| 1 |
/*- |
| 2 |
* Copyright (c) 1985 Sun Microsystems, Inc. |
| 3 |
* Copyright (c) 1980, 1993 |
| 4 |
* The Regents of the University of California. All rights reserved. |
| 5 |
* All rights reserved. |
| 6 |
* |
| 7 |
* Redistribution and use in source and binary forms, with or without |
| 8 |
* modification, are permitted provided that the following conditions |
| 9 |
* are met: |
| 10 |
* 1. Redistributions of source code must retain the above copyright |
| 11 |
* notice, this list of conditions and the following disclaimer. |
| 12 |
* 2. Redistributions in binary form must reproduce the above copyright |
| 13 |
* notice, this list of conditions and the following disclaimer in the |
| 14 |
* documentation and/or other materials provided with the distribution. |
| 15 |
* 3. Neither the name of the University nor the names of its contributors |
| 16 |
* may be used to endorse or promote products derived from this software |
| 17 |
* without specific prior written permission. |
| 18 |
* |
| 19 |
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND |
| 20 |
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| 21 |
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
| 22 |
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE |
| 23 |
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL |
| 24 |
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS |
| 25 |
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) |
| 26 |
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT |
| 27 |
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY |
| 28 |
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF |
| 29 |
* SUCH DAMAGE. |
| 30 |
*/ |
| 31 |
|
| 32 |
#if 0 |
| 33 |
#ifndef lint |
| 34 |
static char sccsid[] = "@(#)lexi.c 8.1 (Berkeley) 6/6/93"; |
| 35 |
#endif /* not lint */ |
| 36 |
#endif |
| 37 |
|
| 38 |
#include "c.h" |
| 39 |
|
| 40 |
/* |
| 41 |
* Here we have the token scanner for indent. It scans off one token and puts |
| 42 |
* it in the global variable "token". It returns a code, indicating the type |
| 43 |
* of token scanned. |
| 44 |
*/ |
| 45 |
|
| 46 |
#include <err.h> |
| 47 |
#include <stdio.h> |
| 48 |
#include <ctype.h> |
| 49 |
#include <stdlib.h> |
| 50 |
#include <string.h> |
| 51 |
#include "indent_globs.h" |
| 52 |
#include "indent_codes.h" |
| 53 |
#include "indent.h" |
| 54 |
|
| 55 |
#define alphanum 1 |
| 56 |
#ifdef undef |
| 57 |
#define opchar 3 |
| 58 |
#endif |
| 59 |
|
| 60 |
struct templ { |
| 61 |
const char *rwd; |
| 62 |
int rwcode; |
| 63 |
}; |
| 64 |
|
| 65 |
/* |
| 66 |
* This table has to be sorted alphabetically, because it'll be used in binary |
| 67 |
* search. For the same reason, string must be the first thing in struct templ. |
| 68 |
*/ |
| 69 |
struct templ specials[] = |
| 70 |
{ |
| 71 |
{"_Bool", 4}, |
| 72 |
{"_Complex", 4}, |
| 73 |
{"_Imaginary", 4}, |
| 74 |
{"auto", 10}, |
| 75 |
{"bool", 4}, |
| 76 |
{"break", 9}, |
