| 1 | %{/* nlmheader.y - parse NLM header specification keywords. |
| 2 | Copyright 1993, 1994, 1995, 1997, 1998, 2001, 2002, 2003, 2007 |
| 3 | Free Software Foundation, Inc. |
| 4 | |
| 5 | This file is part of GNU Binutils. |
| 6 | |
| 7 | This program is free software; you can redistribute it and/or modify |
| 8 | it under the terms of the GNU General Public License as published by |
| 9 | the Free Software Foundation; either version 3 of the License, or |
| 10 | (at your option) any later version. |
| 11 | |
| 12 | This program is distributed in the hope that it will be useful, |
| 13 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 15 | GNU General Public License for more details. |
| 16 | |
| 17 | You should have received a copy of the GNU General Public License |
| 18 | along with this program; if not, write to the Free Software |
| 19 | Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, |
| 20 | MA 02110-1301, USA. */ |
| 21 | |
| 22 | /* Written by Ian Lance Taylor <ian@cygnus.com>. |
| 23 | |
| 24 | This bison file parses the commands recognized by the NetWare NLM |
| 25 | linker, except for lists of object files. It stores the |
| 26 | information in global variables. |
| 27 | |
| 28 | This implementation is based on the description in the NetWare Tool |
| 29 | Maker Specification manual, edition 1.0. */ |
| 30 | |
| 31 | #include "sysdep.h" |
| 32 | #include "safe-ctype.h" |
| 33 | #include "bfd.h" |
| 34 | #include "nlm/common.h" |
| 35 | #include "nlm/internal.h" |
| 36 | #include "bucomm.h" |
| 37 | #include "nlmconv.h" |
| 38 | |
| 39 | /* Information is stored in the structures pointed to by these |
| 40 | variables. */ |
| 41 | |
| 42 | Nlm_Internal_Fixed_Header *fixed_hdr; |
| 43 | Nlm_Internal_Variable_Header *var_hdr; |
| 44 | Nlm_Internal_Version_Header *version_hdr; |
| 45 | Nlm_Internal_Copyright_Header *copyright_hdr; |
| 46 | Nlm_Internal_Extended_Header *extended_hdr; |
| 47 | |
| 48 | /* Procedure named by CHECK. */ |
| 49 | char *check_procedure; |
| 50 | /* File named by CUSTOM. */ |
| 51 | char *custom_file; |
| 52 | /* Whether to generate debugging information (DEBUG). */ |
| 53 | bfd_boolean debug_info; |
| 54 | /* Procedure named by EXIT. */ |
| 55 | char *exit_procedure; |
| 56 | /* Exported symbols (EXPORT). */ |
| 57 | struct string_list *export_symbols; |
| 58 | /* List of files from INPUT. */ |
| 59 | struct string_list *input_files; |
| 60 | /* Map file name (MAP, FULLMAP). */ |
| 61 | char *map_file; |
| 62 | /* Whether a full map has been requested (FULLMAP). */ |
| 63 | bfd_boolean full_map; |
| 64 | /* File named by HELP. */ |
| 65 | char *help_file; |
| 66 | /* Imported symbols (IMPORT). */ |
| 67 | struct string_list *import_symbols; |
| 68 | /* File named by MESSAGES. */ |
| 69 | char *message_file; |
| 70 | /* Autoload module list (MODULE). */ |
| 71 | struct string_list *modules; |
| 72 | /* File named by OUTPUT. */ |
| 73 | char *output_file; |
| 74 | /* File named by SHARELIB. */ |
| 75 | char *sharelib_file; |
| 76 | /* Start procedure name (START). */ |
| 77 | char *start_procedure; |
| 78 | /* VERBOSE. */ |
| 79 | bfd_boolean verbose; |
| 80 | /* RPC description file (XDCDATA). */ |
| 81 | char *rpc_file; |
| 82 | |
| 83 | /* The number of serious errors that have occurred. */ |
| 84 | int parse_errors; |
| 85 | |
| 86 | /* The current symbol prefix when reading a list of import or export |
| 87 | symbols. */ |
| 88 | static char *symbol_prefix; |
| 89 | |
| 90 | /* Parser error message handler. */ |
