
</para> |
| 9032 |
<para> |
| 9033 |
Internally the function uses the <xref linkend="pgtypesdatejulmdy"/> |
| 9034 |
function. |
| 9035 |
</para> |
| 9036 |
</listitem> |
| 9037 |
</varlistentry> |
| 9038 |
|
| 9039 |
<varlistentry id="ecpg-informix-functions-rdefmtdate"> |
| 9040 |
<term><function>rdefmtdate</function></term> |
| 9041 |
<listitem> |
| 9042 |
<para> |
| 9043 |
Use a format mask to convert a character string to a value of type |
| 9044 |
date. |
| 9045 |
<synopsis> |
| 9046 |
int rdefmtdate(date *d, char *fmt, char *str); |
| 9047 |
</synopsis> |
| 9048 |
The function receives a pointer to the date value that should hold the |
| 9049 |
result of the operation (<literal>d</literal>), the format mask to use for |
| 9050 |
parsing the date (<literal>fmt</literal>) and the C char* string containing |
| 9051 |
the textual representation of the date (<literal>str</literal>). The textual |
| 9052 |
representation is expected to match the format mask. However you do not |
| 9053 |
need to have a 1:1 mapping of the string to the format mask. The |
| 9054 |
function only analyzes the sequential order and looks for the literals |
| 9055 |
<literal>yy</literal> or <literal>yyyy</literal> that indicate the |
| 9056 |
position of the year, <literal>mm</literal> to indicate the position of |
| 9057 |
the month and <literal>dd</literal> to indicate the position of the |
| 9058 |
day. |
| 9059 |
</para> |
| 9060 |
<para> |
| 9061 |
The function returns the following values: |
| 9062 |
<itemizedlist> |
| 9063 |
<listitem> |
| 9064 |
<para> |
| 9065 |
0 - The function terminated successfully. |
| 9066 |
</para> |
| 9067 |
</listitem> |
| 9068 |
<listitem> |
| 9069 |
<para> |
| 9070 |
<literal>ECPG_INFORMIX_ENOSHORTDATE</literal> - The date does not contain |
| 9071 |
delimiters between day, month and year. In this case the input |
| 9072 |
string must be exactly 6 or 8 bytes long but isn't. |
| 9073 |
</para> |
| 9074 |
</listitem> |
| 9075 |
<listitem> |
| 9076 |
<para> |
| 9077 |
<literal>ECPG_INFORMIX_ENOTDMY</literal> - The format string did not |
| 9078 |
correctly indicate the sequential order of year, month and day. |
| 9079 |
</para> |
| 9080 |
</listitem> |
| 9081 |
<listitem> |
| 9082 |
<para> |
| 9083 |
<literal>ECPG_INFORMIX_BAD_DAY</literal> - The input string does not |
| 9084 |
contain a valid day. |
| 9085 |
</para> |
| 9086 |
</listitem> |
| 9087 |
<listitem> |
| 9088 |
<para> |
| 9089 |
<literal>ECPG_INFORMIX_BAD_MONTH</literal> - The input string does not |
| 9090 |
contain a valid month. |
| 9091 |
</para> |
| 9092 |
</listitem> |
| 9093 |
<listitem> |
| 9094 |
<para> |
| 9095 |
<literal>ECPG_INFORMIX_BAD_YEAR</literal> - The input string does not |
| 9096 |
contain a valid year. |
| 9097 |
</para> |
| 9098 |
</listitem> |
| 9099 |
</itemizedlist> |
| 9100 |
</para> |
| 9101 |
<para> |
| 9102 |
Internally this function is implemented to use the <xref |
| 9103 |
linkend="pgtypesdatedefmtasc"/> function. See the reference there for a |
| 9104 |
table of example input. |
| 9105 |
</para> |
| 9106 |
</listitem> |
| 9107 |
</varlistentry> |
| 9108 |
|
| 9109 |
<varlistentry id="ecpg-informix-functions-rfmtdate"> |
| 9110 |
<term><function>rfmtdate</function></term> |
| 9111 |
<listitem> |
| 9112 |
<para> |
| 9113 |
Convert a variable of type date to its textual representation using a |
| 9114 |
format mask. |
| 9115 |
<synopsis> |
| 9116 |
int rfmtdate(date d, char *fmt, char *str); |
| 9117 |
</synopsis> |
| 9118 |
The function receives the date to convert (<literal>d</literal>), the format |
| 9119 |
mask (<literal>fmt</literal>) and the string that will hold the textual |
| 9120 |
representation of the date (<literal>str</literal>). |
| 9121 |
</para> |
| 9122 |
<para> |
| 9123 |
On success, 0 is returned and a negative value if an error occurred. |
| 9124 |
</para> |
| 9125 |
<para> |
| 9126 |
Internally this function uses the <xref linkend="pgtypesdatefmtasc"/> |
| 9127 |
function, see the reference there for examples. |
| 9128 |
</para> |
| 9129 |
</listitem> |
| 9130 |
</varlistentry> |
| 9131 |
|
| 9132 |
<varlistentry id="ecpg-informix-functions-rmdyjul"> |
| 9133 |
<term><function>rmdyjul</function></term> |
| 9134 |
<listitem> |
| 9135 |
<para> |
| 9136 |
Create a date value from an array of 3 short integers that specify the |
| 9137 |
day, the month and the year of the date. |
| 9138 |
<synopsis> |
| 9139 |
int rmdyjul(short mdy[3], date *d); |
| 9140 |
</synopsis> |
| 9141 |
The function receives the array of the 3 short integers |
| 9142 |
(<literal>mdy</literal>) and a pointer to a variable of type date that should |
| 9143 |
hold the result of the operation. |
| 9144 |
</para> |
| 9145 |
<para> |
| 9146 |
Currently the function returns always 0. |
| 9147 |
</para> |
| 9148 |
<para> |
| 9149 |
Internally the function is implemented to use the function <xref |
| 9150 |
linkend="pgtypesdatemdyjul"/>. |
| 9151 |
</para> |
| 9152 |
</listitem> |
| 9153 |
</varlistentry> |
| 9154 |
|
| 9155 |
<varlistentry id="ecpg-informix-functions-rdayofweek"> |