| 77 |
{"case", 8}, |
| 78 |
{"char", 4}, |
| 79 |
{"complex", 4}, |
| 80 |
{"const", 4}, |
| 81 |
{"continue", 12}, |
| 82 |
{"default", 8}, |
| 83 |
{"do", 6}, |
| 84 |
{"double", 4}, |
| 85 |
{"else", 6}, |
| 86 |
{"enum", 3}, |
| 87 |
{"extern", 10}, |
| 88 |
{"float", 4}, |
| 89 |
{"for", 5}, |
| 90 |
{"global", 4}, |
| 91 |
{"goto", 9}, |
| 92 |
{"if", 5}, |
| 93 |
{"imaginary", 4}, |
| 94 |
{"inline", 12}, |
| 95 |
{"int", 4}, |
| 96 |
{"long", 4}, |
| 97 |
{"offsetof", 1}, |
| 98 |
{"register", 10}, |
| 99 |
{"restrict", 12}, |
| 100 |
{"return", 9}, |
| 101 |
{"short", 4}, |
| 102 |
{"signed", 4}, |
| 103 |
{"sizeof", 2}, |
| 104 |
{"static", 10}, |
| 105 |
{"struct", 3}, |
| 106 |
{"switch", 7}, |
| 107 |
{"typedef", 11}, |
| 108 |
{"union", 3}, |
| 109 |
{"unsigned", 4}, |
| 110 |
{"void", 4}, |
| 111 |
{"volatile", 4}, |
| 112 |
{"while", 5} |
| 113 |
}; |
| 114 |
|
| 115 |
const char **typenames; |
| 116 |
int typename_count; |
| 117 |
int typename_top = -1; |
| 118 |
|
| 119 |
char chartype[128] = |
| 120 |
{ /* this is used to facilitate the decision of |
| 121 |
* what type (alphanumeric, operator) each |
| 122 |
* character is */ |
| 123 |
0, 0, 0, 0, 0, 0, 0, 0, |
| 124 |
0, 0, 0, 0, 0, 0, 0, 0, |
| 125 |
0, 0, 0, 0, 0, 0, 0, 0, |
| 126 |
0, 0, 0, 0, 0, 0, 0, 0, |
| 127 |
0, 3, 0, 0, 1, 3, 3, 0, |
| 128 |
0, 0, 3, 3, 0, 3, 0, 3, |
| 129 |
1, 1, 1, 1, 1, 1, 1, 1, |
| 130 |
1, 1, 0, 0, 3, 3, 3, 3, |
| 131 |
0, 1, 1, 1, 1, 1, 1, 1, |
| 132 |
1, 1, 1, 1, 1, 1, 1, 1, |
| 133 |
1, 1, 1, 1, 1, 1, 1, 1, |
| 134 |
1, 1, 1, 0, 0, 0, 3, 1, |
| 135 |
0, 1, 1, 1, 1, 1, 1, 1, |
| 136 |
1, 1, 1, 1, 1, 1, 1, 1, |
| 137 |
1, 1, 1, 1, 1, 1, 1, 1, |
| 138 |
1, 1, 1, 0, 3, 0, 3, 0 |
| 139 |
}; |
| 140 |
|
| 141 |
static int |
| 142 |
strcmp_type(const void *e1, const void *e2) |
| 143 |
{ |
| 144 |
return (strcmp(e1, *(const char * const *)e2)); |
| 145 |
} |
| 146 |
|
| 147 |
/* |
| 148 |
* Decide whether "foo(..." is a function definition or declaration. |
| 149 |
* |
| 150 |
* At call, we are looking at the '('. Look ahead to find the first |
| 151 |
* '{', ';' or ',' that is not within parentheses or comments; then |
| 152 |
* it's a definition if we found '{', otherwise a declaration. |
| 153 |
* Note that this rule is fooled by K&R-style parameter declarations, |
| 154 |
* but telling the difference between those and function attributes |
| 155 |
* seems like more trouble than it's worth. This code could also be |
| 156 |
* fooled by mismatched parens or apparent comment starts within string |
| 157 |
* literals, but that seems unlikely in the context it's used in. |
| 158 |
*/ |
| 159 |
static int |
| 160 |
is_func_definition(char *tp) |
| 161 |
{ |
| 162 |
int paren_depth = 0; |
| 163 |
int in_comment = false; |
| 164 |
int in_slash_comment = false; |
| 165 |
int lastc = 0; |
| 166 |
|
| 167 |
/* We may need to look past the end of the current buffer. */ |
| 168 |
lookahead_reset(); |
| 169 |
