Debian Codegraph @

Changes

raw
1 # satisfy kwalitee test
2 use strict;
3
4 =encoding utf8
5
6 DBD::ODBC::Changes - Log of significant changes to the DBD::ODBC
7
8 1.61 2020-01-30
9
10 [BUG FIXES]
11
12 Fix 12blob.t test by pali
13 Fix searching for ODBC libraries in system by pali (#15)
14
15 [ENHANCEMENTS]
16
17 use PERL_NO_GET_CONTEXT for more performance by markusbeth (#13)
18
19 [MISCELLANEOUS]
20
21 Fix travis builds for older Perls by pali
22
23 1.60 2018-10-31
24
25 [BUG FIXES]
26
27 Merged pull request 11 from audun which fixes some issues with the AutoCommit flag
28 on commit and rollback.
29
30 [MISCELLANEOUS]
31
32 Merged pull request 10 from vadz which fixed typo (affecting license) in README.md.
33
34 1.59 2018-08-10
35
36 [BUG FIXES]
37
38 git issue 8. Setting odbc_utf8_on didn't work properly. Thanks to David Wheeler for
39 reporting and helping to debug.
40
41 1.58 2018-03-01
42
43 [MISCELLANEOUS]
44
45 Various changes to the test suite to get better results with Postgres
46
47 1.57 2018-03-01
48
49 [MISCELLANEOUS]
50
51 Merged pull request 6 from genio which allows Makefile.PL argument -u
52 to be set via the environment variable DBD_ODBC_UNICODE
53
54 This version was removed from CPAN because it was uploaded with a nasty
55 bug in the diagnostics code.
56
57 1.56 2016-10-06
58
59 Full release of the 1.53 development series
60
61 One version skipped because of indexing problems.
62
63 1.53_2 2016-02-03
64
65 [MISCELLANEOUS]
66
67 Add new FAQs
68
69 1.53_1 2015-10-16
70
71 [BUG FIXES]
72
73 Strictly speaking this is a bug fix to DBI and not DBD::ODBC but DBI
74 now supports 64 bit row counts where an IV in perl is 64 bits. However, it
75 necessitated changes to DBD::ODBC to pick up the fix. odbc_rows (my workaround
76 since 2012) is still supported but should no longer be required so long as you
77 use this DBD::ODBC and DBI 1.633_92 or above.
78
79 [INTERNALS]
80
81 Removed dbd_st_rows and now setting DBIc_ROW_COUNT.
82
83 [DOCUMENTATION]
84
85 Add tables and table_info section to deviations from the DBI spec.
86
87 [MISCELLANEOUS]
88
89 Change column name in t/rt_101579.t as "method" is a reserved word
90 in. Teradata Thanks to Zhenyi Zhou.
91
92 Remove duplicate dynamic_config from META.yml.
93
94 1.52 2015-04-15
95
96 [MISCELLANEOUS]
97
98 Changes to the test suite to make it run better with Postgres thanks
99 to Greg Sabino Mullane.
100
101 1.51_4 2015-01-18
102
103 [BUG FIXES]
104
105 Numerous errors in the test suite (with SQLite ODBC driver) mostly down to not
106 creating the test table first.
107
108 [MISCELLANEOUS]
109
110 Try and make the test suite run ok for SQLite ODBC driver so I can use it
111 in travis-ci.
112
113 1.51_3 2015-01-17
114
115 [BUG FIXES]
116
117 RT101579 - using bound input parameters for numeric columns (e.g.,
118 SQL_NUMERIC) only works the first time and will quite likely fail
119 with "string data, right truncation" on the second and subsequent
120 calls to execute. Thanks to Laura Cox for finding.
121
122 1.51_2 2014-11-19
123
124 [BUG FIXES]
125
126 The table_info method (ANSI version only) was incorrectly passing
127 the table name for the type argument. I think this bug was
128 introduced last year. Thanks to Nick Gorham for spotting and providing
129 a fix.
130
131 1.51_1 2014-11-14
132
133 [BUG FIXES]
134
135 RT100186 - handle VARBINARY(MAX) parameters with SQL Server native
136 client. Identify "libmsodbcsql*" as the MS ODBC Driver for Linux as
137 there are some specific workarounds for MS Native Client ODBC driver.
138
139 1.50 2014-07-25
140
141 [BUG FIXES]
142
143 The 80_odbc_diags.t test could fail if a driver fails a table does
144 not exist test in the prepare instead of the execute.
145
146 1.49_4 2014-07-08
147
148 [BUG FIXES]
149
150 Fixed sql_type_cast.t test which assumed column aliases which stay
151 lowercase.
152
153 Fixed 87_odbc_lob_read.t test which did not bow out of the test
154 properly if the database was not MS SQL Server.
155
156 [DOCUMENTATION]
157
158 Revised the query notification example and documentation.
159
160 Added a link to a better Query Notification article.
161
162 1.49_3 2014-05-01
163
164 [CHANGE IN BEHAVIOUR]
165
166 As warned years ago, this release removes the odbc_old_unicode attribute.
167 If you have a good reason to use it speak up now before the next non-development
168 release.
169
170 [BUG FIXES]
171
172 Fix rt89255: Fails to create test table for tests using PostgreSQL odbc driver.
173 Change test suite to fallback on PRECISION if COLUMN_SIZE is not found.
174
175 [ENHANCEMENTS]
176
177 Added support for MS SQL Server Query Notification. See the new
178 section in the pod.
179
180 Added a currently undocumented (and experimental)
181 odbc_describe_param method on a statement handle which takes a
182 parameter number as the only argument and returns an array of the
183 data type, parameter size, decimal digits and nullable (as per
184 SQLDescribeParam).
185
186 [DOCUMENTATION]
187
188 Added FAQ on truncated column names with freeTDS.
189
190 [MISCELLANEOUS]
191
192 I have removed the "experimental" tag for odbc_getdiagfield and odbc_getdiagrec
193 methods.
194
195 1.49_2 2014-04-26
196
197 [BUG FIXES]
198
199 Change to data_sources in 1.49_1 could lead to a compile error since
200 data_sources was not returning a value if an error occurred.
201
202 1.49_1 2014-04-25
203
204 [BUG FIXES]
205
206 If you had a lot of DSNs on Windows (more than 280 but it depends on
207 the length of their names) and called the data_sources method it
208 could crash your script. Code internally changed to stop putting the DSNs
209 returned on the stack.
210
211 [CHANGE IN BEHAVIOUR]
212
213 As warned years ago, the private data_sources method has been
214 removed - use DBI one instead.
215
216 [MISCELLANEOUS]
217
218 Added FAQ entry of maximum number of allowed parameters.
219
220 1.48 2014-03-03
221
222 [MISCELLANEOUS]
223
224 Manifest has wrong filename for 90_trace_flags.t
225
226 Forgot to remove warning from ODBC.pm that this is a development
227 release and unicode change when I released 1.47.
228
229 1.47 2014-02-19
230
231 Full release of the 1.46 development releases.
232
233 [MISCELLANEOUS]
234
235 Just some tidying up of dbdimp.c - shouldn't make a difference to anyone.
236
237 Further changes to this change file to make it CPAN::Changes spec.
238 NOTE the changes.cpanhq.com site does not yet support "unknown" for
239 dates.
240
241 1.46_2 2013-12-17
242
243 [BUG FIXES]
244
245 When built with unicode support and odbc_old_unicode is not enabled
246 columns reported as SQL_LONGVARCHAR were not by default bound as
247 SQL_WCHAR and hence were not returned correctly unless the bind was
248 overridden.
249
250 [MISCELLANEOUS]
251
252 Added test 90_trace_flag.t
253
254 1.46_1 2013-11-16
255
256 [CHANGE IN BEHAVIOUR]
257
258 As warned in release 1.45, the binding of unicode parameters to
259 char/varchar columns has changed significantly. If you don't attempt
260 to insert unicode into char/varchar columns or if you only inserted
261 unicode into nchar/nvarchar columns you should see no difference.
262 From this release, unicode data inserted into
263 char/varchar/longvarchar columns is bound as SQL_WCHAR and not
264 whatever the driver reports the parameter as (which is mostly
265 SQL_CHAR).
266
267 Previously if DBD::ODBC received an error or (SQL_SUCCESS_WITH_INFO)
268 from an ODBC API call and then the driver refused to return the
269 error state/text DBD::ODBC would issue its own error saying "Unable
270 to fetch information about the error" and state IM008. That state
271 was wrong and has been changed to HY000.
272
273 [BUG FIXES]
274
275 Some drivers cannot support catalogs and/or schema names in
276 SQLTables. Recent changes set the schema/catalog name to the empty
277 string (good reasons below) which causes "optional feature not
278 implemented" from MS Access (which does not support schemas - even
279 for a simply ping (which uses SQLTables)). Now we call
280 SQLCATALOG_NAME and SQLSCHEMA_USAGE on connect to ascertain support
281 which modifies SQLTables call.
282
283 [MISCELLANEOUS]
284
285 Added test 45_unicode_varchar.t for MS SQL Server only so far.
286
287 1.45 2013-10-28
288
289 [CHANGE IN BEHAVIOUR]
290
291 There is no intentional change in behaviour in this release but I'm
292 adding a warning that the next development release is highly liking
293 to contain some significant unicode changes in behaviour to fix some
294 bugs which have been around for quite a long time now.
295
296 [BUG FIXES]
297
298 If an SQLExecute ODBC API call returned SQL_NO_DATA DBD::ODBC was
299 still calling SQLError (which was a waste of time).
300
301 Since 1.44_1 odbc_out_connect_string stopped returning anything.
302
303 [MISCELLANEOUS]
304
305 Added another link to resources for supplementary characters.
306
307 Added 1 more test to 20SqlServer.t for update statement.
308
309 Small changes to 20SqlServer.t test to skip some tests and note the
310 problem if SQLExecute returns SQL_NO_DATA on a non searched update.
311
312 1.44_4 2013-10-16
313
314 [BUG FIXES]
315
316 Fix method redefinition warnings in threads on Perl >= 5.16 thanks
317 Dagfinn Ilmari Mannsåker
318
319 [MISCELLANEOUS]
320
321 Changed this Changes file to be closer to the version 0.03 change
322 file spec.
323
324 Added t/version.t test.
325
326 Added recommends Test::Version.
327
328 Updates to the odbc_more_results pod to help clarify its use after
329 some confusion was seen in a perlmonks thread.
330
331 1.44_3 2013-10-11
332
333 [CHANGE IN BEHAVIOUR]
334
335 If you attempt to set the ReadOnly attribute and the underlying
336 ODBC driver does not support this (SQL_SUCCESS_WITH_INFO and "option
337 value changed" is returned) a warning is issued. However, until RT
338 89015 "You cannot issue a warning in the STORE method" in DBI is
339 resolved you won't get this warning. As DBI 1.628 it is not
340 resolved. I've only seen the SQLite ODBC driver do this.
341
342 If you set ReadOnly and the underlying ODBC driver does not
343 support this then any subsequent attempts to fetch the ReadOnly
344 attribute will return the value you set.
345
346 [BUG FIXES]
347
348 The 82_table_info test assumed all database and ODBC Drivers
349 supported catalogs and schemas (some don't). Use get_info to
350 find out if catalogs and schemas are supported before
351 running these tests.
352
353 The rt_79190.t could incorrectly fail if your test DSN contained
354 the DRIVER attribute.
355
356 [MISCELLANEOUS]
357
358 Added RedHat spec file to examples courtesy of Michiel Beijen.
359
360 Added "use strict" to FAQ/Changes etc to quieten kwalitee test.
361
362 Added a workaround in the test suite for a probable bug in the
363 postgres ODBC driver which does not return COLUMN_SIZE from
364 SQLGetTypeInfo. It also issues a warning. See
365 http://www.postgresql.org/message-id/[email protected]
366
367 1.44_3 2013-10-11
368
369 [MISCELLANEOUS]
370
371 Skip 70execute_array_native.t test if MS Access - even if behind an
372 ODBC Bridge.
373
374 Fixed some compiler warnings when attempting to print/trace SvCUR.
375
376 1.44_2 2013-09-07
377
378 [BUG FIXES]
379
380 When table_info was called with a '%' for any one of the catalog,
381 schema or type arguments with the rest all '' (the empty string),
382 only a list of catalogs, schemas or types should be returned. It was
383 not doing that as it was changing empty strings to undef/NULL.
384
385 pod for odbc_lob_read had an example only saying lob_read.
386
387 TYPE attribute for odbc_lob_read was actually coded as Type. It is
388 now as documented.
389
390 The example lob_read.pl had the TYPE set to 999 from when I was
391 testing it but it got checked in like this.
392
393 MANIFEST contained column_info.pl but the file was coltest.pl
394
395 [MISCELLANEOUS]
396
397 Fixed RT 86379 - spelling mistakes in ODBC.pm and FAQ - thanks
398 to David Steinbrunner.
399
400 Added 82_table_info.t test.
401
402 Added 87_odbc_log_read.t test.
403
404 1.44_1 2013-06-06
405
406 Moved from subversion to github as svn.perl.org is closing down.
407 Changed docs to show new repository.
408
409 [BUG FIXES]
410
411 Fixed RT 84450 - Database Handle Attribute Fetch broken. Thanks to
412 Stephen Oberholtzer for finding and supplying patch.
413
414 Fixed problem with attributes on bind_col not being sticky. You'll
415 probably only see this if you are using fetchall_arrayref with a
416 slice and setting TYPE or attributes in bind_col first.
417
418 1.43 2013-03-06
419
420 This is a full release of all the 1.42_* development releases.
421
422 plus:
423
424 [BUG FIXES]
425
426 Minor fix to 10handler.t test suite which relied on a native error
427 being true instead of defined.
428
429 1.42_5 2013-01-25
430
431 [BUG FIXES]
432
433 Not all modules used in test code were specified in build_requires.
434
435 1.42_4 2013-01-21
436
437 [ENHANCEMENTS]
438
439 odbc_trace and odbc_trace_file are now full connection attributes
440 so you can set them any time you like, not just in connect.
441
442 1.42_3 2013-01-17
443
444 [ENHANCEMENTS]
445
446 Added odbc_trace_file and odbc_trace attributes to the connect
447 method so you can now enable ODBC API tracing from the connect
448 method instead of having to use the ODBC Driver Manager. These also
449 only enable ODBC API tracing in the application which made the call
450 unlike the ODBC Driver Manager settings.
451
452 1.42_2 2012-12-17
453
454 [MISCELLANEOUS]
455
456 Changed any use of if SvUPGRADE to remove the if test as per email
457 from Dave Mitchell and posting at
458 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2012-12/msg00424.html.
459
460 1.42_1 2012-12-12
461
462 [BUG FIXES]
463
464 DBD::ODBC's ExecDirect method did not return an SQLLEN so if you
465 managed to affect a massive number of rows it would be cast to an
466 int and hence precision lost.
467
468 [CHANGE IN BEHAVIOUR]
469
470 When you called DBI's execute method and odbc_exec_direct was not
471 set (the default) if you managed to affect more rows than would fit
472 into an int you would get the incorrect count (NOTE on 32 bit
473 platforms ODBC's SQLRowCount can only return a 32bit value
474 anyway). You would get whatever casting an SQLLEN to an int would
475 give you. The fix for this needs a change to DBI (see RT 81911) and
476 the change would probably impact every DBD so until then DBD::ODBC
477 will a) warn if an overflow occurs and Warn is set on the handle b)
478 return INT_MAX and c) provide a new statement method odbc_rows which
479 you can use to get the correct value.
480
481 [ENHANCEMENTS]
482
483 New odbc_rows statement method (see above).
484
485 [MISCELLANEOUS]
486
487 New rt_81911.t test case.
488
489 1.42_0 2012-11-28
490
491 [BUG FIXES]
492
493 MS Access requires a longchar column to be bound using SQL_LONGVARCHAR.
494 However, MS Access does not support SQLDescribeParam and we default to
495 SQL_VARCHAR in this case. The point at which we switch to SQL_LONGVARCHAR
496 was defaulted to 4000 (for MS SQL Server). We now default to SQL_LONGVARCHAR
497 for MS Access when data is > 255. This means you can remove those
498 {TYPE => SQL_LONGVARCHAR} from your bind_param calls for longchar columns
499 in MS Access.