| 9156 |
<term><function>rdayofweek</function></term> |
| 9157 |
<listitem> |
| 9158 |
<para> |
| 9159 |
Return a number representing the day of the week for a date value. |
| 9160 |
<synopsis> |
| 9161 |
int rdayofweek(date d); |
| 9162 |
</synopsis> |
| 9163 |
The function receives the date variable <literal>d</literal> as its only |
| 9164 |
argument and returns an integer that indicates the day of the week for |
| 9165 |
this date. |
| 9166 |
<itemizedlist> |
| 9167 |
<listitem> |
| 9168 |
<para> |
| 9169 |
0 - Sunday |
| 9170 |
</para> |
| 9171 |
</listitem> |
| 9172 |
<listitem> |
| 9173 |
<para> |
| 9174 |
1 - Monday |
| 9175 |
</para> |
| 9176 |
</listitem> |
| 9177 |
<listitem> |
| 9178 |
<para> |
| 9179 |
2 - Tuesday |
| 9180 |
</para> |
| 9181 |
</listitem> |
| 9182 |
<listitem> |
| 9183 |
<para> |
| 9184 |
3 - Wednesday |
| 9185 |
</para> |
| 9186 |
</listitem> |
| 9187 |
<listitem> |
| 9188 |
<para> |
| 9189 |
4 - Thursday |
| 9190 |
</para> |
| 9191 |
</listitem> |
| 9192 |
<listitem> |
| 9193 |
<para> |
| 9194 |
5 - Friday |
| 9195 |
</para> |
| 9196 |
</listitem> |
| 9197 |
<listitem> |
| 9198 |
<para> |
| 9199 |
6 - Saturday |
| 9200 |
</para> |
| 9201 |
</listitem> |
| 9202 |
</itemizedlist> |
| 9203 |
</para> |
| 9204 |
<para> |
| 9205 |
Internally the function is implemented to use the function <xref |
| 9206 |
linkend="pgtypesdatedayofweek"/>. |
| 9207 |
</para> |
| 9208 |
</listitem> |
| 9209 |
</varlistentry> |
| 9210 |
|
| 9211 |
<varlistentry id="ecpg-informix-functions-dtcurrent"> |
| 9212 |
<term><function>dtcurrent</function></term> |
| 9213 |
<listitem> |
| 9214 |
<para> |
| 9215 |
Retrieve the current timestamp. |
| 9216 |
<synopsis> |
| 9217 |
void dtcurrent(timestamp *ts); |
| 9218 |
</synopsis> |
| 9219 |
The function retrieves the current timestamp and saves it into the |
| 9220 |
timestamp variable that <literal>ts</literal> points to. |
| 9221 |
</para> |
| 9222 |
</listitem> |
| 9223 |
</varlistentry> |
| 9224 |
|
| 9225 |
<varlistentry id="ecpg-informix-functions-dtcvasc"> |
| 9226 |
<term><function>dtcvasc</function></term> |
| 9227 |
<listitem> |
| 9228 |
<para> |
| 9229 |
Parses a timestamp from its textual representation |
| 9230 |
into a timestamp variable. |
| 9231 |
<synopsis> |
| 9232 |
int dtcvasc(char *str, timestamp *ts); |
| 9233 |
</synopsis> |
| 9234 |
The function receives the string to parse (<literal>str</literal>) and a |
| 9235 |
pointer to the timestamp variable that should hold the result of the |
| 9236 |
operation (<literal>ts</literal>). |
| 9237 |
</para> |
| 9238 |
<para> |
| 9239 |
The function returns 0 on success and a negative value in case of |
| 9240 |
error. |
| 9241 |
</para> |
| 9242 |
<para> |
| 9243 |
Internally this function uses the <xref |
| 9244 |
linkend="pgtypestimestampfromasc"/> function. See the reference there |
| 9245 |
for a table with example inputs. |
| 9246 |
</para> |
| 9247 |
</listitem> |
| 9248 |
</varlistentry> |
| 9249 |
|
| 9250 |
<varlistentry id="ecpg-informix-functions-dtcvfmtasc"> |
| 9251 |
<term><function>dtcvfmtasc</function></term> |
| 9252 |
<listitem> |
| 9253 |
<para> |
| 9254 |
Parses a timestamp from its textual representation |
| 9255 |
using a format mask into a timestamp variable. |
| 9256 |
<synopsis> |
| 9257 |
dtcvfmtasc(char *inbuf, char *fmtstr, timestamp *dtvalue) |
| 9258 |
</synopsis> |
| 9259 |
The function receives the string to parse (<literal>inbuf</literal>), the |
| 9260 |
format mask to use (<literal>fmtstr</literal>) and a pointer to the timestamp |
| 9261 |
variable that should hold the result of the operation |
| 9262 |
(<literal>dtvalue</literal>). |
| 9263 |
</para> |
| 9264 |
<para> |
| 9265 |
This function is implemented by means of the <xref |
| 9266 |
linkend="pgtypestimestampdefmtasc"/> function. See the documentation |
| 9267 |
there for a list of format specifiers that can be used. |
| 9268 |
</para> |
| 9269 |
<para> |
| 9270 |
The function returns 0 on success and a negative value in case of |
| 9271 |
error. |
| 9272 |
</para> |
| 9273 |
</listitem> |
| 9274 |
</varlistentry> |
| 9275 |
|
| 9276 |
<varlistentry id="ecpg-informix-functions-dtsub"> |
| 9277 |
<term><function>dtsub</function></term> |
| 9278 |
<listitem> |
| 9279 |
<para> |
| 9280 |
Subtract one timestamp from another and return a variable of type |
| 9281 |
interval. |
| 9282 |
<synopsis> |
| 9283 |
int dtsub(timestamp *ts1, timestamp *ts2, interval *iv); |
| 9284 |
</synopsis> |
| 9285 |
The function will subtract the timestamp variable that <literal>ts2</literal> |
| 9286 |
points to from the timestamp variable that <literal>ts1</literal> points to |
| 9287 |
and will store the result in the interval variable that <literal>iv</literal> |
| 9288 |
points to. |
| 9289 |
</para> |
| 9290 |
<para> |
| 9291 |
Upon success, the function returns 0 and a negative value if an |
| 9292 |
error occurred. |
| 9293 |
</para> |
| 9294 |
</listitem> |
| 9295 |
</varlistentry> |
| 9296 |
|
| 9297 |
<varlistentry id="ecpg-informix-functions-dttoasc"> |
| 9298 |
<term><function>dttoasc</function></term> |