for (;;) { |
| 170 |
int c; |
| 171 |
|
| 172 |
/* Fetch next character. */ |
| 173 |
if (tp < buf_end) |
| 174 |
c = *tp++; |
| 175 |
else { |
| 176 |
c = lookahead(); |
| 177 |
if (c == EOF) |
| 178 |
break; |
| 179 |
} |
| 180 |
/* Handle comments. */ |
| 181 |
if (in_comment) { |
| 182 |
if (lastc == '*' && c == '/') |
| 183 |
in_comment = false; |
| 184 |
} else if (lastc == '/' && c == '*' && !in_slash_comment) |
| 185 |
in_comment = true; |
| 186 |
else if (in_slash_comment) { |
| 187 |
if (c == '\n') |
| 188 |
in_slash_comment = false; |
| 189 |
} else if (lastc == '/' && c == '/') |
| 190 |
in_slash_comment = true; |
| 191 |
/* Count nested parens properly. */ |
| 192 |
else if (c == '(') |
| 193 |
paren_depth++; |
| 194 |
else if (c == ')') { |
| 195 |
paren_depth--; |
| 196 |
/* |
| 197 |
* If we find unbalanced parens, we must have started inside a |
| 198 |
* declaration. |
| 199 |
*/ |
| 200 |
if (paren_depth < 0) |
| 201 |
return false; |
| 202 |
} else if (paren_depth == 0) { |
| 203 |
/* We are outside any parentheses or comments. */ |
| 204 |
if (c == '{') |
| 205 |
return true; |
| 206 |
else if (c == ';' || c == ',') |
| 207 |
return false; |
| 208 |
} |
| 209 |
lastc = c; |
| 210 |
} |
| 211 |
/* Hit EOF --- for lack of anything better, assume "not a definition". */ |
| 212 |
return false; |
| 213 |
} |
| 214 |
|
| 215 |
int |
| 216 |
lexi(struct parser_state *state) |
| 217 |
{ |
| 218 |
int unary_delim; /* this is set to 1 if the current token |
| 219 |
* forces a following operator to be unary */ |
| 220 |
int code; /* internal code to be returned */ |
| 221 |
char qchar; /* the delimiter character for a string */ |
| 222 |
|
| 223 |
e_token = s_token; /* point to start of place to save token */ |
| 224 |
unary_delim = false; |
| 225 |
state->col_1 = state->last_nl; /* tell world that this token started |
| 226 |
* in column 1 iff the last thing |
| 227 |
* scanned was a newline */ |
| 228 |
state->last_nl = false; |
| 229 |
|
| 230 |
while (*buf_ptr == ' ' || *buf_ptr == '\t') { /* get rid of blanks */ |
| 231 |
state->col_1 = false; /* leading blanks imply token is not in column |
| 232 |
* 1 */ |
| 233 |
if (++buf_ptr >= buf_end) |
| 234 |
fill_buffer(); |
| 235 |
} |
| 236 |
|
| 237 |
/* Scan an alphanumeric token */ |
| 238 |
if (chartype[*buf_ptr & 127] == alphanum || |
| 239 |
(buf_ptr[0] == '.' && isdigit((unsigned char)buf_ptr[1]))) { |
| 240 |
/* |
| 241 |
* we have a character or number |
| 242 |
*/ |
| 243 |
struct templ *p; |
| 244 |
|
| 245 |
if (isdigit((unsigned char)*buf_ptr) || |
| 246 |
(buf_ptr[0] == '.' && isdigit((unsigned char)buf_ptr[1]))) { |
| 247 |
int seendot = 0, |
| 248 |
seenexp = 0, |
| 249 |
seensfx = 0; |
| 250 |
|
| 251 |
/* |
| 252 |
* base 2, base 8, base 16: |
| 253 |
*/ |
| 254 |
if (buf_ptr[0] == '0' && buf_ptr[1] != '.') { |
| 255 |
int len; |
| 256 |
|
| 257 |
if (buf_ptr[1] == 'b' || buf_ptr[1] == 'B') |
| 258 |