500
501 I seem to have introduced a bug in the test suite for MS Access.
502 The last test in the 09bind test binds dates as varchars (by
503 default) and this cannot work in MS Access (it needs to be a
504 timestamp). This test was skipped in the past and the skip got
505 removed.
506
507 [MISCELLANEOUS]
508
509 Steffen Goeldner reported some issues with execute_array in
510 DBD::Oracle where if ArrayTupleStatus was not specified and an error
511 occurred DBD::Oracle did not do the right thing. As I used
512 DBD::Oracle as a base when I wrote execute_for_fetch in DBD::ODBC I
513 added tests to the test suite to ensure these issues did not exist
514 in DBD::ODBC.
515
516 Minor change to sql_type_cast.t test which attempts to insert an
517 integer into a varchar. No databases so far have complained about
518 this until we ran the test against Derby. Changed to use '100'.
519
520 RT 80446 - fix spelling mistake - thanks to Xavier Guimar.
521
522 1.41 2012-10-23
523
524 A full release of the 1.40 development release series.
525
526 1.40_3 2012-10-08
527
528 [BUG FIXES]
529
530 Oops, changes to some rt tests fail when not run to MS SQL Server
531 and they should not be run for other drivers - there was a double
532 done_testing call.
533
534 [CHANGE IN BEHAVIOUR]
535
536 As I warned literally years ago DBD::ODBC's private function
537 DescribeCol has been removed. You can use DBI's statement attributes
538 like NAME, PRECISION etc, instead. All test code has been changed to
539 remove calls to DescribeCol and GetTypeInfo.
540
541 [MISCELLANEOUS]
542
543 New example sqlserver_supplementary_chrs.pl added which shows that
544 in MS SQL Server 2012 you can now store unicode characters
545 over 0xFFFF (ones which are surrogate pairs).
546
547 More documentation for odbc_out_connect_string.
548
549 1.40_2 2012-09-06
550
551 [BUG FIXES]
552
553 Fixed rt 78838 - bind_param does not correctly stringify blessed
554 objects when connected to MS SQL Server
555
556 Fix issue in dbd_bind_ph where if you passed a sql type and were
557 also attempting to change from in to out or vice versa or increasing
558 the size of an output bound param it would not spot this error.
559
560 Allowed the test cases to spot DB2 driver as libXXXdb2.
561
562 [MISCELLANEOUS]
563
564 New test cases added for some rts.
565
566 Added Test::NoWarnings to some tests where it was missing.
567
568 1.40_1 2012-09-04
569
570 [BUG FIXES]
571
572 Debian/Ubuntu have moved unixODBC into /usr/lib/i386-linux-gnu
573 so look in this dir for unixODBC as well - thanks to Meastro for finding.
574
575 Fixed rt 78838
576 I had a sequence point error which is only seen with some compilers
577 as it is sometimes optimized out. It could cause DBD::ODBC to omit
578 adding the UID/PWD to the end of the connection string when using DSN=.
579 Thanks to Zsolt Cserna for spotting it and to ilmari and Leon for
580 explaining it to me.
581
582 Fixed rt 79397
583 Output bound parameters may be incorrectly bound if changed after
584 bind_param_inout is called. If you start with an undef bound param
585 and change it to a defined string/number less than 28 characters
586 before calling execute the original undef will probably be bound.
587 Thanks to runrig on perl monks for providing an example.
588
589 [CHANGE IN BEHAVIOUR]
590
591 If you attempt to bind an rv without amagic DBD::ODBC will now
592 croak - related to rt 78838.
593
594 1.39 2012-07-07
595
596 [BUG FIXES]
597
598 Manifest mentioned 2 files in examples which do not exist - they
599 should have been execute_for_fetch.pl.
600
601 execute_for_fetch.pl example had not be changed since
602 odbc_disable_array_operations became odbc_array_operations.
603
604 1.38_3 2012-06-25
605
606 [BUG FIXES]
607
608 Added encoding line to this file to stop pod test from complaining.
609
610 [DOCUMENTATION]
611
612 Added link to 64 bit ODBC article.
613
614 Fixed some typos in the pod.
615
616 [MISCELLANEOUS]
617
618 Made pod.t an author test.
619
620 1.38_2 2012-05-24
621
622 [ENHANCEMENTS]
623
624 Added support for Oracle TAF (assuming your ODBC driver supports it)
625 - see odbc_taf_callback.
626
627 1.38_1 2012-05-19
628
629 [BUG FIXES]
630
631 Fixed rt 77283. If you overrode the bind type as SQL_INTEGER in a bind_col
632 call AFTER previously binding as another type (or not specifying a type)
633 you would not get the right value back. This also fixes the DiscardString
634 bind_col attribute for SQL_INTEGER binds (however, see below as DiscardString
635 is no longer required for SQL_INTEGER).
636
637 Fixed some format specifiers in trace calls.
638
639 [CHANGE IN BEHAVIOUR]
640
641 DBD::ODBC allowed you to change the bound column type in bind_col
642 after the column was already bound. It now does not allow this
643 and issues a warning.
644
645 You can nolonger override the bound column type (except with
646 SQL_NUMERIC and SQL_DOUBLE). All columns are now bound as either
647 SQL_C_LONG (integer columns) or SQL_C_[W]CHAR (all other columns).
648 If you are calling bind_col with a TYPE => xxx it most likely did
649 not do what you expected and you should examine it carefully with a
650 view to removing it altogether. As a result you no longer have to
651 override the bind type for MS SQL Server XML columns - these will be
652 bound as SQL_C_CHAR or SQL_C_WCHAR depending on whether Unicode is
653 enabled.
654
655 Integer columns are now bound as SQL_C_LONGs and not as before,
656 SQL_C_CHAR. This should not matter to you but if you were adding 0
657 to your integer columns retrieved to make them behave like integers
658 you should nolonger need to do it.
659
660 [OTHER]
661
662 Added some missing SQL_C_xxx types to S_SqlCTypeToString internal
663 function. This only affects tracing.
664
665 Some tests in 08bind were skipped when they did not need to be.
666
667 sql_type_cast tests rewritten due to fixes above.
668
669 1.37 2012-04-07
670
671 A full release of the 1.36 dev releases.
672
673 Please note the changes in behaviour below.
674
675 1.36_2 2012-03-31
676
677 [BUG FIXES]
678
679 * not strictly a bug fix, more a workaround. MS Access mdb driver
680 reports 22 for the size of doubles but then returns an empty string
681 for long doubles that still fit in 22 chrs. rt 69864.
682
683 [CHANGE IN BEHAVIOUR]
684
685 * The odbc_disable_array_operations has been replaced with
686 odbc_array_operations and the default for array operations is off.
687 Sorry, but I warned this was experimental. The
688 ODBC_DISABLE_ARRAY_OPERATIONS env var remains.
689
690 [DOCUMENTATION]
691
692 * Rewrote parts of "Unicode implementation in DBD::ODBC" to describe
693 UTF-16/UCS-2 internal implementation.
694
695 * Changed ordering of some of the pod sections.
696
697 1.36_1 2012-03-21
698
699 [BUG FIXES]
700
701 * Fixed 12blob.t skip count when driver does not have a big enough
702 varchar column to run the test.
703
704 * Work around problems hit during the test suite in DB2 driver which
705 only allows rows of size up to the page size and varchars of 4K.
706
707 * Fix bug in execute_for_fetch where it would ignore the parameters
708 processed when creating the ArrayTupleStatus and hence could
709 attempt to call SQLGetDiagRec on a parameter row which was not
710 executed. See the logs in rt 75687 which show this although this
711 is not a fix for this rt.
712
713 * The code wasn't catching success with info from SQLExecute in
714 execute_for_fetch.
715
716 * Add support for drivers (DB2 in this case) which return
717 SQL_PARAM_DIAG_UNAVAILABLE in bound parameter status array when
718 running execute_for_fetch. Any driver returning doing
719 SQL_PARC_NO_BATCH from SQLGetInfo(SQL_PARAM_ARRAY_ROW_COUNTS)
720 might do this.
721
722 * Fix test code for execute_for_fetch to a) handle drivers doing
723 SQL_PARC_NO_BATCH and b) add "not null" to primary key fields for
724 drivers like DB2 which require it.
725
726 [CHANGE IN BEHAVIOUR]
727
728 * In execute_for_fetch set the parameter status array to all 9999
729 (which is invalid) so we can see if the ODBC driver actually sets
730 them and we can warn if they don't.
731
732 * For freeTDS default odbc_disable_array_operations to 1 as no
733 version of the freeTDS driver can be found that works. I was
734 requested to do this by the dbix-class guys. I may revert this
735 later if freeTDS is fixed.
736
737 * as above for MS Access. It is a shame I cannot find any way of
738 finding out if a driver is capable of array operations.
739
740 [ENHANCEMENTS]
741
742 * execute_for_fetch code now checks the
743 ODBC_DISABLE_ARRAY_OPERATIONS environment variable which can
744 be set to 1 or 0 to override the internal default.
745
746 [DOCUMENTATION]
747
748 * Fixed ColAttributes example in pod which used a $dbh instead of a
749 $sth.
750
751 * Fixed DescribeCol example in pod which used a $dbh instead of a
752 $sth.
753
754 * new FAQ on SQLRowCount, freeTDS and execute_for_fetch
755
756 * Fix typo shown in rt 75860.
757
758 [OTHER]
759
760 * Reduced usage of D_imp_xxx to avoid calls to dbih_getcom2. See
761 thread on dbi-dev at
762 http://www.mail-archive.com/[email protected]/msg06675.html
763
764 * Changed the 70execute_array.t test to run it twice, once using
765 DBI's methods and once using the native one in DBD::ODBC.
766
767 * Made the 2 unicode tests work with DB2 ODBC driver.
768
769 1.35 2012-03-06
770
771 Full release of the 1.34 development releases
772
773 1.34_7 2012-03-02
774
775 [BUG FIXES]
776
777 * Fixed more compiler errors highlighed by a smoker using MS Visual
778 C where some code come before a variable definition.
779
780 1.34_6 2012-02-27
781
782 [BUG FIXES]
783
784 * Fixed some compiler warnings and a compile error highlighed by a
785 smoker using MS Visual C where some code come before a variable
786 definition.
787
788 1.34_5 2012-02-17
789
790 [BUG FIXES]
791
792 * The 40UnicodeRoundTrip tests counts could be 1 off in some cases.
793
794 * Fix for t/03batt.t which could fail a test if the data source had
795 no table - Kenichi Ishigaki
796
797 * If a driver misbehaves during global destruction e.g. SQLFreeStmt
798 fails but no error is available DBD::ODBC issues an error saying
799 an error occurred but no error diagnostics could be found. This is
800 pointless and irritating during global destruction. This stems
801 from a change in 1.28. Thanks to Peter Rabbitson for reporting
802 and suggested fix.
803
804 [CHANGE IN BEHAVIOUR]
805
806 * Prior to this release if you called selectall_* methods with a
807 non-select statement DBD::ODBC would raise an error saying "no
808 select statement currently executing". See RT 68720. After
809 discussions on dbi-dev the concensus seems to be that issuing a
810 warning in this case is better so that is what I've done. As a
811 result t/rt_68720.t has been removed and
812 t/85_selectall_non_select.t has been added.
813
814 [DOCUMENTATION]
815
816 * odbc_getdiagfield was incorrectly named odbc_getdiagrec in the pod
817
818 * add DBI version required for StrictlyTyped and DiscardString to
819 pod
820
821 * Added new FAQ on why a transaction may be committed when
822 AutoCommit is turned off.
823
824 [OTHER]
825
826 * Make examples\odbc_diag.pl more tolerant of drivers which do not
827 handle diagnostic calls properly.
828
829 * Make t/40UnicodeRoundTrip.t work with SQLite - Kenichi Ishigaki
830
831 * Make t/odbc_describe_parameter.t work with SQLite - Kenichi
832 Ishigaki
833
834 * Add 80_odbc_diags.t based on the same file in examples
835
836 1.34_4 2012-02-05
837
838 [BUG FIXES]
839
840 * When odbc_getdiag* methods were added they installed themselves
841 into DBI but did not set IMP_KEEP_ERR so calling them cleared
842 DBI's errors.
843
844 1.34_3 2012-02-03
845
846 [BUG FIXES]
847
848 * Linking against unixODBC was working by accident on most UNIX
849 machines and depended on the order of the files in /usr/lib (or
850 wherever) and what files there were (e.g. an archive or a shared
851 object). Same applied to iODBC but it was more broken especially
852 on machines where libiodbc.so.N.N existed but there was no
853 libiodbc.so which could lead to no adding the shared object at
854 all. I doubt anyone really noticed this but I did eventually on
855 Ubuntu where libiodbc.so.N.N existed but libiodbc.so did not.
856
857 [ENHANCEMENTS]
858
859 * Added experimental odbc_getdiagrec and odbc_getdiagrec methods,
860 examples/odbc_diag.pl and examples/params_in_error.pl.
861
862 [DOCUMENTATION]
863
864 * New FAQ entries.
865
866 1.34_2 2012-01-25
867
868 [BUG FIXES]
869
870 * Fixed rt73734 - debian moved where unixODBC libs are stored.
871
872 * Fixed memory leak of the parameter status array introduced in
873 previous release when execute_for_fetch used. When the statement
874 handle is destroyed the parameter status array was not freed.
875
876 [ENHANCEMENTS]
877
878 * Added environment variable PERL_DBD_ODBC_PREFER_UNIXODBC as a
879 synonym for -x from Rafael Kitover (Caelum).
880
881 [DOCUMENTATION]
882
883 * Add a deviation from DBI spec for type_info_all.
884
885 [OTHER]
886
887 * Added example execute_for_fetch.pl
888
889 1.34_1 2011-12-11
890
891 [ENHANCEMENTS]
892
893 * Added experimental execute_for_fetch support and associated
894 attributes odbc_batch_size and odbc_disable_array_operations.
895
896 1.33 2011-12-01
897
898 This is simply the official release of the 1.32 development
899 releases.
900
901 1.32_5 2011-11-24
902
903 [ENHANCEMENTS]
904
905 * Enable multiple active statement support in 70execute_array.t for
906 drivers we recognise which support MAS.
907
908 * Change column_info to support Unicode catalog/schema/table/column
909 names.
910
911 1.32_4 2011-11-22
912
913 [BUG FIXES]
914
915 * remove debugging printf which output "HERE" in some rare cases.
916 rt 72534 - thanks John Deighan for spotting this.
917
918 * The test 70execute_array.t could fail due to warning being output
919 if the driver does not support Multiple Active Statements.
920
921 [ENHANCEMENTS]
922
923 * Use SQLGetTypeInfoW on unicode builds.
924
925 1.32_3 2011-11-15
926
927 [BUG FIXES]
928
929 * Fix bug in utf16_copy which was not adding a trailing NUL but I'm
930 not sure this affected anyone until I changed table_info this
931 release.
932
933 [ENHANCEMENTS]
934
935 * DBD::ODBC now allows unicode catalog/schema/table parameters to be
936 passed to table_info. Of course they will only reliably work with
937 a supporting Unicode ODBC driver.
938
939 1.32_2 2011-10-22
940
941 [ENHANCEMENTS]
942
943 * Added new odbc_driver_complete attribute allowing the ODBC Driver
944 Manager and ODBC Driver to throw dialogues for incomplete
945 connection strings or expired passwords etc.
946
947 [OTHER]
948
949 * added more examples
950
951 [DOCUMENTATION]
952
953 * new FAQ entries
954
955 * added note saying you cannot pass unicode schema/table/column
956 names to metadata calls like table_info/column_info currently.