| 9299 |
<listitem> |
| 9300 |
<para> |
| 9301 |
Convert a timestamp variable to a C char* string. |
| 9302 |
<synopsis> |
| 9303 |
int dttoasc(timestamp *ts, char *output); |
| 9304 |
</synopsis> |
| 9305 |
The function receives a pointer to the timestamp variable to convert |
| 9306 |
(<literal>ts</literal>) and the string that should hold the result of the |
| 9307 |
operation (<literal>output</literal>). It converts <literal>ts</literal> to its |
| 9308 |
textual representation according to the SQL standard, which is |
| 9309 |
be <literal>YYYY-MM-DD HH:MM:SS</literal>. |
| 9310 |
</para> |
| 9311 |
<para> |
| 9312 |
Upon success, the function returns 0 and a negative value if an |
| 9313 |
error occurred. |
| 9314 |
</para> |
| 9315 |
</listitem> |
| 9316 |
</varlistentry> |
| 9317 |
|
| 9318 |
<varlistentry id="ecpg-informix-functions-dttofmtasc"> |
| 9319 |
<term><function>dttofmtasc</function></term> |
| 9320 |
<listitem> |
| 9321 |
<para> |
| 9322 |
Convert a timestamp variable to a C char* using a format mask. |
| 9323 |
<synopsis> |
| 9324 |
int dttofmtasc(timestamp *ts, char *output, int str_len, char *fmtstr); |
| 9325 |
</synopsis> |
| 9326 |
The function receives a pointer to the timestamp to convert as its |
| 9327 |
first argument (<literal>ts</literal>), a pointer to the output buffer |
| 9328 |
(<literal>output</literal>), the maximal length that has been allocated for |
| 9329 |
the output buffer (<literal>str_len</literal>) and the format mask to |
| 9330 |
use for the conversion (<literal>fmtstr</literal>). |
| 9331 |
</para> |
| 9332 |
<para> |
| 9333 |
Upon success, the function returns 0 and a negative value if an |
| 9334 |
error occurred. |
| 9335 |
</para> |
| 9336 |
<para> |
| 9337 |
Internally, this function uses the <xref |
| 9338 |
linkend="pgtypestimestampfmtasc"/> function. See the reference there for |
| 9339 |
information on what format mask specifiers can be used. |
| 9340 |
</para> |
| 9341 |
</listitem> |
| 9342 |
</varlistentry> |
| 9343 |
|
| 9344 |
<varlistentry id="ecpg-informix-functions-intoasc"> |
| 9345 |
<term><function>intoasc</function></term> |
| 9346 |
<listitem> |
| 9347 |
<para> |
| 9348 |
Convert an interval variable to a C char* string. |
| 9349 |
<synopsis> |
| 9350 |
int intoasc(interval *i, char *str); |
| 9351 |
</synopsis> |
| 9352 |
The function receives a pointer to the interval variable to convert |
| 9353 |
(<literal>i</literal>) and the string that should hold the result of the |
| 9354 |
operation (<literal>str</literal>). It converts <literal>i</literal> to its |
| 9355 |
textual representation according to the SQL standard, which is |
| 9356 |
be <literal>YYYY-MM-DD HH:MM:SS</literal>. |
| 9357 |
</para> |
| 9358 |
<para> |
| 9359 |
Upon success, the function returns 0 and a negative value if an |
| 9360 |
error occurred. |
| 9361 |
</para> |
| 9362 |
</listitem> |
| 9363 |
</varlistentry> |
| 9364 |
|
| 9365 |
<varlistentry id="ecpg-informix-functions-rfmtlong"> |
| 9366 |
<term><function>rfmtlong</function></term> |
| 9367 |
<listitem> |
| 9368 |
<para> |
| 9369 |
Convert a long integer value to its textual representation using a |
| 9370 |
format mask. |
| 9371 |
<synopsis> |
| 9372 |
int rfmtlong(long lng_val, char *fmt, char *outbuf); |
| 9373 |
</synopsis> |
| 9374 |
The function receives the long value <literal>lng_val</literal>, the format |
| 9375 |
mask <literal>fmt</literal> and a pointer to the output buffer |
| 9376 |
<literal>outbuf</literal>. It converts the long value according to the format |
| 9377 |
mask to its textual representation. |
| 9378 |
</para> |
| 9379 |
<para> |
| 9380 |
The format mask can be composed of the following format specifying |
| 9381 |
characters: |
| 9382 |
<itemizedlist> |
| 9383 |
<listitem> |
| 9384 |
<para> |
| 9385 |
<literal>*</literal> (asterisk) - if this position would be blank |
| 9386 |
otherwise, fill it with an asterisk. |
| 9387 |
</para> |
| 9388 |
</listitem> |
| 9389 |
<listitem> |
| 9390 |
<para> |
| 9391 |
<literal>&</literal> (ampersand) - if this position would be |
| 9392 |
blank otherwise, fill it with a zero. |
| 9393 |
</para> |
| 9394 |
</listitem> |
| 9395 |
<listitem> |
| 9396 |
<para> |
| 9397 |
<literal>#</literal> - turn leading zeroes into blanks. |
| 9398 |
</para> |
| 9399 |
</listitem> |
| 9400 |
<listitem> |
| 9401 |
<para> |
| 9402 |
<literal><</literal> - left-justify the number in the string. |
| 9403 |
</para> |
| 9404 |
</listitem> |
| 9405 |
<listitem> |
| 9406 |
<para> |
| 9407 |
<literal>,</literal> (comma) - group numbers of four or more digits |
| 9408 |
into groups of three digits separated by a comma. |
| 9409 |
</para> |
| 9410 |
</listitem> |
| 9411 |
<listitem> |
| 9412 |
<para> |
| 9413 |
<literal>.</literal> (period) - this character separates the |
| 9414 |
whole-number part of the number from the fractional part. |
| 9415 |
</para> |
| 9416 |
</listitem> |
| 9417 |
<listitem> |
| 9418 |
<para> |
| 9419 |
<literal>-</literal> (minus) - the minus sign appears if the number |
| 9420 |