len = strspn(buf_ptr + 2, "01") + 2; |
| 259 |
else if (buf_ptr[1] == 'x' || buf_ptr[1] == 'X') |
| 260 |
len = strspn(buf_ptr + 2, "0123456789ABCDEFabcdef") + 2; |
| 261 |
else |
| 262 |
len = strspn(buf_ptr + 1, "012345678") + 1; |
| 263 |
if (len > 0) { |
| 264 |
CHECK_SIZE_TOKEN(len); |
| 265 |
memcpy(e_token, buf_ptr, len); |
| 266 |
e_token += len; |
| 267 |
buf_ptr += len; |
| 268 |
} |
| 269 |
else |
| 270 |
diag2(1, "Unterminated literal"); |
| 271 |
} |
| 272 |
else /* base 10: */ |
| 273 |
while (1) { |
| 274 |
if (*buf_ptr == '.') { |
| 275 |
if (seendot) |
| 276 |
break; |
| 277 |
else |
| 278 |
seendot++; |
| 279 |
} |
| 280 |
CHECK_SIZE_TOKEN(3); |
| 281 |
*e_token++ = *buf_ptr++; |
| 282 |
if (!isdigit((unsigned char)*buf_ptr) && *buf_ptr != '.') { |
| 283 |
if ((*buf_ptr != 'E' && *buf_ptr != 'e') || seenexp) |
| 284 |
break; |
| 285 |
else { |
| 286 |
seenexp++; |
| 287 |
seendot++; |
| 288 |
*e_token++ = *buf_ptr++; |
| 289 |
if (*buf_ptr == '+' || *buf_ptr == '-') |
| 290 |
*e_token++ = *buf_ptr++; |
| 291 |
} |
| 292 |
} |
| 293 |
} |
| 294 |
|
| 295 |
while (1) { |
| 296 |
CHECK_SIZE_TOKEN(2); |
| 297 |
if (!(seensfx & 1) && (*buf_ptr == 'U' || *buf_ptr == 'u')) { |
| 298 |
*e_token++ = *buf_ptr++; |
| 299 |
seensfx |= 1; |
| 300 |
continue; |
| 301 |
} |
| 302 |
if (!(seensfx & 2) && (strchr("fFlL", *buf_ptr) != NULL)) { |
| 303 |
if (buf_ptr[1] == buf_ptr[0]) |
| 304 |
*e_token++ = *buf_ptr++; |
| 305 |
*e_token++ = *buf_ptr++; |
| 306 |
seensfx |= 2; |
| 307 |
continue; |
| 308 |
} |
| 309 |
break; |
| 310 |
} |
| 311 |
} |
| 312 |
else |
| 313 |
while (chartype[*buf_ptr & 127] == alphanum || *buf_ptr == BACKSLASH) { |
| 314 |
/* fill_buffer() terminates buffer with newline */ |
| 315 |
if (*buf_ptr == BACKSLASH) { |
| 316 |
if (*(buf_ptr + 1) == '\n') { |
| 317 |
buf_ptr += 2; |
| 318 |
if (buf_ptr >= buf_end) |
| 319 |
fill_buffer(); |
| 320 |
} else |
| 321 |
break; |
| 322 |
} |
| 323 |
CHECK_SIZE_TOKEN(1); |
| 324 |
/* copy it over */ |
| 325 |
*e_token++ = *buf_ptr++; |
| 326 |
if (buf_ptr >= buf_end) |
| 327 |
fill_buffer(); |
| 328 |
} |
| 329 |
*e_token = '\0'; |
| 330 |
|
| 331 |
if (s_token[0] == 'L' && s_token[1] == '\0' && |
| 332 |
(*buf_ptr == '"' || *buf_ptr == '\'')) |
| 333 |
return (strpfx); |
| 334 |
|
| 335 |
while (*buf_ptr == ' ' || *buf_ptr == '\t') { /* get rid of blanks */ |
| 336 |
if (++buf_ptr >= buf_end) |
| 337 |
fill_buffer(); |
| 338 |
} |
| 339 |
state->keyword = 0; |
| 340 |
if (state->last_token == structure && !state->p_l_follow) { |
| 341 |
/* if last token was 'struct' and we're not |
| 342 |
* in parentheses, then this token |
| 343 |
* should be treated as a declaration */ |
| 344 |
state->last_u_d = true; |
| 345 |
return (decl); |
| 346 |
} |
| 347 |
/* |
| 348 |
* Operator after identifier is binary unless last token was 'struct' |
| 349 |
*/ |
| 350 |
state->last_u_d = (state->last_token == structure); |
| 351 |
|
| 352 |
p = bsearch(s_token, |