| 91 | #define yyerror(msg) nlmheader_error (msg); |
| 92 | |
| 93 | /* Local functions. */ |
| 94 | static int yylex (void); |
| 95 | static void nlmlex_file_push (const char *); |
| 96 | static bfd_boolean nlmlex_file_open (const char *); |
| 97 | static int nlmlex_buf_init (void); |
| 98 | static char nlmlex_buf_add (int); |
| 99 | static long nlmlex_get_number (const char *); |
| 100 | static void nlmheader_identify (void); |
| 101 | static void nlmheader_warn (const char *, int); |
| 102 | static void nlmheader_error (const char *); |
| 103 | static struct string_list * string_list_cons (char *, struct string_list *); |
| 104 | static struct string_list * string_list_append (struct string_list *, |
| 105 | struct string_list *); |
| 106 | static struct string_list * string_list_append1 (struct string_list *, |
| 107 | char *); |
| 108 | static char *xstrdup (const char *); |
| 109 | |
| 110 | %} |
| 111 | |
| 112 | %union |
| 113 | { |
| 114 | char *string; |
| 115 | struct string_list *list; |
| 116 | }; |
| 117 | |
| 118 | /* The reserved words. */ |
| 119 | |
| 120 | %token CHECK CODESTART COPYRIGHT CUSTOM DATE DEBUG DESCRIPTION EXIT |
| 121 | %token EXPORT FLAG_ON FLAG_OFF FULLMAP HELP IMPORT INPUT MAP MESSAGES |
| 122 | %token MODULE MULTIPLE OS_DOMAIN OUTPUT PSEUDOPREEMPTION REENTRANT |
| 123 | %token SCREENNAME SHARELIB STACK START SYNCHRONIZE |
| 124 | %token THREADNAME TYPE VERBOSE VERSIONK XDCDATA |
| 125 | |
| 126 | /* Arguments. */ |
| 127 | |
| 128 | %token <string> STRING |
| 129 | %token <string> QUOTED_STRING |
| 130 | |
| 131 | /* Typed non-terminals. */ |
| 132 | %type <list> symbol_list_opt symbol_list string_list |
| 133 | %type <string> symbol |
| 134 | |
| 135 | %% |
| 136 | |
| 137 | /* Keywords must start in the leftmost column of the file. Arguments |
| 138 | may appear anywhere else. The lexer uses this to determine what |
| 139 | token to return, so we don't have to worry about it here. */ |
| 140 | |
| 141 | /* The entire file is just a list of commands. */ |
| 142 | |
| 143 | file: |
| 144 | commands |
| 145 | ; |
| 146 | |
| 147 | /* A possibly empty list of commands. */ |
| 148 | |
| 149 | commands: |
| 150 | /* May be empty. */ |
| 151 | | command commands |
| 152 | ; |
| 153 | |
| 154 | /* A single command. There is where most of the work takes place. */ |
| 155 | |
| 156 | command: |
| 157 | CHECK STRING |
| 158 | { |
| 159 | check_procedure = $2; |
| 160 | } |
| 161 | | CODESTART STRING |
| 162 | { |
| 163 | nlmheader_warn (_("CODESTART is not implemented; sorry"), -1); |
| 164 | free ($2); |
| 165 | } |
| 166 | | COPYRIGHT QUOTED_STRING |
| 167 | { |
| 168 | int len; |
| 169 | |
| 170 | strncpy (copyright_hdr->stamp, "CoPyRiGhT=", 10); |
| 171 | len = strlen ($2); |
| 172 | if (len >= NLM_MAX_COPYRIGHT_MESSAGE_LENGTH) |
| 173 | { |
| 174 | nlmheader_warn (_("copyright string is too long"), |
| 175 | NLM_MAX_COPYRIGHT_MESSAGE_LENGTH - 1); |
| 176 | len = NLM_MAX_COPYRIGHT_MESSAGE_LENGTH - 1; |
| 177 | } |
| 178 | copyright_hdr->copyrightMessageLength = len; |
| 179 | strncpy (copyright_hdr->copyrightMessage, $2, len); |
| 180 | copyright_hdr->copyrightMessage[len] = '\0'; |
| 181 | free ($2); |
| 182 | } |
| 183 | | CUSTOM STRING |
| 184 | { |
| 185 | custom_file = $2; |
| 186 | } |
| 187 | | DATE STRING STRING STRING |
| 188 | { |
| 189 | /* We don't set the version stamp here, because we use the |
| 190 | version stamp to detect whether the required VERSION |
| 191 | keyword was given. */ |