is a negative value. |
| 9421 |
</para> |
| 9422 |
</listitem> |
| 9423 |
<listitem> |
| 9424 |
<para> |
| 9425 |
<literal>+</literal> (plus) - the plus sign appears if the number is |
| 9426 |
a positive value. |
| 9427 |
</para> |
| 9428 |
</listitem> |
| 9429 |
<listitem> |
| 9430 |
<para> |
| 9431 |
<literal>(</literal> - this replaces the minus sign in front of the |
| 9432 |
negative number. The minus sign will not appear. |
| 9433 |
</para> |
| 9434 |
</listitem> |
| 9435 |
<listitem> |
| 9436 |
<para> |
| 9437 |
<literal>)</literal> - this character replaces the minus and is |
| 9438 |
printed behind the negative value. |
| 9439 |
</para> |
| 9440 |
</listitem> |
| 9441 |
<listitem> |
| 9442 |
<para> |
| 9443 |
<literal>$</literal> - the currency symbol. |
| 9444 |
</para> |
| 9445 |
</listitem> |
| 9446 |
</itemizedlist> |
| 9447 |
</para> |
| 9448 |
</listitem> |
| 9449 |
</varlistentry> |
| 9450 |
|
| 9451 |
<varlistentry id="ecpg-informix-functions-rupshift"> |
| 9452 |
<term><function>rupshift</function></term> |
| 9453 |
<listitem> |
| 9454 |
<para> |
| 9455 |
Convert a string to upper case. |
| 9456 |
<synopsis> |
| 9457 |
void rupshift(char *str); |
| 9458 |
</synopsis> |
| 9459 |
The function receives a pointer to the string and transforms every |
| 9460 |
lower case character to upper case. |
| 9461 |
</para> |
| 9462 |
</listitem> |
| 9463 |
</varlistentry> |
| 9464 |
|
| 9465 |
<varlistentry id="ecpg-informix-functions-byleng"> |
| 9466 |
<term><function>byleng</function></term> |
| 9467 |
<listitem> |
| 9468 |
<para> |
| 9469 |
Return the number of characters in a string without counting trailing |
| 9470 |
blanks. |
| 9471 |
<synopsis> |
| 9472 |
int byleng(char *str, int len); |
| 9473 |
</synopsis> |
| 9474 |
The function expects a fixed-length string as its first argument |
| 9475 |
(<literal>str</literal>) and its length as its second argument |
| 9476 |
(<literal>len</literal>). It returns the number of significant characters, |
| 9477 |
that is the length of the string without trailing blanks. |
| 9478 |
</para> |
| 9479 |
</listitem> |
| 9480 |
</varlistentry> |
| 9481 |
|
| 9482 |
<varlistentry id="ecpg-informix-functions-ldchar"> |
| 9483 |
<term><function>ldchar</function></term> |
| 9484 |
<listitem> |
| 9485 |
<para> |
| 9486 |
Copy a fixed-length string into a null-terminated string. |
| 9487 |
<synopsis> |
| 9488 |
void ldchar(char *src, int len, char *dest); |
| 9489 |
</synopsis> |
| 9490 |
The function receives the fixed-length string to copy |
| 9491 |
(<literal>src</literal>), its length (<literal>len</literal>) and a pointer to the |
| 9492 |
destination memory (<literal>dest</literal>). Note that you need to reserve at |
| 9493 |
least <literal>len+1</literal> bytes for the string that <literal>dest</literal> |
| 9494 |
points to. The function copies at most <literal>len</literal> bytes to the new |
| 9495 |
location (less if the source string has trailing blanks) and adds the |
| 9496 |
null-terminator. |
| 9497 |
</para> |
| 9498 |
</listitem> |
| 9499 |
</varlistentry> |
| 9500 |
|
| 9501 |
<varlistentry id="ecpg-informix-functions-rgetmsg"> |
| 9502 |
<term><function>rgetmsg</function></term> |
| 9503 |
<listitem> |
| 9504 |
<para> |
| 9505 |
<synopsis> |
| 9506 |
int rgetmsg(int msgnum, char *s, int maxsize); |
| 9507 |
</synopsis> |
| 9508 |
This function exists but is not implemented at the moment! |
| 9509 |
</para> |
| 9510 |
</listitem> |
| 9511 |
</varlistentry> |
| 9512 |
|
| 9513 |
<varlistentry id="ecpg-informix-functions-rtypalign"> |
| 9514 |
<term><function>rtypalign</function></term> |
| 9515 |
<listitem> |
| 9516 |
<para> |
| 9517 |
<synopsis> |
| 9518 |
int rtypalign(int offset, int type); |
| 9519 |
</synopsis> |
| 9520 |
This function exists but is not implemented at the moment! |
| 9521 |
</para> |
| 9522 |
</listitem> |
| 9523 |
</varlistentry> |
| 9524 |
|
| 9525 |
<varlistentry id="ecpg-informix-functions-rtypmsize"> |
| 9526 |
<term><function>rtypmsize</function></term> |
| 9527 |
<listitem> |
| 9528 |
<para> |
| 9529 |
<synopsis> |
| 9530 |
int rtypmsize(int type, int len); |
| 9531 |
</synopsis> |
| 9532 |
This function exists but is not implemented at the moment! |
| 9533 |
</para> |
| 9534 |
</listitem> |
| 9535 |
</varlistentry> |
| 9536 |
|
| 9537 |
<varlistentry id="ecpg-informix-functions-rtypwidth"> |
| 9538 |
<term><function>rtypwidth</function></term> |
| 9539 |
<listitem> |
| 9540 |
<para> |
| 9541 |
<synopsis> |
| 9542 |
int rtypwidth(int sqltype, int sqllen); |
| 9543 |
</synopsis> |
| 9544 |
This function exists but is not implemented at the moment! |
| 9545 |
</para> |
| 9546 |
</listitem> |
| 9547 |
</varlistentry> |
| 9548 |
|
| 9549 |
<varlistentry id="rsetnull"> |
| 9550 |
<term><function>rsetnull</function></term> |
| 9551 |
<listitem> |
| 9552 |
<para> |
| 9553 |
Set a variable to NULL. |
| 9554 |
<synopsis> |
| 9555 |
int rsetnull(int t, char *ptr); |
| 9556 |
</synopsis> |
| 9557 |
The function receives an integer that indicates the type of the |
| 9558 |
variable and a pointer to the variable itself that is cast to a C |