| 353 |
specials, |
| 354 |
sizeof(specials) / sizeof(specials[0]), |
| 355 |
sizeof(specials[0]), |
| 356 |
strcmp_type); |
| 357 |
if (p == NULL) { /* not a special keyword... */ |
| 358 |
char *u; |
| 359 |
|
| 360 |
/* ... so maybe a type_t or a typedef */ |
| 361 |
if ((auto_typedefs && ((u = strrchr(s_token, '_')) != NULL) && |
| 362 |
strcmp(u, "_t") == 0) || (typename_top >= 0 && |
| 363 |
bsearch(s_token, typenames, typename_top + 1, |
| 364 |
sizeof(typenames[0]), strcmp_type))) { |
| 365 |
state->keyword = 4; /* a type name */ |
| 366 |
state->last_u_d = true; |
| 367 |
goto found_typename; |
| 368 |
} |
| 369 |
} else { /* we have a keyword */ |
| 370 |
state->keyword = p->rwcode; |
| 371 |
state->last_u_d = true; |
| 372 |
switch (p->rwcode) { |
| 373 |
case 7: /* it is a switch */ |
| 374 |
return (swstmt); |
| 375 |
case 8: /* a case or default */ |
| 376 |
return (casestmt); |
| 377 |
|
| 378 |
case 3: /* a "struct" */ |
| 379 |
/* FALLTHROUGH */ |
| 380 |
case 4: /* one of the declaration keywords */ |
| 381 |
found_typename: |
| 382 |
if (state->p_l_follow) { |
| 383 |
/* inside parens: cast, param list, offsetof or sizeof */ |
| 384 |
state->cast_mask |= (1 << state->p_l_follow) & ~state->not_cast_mask; |
| 385 |
} |
| 386 |
if (state->last_token == period || state->last_token == unary_op) { |
| 387 |
state->keyword = 0; |
| 388 |
break; |
| 389 |
} |
| 390 |
if (p != NULL && p->rwcode == 3) |
| 391 |
return (structure); |
| 392 |
if (state->p_l_follow) |
| 393 |
break; |
| 394 |
return (decl); |
| 395 |
|
| 396 |
case 5: /* if, while, for */ |
| 397 |
return (sp_paren); |
| 398 |
|
| 399 |
case 6: /* do, else */ |
| 400 |
return (sp_nparen); |
| 401 |
|
| 402 |
case 10: /* storage class specifier */ |
| 403 |
return (storage); |
| 404 |
|
| 405 |
case 11: /* typedef */ |
| 406 |
return (type_def); |
| 407 |
|
| 408 |
default: /* all others are treated like any other |
| 409 |
* identifier */ |
| 410 |
return (ident); |
| 411 |
} /* end of switch */ |
| 412 |
} /* end of if (found_it) */ |
| 413 |
if (*buf_ptr == '(' && state->tos <= 1 && state->ind_level == 0 && |
| 414 |
state->in_parameter_declaration == 0 && state->block_init == 0) { |
| 415 |
if (is_func_definition(buf_ptr)) { |
| 416 |
strncpy(state->procname, token, sizeof state->procname - 1); |
| 417 |
if (state->in_decl) |
| 418 |
state->in_parameter_declaration = 1; |
| 419 |
return (funcname); |
| 420 |
} |
| 421 |
} |
| 422 |
/* |
| 423 |
* The following hack attempts to guess whether or not the current |
| 424 |
* token is in fact a declaration keyword -- one that has been |
| 425 |
* typedefd |
| 426 |
*/ |
| 427 |
else if (!state->p_l_follow && !state->block_init && |
| 428 |
!state->in_stmt && |
| 429 |
((*buf_ptr == '*' && buf_ptr[1] != '=') || |
| 430 |
isalpha((unsigned char)*buf_ptr)) && |
| 431 |
(state->last_token == semicolon || state->last_token == lbrace || |
| 432 |
state->last_token == rbrace)) { |