| 192 | version_hdr->month = nlmlex_get_number ($2); |
| 193 | version_hdr->day = nlmlex_get_number ($3); |
| 194 | version_hdr->year = nlmlex_get_number ($4); |
| 195 | free ($2); |
| 196 | free ($3); |
| 197 | free ($4); |
| 198 | if (version_hdr->month < 1 || version_hdr->month > 12) |
| 199 | nlmheader_warn (_("illegal month"), -1); |
| 200 | if (version_hdr->day < 1 || version_hdr->day > 31) |
| 201 | nlmheader_warn (_("illegal day"), -1); |
| 202 | if (version_hdr->year < 1900 || version_hdr->year > 3000) |
| 203 | nlmheader_warn (_("illegal year"), -1); |
| 204 | } |
| 205 | | DEBUG |
| 206 | { |
| 207 | debug_info = TRUE; |
| 208 | } |
| 209 | | DESCRIPTION QUOTED_STRING |
| 210 | { |
| 211 | int len; |
| 212 | |
| 213 | len = strlen ($2); |
| 214 | if (len > NLM_MAX_DESCRIPTION_LENGTH) |
| 215 | { |
| 216 | nlmheader_warn (_("description string is too long"), |
| 217 | NLM_MAX_DESCRIPTION_LENGTH); |
| 218 | len = NLM_MAX_DESCRIPTION_LENGTH; |
| 219 | } |
| 220 | var_hdr->descriptionLength = len; |
| 221 | strncpy (var_hdr->descriptionText, $2, len); |
| 222 | var_hdr->descriptionText[len] = '\0'; |
| 223 | free ($2); |
| 224 | } |
| 225 | | EXIT STRING |
| 226 | { |
| 227 | exit_procedure = $2; |
| 228 | } |
| 229 | | EXPORT |
| 230 | { |
| 231 | symbol_prefix = NULL; |
| 232 | } |
| 233 | symbol_list_opt |
| 234 | { |
| 235 | export_symbols = string_list_append (export_symbols, $3); |
| 236 | } |
| 237 | | FLAG_ON STRING |
| 238 | { |
| 239 | fixed_hdr->flags |= nlmlex_get_number ($2); |
| 240 | free ($2); |
| 241 | } |
| 242 | | FLAG_OFF STRING |
| 243 | { |
| 244 | fixed_hdr->flags &=~ nlmlex_get_number ($2); |
| 245 | free ($2); |
| 246 | } |
| 247 | | FULLMAP |
| 248 | { |
| 249 | map_file = ""; |
| 250 | full_map = TRUE; |
| 251 | } |
| 252 | | FULLMAP STRING |
| 253 | { |
| 254 | map_file = $2; |
| 255 | full_map = TRUE; |
| 256 | } |
| 257 | | HELP STRING |
| 258 | { |
| 259 | help_file = $2; |
| 260 | } |
| 261 | | IMPORT |
| 262 | { |
| 263 | symbol_prefix = NULL; |
| 264 | } |
| 265 | symbol_list_opt |
| 266 | { |
| 267 | import_symbols = string_list_append (import_symbols, $3); |
| 268 | } |
| 269 | | INPUT string_list |
| 270 | { |
| 271 | input_files = string_list_append (input_files, $2); |
| 272 | } |
| 273 | | MAP |
| 274 | { |
| 275 | map_file = ""; |
| 276 | } |
| 277 | | MAP STRING |
| 278 | { |
| 279 | map_file = $2; |
| 280 | } |
| 281 | | MESSAGES STRING |
| 282 | { |
| 283 | message_file = $2; |
| 284 | } |
| 285 | | MODULE string_list |
| 286 | { |
| 287 | modules = string_list_append (modules, $2); |
| 288 | } |
| 289 | | MULTIPLE |
| 290 | { |
| 291 | fixed_hdr->flags |= 0x2; |
| 292 | } |
| 293 | | OS_DOMAIN |
| 294 | { |
| 295 | fixed_hdr->flags |= 0x10; |
| 296 | } |
| 297 | | OUTPUT STRING |
| 298 | { |
| 299 | if (output_file == NULL) |
| 300 | output_file = $2; |
| 301 | else |
| 302 | nlmheader_warn (_("ignoring duplicate OUTPUT statement"), -1); |
| 303 | } |
| 304 | | PSEUDOPREEMPTION |
| 305 | { |
| 306 | fixed_hdr->flags |= 0x8; |
| 307 | } |
| 308 | | REENTRANT |
| 309 | { |
| 310 | fixed_hdr->flags |= 0x1; |
| 311 | } |
| 312 | | SCREENNAME QUOTED_STRING |
| 313 | { |
| 314 | int len; |
| 315 | |
| 316 | len = strlen ($2); |
| 317 | if (len >= NLM_MAX_SCREEN_NAME_LENGTH) |
| 318 | { |
| 319 | nlmheader_warn (_("screen name is too long"), |
| 320 | NLM_MAX_SCREEN_NAME_LENGTH); |
| 321 | len = NLM_MAX_SCREEN_NAME_LENGTH; |
| 322 | } |
| 323 | var_hdr->screenNameLength = len; |
| 324 | strncpy (var_hdr->screenName, $2, len); |