| 9559 |
char* pointer. |
| 9560 |
</para> |
| 9561 |
<para> |
| 9562 |
The following types exist: |
| 9563 |
<itemizedlist> |
| 9564 |
<listitem> |
| 9565 |
<para> |
| 9566 |
<literal>CCHARTYPE</literal> - For a variable of type <type>char</type> or <type>char*</type> |
| 9567 |
</para> |
| 9568 |
</listitem> |
| 9569 |
<listitem> |
| 9570 |
<para> |
| 9571 |
<literal>CSHORTTYPE</literal> - For a variable of type <type>short int</type> |
| 9572 |
</para> |
| 9573 |
</listitem> |
| 9574 |
<listitem> |
| 9575 |
<para> |
| 9576 |
<literal>CINTTYPE</literal> - For a variable of type <type>int</type> |
| 9577 |
</para> |
| 9578 |
</listitem> |
| 9579 |
<listitem> |
| 9580 |
<para> |
| 9581 |
<literal>CBOOLTYPE</literal> - For a variable of type <type>boolean</type> |
| 9582 |
</para> |
| 9583 |
</listitem> |
| 9584 |
<listitem> |
| 9585 |
<para> |
| 9586 |
<literal>CFLOATTYPE</literal> - For a variable of type <type>float</type> |
| 9587 |
</para> |
| 9588 |
</listitem> |
| 9589 |
<listitem> |
| 9590 |
<para> |
| 9591 |
<literal>CLONGTYPE</literal> - For a variable of type <type>long</type> |
| 9592 |
</para> |
| 9593 |
</listitem> |
| 9594 |
<listitem> |
| 9595 |
<para> |
| 9596 |
<literal>CDOUBLETYPE</literal> - For a variable of type <type>double</type> |
| 9597 |
</para> |
| 9598 |
</listitem> |
| 9599 |
<listitem> |
| 9600 |
<para> |
| 9601 |
<literal>CDECIMALTYPE</literal> - For a variable of type <type>decimal</type> |
| 9602 |
</para> |
| 9603 |
</listitem> |
| 9604 |
<listitem> |
| 9605 |
<para> |
| 9606 |
<literal>CDATETYPE</literal> - For a variable of type <type>date</type> |
| 9607 |
</para> |
| 9608 |
</listitem> |
| 9609 |
<listitem> |
| 9610 |
<para> |
| 9611 |
<literal>CDTIMETYPE</literal> - For a variable of type <type>timestamp</type> |
| 9612 |
</para> |
| 9613 |
</listitem> |
| 9614 |
</itemizedlist> |
| 9615 |
</para> |
| 9616 |
|
| 9617 |
<para> |
| 9618 |
Here is an example of a call to this function: |
| 9619 |
<programlisting><![CDATA[ |
| 9620 |
$char c[] = "abc "; |
| 9621 |
$short s = 17; |
| 9622 |
$int i = -74874; |
| 9623 |
|
| 9624 |
rsetnull(CCHARTYPE, (char *) c); |
| 9625 |
rsetnull(CSHORTTYPE, (char *) &s); |
| 9626 |
rsetnull(CINTTYPE, (char *) &i); |
| 9627 |
]]> |
| 9628 |
</programlisting> |
| 9629 |
</para> |
| 9630 |
</listitem> |
| 9631 |
</varlistentry> |
| 9632 |
|
| 9633 |
<varlistentry id="ecpg-informix-functions-risnull"> |
| 9634 |
<term><function>risnull</function></term> |
| 9635 |
<listitem> |
| 9636 |
<para> |
| 9637 |
Test if a variable is NULL. |
| 9638 |
<synopsis> |
| 9639 |
int risnull(int t, char *ptr); |
| 9640 |
</synopsis> |
| 9641 |
The function receives the type of the variable to test (<literal>t</literal>) |
| 9642 |
as well a pointer to this variable (<literal>ptr</literal>). Note that the |
| 9643 |
latter needs to be cast to a char*. See the function <xref |
| 9644 |
linkend="rsetnull"/> for a list of possible variable types. |
| 9645 |
</para> |
| 9646 |
<para> |
| 9647 |
Here is an example of how to use this function: |
| 9648 |
<programlisting><![CDATA[ |
| 9649 |
$char c[] = "abc "; |
| 9650 |
$short s = 17; |
| 9651 |
$int i = -74874; |
| 9652 |
|
| 9653 |
risnull(CCHARTYPE, (char *) c); |
| 9654 |
risnull(CSHORTTYPE, (char *) &s); |
| 9655 |
risnull(CINTTYPE, (char *) &i); |
| 9656 |
]]> |
| 9657 |
</programlisting> |
| 9658 |
</para> |
| 9659 |
</listitem> |
| 9660 |
</varlistentry> |
| 9661 |
</variablelist> |
| 9662 |
</para> |
| 9663 |
</sect2> |
| 9664 |
|
| 9665 |
<sect2 id="ecpg-informix-constants"> |
| 9666 |
<title>Additional Constants</title> |
| 9667 |
<para> |
| 9668 |
Note that all constants here describe errors and all of them are defined |
| 9669 |
to represent negative values. In the descriptions of the different |
| 9670 |
constants you can also find the value that the constants represent in the |
| 9671 |
current implementation. However you should not rely on this number. You can |
| 9672 |
however rely on the fact all of them are defined to represent negative |
| 9673 |
values. |
| 9674 |
<variablelist> |
| 9675 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-num-overflow"> |
| 9676 |
<term><literal>ECPG_INFORMIX_NUM_OVERFLOW</literal></term> |
| 9677 |
<listitem> |
| 9678 |
<para> |
| 9679 |
Functions return this value if an overflow occurred in a |
| 9680 |
calculation. Internally it is defined as -1200 (the <productname>Informix</productname> |
| 9681 |
definition). |
| 9682 |
</para> |
| 9683 |
</listitem> |
| 9684 |
</varlistentry> |
| 9685 |
|
| 9686 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-num-underflow"> |
| 9687 |
<term><literal>ECPG_INFORMIX_NUM_UNDERFLOW</literal></term> |
| 9688 |
<listitem> |
| 9689 |
<para> |
| 9690 |
Functions return this value if an underflow occurred in a calculation. |
| 9691 |
Internally it is defined as -1201 (the <productname>Informix</productname> definition). |
| 9692 |
</para> |
| 9693 |
</listitem> |
| 9694 |
</varlistentry> |
| 9695 |
|
| 9696 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-divide-zero"> |
