1 2 /* flexdef - definitions file for flex */ 3 4 /* Copyright (c) 1990 The Regents of the University of California. */ 5 /* All rights reserved. */ 6 7 /* This code is derived from software contributed to Berkeley by */ 8 /* Vern Paxson. */ 9 10 /* The United States Government has rights in this work pursuant */ 11 /* to contract no. DE-AC03-76SF00098 between the United States */ 12 /* Department of Energy and the University of California. */ 13 14 /* This file is part of flex. */ 15 16 /* Redistribution and use in source and binary forms, with or without */ 17 /* modification, are permitted provided that the following conditions */ 18 /* are met: */ 19 20 /* 1. Redistributions of source code must retain the above copyright */ 21 /* notice, this list of conditions and the following disclaimer. */ 22 /* 2. Redistributions in binary form must reproduce the above copyright */ 23 /* notice, this list of conditions and the following disclaimer in the */ 24 /* documentation and/or other materials provided with the distribution. */ 25 26 /* Neither the name of the University nor the names of its contributors */ 27 /* may be used to endorse or promote products derived from this software */ 28 /* without specific prior written permission. */ 29 30 /* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR */ 31 /* IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */ 32 /* WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */ 33 /* PURPOSE. */ 34 35 #ifndef FLEXDEF_H 36 #define FLEXDEF_H 1 37 38 #ifdef HAVE_CONFIG_H 39 #include <config.h> 40 #endif 41 42 #include <stdio.h> 43 #include <stdlib.h> 44 #include <stdarg.h> 45 #include <setjmp.h> 46 #include <ctype.h> 47 #include <libgen.h> /* for XPG version of basename(3) */ 48 #include <string.h> 49 50 #ifdef HAVE_ASSERT_H 51 #include <assert.h> 52 #else 53 #define assert(Pred) 54 #endif 55 56 #ifdef HAVE_LIMITS_H 57 #include <limits.h> 58 #endif 59 /* Required: dup() and dup2() in <unistd.h> */ 60 #include <unistd.h> 61 #ifdef HAVE_NETINET_IN_H 62 #include <netinet/in.h> 63 #endif 64 #ifdef HAVE_SYS_PARAMS_H 65 #include <sys/params.h> 66 #endif 67 /* Required: stat() in <sys/stat.h> */ 68 #include <sys/stat.h> 69 /* Required: wait() in <sys/wait.h> */ 70 #include <sys/wait.h> 71 #include <stdbool.h> 72 #include <stdarg.h> 73 /* Required: regcomp(), regexec() and regerror() in <regex.h> */ 74 #include <regex.h> 75 /* Required: strcasecmp() in <strings.h> */ 76 #include <strings.h> 77 #include "flexint.h" 78 79 /* We use gettext. So, when we write strings which should be translated, we mark them with _() */ 80 #ifdef ENABLE_NLS 81 #ifdef HAVE_LOCALE_H 82 #include <locale.h> 83 #endif /* HAVE_LOCALE_H */ 84 #include "gettext.h" 85 #define _(String) gettext (String) 86 #else 87 #define _(STRING) STRING 88 #endif /* ENABLE_NLS */ 89 90 /* Always be prepared to generate an 8-bit scanner. */ 91 #define CSIZE 256 92 93 /* Size of input alphabet - should be size of ASCII set. */ 94 #ifndef DEFAULT_CSIZE 95 #define DEFAULT_CSIZE 128 96 #endif 97 98 /* Maximum line length we'll have to deal with. */ 99 #define MAXLINE 2048 100 101 #ifndef MIN 102 #define MIN(x,y) ((x) < (y) ? (x) : (y)) 103 #endif 104 #ifndef MAX 105 #define MAX(x,y) ((x) > (y) ? (x) : (y)) 106 #endif 107 #ifndef ABS 108 #define ABS(x) ((x) < 0 ? -(x) : (x)) 109 #endif 110 111 /* Whether an integer is a power of two */ 112 #define is_power_of_2(n) ((n) > 0 && ((n) & ((n) - 1)) == 0) 113 114 #define unspecified -1 115 116 /* Special chk[] values marking the slots taking by end-of-buffer and action 117 * numbers. 118 */ 119 #define EOB_POSITION -1 120 #define ACTION_POSITION -2 121 122 /* Number of data items per line for -f output. */ 123 #define NUMDATAITEMS 10 124 125 /* Number of lines of data in -f output before inserting a blank line for 126 * readability. 127 */ 128 #define NUMDATALINES 10 129 130 /* Number of characters to print a line number, i.e., 1 + log10(INT_MAX) */ 131 #define NUMCHARLINES 10 132 133 /* transition_struct_out() definitions. */ 134 #define TRANS_STRUCT_PRINT_LENGTH 14 135 136 /* Returns true if an nfa state has an epsilon out-transition slot 137 * that can be used. This definition is currently not used. 138 */ 139 #define FREE_EPSILON(state) \ 140 (transchar[state] == SYM_EPSILON && \ 141 trans2[state] == NO_TRANSITION && \ 142 finalst[state] != state) 143 144 /* Returns true if an nfa state has an epsilon out-transition character 145 * and both slots are free 146 */ 147 #define SUPER_FREE_EPSILON(state) \ 148 (transchar[state] == SYM_EPSILON && \ 149 trans1[state] == NO_TRANSITION) \ 150 151 /* Maximum number of NFA states that can comprise a DFA state. It's real 152 * big because if there's a lot of rules, the initial state will have a 153 * huge epsilon closure. 