| 325 | var_hdr->screenName[NLM_MAX_SCREEN_NAME_LENGTH] = '\0'; |
| 326 | free ($2); |
| 327 | } |
| 328 | | SHARELIB STRING |
| 329 | { |
| 330 | sharelib_file = $2; |
| 331 | } |
| 332 | | STACK STRING |
| 333 | { |
| 334 | var_hdr->stackSize = nlmlex_get_number ($2); |
| 335 | free ($2); |
| 336 | } |
| 337 | | START STRING |
| 338 | { |
| 339 | start_procedure = $2; |
| 340 | } |
| 341 | | SYNCHRONIZE |
| 342 | { |
| 343 | fixed_hdr->flags |= 0x4; |
| 344 | } |
| 345 | | THREADNAME QUOTED_STRING |
| 346 | { |
| 347 | int len; |
| 348 | |
| 349 | len = strlen ($2); |
| 350 | if (len >= NLM_MAX_THREAD_NAME_LENGTH) |
| 351 | { |
| 352 | nlmheader_warn (_("thread name is too long"), |
| 353 | NLM_MAX_THREAD_NAME_LENGTH); |
| 354 | len = NLM_MAX_THREAD_NAME_LENGTH; |
| 355 | } |
| 356 | var_hdr->threadNameLength = len; |
| 357 | strncpy (var_hdr->threadName, $2, len); |
| 358 | var_hdr->threadName[len] = '\0'; |
| 359 | free ($2); |
| 360 | } |
| 361 | | TYPE STRING |
| 362 | { |
| 363 | fixed_hdr->moduleType = nlmlex_get_number ($2); |
| 364 | free ($2); |
| 365 | } |
| 366 | | VERBOSE |
| 367 | { |
| 368 | verbose = TRUE; |
| 369 | } |
| 370 | | VERSIONK STRING STRING STRING |
| 371 | { |
| 372 | long val; |
| 373 | |
| 374 | strncpy (version_hdr->stamp, "VeRsIoN#", 8); |
| 375 | version_hdr->majorVersion = nlmlex_get_number ($2); |
| 376 | val = nlmlex_get_number ($3); |
| 377 | if (val < 0 || val > 99) |
| 378 | nlmheader_warn (_("illegal minor version number (must be between 0 and 99)"), |
| 379 | -1); |
| 380 | else |
| 381 | version_hdr->minorVersion = val; |
| 382 | val = nlmlex_get_number ($4); |
| 383 | if (val < 0) |
| 384 | nlmheader_warn (_("illegal revision number (must be between 0 and 26)"), |
| 385 | -1); |
| 386 | else if (val > 26) |
| 387 | version_hdr->revision = 0; |
| 388 | else |
| 389 | version_hdr->revision = val; |
| 390 | free ($2); |
| 391 | free ($3); |
| 392 | free ($4); |
| 393 | } |
| 394 | | VERSIONK STRING STRING |
| 395 | { |
| 396 | long val; |
| 397 | |
| 398 | strncpy (version_hdr->stamp, "VeRsIoN#", 8); |
| 399 | version_hdr->majorVersion = nlmlex_get_number ($2); |
| 400 | val = nlmlex_get_number ($3); |
| 401 | if (val < 0 || val > 99) |
| 402 | nlmheader_warn (_("illegal minor version number (must be between 0 and 99)"), |
| 403 | -1); |
| 404 | else |
| 405 | version_hdr->minorVersion = val; |
| 406 | version_hdr->revision = 0; |
| 407 | free ($2); |
| 408 | free ($3); |
| 409 | } |
| 410 | | XDCDATA STRING |
| 411 | { |
| 412 | rpc_file = $2; |
| 413 | } |
| 414 | ; |
| 415 | |
| 416 | /* A possibly empty list of symbols. */ |
| 417 | |
| 418 | symbol_list_opt: |
| 419 | /* Empty. */ |
| 420 | { |
| 421 | $$ = NULL; |
| 422 | } |
| 423 | | symbol_list |
| 424 | { |
| 425 | $$ = $1; |
| 426 | } |
| 427 | ; |
| 428 | |
| 429 | /* A list of symbols in an import or export list. Prefixes may appear |
| 430 | in parentheses. We need to use left recursion here to avoid |
| 431 | building up a large import list on the parser stack. */ |
| 432 | |
| 433 | symbol_list: |
| 434 | symbol |
| 435 | { |
| 436 | $$ = string_list_cons ($1, NULL); |
| 437 | } |
| 438 | | symbol_prefix |
| 439 | { |
| 440 | $$ = NULL; |
| 441 | } |
| 442 | | symbol_list symbol |
| 443 | { |
| 444 | $$ = string_list_append1 ($1, $2); |
| 445 | } |
| 446 | | symbol_list symbol_prefix |
| 447 | { |
| 448 | $$ = $1; |
| 449 | } |
| 450 | ; |
| 451 | |
| 452 | /* A prefix for subsequent symbols. */ |
| 453 | |
| 454 | symbol_prefix: |
| 455 | '(' STRING ')' |