| 433 |
state->keyword = 4; /* a type name */ |
| 434 |
state->last_u_d = true; |
| 435 |
return decl; |
| 436 |
} |
| 437 |
if (state->last_token == decl) /* if this is a declared variable, |
| 438 |
* then following sign is unary */ |
| 439 |
state->last_u_d = true; /* will make "int a -1" work */ |
| 440 |
return (ident); /* the ident is not in the list */ |
| 441 |
} /* end of processing for alphanum character */ |
| 442 |
|
| 443 |
/* Scan a non-alphanumeric token */ |
| 444 |
|
| 445 |
CHECK_SIZE_TOKEN(3); /* things like "<<=" */ |
| 446 |
*e_token++ = *buf_ptr; /* if it is only a one-character token, it is |
| 447 |
* moved here */ |
| 448 |
*e_token = '\0'; |
| 449 |
if (++buf_ptr >= buf_end) |
| 450 |
fill_buffer(); |
| 451 |
|
| 452 |
switch (*token) { |
| 453 |
case '\n': |
| 454 |
unary_delim = state->last_u_d; |
| 455 |
state->last_nl = true; /* remember that we just had a newline */ |
| 456 |
code = (had_eof ? 0 : newline); |
| 457 |
|
| 458 |
/* |
| 459 |
* if data has been exhausted, the newline is a dummy, and we should |
| 460 |
* return code to stop |
| 461 |
*/ |
| 462 |
break; |
| 463 |
|
| 464 |
case '\'': /* start of quoted character */ |
| 465 |
case '"': /* start of string */ |
| 466 |
qchar = *token; |
| 467 |
do { /* copy the string */ |
| 468 |
while (1) { /* move one character or [/<char>]<char> */ |
| 469 |
if (*buf_ptr == '\n') { |
| 470 |
diag2(1, "Unterminated literal"); |
| 471 |
goto stop_lit; |
| 472 |
} |
| 473 |
CHECK_SIZE_TOKEN(2); |
| 474 |
*e_token = *buf_ptr++; |
| 475 |
if (buf_ptr >= buf_end) |
| 476 |
fill_buffer(); |
| 477 |
if (*e_token == BACKSLASH) { /* if escape, copy extra char */ |
| 478 |
if (*buf_ptr == '\n') /* check for escaped newline */ |
| 479 |
++line_no; |
| 480 |
*++e_token = *buf_ptr++; |
| 481 |
++e_token; /* we must increment this again because we |
| 482 |
* copied two chars */ |
| 483 |
if (buf_ptr >= buf_end) |
| 484 |
fill_buffer(); |
| 485 |
} |
| 486 |
else |
| 487 |
break; /* we copied one character */ |
| 488 |
} /* end of while (1) */ |
| 489 |
} while (*e_token++ != qchar); |
| 490 |
stop_lit: |
| 491 |
code = ident; |
| 492 |
break; |
| 493 |
|
| 494 |
case ('('): |
| 495 |
case ('['): |
| 496 |
unary_delim = true; |
| 497 |
code = lparen; |
| 498 |
break; |
| 499 |
|
| 500 |
case (')'): |
| 501 |
case (']'): |
| 502 |
code = rparen; |
| 503 |
break; |
| 504 |
|
| 505 |
case '#': |
| 506 |
unary_delim = state->last_u_d; |
| 507 |
code = preesc; |
| 508 |
break; |
| 509 |
|
| 510 |
case '?': |
| 511 |
unary_delim = true; |
| 512 |
code = question; |
| 513 |
break; |
| 514 |
|
| 515 |
case (':'): |
| 516 |
code = colon; |
| 517 |
unary_delim = true; |
| 518 |
break; |
| 519 |
|
| 520 |
case (';'): |
| 521 |
unary_delim = true; |
| 522 |
code = semicolon; |
| 523 |
break; |
| 524 |
|
| 525 |
case ('{'): |
| 526 |
unary_delim = true; |
| 527 |
|
| 528 |
/* |
| 529 |
* if (state->in_or_st) state->block_init = 1; |