957
958 1.32_1 2011-06-24
959
960 [BUG FIXES]
961
962 * I omitted rt_68720.t from the 1.31 distribution which leads
963 to a warning as it is mentioned in the MANIFEST.
964
965 [OTHER]
966
967 * Changed line endings in README.af and README.unicode to be unix
968 line endings and native eol-style in subversion.
969
970 * Minor changes to Makefile.PL to save the opensuse guys patching.
971
972 * Added unicode_sql.pl and unicode_params.pl examples
973
974 1.31 2011-06-21
975
976 [BUG FIXES]
977
978 Recently introduced test sql_type_cast.t cannot work with DBI less
979 than 1.611.
980
981 Minor change to Makefile.PL to avoid use of unitialised warning on
982 $ENV{LD_LIBRARY_PATH} in warning when it is not set.
983
984 1.30_7 2011-06-15
985
986 [BUG FIXES]
987
988 Some time ago (I don't know when) a few internal errors generated by
989 DBD::ODBC got ignored. There are about 5 of them but I seriously
990 doubt anyone would hit any other than the data truncated error
991 (which is reported by the ODBC driver anyway) and "no select
992 statement currently executing". You can see rt_68720.t in the t
993 directory for an example of the latter.
994
995 [ENHANCEMENTS]
996
997 An extra argument has been added to the sub associated with
998 odbc_err_handler. The arguments passed to the odbc_err_handler are
999 now state (string), error (string), native error code (number) and
1000 the status returned from the last ODBC API. The status will be
1001 SQL_ERROR (-1) for errors or SQL_SUCCESS_WITH_INFO (1) for
1002 informational messages.
1003
1004 1.30_6 2011-06-04
1005
1006 [BUG FIXES]
1007
1008 * When DBD::ODBC calls SQLExecDirect (the do method) it was not
1009 reporting informational diagnostics (SQL_SUCCESS_WITH_INFO) and
1010 not calling the error handler.
1011
1012 Arguably, this is a change in behaviour but one I've struggled to
1013 resolve since in all other cases of warnings DBD::ODBC's error
1014 handler is called. However, DBI does not call its error handler
1015 for warnings so was DBD::ODBC wrong to call it's error in the
1016 first place for warnings? I decided it was better to leave this as
1017 it is but add checking of SQLExecDirect/do. Apart from anything
1018 else if DBD::ODBC does not call its error handler for
1019 informational diagnostics there is no way to retrieve print
1020 statements etc from procedures.
1021
1022 * The odbc_describe_parameter.t test could fail with some versions
1023 of MS SQL Server ODBC Driver. It was down to when
1024 SQLDescribeParameter is disabled, the column_size passed to
1025 SQLBindParameter is 0.
1026
1027 * pod example of odbc_err_handler incorrectly used \$err_handler
1028 instead of \&err_handler.
1029
1030 1.30_5 2011-05-24
1031
1032 [BUG FIXES]
1033
1034 * The change in behavior detailed in 1.30_1 for wide character
1035 binding was still not working properly (see
1036 http://rt.cpan.org/Ticket/Display.html?id=67994). It was working
1037 for SQL_CHAR described columns but not SQL_VARCHAR.
1038
1039 1.30_4 2011-05-18
1040
1041 [BUG FIXES]
1042
1043 * Fix issue described in
1044 http://www.nntp.perl.org/group/perl.dbi.dev/2011/05/msg6567.html.
1045 If you prepare a statement, disconnect and then try and execute the
1046 statement you get an error but it does not tell what is wrong.
1047
1048 [ENHANCEMENTS]
1049
1050 * Added support for StrictlyTyped and DiscardString to the bind_col
1051 method.
1052
1053 [OTHER]
1054
1055 * Minor changes to META.yml for mailing list, dynamic_config,
1056 homepage and keywords.
1057
1058 * The pod was missing = before the heads on a couple of sections in
1059 "Private DBD::ODBC Functions"
1060
1061 * TreatAsLob was incorrectly documented as BindAsLob.
1062
1063 1.30_3 2011-05-17
1064
1065 [BUG FIXES]
1066
1067 * Made the new odbc_describe_parameters work and added test case.
1068
1069 1.30_2 2011-05-16
1070
1071 [ENHANCEMENTS]
1072
1073 * Added the new odbc_describe_parameters attribute.
1074
1075 1.30_1 2011-05-12
1076
1077 [BUG FIXES]
1078
1079 * Fixed some compiler warnings shown with -Wall including some
1080 printf formats that had extra/missing arguments.
1081
1082 * Fixed t/70execute_array.t which was missing an "order by" in the
1083 check_data sub which could cause failures for drivers not
1084 returning the rows in the order they were inserted.
1085
1086 * Minor fix to Makefile.PL to avoid issuing variable used in void
1087 context.
1088
1089 [CHANGE IN BEHAVIOUR]
1090
1091 * DBD::ODBC used to quietly rollback any transactions when
1092 disconnect was called and AutoCommit was off. This can mask a
1093 problem and leads to different behaviour when disconnect is called
1094 vs not calling disconnect (where you get a warning). This release
1095 issues a warning if you call disconnect and a transaction is in
1096 progress then it is rolled back.
1097
1098 * DBD::ODBC used to bind char/varchar/longvarchar columns as SQL_CHAR
1099 meaning that in the unicode build of DBD::ODBC the bound column
1100 data would be returned 8bit in whatever character-set (codepage) the
1101 data was in, in the database. This was inconvenient and arguably a
1102 mistake. Columns like nchar/nvarchar etc were bound as SQL_WCHAR and
1103 returned as Unicode. This release changes the behaviour in a unicode
1104 build of DBD::ODBC to bind all char columns as SQL_WCHAR. This may
1105 inconvenience a few people who expected 8bit chars back, knew the
1106 char set and decoded them (sorry). See odbc_old_unicode to return
1107 to old behaviour.
1108
1109 [ENHANCEMENTS]
1110
1111 * added -w option to Makefile.PL to add "-Wall" to CCFLAGS and
1112 -fno-strict-aliasing so I can find warnings.
1113
1114 * Cope with broken ODBC drivers that describe a parameter as SQL
1115 type 0.
1116
1117 [OTHER]
1118
1119 * Add "disconnect and transactions" to pod describing what DBD::ODBC
1120 does if you call disconnect with an outstanding transaction.
1121
1122 * Reorganised FAQ for bound parameter issues and added a lot on
1123 bound parameter problems.
1124
1125 * Added new FAQ entry for Firebird
1126
1127 * Removed some unused variables and added some missing function
1128 prototypes
1129
1130 1.29 2011-03-08
1131
1132 * An official release of the 1.28 development releases.
1133
1134 [NOTE]
1135
1136 * The 1.28 development releases made a change which causes a generic
1137 error to be reported when an ODBC call fails but an error message
1138 is not retrieved from the ODBC Driver. It appears this has caught
1139 out a few ODBC drivers - notably freeTDS and Firebird. You now may
1140 see errors that were missed before e.g., DBIx::Class's tests for
1141 Firebird now errors test 21 "outer txn rolled back" (and others)
1142 because SQLRowCount returns an error after "ROLLBACK TO SAVEPOINT
1143 savepoint_0"; before this error was missed.
1144
1145 1.28_5 2011-03-06
1146
1147 [BUG FIXES]
1148
1149 * rt_59621.t had wrong skip count
1150
1151 * Fixed missing SQL_MAX_TABLE_NAME_LEN definition from test.
1152
1153 * Fixed problem with some drivers which batch "insert;select" where
1154 SQLMoreResults is not required and an extra describe is done.
1155
1156 * Fixed "select 1" in 02simple.t for Firebird ODBC Driver.
1157
1158 * disconnect call added to 70execute_array.t was in the wrong place.
1159
1160 * In non-unicode mode we bind strings as SQL_CHAR but the driver may
1161 have described them as SQL_WCHAR and we were not doing ChopBlanks
1162 processing in that case.
1163
1164 [REQUIREMENTS]
1165
1166 * Now needs Test::Simple 0.90.
1167
1168 [OTHER]
1169
1170 * Added dml_counts.pl example
1171
1172 * worked around a problem in freeTDS in the 20SqlServer.t test
1173 provided by Ralph Doncaster.
1174
1175 * Changed test rt_62033.t to try and make it work with freeTDS - I
1176 failed. It now skips attempts to fetch when the insert fails.
1177
1178 * Worked around problem in Firebird ODBC driver which reports
1179 timestamps have a display size of 24 characters but then can
1180 return 29 which could lead to data truncation errors.
1181 See http://tracker.firebirdsql.org/browse/ODBC-112
1182
1183 * Worked around problem in Firebird ODBC driver which reports
1184 VARCHARs have a maximum length of 32765 but in fact it is 4000.
1185 See http://tracker.firebirdsql.org/browse/ODBC-111
1186
1187 * Improvements in tracing to take account of DBI's neatsvpv lops 5
1188 characters off maxsize on a string.
1189
1190 1.28_4 2011-02-24
1191
1192 [BUG FIXES]
1193
1194 * Fixed compilation problems with DBIf_TRACE_TXN
1195
1196 * Added missing disconnect call to 70execute_array.t
1197
1198 1.28_3 2011-02-22
1199
1200 [BUG FIXES]
1201
1202 * Fixed MANIFEST in 1.28_2 which named 2 files incorrectly.
1203
1204 * Fixed use of PREREQ_PRINT in versions of ExtUtils::MakeMaker which
1205 don't support it.
1206
1207 * Fixed a check on LD_LIBRARY_PATH for unixODBC which could report
1208 you've not set LD_LIBRARY_PATH correctly.
1209
1210 [ENHANCEMENTS]
1211
1212 * Added Perl and ExtUtils::MakeMaker version output to build process.
1213
1214 * Added support for DBI's new trace flags ENC, CON, TXN and
1215 DBD. From DBI 1.617 you should be able to use: DBI_TRACE=DBD to
1216 ONLY get DBD::ODBC tracing without DBI tracing ENC and CON DBI
1217 flags are synonymous with DBD::ODBC's odbcconnection and
1218 odbcunicode trace flags which you can still use for now.
1219
1220 [OTHER]
1221
1222 * From now on I'm changing the way the Changes file is written as
1223 per article at
1224 http://blog.urth.org/2011/01/changes-file-how-and-how-not-to.html
1225
1226 * Some broken drivers (freeTDS in this case) can return SQL_ERROR
1227 from an ODBC API function and then SQLError does not return
1228 error details. In this case set a generic error saying an error
1229 occurred but we could not retrieve it.
1230
1231 * Added FAQ entry on MS Access and text columns limited to 255 chrs.
1232
1233 * Added 70execute_array.t test.
1234
1235 1.28_2 2011-01-24
1236
1237 Added -x argument to Makefile.PL saying you prefer unixODBC over
1238 iODBC driver managers as a) we need to look for iODBC first on some
1239 platforms to detect iODBC and b) some platforms (Debian/Ubuntu)
1240 people have both binary packages installed but only unixODBC dev
1241 package.
1242
1243 Patch from Rafael Kitover (Caelum) for better Cygwin handling.
1244
1245 Minor change to data sources test to cope with someone having no
1246 data sources and using a DSN-less connection for testing.
1247
1248 Fixed MARS test when a DSN-less connection used for testing - thanks
1249 to Rafael Kitover (Caelum) for spotting this.
1250
1251 pod patch for "CPAN Testers Reporting" to point at cpan testers wiki
1252 from Rafael Kitover (Caelum).
1253
1254 Fixed some broken links in the FAQ.
1255
1256 Add a multiple active statement document link to random links and
1257 the FAQ entry.
1258
1259 A call to SQLNumResultCols was not checked to test it succeeded.
1260 Not seen anyone run into this as yet.
1261
1262 1.28_1 2010-12-29
1263
1264 Rewrote documentation on odbc_SQL_ROWSET_SIZE and added loads of
1265 notes so I don't have to go through a huge irc conversation with
1266 ribasushi again.
1267
1268 Workaround bug in DBI (prior to 1.616) which mistakenly issues a
1269 FETCH on any attribute passed to the connect method and sometimes
1270 driver managers error on SQL_ROWSET_SIZE in SQLGetConnectAttr.
1271
1272 ChopBlanks was not working on UCS-2 encoded data written into bound
1273 columns and was also sometimes reading off the end of the bound
1274 array.
1275
1276 Minor FAQ changes:
1277 Added an additional way to read MS Access dbs from Unix
1278 Clarified versions for MARS_Connection
1279 updates to cancel_big_fetch.pl
1280
1281 Updated TO_DO with more stuff to do
1282
1283 Improved tracing output
1284
1285 Tidied up some of the examples
1286
1287 1.27 2010-12-29
1288
1289 Official release of the 1.26 development releases.
1290
1291 1.26_4 2010-12-14
1292
1293 Fixed bug highlighted by investigation into rt 62033 where the
1294 active flag was sometimes not set on the statement after
1295 SQLMoreResults indicates a result-set exists.
1296
1297 Fix rt 63550 reported by John Corcoran where if a do method call
1298 fails the SQL C<Statement> is not available in an error handler as
1299 we never created a DBI statement in the first place. Added a note
1300 to "do" deviations pod.
1301
1302 Minor fix to head at wrong level in the pod.
1303
1304 Fix a possible snprintf buffer overflow in GetTypeInfo when
1305 the type is specified and it is negative.
1306
1307 1.26_3 2010-11-18
1308
1309 Fixed rt 63108. The change to column binding in 1.24_2 was not
1310 complete and causes bound columns to be rebound on each execute
1311 wasting time and leaking memory. Found, diagnosed and proposed fix
1312 by Steve Bentley.
1313
1314 1.26_2 2010-11-09
1315
1316 Fixed bug found by frew where an snprintf can overflow when binding
1317 a lot of parameters.
1318
1319 1.26_1 2010-10-24
1320
1321 There are over 200 block changes in the source code for this release
1322 from the previous one (see below). If you are using DBD::ODBC in
1323 production you should not upgrade without testing this release first
1324 as it introduces a lot of internal changes. DBD::ODBC has now gone
1325 entirely ODBC 3 and relies on an ODBC Driver Manager to map calls
1326 to ODBC 2.0 drivers (why are you still using ODBC 2.0 drivers?).
1327 From now on DBD::ODBC needs to be linked with an ODBC Driver Manager
1328 and I recommend unixODBC on UNIX and the MS ODBC Driver Manager
1329 on Windows. There are really good reasons for this but mainly it
1330 is because an ODBC Driver Manager will map ODBC 2.0 calls to an
1331 ODBC 3.0 driver and vice versa and handle UNICODE transparently.
1332
1333 Bumped Test::Simple requirement up to 0.90 so we can use done_testing
1334 and note.
1335
1336 Bump Perl requirement to 5.8 as per DBI.
1337
1338 Workaround a bug in mdbtools which fails to set the out connection
1339 string in a SQLDriverConnect call. This can lead to:
1340
1341 Fixed panic: sv_setpvn called with negative strlen at
1342 blib/lib/DBD/ODBC.pm line 107.
1343
1344 Added rt_61370.t for rt 61370.
1345
1346 Removed last remaining sprintf calls and replaced with snprintf.
1347
1348 Changed the point at which DBD::ODBC switches from VARCHAR to
1349 LONGVARCHAR or WVARCHAR to WLONGVARCHAR when SQLDesribeParam fails. It
1350 was 4000 and is now 2000 for unicode builds. Works around a daft issue
1351 in the MS SQL Server driver which will not allow 'x' x 2001 converted to
1352 wide characters to be inserted into a varchar(8000).
1353
1354 Minor change to Makefile.PL to print out found libs for iODBC and
1355 unixODBC.
1356
1357 Added some FAQs for problems with iODBC and a recent bug in DBI.
1358
1359 Added FAQ on my_snprintf problem.
1360
1361 Fixed some pod errors in FAQ document.
1362
1363 Fixed warning on 64 bit Strawberry Perl when compiling dbdimp.c for
1364 cast from ptr to UDWORD.