| 456 | { |
| 457 | if (symbol_prefix != NULL) |
| 458 | free (symbol_prefix); |
| 459 | symbol_prefix = $2; |
| 460 | } |
| 461 | ; |
| 462 | |
| 463 | /* A single symbol. */ |
| 464 | |
| 465 | symbol: |
| 466 | STRING |
| 467 | { |
| 468 | if (symbol_prefix == NULL) |
| 469 | $$ = $1; |
| 470 | else |
| 471 | { |
| 472 | $$ = xmalloc (strlen (symbol_prefix) + strlen ($1) + 2); |
| 473 | sprintf ($$, "%s@%s", symbol_prefix, $1); |
| 474 | free ($1); |
| 475 | } |
| 476 | } |
| 477 | ; |
| 478 | |
| 479 | /* A list of strings. */ |
| 480 | |
| 481 | string_list: |
| 482 | /* May be empty. */ |
| 483 | { |
| 484 | $$ = NULL; |
| 485 | } |
| 486 | | STRING string_list |
| 487 | { |
| 488 | $$ = string_list_cons ($1, $2); |
| 489 | } |
| 490 | ; |
| 491 | |
| 492 | %% |
| 493 | |
| 494 | /* If strerror is just a macro, we want to use the one from libiberty |
| 495 | since it will handle undefined values. */ |
| 496 | #undef strerror |
| 497 | extern char *strerror PARAMS ((int)); |
| 498 | |
| 499 | /* The lexer is simple, too simple for flex. Keywords are only |
| 500 | recognized at the start of lines. Everything else must be an |
| 501 | argument. A comma is treated as whitespace. */ |
| 502 | |
| 503 | /* The states the lexer can be in. */ |
| 504 | |
| 505 | enum lex_state |
| 506 | { |
| 507 | /* At the beginning of a line. */ |
| 508 | BEGINNING_OF_LINE, |
| 509 | /* In the middle of a line. */ |
| 510 | IN_LINE |
| 511 | }; |
| 512 | |
| 513 | /* We need to keep a stack of files to handle file inclusion. */ |
| 514 | |
| 515 | struct input |
| 516 | { |
| 517 | /* The file to read from. */ |
| 518 | FILE *file; |
| 519 | /* The name of the file. */ |
| 520 | char *name; |
| 521 | /* The current line number. */ |
| 522 | int lineno; |
| 523 | /* The current state. */ |
| 524 | enum lex_state state; |
| 525 | /* The next file on the stack. */ |
| 526 | struct input *next; |
| 527 | }; |
| 528 | |
| 529 | /* The current input file. */ |
| 530 | |
| 531 | static struct input current; |
| 532 | |
| 533 | /* The character which introduces comments. */ |
| 534 | #define COMMENT_CHAR '#' |
| 535 | \f |
| 536 | /* Start the lexer going on the main input file. */ |
| 537 | |
| 538 | bfd_boolean |
| 539 | nlmlex_file (const char *name) |
| 540 | { |
| 541 | current.next = NULL; |
| 542 | return nlmlex_file_open (name); |
| 543 | } |
| 544 | |
| 545 | /* Start the lexer going on a subsidiary input file. */ |
| 546 | |
| 547 | static void |
| 548 | nlmlex_file_push (const char *name) |
| 549 | { |
| 550 | struct input *push; |
| 551 | |
| 552 | push = (struct input *) xmalloc (sizeof (struct input)); |
| 553 | *push = current; |
| 554 | if (nlmlex_file_open (name)) |
| 555 | current.next = push; |
| 556 | else |
| 557 | { |
| 558 | current = *push; |
| 559 | free (push); |
| 560 | } |
| 561 | } |
| 562 | |
| 563 | /* Start lexing from a file. */ |
| 564 | |
| 565 | static bfd_boolean |
| 566 | nlmlex_file_open (const char *name) |
| 567 | { |
| 568 | current.file = fopen (name, "r"); |
| 569 | if (current.file == NULL) |
| 570 | { |
| 571 | fprintf (stderr, "%s:%s: %s\n", program_name, name, strerror (errno)); |
| 572 | ++parse_errors; |
| 573 | return FALSE; |
| 574 | } |
| 575 | current.name = xstrdup (name); |
| 576 | current.lineno = 1; |
| 577 | current.state = BEGINNING_OF_LINE; |
| 578 | return TRUE; |
| 579 | } |
| 580 | \f |
| 581 | /* Table used to turn keywords into tokens. */ |
| 582 | |
| 583 | struct keyword_tokens_struct |
| 584 | { |