| 530 |
*/ |
| 531 |
/* ? code = state->block_init ? lparen : lbrace; */ |
| 532 |
code = lbrace; |
| 533 |
break; |
| 534 |
|
| 535 |
case ('}'): |
| 536 |
unary_delim = true; |
| 537 |
/* ? code = state->block_init ? rparen : rbrace; */ |
| 538 |
code = rbrace; |
| 539 |
break; |
| 540 |
|
| 541 |
case 014: /* a form feed */ |
| 542 |
unary_delim = state->last_u_d; |
| 543 |
state->last_nl = true; /* remember this so we can set 'state->col_1' |
| 544 |
* right */ |
| 545 |
code = form_feed; |
| 546 |
break; |
| 547 |
|
| 548 |
case (','): |
| 549 |
unary_delim = true; |
| 550 |
code = comma; |
| 551 |
break; |
| 552 |
|
| 553 |
case '.': |
| 554 |
unary_delim = false; |
| 555 |
code = period; |
| 556 |
break; |
| 557 |
|
| 558 |
case '-': |
| 559 |
case '+': /* check for -, +, --, ++ */ |
| 560 |
code = (state->last_u_d ? unary_op : binary_op); |
| 561 |
unary_delim = true; |
| 562 |
|
| 563 |
if (*buf_ptr == token[0]) { |
| 564 |
/* check for doubled character */ |
| 565 |
*e_token++ = *buf_ptr++; |
| 566 |
/* buffer overflow will be checked at end of loop */ |
| 567 |
if (state->last_token == ident || state->last_token == rparen) { |
| 568 |
code = (state->last_u_d ? unary_op : postop); |
| 569 |
/* check for following ++ or -- */ |
| 570 |
unary_delim = false; |
| 571 |
} |
| 572 |
} |
| 573 |
else if (*buf_ptr == '=') |
| 574 |
/* check for operator += */ |
| 575 |
*e_token++ = *buf_ptr++; |
| 576 |
else if (*buf_ptr == '>') { |
| 577 |
/* check for operator -> */ |
| 578 |
*e_token++ = *buf_ptr++; |
| 579 |
unary_delim = false; |
| 580 |
code = unary_op; |
| 581 |
state->want_blank = false; |
| 582 |
} |
| 583 |
break; /* buffer overflow will be checked at end of |
| 584 |
* switch */ |
| 585 |
|
| 586 |
case '=': |
| 587 |
if (state->in_or_st) |
| 588 |
state->block_init = 1; |
| 589 |
#ifdef undef |
| 590 |
if (chartype[*buf_ptr & 127] == opchar) { /* we have two char assignment */ |
| 591 |
e_token[-1] = *buf_ptr++; |
| 592 |
if ((e_token[-1] == '<' || e_token[-1] == '>') && e_token[-1] == *buf_ptr) |
| 593 |
*e_token++ = *buf_ptr++; |
| 594 |
*e_token++ = '='; /* Flip =+ to += */ |
| 595 |
*e_token = 0; |
| 596 |
} |
| 597 |
#else |
| 598 |
if (*buf_ptr == '=') {/* == */ |
| 599 |
*e_token++ = '='; /* Flip =+ to += */ |
| 600 |
buf_ptr++; |
| 601 |
*e_token = 0; |
| 602 |
} |
| 603 |
#endif |
| 604 |
code = binary_op; |
| 605 |
unary_delim = true; |
| 606 |
break; |
| 607 |
/* can drop thru!!! */ |
| 608 |
|
| 609 |
case '>': |
| 610 |
case '<': |
| 611 |
case '!': /* ops like <, <<, <=, !=, etc */ |
| 612 |
if (*buf_ptr == '>' || *buf_ptr == '<' || *buf_ptr == '=') { |
| 613 |
*e_token++ = *buf_ptr; |
| 614 |
if (++buf_ptr >= buf_end) |
| 615 |
fill_buffer(); |
| 616 |
} |
| 617 |
if (*buf_ptr == '=') |
| 618 |
*e_token++ = *buf_ptr++; |
| 619 |
code = (state->last_u_d ? unary_op : binary_op); |
| 620 |
unary_delim = true; |
| 621 |
break; |
| 622 |
|
| 623 |
case '*': |
| 624 |