1365
1366 Moved to_do items from Changes to TO_DO.
1367
1368 Reformatted this file to save Jens work.
1369
1370 Changed calls to SQLTransact (ODBC 2.0) to SQLEndTran (ODBC 3.0).
1371 There really shouldn't be any ODBC 2.0 drivers still around but even
1372 if there are, the ODBC driver manager should do the mapping for us.
1373
1374 Changed calls to SQLGetConnectOption/SQLSetConnectOption to
1375 to SQLGetConnectAttr/SQLSetConnectAttr for ODBC 3.0.
1376
1377 Changed calls to SQLColAttributes (ODBC 2.0) to SQLColAttribute
1378 (ODBC 3.0).
1379
1380 Bumped requirement on DBI to 1.609 because that is the first version
1381 to contain a dbipport.h which defined my_snprintf - see
1382 https://rt.cpan.org/Public/Bug/Display.html?id=62346 and
1383 http://www.nntp.perl.org/group/perl.dbi.dev/2010/10/msg6335.html
1384
1385 Various small changes to dbdimp.c which should have no effect other
1386 than to make it leaner:
1387
1388 Removed all dTHR occurrences from dbdimp.c as it is a NOOP since 5.8
1389 and we need 5.8 at least.
1390 Removed dbd_param_err as it was not used
1391 Removed odbc_get_query_timeout as it was never compiled
1392 Removed eod field from statement as it was never used
1393 Removed a load of commented out code
1394 Replaced some SvPV calls with SvPV_nolen when we didn't used the
1395 length
1396 Removed some silly code from dbd_db_execdirect which attempted to
1397 return -3 - don't think it could ever get run.
1398 Minor tracing output tidy up
1399 Removed dbd_caution as it was no used
1400 Localised more variable declarations
1401
1402 1.25 2010-09-22
1403
1404 Official release of 1.25 combining all the changes in the 1.24_x
1405 development releases.
1406
1407 1.24_6 2010-09-16
1408
1409 rt 61370 - default XML type parameters in SQL Server to SQL_WCHAR so
1410 they accept unicode strings.
1411
1412 1.24_5 2010-09-15
1413
1414 Fixed missing SvSETMAGIC on a bound scalar which was causing length() to
1415 return the wrong result - see http://www.perlmonks.org/?node_id=860211
1416 and a big thank you to Perl Monks and in particular ikegami.
1417
1418 Changed bind_col so it actually pays attention to the TYPE attribute as
1419 you could not override the bind type of a bound column before.
1420
1421 1.24_4 2010-09-08
1422
1423 Left a sv_undef in - thanks smoke testers for finding that.
1424
1425 Change sprintf to snprintf for safety.
1426
1427 1.24_3 2010-09-06
1428
1429 Added note from Robert Freimuth for obtaining the last insert ID in MS
1430 Access.
1431
1432 Changed all &sv_yes/&sv_no occurrances in XS to PL_sv_yes/PL_sv_no as
1433 the originals have now gone from blead Perl.
1434
1435 Minor change to fix missing newline in trace output.
1436
1437 Added a FAQ entry for how "use regional settings" in MS SQL Server
1438 breaks things.
1439
1440 1.24_2 2010-07-23
1441
1442 Fix rt57957 reported by Marc Prewitt. DBD::ODBC was not ignoring named
1443 placeholders and ? inside comments. Comments are deemed as text between
1444 "/*" and "*/" (if not in a literal) and line comments begin with "--".
1445
1446 Added a FAQ on procedures not completing in MS SQL Server. Thanks to
1447 Peter Rabbitson for hitting this problem and reminding me I'd seen it a
1448 couple of times before.
1449
1450 Added a FAQ on equality comparisons with ntext columns.
1451
1452 Added pod for last_insert_id which is not currently supported.
1453
1454 Fix bug where if SQLMoreResults was called and failed it was not
1455 reported.
1456
1457 Removed some unused fields from the fbh structure which should save a
1458 little memory for each column in a result-set.
1459
1460 Started adding support for DBI's DiscardString and StrictlyTyped but not
1461 complete yet so don't use them yet.
1462
1463 Added experimental odbc_lob_read method - see pod. Thanks to tye and
1464 ikegami on perlmonks for pointing out some problems with my initial
1465 implementation.
1466
1467 Moved the binding of columns to the first call to fetch instead of after
1468 execute is called as it prevents bind_col overrriding the type used for
1469 binding and I needed it to support odbc_lob_read. This may have
1470 undesired affects so any testing of this release would be appreciated.
1471
1472 Added bind_col method so DBD::ODBC can support attributes on a bind_col
1473 call.
1474
1475 Removed support for DBI's blob_read - it was totally flawed and did not
1476 work at all. May replace in the future.
1477
1478 Added support for MS SQL Server XML type (SQL type -152). See rt 59621.
1479
1480 Added note on do method implementation in DBD::ODBC and how some may
1481 consider it to deviate from the DBI specification.
1482
1483 1.24_1 2010-05-27
1484
1485 Corrected pod and private attributes for the odbc_SQL_DRIVER_ODBC_VER
1486 attribute which was documented as SQL_DRIVER_ODBC_VER.
1487
1488 Added FAQ on pauses on statement destruction when all the result-set has
1489 not been retrieved (mostly freeTDS and MS SQL Server ODBC Driver).
1490
1491 Fixed bug introduced in 1.24 where if you are using MS SQL Server, and
1492 you are preparing, binding placeholders and re-executing multiple times
1493 you may get a "Invalid character value for cast specification"
1494 error. Thanks to anonymous for spotting this and producing a standalone
1495 example of the problem that made it so much easier to find.
1496
1497 1.24 2010-05-14
1498
1499 Minor change in Makefile.PL to only use NO_META if ExtUtils::MakeMaker
1500 is at least at version 6.10. Reported by Chunmei Wu.
1501
1502 Minor change to test rt_50852 which had wrong skip count.
1503
1504 1.23_5 2010-05-06
1505
1506 Added advice from Jan Dubois (ActiveState) on building DBD::ODBC for
1507 ActivePerl (see README.windows).
1508
1509 rt56692. Fix spelling mistake in DBD::ODBC pod - thanks to Ansgar
1510 Burchardt.
1511
1512 Added a 7th way to help documentation - become a tester.
1513
1514 Hopefully fixed problems building on windows 32 bit platforms that have
1515 old sql header files not mentioning SQLLEN/SQLULEN.
1516
1517 1.23_4 2010-04-13
1518
1519 Added more FAQs.
1520
1521 Small optimization to remove calls to SQLError when tracing is not
1522 turned on. This was a bug. We only need to call SQLError when SQLExecute
1523 succeeds if there is an error handler or if tracing is enabled. The test
1524 was for tracing disabled!
1525
1526 Large experimental change primarily affecting MS SQL Server users but it
1527 does impact on other drivers too. Firstly, for MS SQL Server users we no
1528 longer SQLFreeStmt(SQL_RESET_PARAMS) and rebind bound parameters as it
1529 is causing the MS SQL Server ODBC driver to re-prepare the SQL.
1530 Secondly (for all drivers) we no longer call SQLBindParameter again IF
1531 all the arguments to it are the same as the previous call. If you find
1532 something not working you better let me know as this is such a speed up
1533 I'm going to go with this unless anyone complains.
1534
1535 Minor change to avoid a double call to SQLGetInfo for SQL_DBMS_NAME
1536 immediately after connection.
1537
1538 Small change for rt 55736 (reported by Matthew Kidd) to not assume a
1539 parameter is varXXX(max) if SQLDescribeParam failed in the Microsoft
1540 Native Client driver.
1541
1542 1.23_3 2010-03-24
1543
1544 Minor changes to Makefile.PL and dbdimp.c to remove some compiler
1545 warnings.
1546
1547 Fix some calls to SQLMoreResults which were not passing informational
1548 messages on to DBI's set_err. As you could not see all the informational
1549 messages from procedures, only the first.
1550
1551 Fix minor issue in 02simple test which printed the Perl subversion
1552 before the version.
1553
1554 Changes to 20SqlServer.t to fix a few typos and make table names
1555 consistent wrt to case - (as someone had turned on case-sensitivity in
1556 SQL Server) Similar changes in rt_38977.t and rt_50852.t
1557
1558 1.23_2 2010-01-26
1559
1560 Fixed bug in Makefile.PL which could fail to find unixODBC/iODBC header
1561 files but not report it as a problem. Thanks to Thomas J. Dillman and
1562 his smoker for finding this.
1563
1564 Fixed some compiler warnings in dbdimp.c output by latest gcc wrt to
1565 format specifiers in calls to PerlIO_printf.
1566
1567 Added the odbc_force_bind_type attribute to help sort out problems with
1568 ODBC Drivers which support SQLDescribeParam but describe the parameters
1569 incorrectly (see rt 50852). Test case also added as rt_50852.t.
1570
1571 1.23_1 2009-10-21
1572
1573 makefile.PL changes:
1574 some formatting changes to output warn if unixodbc headers are not
1575 found that the unixodbc-dev package is not installed use $arext
1576 instead of "a"
1577 pattern match for pulling libodbc.* changed
1578 warn if DBI_DSN etc not defined
1579 change odbc_config output for stderr to /dev/null
1580 missing / on /usr/local wheb finding find_dm_hdr_files()
1581
1582 New FAQ entries from Oystein Torget for bind parameter bugs in SQL
1583 Server.
1584
1585 rt_46597.rt - update on wrong table
1586
1587 Copied dbivport.h from the latest DBI distribution into DBD::ODBC.
1588
1589 Added if_you_are_taking_over_this_code.txt.
1590
1591 Add latest Devel::PPPort ppport.h to DBD::ODBC and followed all
1592 recommendations for changes to dbdimp.c.
1593
1594 Added change to Makefile.PL provided by Shawn Zong to make
1595 Windows/Cygwin work again.
1596
1597 Minor change to Makefile.PL to output env vars to help in debugging
1598 peoples build failures.
1599
1600 Added odbc_utf8_on attribute to dbh and sth handles to mark all strings
1601 coming from the database as utf8. This is for Aster (based on
1602 PostgreSQL) which returns all strings as UTF-8 encoded unicode. Thanks
1603 to Noel Burton-Krahn.
1604
1605 1.23 2009-09-11
1606
1607 Only a readme change and version bumped to 1.23. This is a full release
1608 of all the 1.22_x development releases.
1609
1610 1.22_3 2009-08-19
1611
1612 Fix skip count in rt_38977.t and typo in ok call.
1613
1614 Workaround a bug in unixODBC 2.2.11 which can write off the end of the
1615 string buffer passed to SQLColAttributes.
1616
1617 Fix skip count in rt_null_nvarchar.t test for non SQL Server drivers.
1618
1619 Fix test in 02simple.t which reported a fail if you have no ODBC
1620 datasources.
1621
1622 In 99_yaml.t pick up the yaml spec version from the meta file instead of
1623 specifying it.
1624
1625 Change calls to SQLPrepare which passed in the string length of the SQL
1626 to use SQL_NTS because a) they are null terminated and more importantly
1627 b) unixODBC contains a bug in versions up to 2.2.16 which can overwrite
1628 the stack by 1 byte if the string length is specified and not built with
1629 iconv support and converting the SQL from ASCII to Unicode.
1630
1631 Fixed bug in ping method reported by Lee Anne Lester where it dies if
1632 used after the connection is closed.
1633
1634 A great deal of changes to Makefile.PL to improve the automatic
1635 detection and configuration for ODBC driver managers - especially on
1636 64bit platforms. See rt47650 from Marten Lehmann which started it all
1637 off.
1638
1639 Add changes from Chris Clark for detecting IngresCLI.
1640
1641 Fix for rt 48304. If you are using a Microsoft SQL Server database and
1642 nvarchar(max) you could not insert values between 4001 and 8000
1643 (inclusive) in size. A test was added to the existing rt_38977.t test.
1644 Thanks to Michael Thomas for spotting this.
1645
1646 Added FAQ on UTF-8 encoding and IBM iSeries ODBC driver.
1647
1648 Add support for not passing usernames and passwords in call to connect.
1649 Previously DBD::ODBC would set an unspecified username/password to '' in
1650 ODBC.pm before calling one of the login_xxx functions. This allows the
1651 driver to pull the username/password from elsewhere e.g., like the
1652 odbc.ini file.
1653
1654 1.22_1 2009-06-16
1655
1656 Applied a slightly modified version of patch from Jens Rehsack to
1657 improve support for finding the iODBC driver manager.
1658
1659 A UNICODE enabled DBD::ODBC (the default on Windows) did not handle
1660 UNICODE usernames and passwords in the connect call properly.
1661
1662 Updated "Attribution" in ODBC.pm.
1663
1664 Unicode support is no longer experimental hence warning and prompt
1665 removed from the Makefile.PL.
1666
1667 old_ping method removed.
1668
1669 Fixed bug in 02simple.t test which is supposed to check you have at
1670 least one data source defined. Unfortunately, it was checking you had
1671 more than 1 data source defined.
1672
1673 rt_null_varchar had wrong skip count meaning non-sql-server drivers or
1674 sql server drivers too old skipped 2 tests more than were planned.
1675
1676 1.22 2009-06-10
1677
1678 Fixed bug which led to "Use of uninitialized value in subroutine entry"
1679 warnings when writing a NULL into a NVARCHAR with a unicode-enabled
1680 DBD::ODBC. Thanks to Jirka Novak and Pavel Richter who found, reported
1681 and patched a fix.
1682
1683 Fixed serious bug in unicode_helper.c for utf16_len which I'm ashamed
1684 to say was using an unsigned short to return the length. This meant you
1685 could never have UTF16 strings of more than ~64K without risking
1686 serious problems. The DBD::ODBC test code actually got a
1687
1688 *** glibc detected *** /usr/bin/perl: double free or corruption
1689 (out): 0x406dd008 ***
1690
1691 If you use a UNICODE enabled DBD::ODBC (the default on Windows) and
1692 unicode strings larger than 64K you should definitely upgrade now.
1693
1694 1.21_1 2009-06-02
1695
1696 Fixed bug referred to in rt 46597 reported by taioba and identified by
1697 Tim Bunce. In Calls to bind_param for a given statement handle if you
1698 specify a SQL type to bind as, this should be "sticky" for that
1699 parameter. That means if you do:
1700
1701 $sth->bind_param(1, $param, DBI::SQL_LONGVARCHAR)
1702
1703 and follow it up with execute calls that also specify the parameter:
1704
1705 $sth->execute("a param");
1706
1707 then the parameter should stick with the SQL_LONGVARCHAR type and not
1708 revert to the default parameter type. The DBI docs (from 1.609) make it
1709 clear the parameter type is sticky for the duration of the statement but
1710 some DBDs allow the parameter to be rebound with a different type -
1711 DBD::ODBC is one of those drivers.
1712
1713 1.21 2009-04-27
1714
1715 Change 02simple test to output Perl, DBI and DBD::ODBC versions.
1716
1717 Fixed bug where if ODBC driver supports SQLDescribeParam and it succeeds
1718 for a parameterised query but you override the parameter type, DBD::ODBC
1719 was still using the size returned by SQLDescribeParam. Thanks to Brian
1720 Becker for finding, diagnosing and fixing this issue.
1721
1722 Added FAQ entry about SQL Server and calling procedures with named
1723 parameters out of order.
1724
1725 Added test_results.txt containing some supplied make test results.
1726
1727 1.20 2009-04-20
1728
1729 Fix bug in handling of SQL_WLONGVARCHAR when not built with unicode
1730 support. The column was not identified as a long column and hence the
1731 size of the column was not restricted to LongReadLen. Can cause
1732 DBD::ODBC to attempt to allocate a huge amount of memory.
1733
1734 Minor changes to Makefile.PL to help diagnose how it decided which
1735 driver manager to use and where it was found.
1736
1737 Offer suggestion to debian-based systems when some of unixODBC is found
1738 (the bin part) but the development part is missing.
1739
1740 In 20SqlServer.t attempt to drop any procedures we created if they still
1741 exist at the end of the test. Reported by Michael Higgins.