| 585 | const char *keyword; |
| 586 | int token; |
| 587 | }; |
| 588 | |
| 589 | static struct keyword_tokens_struct keyword_tokens[] = |
| 590 | { |
| 591 | { "CHECK", CHECK }, |
| 592 | { "CODESTART", CODESTART }, |
| 593 | { "COPYRIGHT", COPYRIGHT }, |
| 594 | { "CUSTOM", CUSTOM }, |
| 595 | { "DATE", DATE }, |
| 596 | { "DEBUG", DEBUG }, |
| 597 | { "DESCRIPTION", DESCRIPTION }, |
| 598 | { "EXIT", EXIT }, |
| 599 | { "EXPORT", EXPORT }, |
| 600 | { "FLAG_ON", FLAG_ON }, |
| 601 | { "FLAG_OFF", FLAG_OFF }, |
| 602 | { "FULLMAP", FULLMAP }, |
| 603 | { "HELP", HELP }, |
| 604 | { "IMPORT", IMPORT }, |
| 605 | { "INPUT", INPUT }, |
| 606 | { "MAP", MAP }, |
| 607 | { "MESSAGES", MESSAGES }, |
| 608 | { "MODULE", MODULE }, |
| 609 | { "MULTIPLE", MULTIPLE }, |
| 610 | { "OS_DOMAIN", OS_DOMAIN }, |
| 611 | { "OUTPUT", OUTPUT }, |
| 612 | { "PSEUDOPREEMPTION", PSEUDOPREEMPTION }, |
| 613 | { "REENTRANT", REENTRANT }, |
| 614 | { "SCREENNAME", SCREENNAME }, |
| 615 | { "SHARELIB", SHARELIB }, |
| 616 | { "STACK", STACK }, |
| 617 | { "STACKSIZE", STACK }, |
| 618 | { "START", START }, |
| 619 | { "SYNCHRONIZE", SYNCHRONIZE }, |
| 620 | { "THREADNAME", THREADNAME }, |
| 621 | { "TYPE", TYPE }, |
| 622 | { "VERBOSE", VERBOSE }, |
| 623 | { "VERSION", VERSIONK }, |
| 624 | { "XDCDATA", XDCDATA } |
| 625 | }; |
| 626 | |
| 627 | #define KEYWORD_COUNT (sizeof (keyword_tokens) / sizeof (keyword_tokens[0])) |
| 628 | \f |
| 629 | /* The lexer accumulates strings in these variables. */ |
| 630 | static char *lex_buf; |
| 631 | static int lex_size; |
| 632 | static int lex_pos; |
| 633 | |
| 634 | /* Start accumulating strings into the buffer. */ |
| 635 | #define BUF_INIT() \ |
| 636 | ((void) (lex_buf != NULL ? lex_pos = 0 : nlmlex_buf_init ())) |
| 637 | |
| 638 | static int |
| 639 | nlmlex_buf_init (void) |
| 640 | { |
| 641 | lex_size = 10; |
| 642 | lex_buf = xmalloc (lex_size + 1); |
| 643 | lex_pos = 0; |
| 644 | return 0; |
| 645 | } |
| 646 | |
| 647 | /* Finish a string in the buffer. */ |
| 648 | #define BUF_FINISH() ((void) (lex_buf[lex_pos] = '\0')) |
| 649 | |
| 650 | /* Accumulate a character into the buffer. */ |
| 651 | #define BUF_ADD(c) \ |
| 652 | ((void) (lex_pos < lex_size \ |
| 653 | ? lex_buf[lex_pos++] = (c) \ |
| 654 | : nlmlex_buf_add (c))) |
| 655 | |
| 656 | static char |
| 657 | nlmlex_buf_add (int c) |
| 658 | { |
| 659 | if (lex_pos >= lex_size) |
| 660 | { |
| 661 | lex_size *= 2; |
| 662 | lex_buf = xrealloc (lex_buf, lex_size + 1); |
| 663 | } |
| 664 | |
| 665 | return lex_buf[lex_pos++] = c; |
| 666 | } |
| 667 | \f |
| 668 | /* The lexer proper. This is called by the bison generated parsing |
| 669 | code. */ |
| 670 | |
| 671 | static int |
| 672 | yylex (void) |
| 673 | { |
| 674 | int c; |
| 675 | |
| 676 | tail_recurse: |
| 677 | |
| 678 | c = getc (current.file); |
| 679 | |
| 680 | /* Commas are treated as whitespace characters. */ |
| 681 | while (ISSPACE (c) || c == ',') |
| 682 | { |
| 683 | current.state = IN_LINE; |
| 684 | if (c == '\n') |
| 685 | { |
| 686 | ++current.lineno; |
| 687 | current.state = BEGINNING_OF_LINE; |
| 688 | } |
| 689 | c = getc (current.file); |
| 690 | } |
| 691 | |
| 692 | /* At the end of the file we either pop to the previous file or |
| 693 | finish up. */ |
| 694 | if (c == EOF) |
| 695 | { |
| 696 | fclose (current.file); |
| 697 | free (current.name); |
| 698 | if (current.next == NULL) |
| 699 | return 0; |
| 700 | else |
| 701 | { |
| 702 | struct input *next; |
| 703 | |