unary_delim = true; |
| 625 |
if (!state->last_u_d) { |
| 626 |
if (*buf_ptr == '=') |
| 627 |
*e_token++ = *buf_ptr++; |
| 628 |
code = binary_op; |
| 629 |
break; |
| 630 |
} |
| 631 |
while (*buf_ptr == '*' || isspace((unsigned char)*buf_ptr)) { |
| 632 |
if (*buf_ptr == '*') { |
| 633 |
CHECK_SIZE_TOKEN(1); |
| 634 |
*e_token++ = *buf_ptr; |
| 635 |
} |
| 636 |
if (++buf_ptr >= buf_end) |
| 637 |
fill_buffer(); |
| 638 |
} |
| 639 |
code = unary_op; |
| 640 |
break; |
| 641 |
|
| 642 |
default: |
| 643 |
if (token[0] == '/' && *buf_ptr == '*') { |
| 644 |
/* it is start of comment */ |
| 645 |
*e_token++ = '*'; |
| 646 |
|
| 647 |
if (++buf_ptr >= buf_end) |
| 648 |
fill_buffer(); |
| 649 |
|
| 650 |
code = comment; |
| 651 |
unary_delim = state->last_u_d; |
| 652 |
break; |
| 653 |
} |
| 654 |
while (*(e_token - 1) == *buf_ptr || *buf_ptr == '=') { |
| 655 |
/* |
| 656 |
* handle ||, &&, etc, and also things as in int *****i |
| 657 |
*/ |
| 658 |
CHECK_SIZE_TOKEN(1); |
| 659 |
*e_token++ = *buf_ptr; |
| 660 |
if (++buf_ptr >= buf_end) |
| 661 |
fill_buffer(); |
| 662 |
} |
| 663 |
code = (state->last_u_d ? unary_op : binary_op); |
| 664 |
unary_delim = true; |
| 665 |
|
| 666 |
|
| 667 |
} /* end of switch */ |
| 668 |
if (buf_ptr >= buf_end) /* check for input buffer empty */ |
| 669 |
fill_buffer(); |
| 670 |
state->last_u_d = unary_delim; |
| 671 |
CHECK_SIZE_TOKEN(1); |
| 672 |
*e_token = '\0'; /* null terminate the token */ |
| 673 |
return (code); |
| 674 |
} |
| 675 |
|
| 676 |
void |
| 677 |
alloc_typenames(void) |
| 678 |
{ |
| 679 |
|
| 680 |
typenames = (const char **)malloc(sizeof(typenames[0]) * |
| 681 |
(typename_count = 16)); |
| 682 |
if (typenames == NULL) |
| 683 |
err(1, NULL); |
| 684 |
} |
| 685 |
|
| 686 |
void |
| 687 |
add_typename(const char *key) |
| 688 |
{ |
| 689 |
int comparison; |
| 690 |
const char *copy; |
| 691 |
|
| 692 |
if (typename_top + 1 >= typename_count) { |
| 693 |
typenames = realloc((void *)typenames, |
| 694 |
sizeof(typenames[0]) * (typename_count *= 2)); |
| 695 |
if (typenames == NULL) |
| 696 |
err(1, NULL); |
| 697 |
} |
| 698 |
if (typename_top == -1) |
| 699 |
typenames[++typename_top] = copy = strdup(key); |
| 700 |
else if ((comparison = strcmp(key, typenames[typename_top])) >= 0) { |
| 701 |
/* take advantage of sorted input */ |
| 702 |
if (comparison == 0) /* remove duplicates */ |
| 703 |
return; |
| 704 |
typenames[++typename_top] = copy = strdup(key); |
| 705 |
} |
| 706 |
else { |
| 707 |
int p; |
| 708 |
|
| 709 |
for (p = 0; (comparison = strcmp(key, typenames[p])) > 0; p++) |
| 710 |
/* find place for the new key */; |
| 711 |
if (comparison == 0) /* remove duplicates */ |
| 712 |
return; |
| 713 |
memmove(&typenames[p + 1], &typenames[p], |
| 714 |
sizeof(typenames[0]) * (++typename_top - p)); |
| 715 |
typenames[p] = copy = strdup(key); |
| 716 |
} |
| 717 |
|
| 718 |
if (copy == NULL) |
| 719 |
err(1, NULL); |
| 720 |
} |