1742
1743 In 12blob.t separate code to delete test table into sub and call at
1744 being and end, handle failures from prepare there were two ENDs.
1745
1746 In ODBCTEST.pm when no acceptable test column type is found output all
1747 the found types and BAIL_OUT the entire test.
1748
1749 Skip rt_39841.t unless actually using the SQL Server ODBC driver or
1750 native client.
1751
1752 Handle drivers which return 0 for SQL_MAX_COLUMN_NAME_LEN.
1753
1754 Double the buffer size used for column names if built with unicode.
1755
1756 1.19 2009-04-02
1757
1758 Some minor diagnostic output during tests when running against freeTDS
1759 to show we know of issues in freeTDS.
1760
1761 Fixed issue in 20SqlServer.t where the connection string got set with
1762 two consecutive semi-colons. Most drivers don't mind this but freeTDS
1763 ignores everything after that point in the connection string.
1764
1765 Quieten some delete table output during tests.
1766
1767 Handle connect failures in 20SqlServer.t in the multiple active
1768 statement tests.
1769
1770 In 02simple.t cope with ODBC drivers or databases that do not need a
1771 username or password (MS Access).
1772
1773 In 20SqlServer.t fix skip count and an erroneous assignment for
1774 driver_name.
1775
1776 Change some if tests to Test::More->is tests in 02simple.t.
1777
1778 Fix "invalid precision" error during tests with the new ACEODBC.DLL MS
1779 Access driver. Same workaround applied for the old MS Access driver
1780 (ODBCJT32.DLL) some time ago.
1781
1782 Fix out of memory error during tests against the new MS Access driver
1783 (ACEODBC.DLL). The problem appears to be that the new Access driver
1784 reports ridiculously large parameter sizes for "select ?" queries and
1785 there are some of these in the unicode round trip test.
1786
1787 Fixed minor typo in Makefile.PL - diagnostic message mentioned "ODBC
1788 HOME" instead of ODBCHOME.
1789
1790 12blob.t test somehow got lost from MANIFEST - replaced. Also changed
1791 algorithm to get a long char type column as some MS Access drivers only
1792 show SQL_WLONGVARCHAR type in unicode.
1793
1794 Added diagnostic output to 02simple.t to show the state of
1795 odbc_has_unicode.
1796
1797 1.18_4 2009-03-13
1798
1799 A mistake in the MANIFEST lead to the rt_43384.t test being omitted.
1800
1801 Brian Becker reported the tables PERL_DBD_39897 and PERL_DBD_TEST are
1802 left behind after testing. I've fixed the former but not the latter yet.
1803
1804 Yet another variation on the changes for rt 43384. If the parameter is
1805 bound specifically as SQL_VARCHAR, you got invalid precision
1806 error. Thanks to Øystein Torget for finding this and helping me verify
1807 the fix.
1808
1809 If you attempt to insert large amounts of data into MS Access (which
1810 does not have SQLDescribeParam) you can get an invalid precision error
1811 which can be worked around by setting the bind type to
1812 SQL_LONGVARCHAR. This version does that for you.
1813
1814 08bind2.t had a wrong skip count.
1815
1816 12blob.t had strict commented out and GetTypeInfo was not quoted. Also
1817 introduced a skip if the execute fails as it just leads to more obvious
1818 failures.
1819
1820 In dbdimp.c/rebind_ph there was a specific workaround for SQL Server
1821 which was not done after testing if we are using SQL Server - this was
1822 breaking tests to MS Access.
1823
1824 1.18_2 2009-03-09
1825
1826 Added yet another workaround for the SQL Server Native Client driver
1827 version 2007.100.1600.22 and 2005.90.1399.00 (driver version 09.00.1399)
1828 which leads to HY104, "Invalid precision value" in the rt_39841.t test.
1829
1830 1.18_1 2009-03-06
1831
1832 Fixed bug reported by Toni Salomäki leading to a describe failed error
1833 when calling procedures with no results. Test cases added to
1834 20SqlServer.t.
1835
1836 Fixed bug rt 43384 reported by Øystein Torget where you cannot insert
1837 more than 127 characters into a Microsoft Access text(255) column when
1838 DBD::ODBC is built in unicode mode.
1839
1840 1.18 2009-01-16
1841
1842 Major release of all the 1.17 development releases below.
1843
1844 1.17_3 2008-12-19
1845
1846 Reinstated the answer in the FAQ for "Why do I get invalid value for
1847 cast specification" which had got lost - thanks to EvanCarroll in
1848 rt41663.
1849
1850 rt 41713. Applied patch from JHF Remmelzwaal to handle ODBC drivers
1851 which do not support SQL_COLUMN_DISPLAY_SIZE and SQL_COLUMN_LENGTH
1852 attributes in the SQLColAttributes calls after SQLTables and
1853 SQLColumns. Specifically, the driver he was using was the "Infor
1854 Integration ODBC driver".
1855
1856 Added notes from JHF Remmelzwaal on resolving some problems he came
1857 across building DBD::ODBC on Windows with Visual Studio 6.0 and SDK Feb
1858 2003.
1859
1860 New odbc_column_display_size attribute for when drivers does not return
1861 a display size.
1862
1863 Loads of tracing changes to make it easier for me to debug problems.
1864
1865 Fixed bug in tracing of dbd_execute when parameter is char but undef
1866 which was leading to an access violation on Windows when tracing
1867 enabled.
1868
1869 Minor changes to diagnostic output in some rt tests.
1870
1871 One of the rt tests was not skipping the correct number of tests if the
1872 driver was not SQL Server.
1873
1874 1.17_2 2008-11-17
1875
1876 Changed ParamTypes attribute to be as specification i.e.,
1877
1878 {
1879 parameter_1 => {TYPE => sql_type}
1880 parameter_2 => {TYPE => sql_type}
1881 ...
1882 }
1883
1884 and changed the tests in 07bind.t to reflect this.
1885
1886 A few minor perlcritic changes to ODBC.pm.
1887
1888 Added 99_yaml.t test to check META.yml.
1889
1890 Added patch from Spicy Jack to workaround problems with Strawberry Perl
1891 setting INC on the command line when running Makefile.PL.
1892
1893 1.17_1 2008-10-10
1894
1895 Missing newline from end of META.yml upsets cpan
1896
1897 Add code to Makefile.PL to spot command line containing INC, outline
1898 problem and resolution and not generate Makefile to avoid cpan-testers
1899 failures.
1900
1901 Loads of pod formatting changes including a section in the wrong place
1902
1903 New kwalitee test
1904
1905 Fix rt 39841. Bug in SQL Server ODBC driver which describes parameters
1906 by rearranging your SQL to do a select on the columns matching the
1907 parameters. Sometimes it gets this wrong and ends up describing the
1908 wrong column. This can lead to a varchar(10) being described with a
1909 column-size less than 10 and hence you get data truncation on execute.
1910
1911 Added a test case for rt 39841.
1912
1913 Fix rt 39897. 1.17 added support for varchar(max) in SQL Server but it
1914 broke support for SQL_VARCHAR columns in that they had LongReadLen and
1915 LongTruncOk applied to them. This means that in 1.16 you could retrieve
1916 a SQL_VARCHAR column without worrying about how long it was but in 1.17
1917 if the same column was greater than 80 characters then you would get a
1918 truncated error. The only way the around this was to set LongTruncOk or
1919 LongReadLen.
1920
1921 Added a test case for rt 39897.
1922
1923 1.17 2008-09-22
1924
1925 In the absence of any bug reports since 1.16_4 this is the official 1.17
1926 release. See below for changes since 1.16.
1927
1928 Minor pod changes.
1929
1930 Added support for ParamTypes (see DBI spec) and notes in DBD::ODBC pod.
1931
1932 1.16_4 2008-09-12
1933
1934 Small change to Makefile.PL to work around problem in darwin 8 with
1935 iODBC which leads to "Symbol not found: _SQLGetPrivateProfileString"
1936 errors.
1937
1938 Added new [n]varXXX(max) column type tests to 20SqlServer.t.
1939
1940 Fixed support for SQL_WCHAR and SQL_WVARCHAR support in non-unicode
1941 build. These types had ended up only being included for unicode builds.
1942
1943 More changes to ODBC pod to 1) encourage people to use CPAN::Reporter,
1944 2) rework contributing section, 3) mention DBIx::Log4perl 4) add a BUGS
1945 section 5) add a "ODBC Support in ODBC Drivers" section etc.
1946
1947 Changed default fallback parameter bind type to SQL_WVARCHAR for unicode
1948 builds. This affects ODBC drivers which don't have
1949 SQLDescribeParam. Problem reported by Vasili Galka with MS Access when
1950 reading unicode data from one table and inserting it into another table.
1951 The read data was unicode but it was inserted as SQL_CHARs because
1952 SQLDescribeParam does not exist in MS Access so we fallback to either a
1953 default bind type (which was SQL_VARCHAR) or whatever was specified in
1954 the bind_param call.
1955
1956 Fixed bug in 20SqlServer.t when DBI_DSN is defined including "DSN=".
1957
1958 1.16_3 2008-09-03
1959
1960 Changed Makefile.PL to add "-framework CoreFoundation" to linker line on
1961 OSX/darwin.
1962
1963 Disallow building with iODBC if it is a unicode build.
1964
1965 More tracing for odbcconnect flag.
1966
1967 Fix bug in out connection string handling that attempted to use an out
1968 connection string when SQLDriverConnect[W] fails.
1969
1970 Fixed yet more test count problems due to Test::NoWarnings not being
1971 installed.
1972
1973 Skip private_attribute_info tests if DBI < 1.54
1974
1975 About a 30% rewrite of bound parameter code which started with an
1976 attempt to support the new VARBINARY(MAX) and VARCHAR(MAX) columns in
1977 SQL Server when the parameter length is > 400K in size (see elsewhere in
1978 this Changelog). This is a seriously big change to DBD::ODBC to attempt
1979 to be a bit more clever in its handling of drivers which either do not
1980 support SQLDescribeParam or do support SQLDescribeParam but cannot
1981 describe all parameters e.g., MS SQL Server ODBC driver cannot describe
1982 "select ?, LEN(?)". If you specify the bound parameter type in your
1983 calls to bind_param and run them to an ODBC driver which supports
1984 SQLDescribeParam you may want to check carefully and probably remove the
1985 parameter type from the bind_param method call.
1986
1987 Added rt_38977.t test to test suite to test varchar(max) and
1988 varbinary(max) columns in SQL Server.
1989
1990 Moved most of README.unicode to ODBC.pm pod.
1991
1992 Added workaround for problem with the Microsoft SQL Server driver when
1993 attempting to insert more than 400K into a varbinary(max) or
1994 varchar(max) column. Thanks to Julian Lishev for finding the problem and
1995 identifying 2 possible solutions.
1996
1997 1.16_2 2008-09-02
1998
1999 Removed szDummyBuffer field from imp_fbh_st and code in dbd_describe
2000 which clears it. It was never used so this was a waste of time.
2001
2002 Changed the remaining calls to SQLAllocEnv, SQLAllocConnect and
2003 SQLAllocStmt and their respective free calls to the ODBC 3.0
2004 SQLAllocHandle and SQLFreeHandle equivalents.
2005
2006 Rewrote ColAttributes code to understand string and numeric attributes
2007 rather than trying to guess by what the driver returns. If you see any
2008 change in behaviour in ColAttributes calls you'll have to let me know as
2009 there were a number of undocumented workarounds for drivers.
2010
2011 Unicode build of DBD::ODBC now supports:
2012 column names
2013 The retrieval of unicode column names
2014 SQL strings
2015 Unicode in prepare strings (but not unicode parameter names) e.g.,
2016 select unicode_column from unicode_table
2017 is fine but
2018 select * from table where column = :unicode_param_name
2019 is not so stick to ascii parameter names if you use named
2020 parameters.
2021 Unicode SQL strings passed to the do method are supported.
2022 SQL strings passed to DBD::ODBC when the odbc_exec_direct attribute
2023 is set will not be passed as unicode strings - this is a
2024 limitation of the odbc_exec_direct attribute.
2025 connection strings
2026 True unicode connection string support will require a new version of
2027 DBI (post 1.607).
2028 Note that even though unicode connection strings are not
2029 supported currently DBD::ODBC has had to be changed to call
2030 SQLDriverConnectW/SQLConnectW to indicate to the driver manager it's
2031 intention to use some of the ODBC wide APIs. This only affects
2032 DBD::ODBC when built for unicode.
2033 odbcunicode trace flag
2034 There is a new odbcunicode trace flag to enable unicode-specific
2035 tracing.
2036
2037 Skipped 40Unicode.t test if the ODBC driver is Oracle's ODBC as I cannot
2038 make it work.
2039
2040 Changes internally to use sv_utf8_decode (where defined) instead of
2041 setting utf8 flag.
2042
2043 Fix problems in the test when Test::NoWarnings is not installed.
2044
2045 Removed some unused variables that were leading to compiler warnings.
2046
2047 Changed a lot of tracing to use new odbcconnection flag
2048
2049 Changed to use dbd_db_login6_sv if DBI supports it.
2050
2051 Commented out a diag in 20SqlServer.t that was leading to confusion.
2052
2053 Added diag to 20SqlServer.t in mars test to explain why it may fail.
2054
2055 Various pod changes for clarification and to note odbc_err_handler is
2056 deprecated.
2057
2058 Removed odbcdev trace flag - it was not really used.
2059
2060 New odbc_out_connect_string attribute for connections which returns the
2061 ODBC driver out connection string.
2062
2063 1.16_1 2008-08-28
2064
2065 Fixed bug in odbc_cancel which was checking the ACTIVE flag was on
2066 before calling SQLCancel. Non-select statements can also be cancelled so
2067 this was wrong. Thanks to Dean Arnold for spotting.
2068
2069 Minor changes to test 01base to attempt to catch install_driver failing,
2070 report it as a fail and skip other tests.
2071
2072 Fixed bug reported by James K. Lowden with asynchronous mode and
2073 SQLParamData where the code was not expecting SQL_STILL_EXECUTING and
2074 did not handle it
2075
2076 Added odbc_putdata_start attribute to determine when to start using
2077 SQLPutData on lobs.
2078
2079 Fixed bug in lob inserts where decimal_digits was being set to the size
2080 of the bound lob unnecessarily.
2081
2082 Minor change to connect/login code to delete driver-specific attributes
2083 passed to connect so they do not have to be processed again when DBI
2084 calls STORE with them.
2085
2086 New 12blob.t test.
2087
2088 A lot of code tidy up but not expecting any real benefit or detriment
2089 when using DBD::ODBC.
2090
2091 Fixed retrieving [n]varchar(max) columns which were only returning 1
2092 byte - thanks to Fumiaki Yoshimatsu and perl monks for finding it. See
2093 http://markmail.org/message/fiym5r7q22oqlzsf#query:Fumiaki Yoshimatsu
2094 odbc+page:1+mid:fiym5r7q22oqlzsf+state:results
2095
2096 Various minor changes to get the CPANTS kwalitee score up.
2097 fixed pod issues in FAQ.pm
2098 moved mytest dir to examples
2099 added generated_by and requires perl version to META.yml
2100 added pod and pod-coverage tests
2101 removed executable flag from Makefile.PL
2102 added use warnings to some modules and tests
2103 fixed pod errors in ODBC.pm
2104 added AUTHOR and LICENSE section to ODBC.pm
2105 added Test::NoWarnings to tests
2106
2107 Added support for setting the new(ish) DBI ReadOnly attribute on a
2108 connection. See notes in pod.
2109
2110 Changes to test suite to work around problems in Oracle's instant client
2111 11r1 ODBC driver for Linux (SQLColAttributes problems - see 02simple.t).
2112
2113 New tests in 30Oracle.t for oracle procedures.
2114
2115 1.16 2008-05-13
2116
2117 [TESTS]
2118
2119 Small change to the last test in 10handler.t to cope with the prepare
2120 failing instead of the execute failing - spotted by Andrei Kovalevski
2121 with the ODBCng Postgres driver.