| 704 | next = current.next; |
| 705 | current = *next; |
| 706 | free (next); |
| 707 | goto tail_recurse; |
| 708 | } |
| 709 | } |
| 710 | |
| 711 | /* A comment character always means to drop everything until the |
| 712 | next newline. */ |
| 713 | if (c == COMMENT_CHAR) |
| 714 | { |
| 715 | do |
| 716 | { |
| 717 | c = getc (current.file); |
| 718 | } |
| 719 | while (c != '\n'); |
| 720 | ++current.lineno; |
| 721 | current.state = BEGINNING_OF_LINE; |
| 722 | goto tail_recurse; |
| 723 | } |
| 724 | |
| 725 | /* An '@' introduces an include file. */ |
| 726 | if (c == '@') |
| 727 | { |
| 728 | do |
| 729 | { |
| 730 | c = getc (current.file); |
| 731 | if (c == '\n') |
| 732 | ++current.lineno; |
| 733 | } |
| 734 | while (ISSPACE (c)); |
| 735 | BUF_INIT (); |
| 736 | while (! ISSPACE (c) && c != EOF) |
| 737 | { |
| 738 | BUF_ADD (c); |
| 739 | c = getc (current.file); |
| 740 | } |
| 741 | BUF_FINISH (); |
| 742 | |
| 743 | ungetc (c, current.file); |
| 744 | |
| 745 | nlmlex_file_push (lex_buf); |
| 746 | goto tail_recurse; |
| 747 | } |
| 748 | |
| 749 | /* A non-space character at the start of a line must be the start of |
| 750 | a keyword. */ |
| 751 | if (current.state == BEGINNING_OF_LINE) |
| 752 | { |
| 753 | BUF_INIT (); |
| 754 | while (ISALNUM (c) || c == '_') |
| 755 | { |
| 756 | BUF_ADD (TOUPPER (c)); |
| 757 | c = getc (current.file); |
| 758 | } |
| 759 | BUF_FINISH (); |
| 760 | |
| 761 | if (c != EOF && ! ISSPACE (c) && c != ',') |
| 762 | { |
| 763 | nlmheader_identify (); |
| 764 | fprintf (stderr, _("%s:%d: illegal character in keyword: %c\n"), |
| 765 | current.name, current.lineno, c); |
| 766 | } |
| 767 | else |
| 768 | { |
| 769 | unsigned int i; |
| 770 | |
| 771 | for (i = 0; i < KEYWORD_COUNT; i++) |
| 772 | { |
| 773 | if (lex_buf[0] == keyword_tokens[i].keyword[0] |
| 774 | && strcmp (lex_buf, keyword_tokens[i].keyword) == 0) |
| 775 | { |
| 776 | /* Pushing back the final whitespace avoids worrying |
| 777 | about \n here. */ |
| 778 | ungetc (c, current.file); |
| 779 | current.state = IN_LINE; |
| 780 | return keyword_tokens[i].token; |
| 781 | } |
| 782 | } |
| 783 | |
| 784 | nlmheader_identify (); |
| 785 | fprintf (stderr, _("%s:%d: unrecognized keyword: %s\n"), |
| 786 | current.name, current.lineno, lex_buf); |
| 787 | } |
| 788 | |
| 789 | ++parse_errors; |
| 790 | /* Treat the rest of this line as a comment. */ |
| 791 | ungetc (COMMENT_CHAR, current.file); |
| 792 | goto tail_recurse; |
| 793 | } |
| 794 | |
| 795 | /* Parentheses just represent themselves. */ |
| 796 | if (c == '(' || c == ')') |
| 797 | return c; |
| 798 | |
| 799 | /* Handle quoted strings. */ |
| 800 | if (c == '"' || c == '\'') |
| 801 | { |
| 802 | int quote; |
| 803 | int start_lineno; |
| 804 | |
| 805 | quote = c; |
| 806 | start_lineno = current.lineno; |
| 807 | |
| 808 | c = getc (current.file); |
| 809 | BUF_INIT (); |
| 810 | while (c != quote && c != EOF) |
| 811 | { |
| 812 | BUF_ADD (c); |
| 813 | if (c == '\n') |
| 814 | ++current.lineno; |
| 815 | c = getc (current.file); |
| 816 | } |
| 817 | BUF_FINISH (); |
| 818 | |
| 819 | if (c == EOF) |
| 820 | { |
| 821 | nlmheader_identify (); |
| 822 | fprintf (stderr, _("%s:%d: end of file in quoted string\n"), |
| 823 | current.name, start_lineno); |
| 824 | ++parse_errors; |
| 825 | } |
| 826 | |
| 827 | /* FIXME: Possible memory leak. */ |
| 828 | yylval.string = xstrdup (lex_buf); |
| 829 | return QUOTED_STRING; |
| 830 | } |
| 831 | |
| 832 | /* Gather a generic argument. */ |