| 9697 |
<term><literal>ECPG_INFORMIX_DIVIDE_ZERO</literal></term> |
| 9698 |
<listitem> |
| 9699 |
<para> |
| 9700 |
Functions return this value if an attempt to divide by zero is |
| 9701 |
observed. Internally it is defined as -1202 (the <productname>Informix</productname> definition). |
| 9702 |
</para> |
| 9703 |
</listitem> |
| 9704 |
</varlistentry> |
| 9705 |
|
| 9706 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-bad-year"> |
| 9707 |
<term><literal>ECPG_INFORMIX_BAD_YEAR</literal></term> |
| 9708 |
<listitem> |
| 9709 |
<para> |
| 9710 |
Functions return this value if a bad value for a year was found while |
| 9711 |
parsing a date. Internally it is defined as -1204 (the <productname>Informix</productname> |
| 9712 |
definition). |
| 9713 |
</para> |
| 9714 |
</listitem> |
| 9715 |
</varlistentry> |
| 9716 |
|
| 9717 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-bad-month"> |
| 9718 |
<term><literal>ECPG_INFORMIX_BAD_MONTH</literal></term> |
| 9719 |
<listitem> |
| 9720 |
<para> |
| 9721 |
Functions return this value if a bad value for a month was found while |
| 9722 |
parsing a date. Internally it is defined as -1205 (the <productname>Informix</productname> |
| 9723 |
definition). |
| 9724 |
</para> |
| 9725 |
</listitem> |
| 9726 |
</varlistentry> |
| 9727 |
|
| 9728 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-bad-day"> |
| 9729 |
<term><literal>ECPG_INFORMIX_BAD_DAY</literal></term> |
| 9730 |
<listitem> |
| 9731 |
<para> |
| 9732 |
Functions return this value if a bad value for a day was found while |
| 9733 |
parsing a date. Internally it is defined as -1206 (the <productname>Informix</productname> |
| 9734 |
definition). |
| 9735 |
</para> |
| 9736 |
</listitem> |
| 9737 |
</varlistentry> |
| 9738 |
|
| 9739 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-enoshortdate"> |
| 9740 |
<term><literal>ECPG_INFORMIX_ENOSHORTDATE</literal></term> |
| 9741 |
<listitem> |
| 9742 |
<para> |
| 9743 |
Functions return this value if a parsing routine needs a short date |
| 9744 |
representation but did not get the date string in the right length. |
| 9745 |
Internally it is defined as -1209 (the <productname>Informix</productname> definition). |
| 9746 |
</para> |
| 9747 |
</listitem> |
| 9748 |
</varlistentry> |
| 9749 |
|
| 9750 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-date-convert"> |
| 9751 |
<term><literal>ECPG_INFORMIX_DATE_CONVERT</literal></term> |
| 9752 |
<listitem> |
| 9753 |
<para> |
| 9754 |
Functions return this value if an error occurred during date |
| 9755 |
formatting. Internally it is defined as -1210 (the |
| 9756 |
<productname>Informix</productname> definition). |
| 9757 |
</para> |
| 9758 |
</listitem> |
| 9759 |
</varlistentry> |
| 9760 |
|
| 9761 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-out-of-memory"> |
| 9762 |
<term><literal>ECPG_INFORMIX_OUT_OF_MEMORY</literal></term> |
| 9763 |
<listitem> |
| 9764 |
<para> |
| 9765 |
Functions return this value if memory was exhausted during |
| 9766 |
their operation. Internally it is defined as -1211 (the |
| 9767 |
<productname>Informix</productname> definition). |
| 9768 |
</para> |
| 9769 |
</listitem> |
| 9770 |
</varlistentry> |
| 9771 |
|
| 9772 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-enotdmy"> |
| 9773 |
<term><literal>ECPG_INFORMIX_ENOTDMY</literal></term> |
| 9774 |
<listitem> |
| 9775 |
<para> |
| 9776 |
Functions return this value if a parsing routine was supposed to get a |
| 9777 |
format mask (like <literal>mmddyy</literal>) but not all fields were listed |
| 9778 |
correctly. Internally it is defined as -1212 (the <productname>Informix</productname> definition). |
| 9779 |
</para> |
| 9780 |
</listitem> |
| 9781 |
</varlistentry> |
| 9782 |
|
| 9783 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-bad-numeric"> |
| 9784 |
<term><literal>ECPG_INFORMIX_BAD_NUMERIC</literal></term> |
| 9785 |
<listitem> |
| 9786 |
<para> |
| 9787 |
Functions return this value either if a parsing routine cannot parse |
| 9788 |
the textual representation for a numeric value because it contains |
| 9789 |
errors or if a routine cannot complete a calculation involving numeric |
| 9790 |
variables because at least one of the numeric variables is invalid. |
| 9791 |
Internally it is defined as -1213 (the <productname>Informix</productname> definition). |
| 9792 |
</para> |
| 9793 |
</listitem> |
| 9794 |
</varlistentry> |
| 9795 |
|
| 9796 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-bad-exponent"> |
| 9797 |
<term><literal>ECPG_INFORMIX_BAD_EXPONENT</literal></term> |
| 9798 |
<listitem> |
| 9799 |
<para> |
| 9800 |
Functions return this value if a parsing routine cannot parse |
| 9801 |
an exponent. Internally it is defined as -1216 (the |
| 9802 |
<productname>Informix</productname> definition). |
| 9803 |
</para> |
| 9804 |
</listitem> |
| 9805 |
</varlistentry> |
| 9806 |
|
| 9807 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-bad-date"> |
| 9808 |
<term><literal>ECPG_INFORMIX_BAD_DATE</literal></term> |