2122
2123 Changed the 20SqlServer.t test to specifically disable MARS for the test
2124 to check multiple active statements and added a new test to check that
2125 when MARS_Connection is enabled multiple active statements are allowed.
2126
2127 Changed the 09multi.t test to use ; as a SQL statement seperator instead
2128 of a newline.
2129
2130 A few minor "use of unitialised" fixes in tests when a test fails.
2131
2132 In 02simple.t Output DBMS_NAME/VER, DRIVER_NAME/VER as useful debugging
2133 aid when cpan testers report a fail.
2134
2135 2 new tests for odbc_query_timeout added to 03dbatt.t.
2136
2137 Changed 02simple.t test which did not work for Oracle due to a "select
2138 1" in the test. Test changed to do "select 1 from dual" for Oracle.
2139
2140 New tests for numbered and named placeholders.
2141
2142 [DOCUMENTATION]
2143
2144 Added references to DBD::ODBC ohloh listing and markmail archives.
2145
2146 Added Tracing sections.
2147
2148 Added "Deviations from the DBI specification" section.
2149
2150 Moved the FAQ entries from ODBC.pm to new FAQ document. You can view the
2151 FAQ with perldoc DBD::ODBC::FAQ.
2152
2153 Added provisional README.windows document.
2154
2155 Rewrote pod for odbc_query_timeout.
2156
2157 Added a README.osx.
2158
2159 [INTERNAL CHANGES]
2160
2161 More tracing in dbdimp.c for named parameters.
2162
2163 #ifdeffed out odbc_get_primary_keys in dbdimp.c as it is no longer used.
2164 $h->func($catalog, $schema, $table, 'GetPrimaryKeys') ends up in
2165 dbdimp.c/dbd_st_primary_keys now.
2166
2167 Reformatted dbdimp.c to avoid going over 80 columns.
2168
2169 Tracing changed. Levels reviewed and changed in many cases avoiding
2170 levels 1 and 2 which are reserved for DBI. Now using DBIc_TRACE macro
2171 internally. Also tracing SQL when 'SQL' flag set.
2172
2173 [BUILD CHANGES]
2174
2175 Changes to Makefile.PL to fix a newly introduced bug with 'tr', remove
2176 easysoft OOB detection and to try and use odbc_config and odbcinst if we
2177 find them to aid automatic configuration. This latter change also adds
2178 "odbc_config --cflags" to the CC line when building DBD::ODBC.
2179
2180 Avoid warning when testing ExtUtils::MakeMaker version and it is a test
2181 release with an underscore in the version.
2182
2183 [ENHANCEMENTS]
2184
2185 Added support for parse_trace_flag and parse_trace_flags methods and
2186 defined a DBD::ODBC private flag 'odbcdev' as a test case.
2187
2188 Add support for the 'SQL' trace type. Added private trace type odbcdev
2189 as an experimental start.
2190
2191 Change odbc_query_timeout attribute handling so if it is set to 0 after
2192 having set it to a non-zero value the default of no time out is
2193 restored.
2194
2195 Added support for DBI's statistics_info method.
2196
2197 [BUG FIXES]
2198
2199 Fix bug in support for named placeholders leading to error "Can't rebind
2200 placeholder" when there is more than one named placeholder.
2201
2202 Guard against scripts attempting to use a named placeholder more than
2203 once in a single SQL statement.
2204
2205 If you called some methods after disconnecting (e.g., prepare/do and any
2206 of the DBD::ODBC specific methods via "func") then no error was
2207 generated.
2208
2209 Fixed issue with use of true/false as fields names in structure on MAC
2210 OS X 10.5 (Leopard) thanks to Hayden Stainsby.
2211
2212 Remove tracing of bound wide characters as it relies on null-terminated
2213 strings that don't exist.
2214
2215 Fix issue causing a problem with repeatedly executing a stored procedure
2216 which returns no result-set. SQLMoreResults was only called on the first
2217 execute and some drivers (SQL Server) insist a procedure is not finished
2218 until SQLMoreResults returns SQL_NO_DATA.
2219
2220 1.15 2008-01-29
2221
2222 1.15 final release.
2223
2224 Fixed bug reported by Toni Salomaki where DBD::ODBC may call
2225 SQLNumResultCols after SQLMoreResults returns SQL_NO_DATA. It led to the
2226 error:
2227 Describe failed during DBI::st=HASH(0x19c2048)->FETCH(NUM_OF_FIELDS,0)
2228 when NUM_OF_FIELDS was referenced in the Perl script.
2229
2230 Updated odbc_exec_direct documentation to describe its requirement when
2231 creating temporary objects in SQL Server.
2232
2233 Added FAQ on SQL Server temporary tables.
2234
2235 Fixed bug in dbdimp.c which was using SQL_WCHAR without testing it was
2236 defined - thanks Jergen Dutch.
2237
2238 Fixed use of "our" in UCHelp.pm which fails on older Perls.
2239
2240 Minor changes to 02simple.t and 03dbatt.t to fix diagnostics output and
2241 help debug DBD which does not handle long data properly.
2242
2243 Further changes to Makefile.PL to avoid change in behaviour of
2244 ExtUtils::MakeMaker wrt order of execution of PREREQ_PM and CONFIGURE.
2245 Now if DBI::DBD is not installed we just warn and exit 0 to avoid a
2246 cpan-testers failure.
2247
2248 1.15_2 2007-11-14
2249
2250 Fix bug in DBD::ODBC's private function data_sources which was returning
2251 data sources prefixed with "DBI:ODBC" instead of "dbi:ODBC".
2252
2253 If you don't have at least DBI 1.21 it is now a fatal error instead of
2254 just a warning.
2255
2256 DBI->connect changed so informational diagnostics like "Changed database
2257 context to 'master'" from SQL Server are available in
2258 errstr/state. These don't cause DBI->connect to die but you can test
2259 $h->err eq "" after connect and obtain the informational diagnostics
2260 from errstr/state if you want them.
2261
2262 Fixed problem in 41Unicode.t where utf8 was used before testing we had a
2263 recent enough Perl - thank you cpan testers.
2264
2265 Changed "our" back to "my" in Makefile.PL - thank you cpan testers.
2266
2267 Removed all calls to DBIh_EVENT2 in dbdimp.c as it is no longer used
2268 (see posts on dbi-dev).
2269
2270 Changed text output when a driver manager is not found to stop referring
2271 to iodbcsrc which is no longer included with DBD::ODBC.
2272
2273 Changed Makefile.PL to attempt to find unixODBC if -o or ODBCHOME not
2274 specified.
2275
2276 Updated META.yml based on new 1.2 spec.
2277
2278 Changed Makefile.PL so if an ODBC driver manager is not found then we
2279 issue warning and exit cleanly but not generating a Makefile. This
2280 should stop cpan-testers from flagging a fail because they haven't got
2281 an ODBC driver manager.
2282
2283 Changed Makefile.PL so it no longer "use"s DBI/DBI::DBD because this
2284 makes cpan-testers log a fail if DBI is not installed. Changed to put
2285 the DBI::DBD use in the CONFIGURE sub so PREREQ_PM will filter out
2286 machines where DBI is not installed.
2287
2288 Fix a possible typo, used once in 10handler.t.
2289
2290 1.15_1 2007-11-06
2291
2292 Minor changes to 20SqlServer.t test for SQL Server 2008 (Katmai).
2293 Timestamps now return an extra 4 digits of precision (all 0000) and the
2294 driver reported in dbcc messages has a '.' in the version which was not
2295 handled.
2296
2297 New FAQ entry and test code for "Datetime field overflow" problem in
2298 Oracle.
2299
2300 Changed all ODBC code to use new SQLLEN/SQLULEN types where Microsoft's
2301 headers indicate, principally so DBD::ODBC builds and works on
2302 win64. NOTE: you will need an ODBC Driver Manager on UNIX which knows
2303 SQLLEN/SQLULEN types. The unixODBC driver manager uses SQLLEN/SQLULEN in
2304 versions from at least 2.2.12. Thanks to Nelson Oliveira for finding,
2305 patching and testing this and then fixing problems with bound parameters
2306 on 64 bit Windows.
2307
2308 Added private_attribute_info method DBI introduced (see DBI docs) and
2309 test cases to 02simple.t.
2310
2311 Fairly major changes to dbd_describe in dbdimp.c to reduce ODBC calls by
2312 1 SQLDescribeCol call per column when describing result sets. Instead of
2313 calculating the amount of memory required to hold every column name we
2314 work on the basis that (num_columns + 1) * SQL_MAX_COLUMN_NAME_LEN can
2315 hold all column names. However, to avoid using a large amount of memory
2316 unnecessarily if an ODBC driver supports massive column name lengths the
2317 maximum size per column is restricted to 256.
2318
2319 Changed to avoid using explicit use of DBIc_ERRXXX in favour of newish
2320 (ok, DBD::ODBC is a bit behind the times in this respect)
2321 DBIh_SET_ERR_CHAR. This involved a reworking or the error handling and
2322 although all test cases still pass I cannot guarantee it has no other
2323 effects - please let me know if you spot differences in error messages.
2324
2325 Fixed bug in 20SqlServer test for multiple results that was passing but
2326 for the wrong reason (namely, that the odbc_err_handler was being called
2327 when it should not have been).
2328
2329 Fixed bug in odbc_err_handler that prevented it from being reset so you
2330 don't have an error handler. Looks like the problem was in
2331 dbd_db_STORE_attrib where "if(valuesv == &PL_sv_undef)" was used to
2332 detect undef and I think it should be "if (!SvOK(valuesv))".
2333
2334 Improvements to odbc_err_handler documentation.
2335
2336 Added 10handler.t test cases.
2337
2338 More tests in 02simple.t to check NUM_OF_FIELDS and NAMES_uc.
2339
2340 Bit of a tidy up:
2341
2342 Removed some unused variable declarations from dbdimp.c.
2343
2344 Lots of changes to DBD::ODBC tracing, particularly in dbd_describe, and
2345 dbd_error2 and login6.
2346
2347 Removed a lot of tracing code in comments or #if 0 as it never gets
2348 built.
2349
2350 Changed dual tests on SQL_SUCCESS and SQL_SUCCESS_WITH_INFO to use
2351 SQL_SUCCEEDED.
2352
2353 1.14 2007-07-17
2354
2355 Fix bug reported where ping crashes after disconnect thanks to Steffen
2356 Goeldner.
2357
2358 Fix bug in dbd_bind_ph which leads to the error Can't change param 1
2359 maxlen (51->50) after first bind in the 20SqlServer test. This is caused
2360 by svGROW in Perl 5.8.8 being changed to possibly grow by more than you
2361 asked (e.g. up to the next longword boundary).
2362
2363 Fix problem with binding undef as an output parameter. Reported by
2364 Stephen More with IBM's ODBC driver for iSeries.
2365
2366 Removed comment delimiters in comments in dbdimp.h leading to warnings.
2367
2368 Removed some unused variable declarations leading to warnings.
2369
2370 Removed PerlIO_flush calls as it is believed they are not required.
2371
2372 Add logging for whether SQLDescribeParam is supported.
2373
2374 Fixed use of unitialised variable in dbd_bind_ph where an undef is bound
2375 and tracing is enabled.
2376
2377 Fixed issue with TRACESTATUS change in 20SqlServer.t tests 28, 31, 32
2378 and 33 leading to those tests failing when testing with SQL Server 2005
2379 or Express.
2380
2381 Many compiler warnings fixed - especially for incompatible types.
2382
2383 Add provisional Unicode support - thanks to Alexander Foken. This change
2384 is very experimental (especially on UNIX). Please see ODBC.pm
2385 documentation. Also see README.unicode and README.af. New database
2386 attribute odbc_has_unicode to test if DBD::ODBC was built with UNICODE
2387 support. New tests for Unicode. New requirement for Perl 5.8.1 if
2388 Unicode support required. New -[no]u argument to Makefile.PL. New
2389 warning in Makefile.PL if Unicode support built for UNIX.
2390
2391 Fix use of unitialised var in Makefile.PL.
2392
2393 Fix use of scalar with no effect on Makefile.PL
2394
2395 Added warning to Makefile.PL about building/running with LANG using
2396 UTF8.
2397
2398 Added warning to Makefile.PL about using thread-safe ODBC drivers.
2399
2400 Updated MANIFEST to include more test code from mytest and remove
2401 MANIFEST.SKIP etc.
2402
2403 Removed calls to get ODBC errors when SQLMoreResults returns
2404 SQL_NO_DATA. These are a waste of time since SQL_NO_DATA is expected
2405 and there is no error diagnostic to retrieve.
2406
2407 Changes to test 17 of 02simple.t which got "not ok 17 - Col count
2408 matches correct col count" errors with some Postgres ODBC
2409 drivers. Caused by test expecting column names to come back
2410 uppercase. Fixes by uppercasing returned column names.
2411
2412 Changes to tests in 03batt.t which correctly expects an ODBC 3 driver to
2413 return the column names in SQLTables result-set as per ODBC 3.0
2414 spec. Postgres which reports itself as an ODBC 3.0 driver seems to
2415 return the ODBC 2 defined column names. Changed tests to catch ODBC 2.0
2416 names are pass test put issue warning.
2417
2418 For postgres skip test (with warning) checking $sth->{NAME} returns
2419 empty listafter execute on update
2420
2421 Updated FAQ, added a few more questions etc.
2422
2423 DBD::ODBC requires at least 5.6.0 of Perl.
2424
2425 Many updates to pod documentation.
2426
2427 Removed some dead HTTP links in the pod I could not find equivalents for
2428 - let me know if you have working replacements for ones removed
2429
2430 Add some HTTP links to useful tutorials on DBD::ODBC
2431
2432 1.13 2004-11-08
2433
2434 Fix inconsistency/bug with odbc_exec_direct vs. odbc_execdirect
2435 settings. Now made consistent with odbc_exec_direct. For now, will
2436 still look for odbc_execdirect in prepare, but not as DBH attribute as a
2437 backup (which is what it was doing), but that support will be dropped at
2438 some time in the future. Please use odbc_exec_direct from now on...
2439
2440 Fix handling of print statements for SQL Server thanks to Martin Evans!
2441 Thanks for all your work on this!
2442
2443 Due to bug in SQL Server, you must use odbc_exec_direct. See
2444 t/20SqlServer.t for example. You will need to call
2445 $sth->{odbc_more_results} to clear out any trailing messages.
2446
2447 Change tests to use Test::More. Whew, that's much nicer!
2448
2449 Fix Oracle integral/numeric output params so that warning not printed
2450 about value not being numeric (even though it is!)
2451
2452 1.12 2004-10-26
2453
2454 Fix bug with odbc_execdirect attributed thanks to Martin Evans
2455
2456 Fix bug(s) with odbc_query_timeout and tested with SQL*Server. Oracle
2457 tests failed with setting timeout. Probably not handled by Oracle's
2458 ODBC driver
2459
2460 1.11 2004-10-11
2461
2462 Added odbc_timeout, but untested
2463
2464 1.10 2004-09-08
2465
2466 Fixed bug in Makefile.PL.
2467
2468 Added pod.t test, taken from DBI.
2469
2470 Fixed various small POD issues, discovered during the pod test.
2471
2472 Fixed bug in bind_param_inout
2473
2474 1.09 2004-03-10
2475
2476 Duh. I forgot to add new dbivport.h to MANIFEST and SVN before
2477 submitting. Fixed.
2478
2479 1.08 2004-03-06
2480
2481 Added check in Makefile.PL to detect if the environment variable LANG is
2482 Set. If so, prints a warning about potential makefile generation
2483 issues.
2484
2485 Change to use dbivport.h per new DBI spec.
2486
2487 Add ability to set the cursor type during the connect. This may allow
2488 some servers which do not support multiple concurrent statements to
2489 permit them -- tested with SQL Server. Thanks to Martin Busik! See
2490 odbc_cursortype information in the ODBC POD.