| 833 | BUF_INIT (); |
| 834 | while (! ISSPACE (c) |
| 835 | && c != ',' |
| 836 | && c != COMMENT_CHAR |
| 837 | && c != '(' |
| 838 | && c != ')') |
| 839 | { |
| 840 | BUF_ADD (c); |
| 841 | c = getc (current.file); |
| 842 | } |
| 843 | BUF_FINISH (); |
| 844 | |
| 845 | ungetc (c, current.file); |
| 846 | |
| 847 | /* FIXME: Possible memory leak. */ |
| 848 | yylval.string = xstrdup (lex_buf); |
| 849 | return STRING; |
| 850 | } |
| 851 | \f |
| 852 | /* Get a number from a string. */ |
| 853 | |
| 854 | static long |
| 855 | nlmlex_get_number (const char *s) |
| 856 | { |
| 857 | long ret; |
| 858 | char *send; |
| 859 | |
| 860 | ret = strtol (s, &send, 10); |
| 861 | if (*send != '\0') |
| 862 | nlmheader_warn (_("bad number"), -1); |
| 863 | return ret; |
| 864 | } |
| 865 | |
| 866 | /* Prefix the nlmconv warnings with a note as to where they come from. |
| 867 | We don't use program_name on every warning, because then some |
| 868 | versions of the emacs next-error function can't recognize the line |
| 869 | number. */ |
| 870 | |
| 871 | static void |
| 872 | nlmheader_identify (void) |
| 873 | { |
| 874 | static int done; |
| 875 | |
| 876 | if (! done) |
| 877 | { |
| 878 | fprintf (stderr, _("%s: problems in NLM command language input:\n"), |
| 879 | program_name); |
| 880 | done = 1; |
| 881 | } |
| 882 | } |
| 883 | |
| 884 | /* Issue a warning. */ |
| 885 | |
| 886 | static void |
| 887 | nlmheader_warn (const char *s, int imax) |
| 888 | { |
| 889 | nlmheader_identify (); |
| 890 | fprintf (stderr, "%s:%d: %s", current.name, current.lineno, s); |
| 891 | if (imax != -1) |
| 892 | fprintf (stderr, " (max %d)", imax); |
| 893 | fprintf (stderr, "\n"); |
| 894 | } |
| 895 | |
| 896 | /* Report an error. */ |
| 897 | |
| 898 | static void |
| 899 | nlmheader_error (const char *s) |
| 900 | { |
| 901 | nlmheader_warn (s, -1); |
| 902 | ++parse_errors; |
| 903 | } |
| 904 | |
| 905 | /* Add a string to a string list. */ |
| 906 | |
| 907 | static struct string_list * |
| 908 | string_list_cons (char *s, struct string_list *l) |
| 909 | { |
| 910 | struct string_list *ret; |
| 911 | |
| 912 | ret = (struct string_list *) xmalloc (sizeof (struct string_list)); |
| 913 | ret->next = l; |
| 914 | ret->string = s; |
| 915 | return ret; |
| 916 | } |
| 917 | |
| 918 | /* Append a string list to another string list. */ |
| 919 | |
| 920 | static struct string_list * |
| 921 | string_list_append (struct string_list *l1, struct string_list *l2) |
| 922 | { |
| 923 | register struct string_list **pp; |
| 924 | |
| 925 | for (pp = &l1; *pp != NULL; pp = &(*pp)->next) |
| 926 | ; |
| 927 | *pp = l2; |
| 928 | return l1; |
| 929 | } |
| 930 | |
| 931 | /* Append a string to a string list. */ |
| 932 | |
| 933 | static struct string_list * |
| 934 | string_list_append1 (struct string_list *l, char *s) |
| 935 | { |
| 936 | struct string_list *n; |
| 937 | register struct string_list **pp; |
| 938 | |
| 939 | n = (struct string_list *) xmalloc (sizeof (struct string_list)); |
| 940 | n->next = NULL; |
| 941 | n->string = s; |
| 942 | for (pp = &l; *pp != NULL; pp = &(*pp)->next) |
| 943 | ; |
| 944 | *pp = n; |
| 945 | return l; |
| 946 | } |
| 947 | |
| 948 | /* Duplicate a string in memory. */ |
| 949 | |
| 950 | static char * |
| 951 | xstrdup (const char *s) |
| 952 | { |
| 953 | unsigned long len; |
| 954 | char *ret; |
| 955 | |
| 956 | len = strlen (s); |
| 957 | ret = xmalloc (len + 1); |
| 958 | strcpy (ret, s); |
| 959 | return ret; |
| 960 | } |