| 9809 |
<listitem> |
| 9810 |
<para> |
| 9811 |
Functions return this value if a parsing routine cannot parse |
| 9812 |
a date. Internally it is defined as -1218 (the |
| 9813 |
<productname>Informix</productname> definition). |
| 9814 |
</para> |
| 9815 |
</listitem> |
| 9816 |
</varlistentry> |
| 9817 |
|
| 9818 |
<varlistentry id="ecpg-informix-constants-ecpg-informix-extra-chars"> |
| 9819 |
<term><literal>ECPG_INFORMIX_EXTRA_CHARS</literal></term> |
| 9820 |
<listitem> |
| 9821 |
<para> |
| 9822 |
Functions return this value if a parsing routine is passed extra |
| 9823 |
characters it cannot parse. Internally it is defined as -1264 (the |
| 9824 |
<productname>Informix</productname> definition). |
| 9825 |
</para> |
| 9826 |
</listitem> |
| 9827 |
</varlistentry> |
| 9828 |
</variablelist> |
| 9829 |
</para> |
| 9830 |
</sect2> |
| 9831 |
</sect1> |
| 9832 |
|
| 9833 |
<sect1 id="ecpg-oracle-compat"> |
| 9834 |
<title><productname>Oracle</productname> Compatibility Mode</title> |
| 9835 |
<para> |
| 9836 |
<command>ecpg</command> can be run in a so-called <firstterm>Oracle |
| 9837 |
compatibility mode</firstterm>. If this mode is active, it tries to |
| 9838 |
behave as if it were Oracle <productname>Pro*C</productname>. |
| 9839 |
</para> |
| 9840 |
|
| 9841 |
<para> |
| 9842 |
Specifically, this mode changes <command>ecpg</command> in three ways: |
| 9843 |
|
| 9844 |
<itemizedlist> |
| 9845 |
<listitem> |
| 9846 |
<para> |
| 9847 |
Pad character arrays receiving character string types with |
| 9848 |
trailing spaces to the specified length |
| 9849 |
</para> |
| 9850 |
</listitem> |
| 9851 |
|
| 9852 |
<listitem> |
| 9853 |
<para> |
| 9854 |
Zero byte terminate these character arrays, and set the indicator |
| 9855 |
variable if truncation occurs |
| 9856 |
</para> |
| 9857 |
</listitem> |
| 9858 |
|
| 9859 |
<listitem> |
| 9860 |
<para> |
| 9861 |
Set the null indicator to <literal>-1</literal> when character |
| 9862 |
arrays receive empty character string types |
| 9863 |
</para> |
| 9864 |
</listitem> |
| 9865 |
</itemizedlist> |
| 9866 |
</para> |
| 9867 |
</sect1> |
| 9868 |
|
| 9869 |
<sect1 id="ecpg-develop"> |
| 9870 |
<title>Internals</title> |
| 9871 |
|
| 9872 |
<para> |
| 9873 |
This section explains how <application>ECPG</application> works |
| 9874 |
internally. This information can occasionally be useful to help |
| 9875 |
users understand how to use <application>ECPG</application>. |
| 9876 |
</para> |
| 9877 |
|
| 9878 |
<para> |
| 9879 |
The first four lines written by <command>ecpg</command> to the |
| 9880 |
output are fixed lines. Two are comments and two are include |
| 9881 |
lines necessary to interface to the library. Then the |
| 9882 |
preprocessor reads through the file and writes output. Normally |
| 9883 |
it just echoes everything to the output. |
| 9884 |
</para> |
| 9885 |
|
| 9886 |
<para> |
| 9887 |
When it sees an <command>EXEC SQL</command> statement, it |
| 9888 |
intervenes and changes it. The command starts with <command>EXEC |
| 9889 |
SQL</command> and ends with <command>;</command>. Everything in |
| 9890 |
between is treated as an <acronym>SQL</acronym> statement and |
| 9891 |
parsed for variable substitution. |
| 9892 |
</para> |
| 9893 |
|
| 9894 |
<para> |
| 9895 |
Variable substitution occurs when a symbol starts with a colon |
| 9896 |
(<literal>:</literal>). The variable with that name is looked up |
| 9897 |
among the variables that were previously declared within a |
| 9898 |
<literal>EXEC SQL DECLARE</literal> section. |
| 9899 |
</para> |
| 9900 |
|
| 9901 |
<para> |
| 9902 |
The most important function in the library is |
| 9903 |
<function>ECPGdo</function>, which takes care of executing most |
| 9904 |
commands. It takes a variable number of arguments. This can easily |
| 9905 |
add up to 50 or so arguments, and we hope this will not be a |
| 9906 |
problem on any platform. |
| 9907 |
</para> |
| 9908 |
|
| 9909 |
<para> |
| 9910 |
The arguments are: |
| 9911 |
|
| 9912 |
<variablelist> |
| 9913 |
<varlistentry id="ecpg-develop-line-number"> |
| 9914 |
<term>A line number</term> |
| 9915 |
<listitem> |
| 9916 |
<para> |
| 9917 |
This is the line number of the original line; used in error |
| 9918 |
messages only. |
| 9919 |
</para> |
| 9920 |
</listitem> |
| 9921 |
</varlistentry> |
| 9922 |
|
| 9923 |
<varlistentry id="ecpg-develop-string"> |
| 9924 |
<term>A string</term> |
| 9925 |
<listitem> |
| 9926 |
<para> |
| 9927 |
This is the <acronym>SQL</acronym> command that is to be issued. |
| 9928 |
It is modified by the input variables, i.e., the variables that |
| 9929 |
where not known at compile time but are to be entered in the |
| 9930 |
command. Where the variables should go the string contains |
| 9931 |
<literal>?</literal>. |
| 9932 |
</para> |
| 9933 |
</listitem> |
| 9934 |
</varlistentry> |
| 9935 |
|
| 9936 |
<varlistentry id="ecpg-develop-input-variables"> |
| 9937 |
<term>Input variables</term> |
| 9938 |
<listitem> |
| 9939 |
<para> |
| 9940 |
Every input variable causes ten arguments to be created. (See below.) |