2491
2492 1.07 2004-02-19
2493
2494 Added to Subversion version control hosted by perl.org. Thanks Robert!
2495 See ODBC.pm POD for more information.
2496
2497 Added contributing section to ODBC.pm POD -- see more details there!
2498
2499 Added parameter to odbc_errhandler for the NativeError -- thanks to
2500 Martin Busik.
2501
2502 Fix for Makefile.PL not having tab in front of $(NOOP) (Finally).
2503
2504 Fix for SQLForeignKeys thanks to Kevin Shepherd.
2505
2506 1.06 2003-06-19
2507
2508 Fixed test in t/02simple.t to skip if the DSN defined by the user has
2509 DSN= in it.
2510
2511 Added tests for wrong DSN, ensuring the DBI::errstr is appropriately
2512 set.
2513
2514 Fixed small issue in Makefile.PL for Unix systems thanks to H.Merijn
2515 Brand.
2516
2517 Update to NOT copy user id and password to connect string if UID or PWD
2518 parameter in connect string.
2519
2520 Updated Makefile.PL for dmake, per patch by Steffen Goldner. Thanks
2521 Steffen!
2522
2523 1.05 2003-03-14
2524
2525 Cleaned up Makefile.PL and added Informix support thanks to Jonathan
2526 Leffler (see README.informix)
2527
2528 Added nicer error message when attempting to do anything while the
2529 database is disconnected.
2530
2531 Fixed fetchrow_hashref('NAME_uc | NAME_lc') with odbc_more_results.
2532
2533 Added exporter to allow perl -MDBD::ODBC=9999 command line to determine
2534 version
2535
2536 Fixed for building with DBI 1.33 and greater
2537
2538 Removed all C++ style comments
2539
2540 Ensured files are in Unix format, with the exception of the README type
2541 information and Makefile.PL
2542
2543 1.04 2003-01-24
2544
2545 It seems that case insensitive string comparison with a limit causes
2546 problems for multiple platforms. strncmpi, strncasecmp, _strcmpin are
2547 all functions hit and it seems to be a hit-or-miss. Hence, I rewrote it
2548 to upper case the string then do strncmp, which should be safe...sheesh.
2549 A simple thing turned into a headache...
2550
2551 1.03 2003-01-17
2552
2553 Add automatic detection of DRIVER= or DSN= to add user id and password
2554 to connect string.
2555
2556 1.02 2003-01-06
2557
2558 Fix to call finish() automatically if execute is re-called in a loop
2559 (and test in t/02simple.t to ensure it's fixed)
2560
2561 Augmented error message when longs are truncated to help users determine
2562 where to look for help.
2563
2564 Fixes for build under Win32 with Perl5.8.
2565
2566 1.01 2002-12-09
2567
2568 Forgot to fix require DBI 1.201 in ODBC.pm to work for perl 5.8. Fixed
2569
2570 1.00 2002-12-08
2571
2572 (Please see all changes since version 0.43)
2573
2574 Updated Makefile.PL to handle SQL_Wxxx types correctly with unixODBC and
2575 linking directly with EasySoft OOB. Note that I could not find where
2576 iODBC defines SQL_WLONG_VARCHAR, so I'm not sure it's fixed on all other
2577 platforms. Should not have been a problem under Win32...
2578
2579 Found that the fix in _18 was only enabled if debug enabled and it broke
2580 something else. removed the fix.
2581
2582 Updated Makefile.PL to use DBI version 1.21 instead of 1.201 to
2583 facilitate builds under latest development versions of Perl.
2584
2585 Updated code to use the *greater* of the column display size and the
2586 column length for allocating column buffers. This *should* workaround a
2587 problem with DBD::ODBC and the Universe database.
2588
2589 Added code thanks to Michael Riber to handle SQLExecDirect instead of
2590 SQLPrepare. There are two ways to get this:
2591 $dbh->prepare($sql, { odbc_execdirect => 1});
2592 and
2593 $dbh->{odbc_execdirect} = 1;
2594
2595 When $dbh->prepare() is called with the attribute "ExecDirect" set to a
2596 non-zero value dbd_st_prepare do NOT call SQLPrepare, but set the sth
2597 flag odbc_exec_direct to 1.
2598
2599 Fixed numeric value binding when binding non-integral values. Now lets
2600 the driver or the database handle the conversion.
2601
2602 Fixed makefile.pl generation of makefile to force the ODBC directory
2603 first in the include list to help those installing ODBC driver managers
2604 on systems which already have ODBC drivers in their standard include
2605 path.
2606
2607 0.45_18 2002-09-26
2608
2609 Updated MANIFEST to include more of the mytest/* files (examples, tests)
2610
2611 Fixed problem when attempting to get NUM_OF_FIELDS after execute returns
2612 no rows/columns.
2613
2614 0.45_17 2002-08-26
2615
2616 More fixes for multiple result sets. Needed to clear the DBIc_FIELDS_AV
2617 when re-executing the multiple-result set stored procedure/query.
2618
2619 0.45_16 2002-08-26
2620
2621 Updated to fix output parameters with multiple result sets. The output
2622 parameters are not set until the last result set has been retrieved.
2623
2624 0.45_15 2002-08-20
2625
2626 Updated for new DBIc_STATE macros (all debug, as it turned out) to be
2627 thread safer in the long run
2628
2629 Updated for the new DBIc_LOGFP macros
2630
2631 Added CLONE method
2632
2633 Fix for SQL Server where multiple result sets being returned from a
2634 stored proc, where one of the result sets was empty (insert/update).
2635
2636 Added new attribute odbc_force_rebind, which forces DBD::ODBC to check
2637 recheck for new result sets every execute. This is only really
2638 necessary if you have a stored procedure which returns different result
2639 sets with each execute, given the same "prepare". Many times this will
2640 be automatically set by DBD::ODBC, however, if there is only one result
2641 set in the stored proc, but it can differ with each call, then DBD::ODBC
2642 will not know to set it.
2643
2644 Updated the DBD::ODBC POD documentation to document DBD::ODBC private
2645 attributes and usage.
2646
2647 0.45_14 2002-08-13
2648
2649 Added support to handle (better) DBI begin_work().
2650
2651 Fix for binding undef parameters on SQL Server.
2652
2653 Fix bug when connecting twice in the same script. Trying to set the
2654 environment ODBC version twice against the same henv caused an error.
2655
2656 0.45_13 2002-08-09
2657
2658 Workaround problem with iODBC where SQLAllocHandleStd is not present in
2659 iODBC.
2660
2661 Made Changes file accessible via perldoc DBD::ODBC::Changes. In the
2662 near future the change log will be removed from here and put in changes
2663 to tidy up a bit.
2664
2665 0.45_12 2002-08-09
2666
2667 Fixed global destruction access violation (which was seemingly random).
2668
2669 0.45_11 2002-08-08
2670
2671 Updated manifest to include more samples.
2672
2673 Working on checking for leaks on Linux, where I might get more
2674 information about the process memory.
2675
2676 Working on fixing problems with MS SQL Server binding parameters. It
2677 seems that SQLServer gets "confused" if you bind a NULL first. In
2678 "older" (SQLServer 2000 initial release) versions of the driver, it
2679 would truncate char fields. In "newer" versions of the SQL Server
2680 driver, it seems to only truncate dates (actually, round them to the
2681 nearest minute). If you have problems in the SQL Server tests, please
2682 upgrade your driver to the latest version on Microsoft's website (MDAC
2683 2.7 or above) http://www.microsoft.com/data
2684
2685 0.45_10 2002-07-30
2686
2687 Added database specific tests to ensure things are working. Some of the
2688 tests may not work for all people or may not be desirable. I have tried
2689 to keep them as safe as possible, but if they don't work, please let me
2690 know.
2691
2692 Added support for the internal function GetFunctions to handle ODBC 3's
2693 SQL_API_ODBC3_ALL_FUNCTIONS. Would have caused a memory overwrite on
2694 the stack if it was called.
2695
2696 0.45_9 2002-07-30
2697
2698 Fixed bug in procedure handling for SQLServer. Was not re-describing
2699 the result sets if the SQLMoreResults in the execute needs to be called.
2700
2701 0.45_8 2002-07-25
2702
2703 Fixed bug in tracing code when binding an undef parameter which did not
2704 happen to have a valid buffer with tracing level >= 2
2705
2706 Fixed bug when binding undef after a valid data bind on a timestamp.
2707 The Scale value was being calculated based upon the string that had been
2708 bound prior to the bind of the undef and if that had a sub-second value,
2709 then the scale would be set to the wrong value...I.e.
2710 bind_param(1, '2000-05-17 00:01:00.250', SQL_TYPE_TIMESTAMP) then
2711 execute
2712 bind_param(1, undef, SQL_TYPE_TIMESTAMP) then
2713
2714 Fixed SQL Server issue when binding a null and the length was set to 0
2715 instead of 1
2716
2717 0.45_7 2002-07-25
2718
2719 Adding support for array binding, but not finished.
2720
2721 Fixed bug where SqlServer Stored procedures which perform INSERT would
2722 not correctly return a result set. Thanks to Joe Tebelskis for finding
2723 it and Martin Evans for supplying a fix.
2724
2725 Fixed bug where binding the empty string would cuase a problem. Fixed
2726 and added test in t/07bind.t.
2727
2728 0.45_6 2002-07-24
2729
2730 Added support for new DBI ParamValues feature.
2731
2732 0.45_5 2002-07-23
2733
2734 Added odbc_err_handler and odbc_async_exec thanks to patches by David
2735 L. Good. See example in mytest/testerrhandler.pl
2736
2737 Here's the notes about it:
2738
2739 I've implemented two separate functions. The first is an "error
2740 handler" similar to that in DBD::Sybase. The error handler can be
2741 used to intercept error and status messages from the server. It is
2742 the only way (at least currently) that you can retrieve non-error
2743 status messages when execution is successful.
2744
2745 To use the error handler, set the "odbc_err_handler" attribute on your
2746 database handle to a reference to a subroutine that will act as the
2747 error handler. This subroutine will be passed two args, the SQLSTATE
2748 and the error message. If the subroutine returns 0, the error message
2749 will be otherwise ignored. If it returns non-zero, the error message
2750 will be processed normally.
2751
2752 The second function implemented is asynchronous execution. It's only
2753 useful for retrieving server messages with an error handler during an
2754 execute() that takes a long time (such as a DBCC on a large database)
2755 ODBC doesn't have the concept of a callback routine like Sybase's
2756 DBlib/CTlib does, so asynchronous execution is needed to be able to
2757 get the server messages before the SQL statement is done processing.
2758
2759 To use asynchronous execution, set the "odbc_async_exec" attribute on
2760 your database handle to 1. Not all ODBC drivers support asynchronous
2761 execution. To see if yours does, set odbc_async_exec to 1 and then
2762 check it's value. If the value is 1, your ODBC driver can do
2763 asynchronous execution. If the value is 0, your ODBC driver cannot.
2764
2765 0.45_4 2002-07-22
2766
2767 More fixes for DB2 tests and timestamp handling.
2768
2769 0.45_3 2002-07-22
2770
2771 Changes to internal timestamp type handling and test structure to ensure
2772 tests work for all platforms. DB2 was giving me fits due to bad
2773 assumptions. Thanks to Martin Evans (again) for help in identifying the
2774 problems and helping research solutions. This includes the
2775 scale/precision values to correctly store full timestamps.
2776
2777 0.45_2 2002-07-19
2778
2779 Moving API usage to ODBC 3.0 specifications. With lots of help from
2780 Martin Evans (again!). Thanks Martin!!!!!
2781
2782 0.44 2002-07-18
2783
2784 .44 was never officially released.
2785
2786 Fix for do() and execute to handle DB2 correctly. Patch/discovery
2787 thanks to Martin Evans.
2788
2789 Partly moving towards defaulting to ODBC 3.x standards.
2790
2791 0.43 2002-07-18
2792
2793 Fix for FoxPro (and potentially other) Drivers!!!!!
2794
2795 Add support for DBI column_info
2796
2797 Fix for binding undef value which comes from dereferencing hash
2798
2799 Fix to make all bound columns word (int) aligned in the buffer.
2800
2801 0.42 2002-07-08
2802
2803 Added patches to the tests to support ActiveState's automated build
2804 process.
2805
2806 Fix ping() to try SQLTables for a test, instead of a strange query.
2807
2808 0.41 2002-04-15
2809
2810 Fixed problem where SQLDescribeParam would fail (probably bug in ODBC
2811 driver). Now reverts to SQL_VARCHAR if that happens, instead of failing
2812 the query.
2813
2814 Fixed error report when using Oracle's driver. There is a known
2815 problem. Left the error on the test, but added warning indicating it's
2816 a known problem.
2817
2818 0.40 2002-04-12
2819
2820 Most significant change is the change in the default binding type which
2821 forces DBD::ODBC to attempt to determine the bind type if one is not
2822 passed. I decided to make this the default behaviour to make things as
2823 simple as possible.
2824
2825 Fixed connection code put in 0.39 to work correctly.
2826
2827 Two minor patches for building, one for Cygwin one if both iODBC and
2828 unixODBC libraries are installed. Probably need better command line on
2829 this, but if someone has the problem, please let me know (and hopefully
2830 send a patch with it).
2831
2832 0.39 2002-03-12
2833
2834 See mytest/longbin.pl for demonstration of inserting and retrieving long
2835 binary files to/from the db. Uses MD5 algorithm to verify data. Please
2836 do some similar test(s) with your database before using it in
2837 production. The various bind types are different for each database!
2838
2839 Finally removed distribution of old iODBC. See www.iodbc.org or
2840 www.unixodbc.org for newer/better versions of the ODBC driver manager
2841 for Unix (and others?).
2842
2843 Added ability to force ODBC environment version.
2844
2845 Fix to SQLColAttributes.
2846
2847 Changes to connect sequence to provide better error messages for those
2848 using DSN-less connections.
2849
2850 0.38 2002-02-12
2851
2852 Fixed do function (again) thanks to work by Martin Evans.
2853
2854 0.37 2002-02-10
2855
2856 Patches for get_info where return type is string. Patches thanks to
2857 Steffen Goldner. Thanks Steffen!
2858
2859 Patched get_info to NOT attempt to get data for SQL_DRIVER_HSTMT and
2860 SQL_DRIVER_HDESC as they expect data in and have limited value (IMHO).
2861
2862 Further fixed build for ODBC 2.x drivers. The new SQLExecDirect code
2863 had SQLAllocHandle which is a 3.x function, not a 2.x function. Sigh.
2864 I should have caught that the first time. Signed, the Mad-and-
2865 not-thorough-enough-patcher.
2866
2867 Additionally, a random core dump occurred in the tests, based upon the
2868 new SQLExecDirect code. This has been fixed.
2869
2870 0.36 2002-02-10
2871
2872 Fixed build for ODBC 2.x drivers. The new SQLExecDirect code had
2873 SQLFreeHandle which is a 3.x function, not a 2.x function.
2874
2875 0.35 2002-02-09
2876
2877 Fixed (finally) multiple result sets with differing numbers of columns.
2878 The final fix was to call SQLFreeStmt(SQL_UNBIND) before repreparing the
2879 statement for the next query.
2880
2881 Added more to the multi-statement tests to ensure the data retrieved was
2882 what was expected.
2883
2884 Now, DBD::ODBC overrides DBI's do to call SQLExecDirect for simple
2885 statements (those without parameters). Please advise if you run into
2886 problems. Hopefully, this will provide some small speed improvement for
2887 simple "do" statements. You can also call $dbh->func($stmt,
2888 ExecDirect). I'm not sure this has great value unless you need to
2889 ensure SQLExecDirect is being called. Patches thanks to Merijn Broeren.
2890 Thanks Merijn!
2891
2892 0.34 2002-02-07
2893
2894 Further revamped tests to attempt to determine if SQLDescribeParam will
2895 work to handle the binding types. The t/08bind.t attempts to determine
2896 if SQLDescribeParam is supported. note that Oracle's ODBC driver under
2897 NT doesn't work correctly when binding dates using the ODBC date
2898 formatting {d } or {ts }. So, test #3 will fail in t/08bind.t
2899
2900 New support for primary_key_info thanks to patches by Martin Evans.
2901
2902 New support for catalog, schema, table and table_type in table_info
2903 thanks to Martin Evans. Thanks Martin for your work and your continuing
2904 testing, suggestions and general support!
2905
2906 Support for upcoming dbi get_info.
2907
2908 0.33_3 2002-02-04
2909
2910 Revamped tests to include tests for multiple result sets. The tests are
2911 ODBC driver platform specific and will be skipped for drivers which do
2912 not support multiple result sets.
2913
2914 0.33_2 2002-02-04
2915
2916 Finally tested new binding techniques with SQL Server 2000, but there is
2917 a nice little bug in their MDAC and ODBC drivers according to the
2918 knowledge base article # Q273813, titled
2919
2920 "FIX: "Incorrect Syntax near the Keyword 'by' "
2921 Error Message with Column Names of "C", "CA" or "CAS" (Q273813)
2922
2923 DBD::ODBC now does not name any of the columns A, B, C, or D they are
2924 now COL_A, COL_B, COL_C, COL_D.
2925
2926 *** NOTE: *** I AM STRONGLY CONSIDERING MAKING THE NEW BINDING the
2927 default for future versions. I do not believe it will break much
2928 existing code (if any) as anyone binding to non VARCHAR (without the
2929 ODBC driver doing a good conversion from the VARCHAR) will have a
2930 problem. It may be subtle, however, since much code will work, but
2931 say, binding dates may not with some drivers.
2932
2933 Please comment soon...
2934
2935 0.33_1 2002-02-04
2936
2937 *** WARNING: ***
2938
2939 Changes to the binding code to allow the use of SQLDescribeParam to
2940 determine if the type of column being bound. This is experimental and
2941 activated by setting
2942
2943 $dbh->{odbc_default_bind_type} = 0; # before creating the query...
2944
2945 Currently the default value of odbc_default_bind_type = SQL_VARCHAR
2946 which mimicks the current behaviour. If you set odbc_default_bind_type
2947 to 0, then SQLDescribeParam will be called to determine the columen
2948 type. Not ALL databases handle this correctly. For example, Oracle
2949 returns SQL_VARCHAR for all types and attempts to convert to the correct
2950 type for us. However, if you use the ODBC escaped date/time format such
2951 as: {ts '1998-05-13 00:01:00'} then Oracle complains. If you bind this
2952 with a SQL_TIMESTAMP type, however, Oracle's ODBC driver will parse the
2953 time/date correctly. Use at your own risk!
2954
2955 Fix to dbdimp.c to allow quoted identifiers to begin/end with either "
2956 or '. The following will not be treated as if they have a bind
2957 placeholder:
2958
2959 "isEstimated?"
2960 '01-JAN-1987 00:00:00'
2961 'Does anyone insert a ?'
2962
2963 0.32 2002-01-22
2964
2965 More SAP patches to Makfile.PL to eliminate the call to Data Sources
2966
2967 A patch to the test (for SAP and potentially others), to allow fallback
2968 to SQL_TYPE_DATE in the tests
2969
2970 0.31 2002-01-18
2971
2972 Added SAP patches to build directly against SAP driver instead of driver
2973 manager thanks to Flemming Frandsen (thanks!)
2974
2975 Added support to fix ping for Oracle8. May break other databases, so
2976 please report this as soon as possible. The downside is that we need to
2977 actually execute the dummy query.
2978
2979 0.30 2002-01-08
2980
2981 Added ping patch for Solid courtesy of Marko Asplund
2982
2983 Updated disconnect to rollback if autocommit is not on. This should
2984 silence some errors when disconnecting.
2985
2986 Updated SQL_ROWSET_SIZE attribute. Needed to force it to
2987 odbc_SQL_ROWSET_SIZE to obey the DBI rules.
2988
2989 Added odbc_SQL_DRIVER_ODBC_VER, which obtains the version of the Driver
2990 upon connect. This internal capture of the version is a read-only
2991 attribute and is used during array binding of parameters.
2992
2993 Added odbc_ignore_named_placeholders attribute to facilicate creating
2994 triggers within SAPDB and Oracle, to name two. The syntax in these DBs
2995 is to allow use of :old and :new to access column values before and
2996 after updates. Example:
2997
2998 $dbh->{odbc_ignore_named_placeholders} = 1;
2999 # set it for all future statements
3000 # ignores :foo, :new, etc, but not :1 or ?
3001 $dbh->do("create or replace etc :new.D = sysdate etc");
3002
3003 0.29 2001-08-22
3004
3005 Cygwin patches from Neil Lunn (untested by me). Thanks Neil!
3006
3007 SQL_ROWSET_SIZE attribute patch from Andrew Brown
3008
3009 There are only 2 additional lines allowing for the setting of
3010 SQL_ROWSET_SIZE as db handle option.
3011
3012 The purpose to my madness is simple. SqlServer (7 anyway) by default
3013 supports only one select statement at once (using std ODBC cursors).
3014 According to the SqlServer documentation you can alter the default
3015 setting of three values to force the use of server cursors - in which
3016 case multiple selects are possible.
3017
3018 The code change allows for:
3019 $dbh->{SQL_ROWSET_SIZE} = 2; # Any value > 1
3020
3021 For this very purpose.
3022
3023 The setting of SQL_ROWSET_SIZE only affects the extended fetch command
3024 as far as I can work out and thus setting this option shouldn't affect
3025 DBD::ODBC operations directly in any way.
3026
3027 Andrew
3028
3029 VMS and other patches from Martin Evans (thanks!)
3030
3031 [1] a fix for Makefile.PL to build DBD::ODBC on OpenVMS.
3032 [2] fix trace message coredumping after SQLDriverConnect
3033 [3] fix call to SQLCancel which fails to pass the statement handle
3034 properly.
3035 [4] consume diagnostics after SQLDriverConnect/SQLConnect call or they
3036 remain until the next error occurs and it then looks confusing (this
3037 is due to ODBC spec for SQLError). e.g. test 02simple returns a data
3038 truncated error only now instead of all the informational diags that
3039 are left from the connect call, like the "database changed", "language
3040 changed" messages you get from MS SQL Server.
3041
3042 Replaced C++ style comments with C style to support more platforms more
3043 easily.
3044
3045 Fixed bug which use the single quote (') instead of a double quote (")
3046 for "literal" column names. This helped when having a colon (:) in the
3047 column name.
3048
3049 Fixed bug which would cause DBD::ODBC to core-dump (crash) if DBI
3050 tracing level was greater than 3.
3051
3052 Fixed problem where ODBC.pm would have "use of uninitialized variable"
3053 if calling DBI's type_info.
3054
3055 Fixed problem where ODBC.xs *may* have an overrun when calling
3056 SQLDataSources.
3057
3058 Fixed problem with DBI 1.14, where fprintf was being called instead of
3059 PerlIO_printf for debug information
3060
3061 Fixed problem building with unixODBC per patch from Nick Gorham
3062
3063 Added ability to bind_param_inout() via patches from Jeremy Cooper.
3064 Haven't figured out a good, non-db specific way to test. My current
3065 test platform attempts to determine the connected database type via ugly
3066 hacks and will test, if it thinks it can. Feel free to patch and send
3067 me something...Also, my current Oracle ODBC driver fails miserably and
3068 dies.
3069
3070 Updated t/02simple.t to not print an error, when there is not one.
3071
3072 0.28 2000-03-23
3073
3074 Added support for SQLSpecialColumns thanks to patch provided by Martin
3075 J. Evans
3076
3077 Fixed bug introduced in 0.26 which was introduced of SQLMoreResults was
3078 not supported by the driver.
3079
3080 0.27 2000-03-08
3081
3082 Examined patch for ping method to repair problem reported by Chris
3083 Bezil. Thanks Chris!
3084
3085 Added simple test for ping method working which should identify this in
3086 the future.
3087
3088 0.26 2000-03-05
3089
3090 Put in patch for returning only positive rowcounts from dbd_st_execute.
3091 The original patch was submitted by Jon Smirl and put back in by David
3092 Good. Reasoning seems sound, so I put it back in. However, any
3093 databases that return negative rowcounts for specific reasons, will no
3094 longer do so.
3095
3096 Put in David Good's patch for multiple result sets. Thanks David! See
3097 mytest\moreresults.pl for an example of usage.
3098
3099 Added readme.txt in iodbcsrc explaining an issue there with iODBC 2.50.3
3100 and C<data_sources>.
3101
3102 Put in rudimentary cancel support via SQLCancel. Call $sth->cencel to
3103 utilize. However, it is largely untested by me, as I do not have a good
3104 sample for this yet. It may come in handy with threaded perl, someday
3105 or it may work in a signal handler.
3106
3107 0.25 2000-03-04
3108
3109 Added conditional compilation for SQL_WVARCHAR and SQL_WLONGVARCHAR. If
3110 they are not defined by your driver manager, they will not be compiled
3111 in to the code. If you would like to support these types on some
3112 platforms, you may be able to #define SQL_WVARCHAR (-9) #define
3113 SQL_WLONGVARCHAR (-10)
3114
3115 Added more long tests with binding in t\09bind.t. Note use of
3116 bind_param!
3117
3118 0.24 2000-02-24
3119
3120 Fixed Test #13 in 02simple.t. Would fail, improperly, if there was only
3121 one data source defined.
3122
3123 Fixed (hopefully) SQL Server 7 and ntext type "Out of Memory!" errors
3124 via patch from Thomas Lowery. Thanks Thomas!
3125
3126 Added more support for Solid to handle the fact that it does not support
3127 data_sources nor SQLDriverConnect. Patch supplied by Samuli Karkkainen
3128 [[email protected]]. Thanks! It's untested by me, however.
3129
3130 Added some information from Adam Curtin about a bug in iodbc 2.50.3's
3131 data_sources. See iodbcsrc\readme.txt.
3132
3133 Added information in this pod from Stephen Arehart regarding DSNLess
3134 connections.
3135
3136 Added fix for sp_prepare/sp_execute bug reported by Paul G. Weiss.
3137
3138 Added some code for handling a hint on disconnect where the user gets an
3139 error for not committing.
3140
3141 0.22 1999-09-08
3142
3143 Fixed for threaded perl builds. Note that this was tested only on
3144 Win32, with no threads in use and using DBI 1.13. Note, for
3145 ActiveState/PERL_OBJECT builds, DBI 1.13_01 is required as of 9/8/99.
3146 If you are using ActiveState's perl, this can be installed by using PPM.
3147
3148 0.21 Unknown
3149
3150 Thanks to all who provided patches!
3151
3152 Added ability to connect to an ODBC source without prior creation of
3153 DSN. See mytest/contest.pl for example with MS Access. (Also note that
3154 you will need documentation for your ODBC driver -- which, sadly, can be
3155 difficult to find).
3156
3157 Fixed case sensitivity in tests.
3158
3159 Hopefully fixed test #4 in t/09bind.t. Updated it to insert the date
3160 column and updated it to find the right type of the column. However, it
3161 doesn't seem to work on my Linux test machine, using the OpenLink
3162 drivers with MS-SQL Server (6.5). It complains about binding the date
3163 time. The same test works under Win32 with SQL Server 6.5, Oracle 8.0.3
3164 and MS Access 97 ODBC drivers. Hmmph.
3165
3166 Fixed some binary type issues (patches from Jon Smirl)
3167
3168 Added SQLStatistics, SQLForeignKeys, SQLPrimaryKeys (patches from Jon
3169 Smirl) Thanks (again), Jon, for providing the build_results function to
3170 help reduce duplicate code!
3171
3172 Worked on LongTruncOk for Openlink drivers.
3173
3174 Note: those trying to bind variables need to remember that you should
3175 use the following syntax:
3176
3177 use DBI;
3178 ...
3179 $sth->bind_param(1, $str, DBI::SQL_LONGVARCHAR);
3180
3181 Added support for unixodbc (per Nick Gorham)
3182
3183 Added support for OpenLinks udbc (per Patrick van Kleef)
3184
3185 Added Support for esodbc (per Martin Evans)
3186
3187 Added Support for Easysoft (per Bob Kline)
3188
3189 Changed table_info to produce a list of views, too.
3190
3191 Fixed bug in SQLColumns call.
3192
3193 Fixed blob handling via patches from Jochen Wiedmann.
3194
3195 Added data_sources capability via snarfing code from DBD::Adabas (Jochen
3196 Wiedmann)
3197
3198 0.20 1998-08-14
3199
3200 SQLColAttributes fixes for SQL Server and MySQL. Fixed tables method by
3201 renaming to new table_info method. Added new tyoe_info_all method.
3202 Improved Makefile.PL support for Adabase.
3203
3204 0.19 Unknown
3205
3206 Added iODBC source code to distribution.Fall-back to using iODBC header
3207 files in some cases.
3208
3209 0.18 Unknown
3210
3211 Enhancements to build process. Better handling of errors in error
3212 handling code.
3213
3214 0.17 Unknown
3215
3216 This release is mostly due to the good work of Jeff Urlwin. My eternal
3217 thanks to you Jeff.
3218
3219 Fixed "SQLNumResultCols err" on joins and 'order by' with some drivers
3220 (see Microsoft Knowledge Base article #Q124899). Thanks to Paul
3221 O'Fallon for that one.
3222
3223 Added more (probably incomplete) support for unix ODBC in Makefile.PL
3224
3225 Increased default SQL_COLUMN_DISPLAY_SIZE and SQL_COLUMN_LENGTH to 2000
3226 for drivers that don't provide a way to query them dynamically. Was 100!
3227
3228 When fetch reaches the end-of-data it automatically frees the internal
3229 ODBC statement handle and marks the DBI statement handle as inactive
3230 (thus an explicit 'finish' is *not* required).
3231
3232 Also:
3233
3234 LongTruncOk for Oracle ODBC (where fbh->datalen < 0)
3235
3236 Added tracing into SQLBindParameter (help diagnose oracle odbc bug)
3237
3238 Fixed/worked around bug/result from Latest Oracle ODBC driver where in
3239 SQLColAttribute cbInfoValue was changed to 0 to indicate fDesc had a value
3240
3241 Added work around for compiling w/ActiveState PRK (PERL_OBJECT)
3242
3243 Updated tests to include date insert and type
3244
3245 Added more "backup" SQL_xxx types for tests
3246
3247 Updated bind test to test binding select
3248
3249 NOTE: bind insert fails on Paradox driver (don't know why)
3250
3251 Added support for: (see notes below)
3252
3253 SQLGetInfo via $dbh->func(xxx, GetInfo)
3254 SQLGetTypeInfo via $dbh->func(xxx, GetTypeInfo)
3255 SQLDescribeCol via $sth->func(colno, DescribeCol)
3256 SQLColAttributes via $sth->func(xxx, colno, ColAttributes)
3257 SQLGetFunctions via $dbh->func(xxx, GetFunctions)
3258 SQLColumns via $dbh->func(catalog, schema, table, column, 'columns')
3259
3260 Fixed $DBI::err to reflect the real ODBC error code which is a 5 char
3261 code, not necessarily numeric.
3262
3263 Fixed fetches when LongTruncOk == 1.
3264
3265 Updated tests to pass more often (hopefully 100% <G>)
3266
3267 Updated tests to test long reading, inserting and the LongTruncOk attribute.
3268
3269 Updated tests to be less driver specific.
3270
3271 They now rely upon SQLGetTypeInfo I<heavily> in order to create the
3272 tables. The test use this function to "ask" the driver for the name of
3273 the SQL type to correctly create long, varchar, etc types. For example,
3274 in Oracle the SQL_VARCHAR type is VARCHAR2, while MS Access uses TEXT
3275 for the SQL Name. Again, in Oracle the SQL_LONGVARCHAR is LONG, while
3276 in Access it's MEMO. The tests currently handle this correctly (at
3277 least with Access and Oracle, MS SQL server will be tested also).