| 9941 |
</para> |
| 9942 |
</listitem> |
| 9943 |
</varlistentry> |
| 9944 |
|
| 9945 |
<varlistentry id="ecpg-develop-ecpgt-eoit"> |
| 9946 |
<term><parameter>ECPGt_EOIT</parameter></term> |
| 9947 |
<listitem> |
| 9948 |
<para> |
| 9949 |
An <type>enum</type> telling that there are no more input |
| 9950 |
variables. |
| 9951 |
</para> |
| 9952 |
</listitem> |
| 9953 |
</varlistentry> |
| 9954 |
|
| 9955 |
<varlistentry id="ecpg-develop-output-variables"> |
| 9956 |
<term>Output variables</term> |
| 9957 |
<listitem> |
| 9958 |
<para> |
| 9959 |
Every output variable causes ten arguments to be created. |
| 9960 |
(See below.) These variables are filled by the function. |
| 9961 |
</para> |
| 9962 |
</listitem> |
| 9963 |
</varlistentry> |
| 9964 |
|
| 9965 |
<varlistentry id="ecpg-develop-ecpgt-eort"> |
| 9966 |
<term><parameter>ECPGt_EORT</parameter></term> |
| 9967 |
<listitem> |
| 9968 |
<para> |
| 9969 |
An <type>enum</type> telling that there are no more variables. |
| 9970 |
</para> |
| 9971 |
</listitem> |
| 9972 |
</varlistentry> |
| 9973 |
</variablelist> |
| 9974 |
</para> |
| 9975 |
|
| 9976 |
<para> |
| 9977 |
For every variable that is part of the <acronym>SQL</acronym> |
| 9978 |
command, the function gets ten arguments: |
| 9979 |
|
| 9980 |
<orderedlist> |
| 9981 |
<listitem> |
| 9982 |
<para> |
| 9983 |
The type as a special symbol. |
| 9984 |
</para> |
| 9985 |
</listitem> |
| 9986 |
|
| 9987 |
<listitem> |
| 9988 |
<para> |
| 9989 |
A pointer to the value or a pointer to the pointer. |
| 9990 |
</para> |
| 9991 |
</listitem> |
| 9992 |
|
| 9993 |
<listitem> |
| 9994 |
<para> |
| 9995 |
The size of the variable if it is a <type>char</type> or <type>varchar</type>. |
| 9996 |
</para> |
| 9997 |
</listitem> |
| 9998 |
|
| 9999 |
<listitem> |
| 10000 |
<para> |
| 10001 |
The number of elements in the array (for array fetches). |
| 10002 |
</para> |
| 10003 |
</listitem> |
| 10004 |
|
| 10005 |
<listitem> |
| 10006 |
<para> |
| 10007 |
The offset to the next element in the array (for array fetches). |
| 10008 |
</para> |
| 10009 |
</listitem> |
| 10010 |
|
| 10011 |
<listitem> |
| 10012 |
<para> |
| 10013 |
The type of the indicator variable as a special symbol. |
| 10014 |
</para> |
| 10015 |
</listitem> |
| 10016 |
|
| 10017 |
<listitem> |
| 10018 |
<para> |
| 10019 |
A pointer to the indicator variable. |
| 10020 |
</para> |
| 10021 |
</listitem> |
| 10022 |
|
| 10023 |
<listitem> |
| 10024 |
<para> |
| 10025 |
0 |
| 10026 |
</para> |
| 10027 |
</listitem> |
| 10028 |
|
| 10029 |
<listitem> |
| 10030 |
<para> |
| 10031 |
The number of elements in the indicator array (for array fetches). |
| 10032 |
</para> |
| 10033 |
</listitem> |
| 10034 |
|
| 10035 |
<listitem> |
| 10036 |
<para> |
| 10037 |
The offset to the next element in the indicator array (for |
| 10038 |
array fetches). |
| 10039 |
</para> |
| 10040 |
</listitem> |
| 10041 |
</orderedlist> |
| 10042 |
</para> |
| 10043 |
|
| 10044 |
<para> |
| 10045 |
Note that not all SQL commands are treated in this way. For |
| 10046 |
instance, an open cursor statement like: |
| 10047 |
<programlisting> |
| 10048 |
EXEC SQL OPEN <replaceable>cursor</replaceable>; |
| 10049 |
</programlisting> |
| 10050 |
is not copied to the output. Instead, the cursor's |
| 10051 |
<command>DECLARE</command> command is used at the position of the <command>OPEN</command> command |
| 10052 |
because it indeed opens the cursor. |
| 10053 |
</para> |
| 10054 |
|
| 10055 |
<para> |
| 10056 |
Here is a complete example describing the output of the |
| 10057 |
preprocessor of a file <filename>foo.pgc</filename> (details might |
| 10058 |
change with each particular version of the preprocessor): |
| 10059 |
<programlisting> |
| 10060 |
EXEC SQL BEGIN DECLARE SECTION; |
| 10061 |
int index; |
| 10062 |
int result; |
| 10063 |
EXEC SQL END DECLARE SECTION; |
| 10064 |
... |
| 10065 |
EXEC SQL SELECT res INTO :result FROM mytable WHERE index = :index; |
| 10066 |
</programlisting> |
| 10067 |
is translated into: |
| 10068 |
<programlisting><![CDATA[ |
| 10069 |
/* Processed by ecpg (2.6.0) */ |
| 10070 |
/* These two include files are added by the preprocessor */ |
| 10071 |
#include <ecpgtype.h>; |
| 10072 |
#include <ecpglib.h>; |
| 10073 |
|
| 10074 |
/* exec sql begin declare section */ |
| 10075 |
|
| 10076 |
#line 1 "foo.pgc" |
| 10077 |
|
| 10078 |
int index; |
| 10079 |
int result; |
| 10080 |
/* exec sql end declare section */ |
| 10081 |
... |
| 10082 |
ECPGdo(__LINE__, NULL, "SELECT res FROM mytable WHERE index = ? ", |
| 10083 |
ECPGt_int,&(index),1L,1L,sizeof(int), |
| 10084 |
ECPGt_NO_INDICATOR, NULL , 0L, 0L, 0L, ECPGt_EOIT, |
| 10085 |
ECPGt_int,&(result),1L,1L,sizeof(int), |
| 10086 |
ECPGt_NO_INDICATOR, NULL , 0L, 0L, 0L, ECPGt_EORT); |
| 10087 |
#line 147 "foo.pgc" |
| 10088 |
]]> |
| 10089 |
</programlisting> |
| 10090 |
(The indentation here is added for readability and not |
| 10091 |
something the preprocessor does.) |
| 10092 |
</para> |
| 10093 |
</sect1> |
| 10094 |
</chapter> |