154 */ 155 #define INITIAL_MAX_DFA_SIZE 750 156 #define MAX_DFA_SIZE_INCREMENT 750 157 158 159 /* A note on the following masks. They are used to mark accepting numbers 160 * as being special. As such, they implicitly limit the number of accepting 161 * numbers (i.e., rules) because if there are too many rules the rule numbers 162 * will overload the mask bits. Fortunately, this limit is \large/ (0x2000 == 163 * 8192) so unlikely to actually cause any problems. A check is made in 164 * new_rule() to ensure that this limit is not reached. 165 */ 166 167 /* Mask to mark a trailing context accepting number. */ 168 #define YY_TRAILING_MASK 0x2000 169 170 /* Mask to mark the accepting number of the "head" of a trailing context 171 * rule. 172 */ 173 #define YY_TRAILING_HEAD_MASK 0x4000 174 175 /* Maximum number of rules, as outlined in the above note. */ 176 #define MAX_RULE (YY_TRAILING_MASK - 1) 177 178 179 /* NIL must be 0. If not, its special meaning when making equivalence classes 180 * (it marks the representative of a given e.c.) will be unidentifiable. 181 */ 182 #define NIL 0 183 184 #define JAM -1 /* to mark a missing DFA transition */ 185 #define NO_TRANSITION NIL 186 #define UNIQUE -1 /* marks a symbol as an e.c. representative */ 187 #define INFINITE_REPEAT -1 /* for x{5,} constructions */ 188 189 #define INITIAL_MAX_CCLS 100 /* max number of unique character classes */ 190 #define MAX_CCLS_INCREMENT 100 191 192 /* Size of table holding members of character classes. */ 193 #define INITIAL_MAX_CCL_TBL_SIZE 500 194 #define MAX_CCL_TBL_SIZE_INCREMENT 250 195 196 #define INITIAL_MAX_RULES 100 /* default maximum number of rules */ 197 #define MAX_RULES_INCREMENT 100 198 199 #define INITIAL_MNS 2000 /* default maximum number of nfa states */ 200 #define MNS_INCREMENT 1000 /* amount to bump above by if it's not enough */ 201 202 #define INITIAL_MAX_DFAS 1000 /* default maximum number of dfa states */ 203 #define MAX_DFAS_INCREMENT 1000 204 205 #define JAMSTATE -32766 /* marks a reference to the state that always jams */ 206 207 /* Maximum number of NFA states. */ 208 #define MAXIMUM_MNS 31999 209 #define MAXIMUM_MNS_LONG 1999999999 210 211 /* Enough so that if it's subtracted from an NFA state number, the result 212 * is guaranteed to be negative. 213 */ 214 #define MARKER_DIFFERENCE (maximum_mns+2) 215 216 /* Maximum number of nxt/chk pairs for non-templates. */ 217 #define INITIAL_MAX_XPAIRS 2000 218 #define MAX_XPAIRS_INCREMENT 2000 219 220 /* Maximum number of nxt/chk pairs needed for templates. */ 221 #define INITIAL_MAX_TEMPLATE_XPAIRS 2500 222 #define MAX_TEMPLATE_XPAIRS_INCREMENT 2500 223 224 #define SYM_EPSILON (CSIZE + 1) /* to mark transitions on the symbol epsilon */ 225 226 #define INITIAL_MAX_SCS 40 /* maximum number of start conditions */ 227 #define MAX_SCS_INCREMENT 40 /* amount to bump by if it's not enough */ 228 229 #define ONE_STACK_SIZE 500 /* stack of states with only one out-transition */ 230 #define SAME_TRANS -1 /* transition is the same as "default" entry for state */ 231 232 /* The following percentages are used to tune table compression: 233 234 * The percentage the number of out-transitions a state must be of the 235 * number of equivalence classes in order to be considered for table 236 * compaction by using protos. 237 */ 238 #define PROTO_SIZE_PERCENTAGE 15 239 240 /* The percentage the number of homogeneous out-transitions of a state 241 * must be of the number of total out-transitions of the state in order 242 * that the state's transition table is first compared with a potential 243 * template of the most common out-transition instead of with the first 244 * proto in the proto queue. 245 */ 246 #define CHECK_COM_PERCENTAGE 50 247 248 /* The percentage the number of differences between a state's transition 249 * table and the proto it was first compared with must be of the total 250 * number of out-transitions of the state in order to keep the first 251 * proto as a good match and not search any further. 252 */ 253 #define FIRST_MATCH_DIFF_PERCENTAGE 10 254 255 /* The percentage the number of differences between a state's transition 256 * table and the most similar proto must be of the state's total number 257 * of out-transitions to use the proto as an acceptable close match. 258 */ 259 #define ACCEPTABLE_DIFF_PERCENTAGE 50 260 261 /* The percentage the number of homogeneous out-transitions of a state 262 * must be of the number of total out-transitions of the state in order 263 * to consider making a template from the state. 264 */ 265 #define TEMPLATE_SAME_PERCENTAGE 60 266 267 /* The percentage the number of differences between a state's transition 268 * table and the most similar proto must be of the state's total number 269 * of out-transitions to create a new proto from the state. 270 */ 271 #define NEW_PROTO_DIFF_PERCENTAGE 20 272 273 /* The percentage the total number of out-transitions of a state must be 274 * of the number of equivalence classes in order to consider trying to 275 * fit the transition table into "holes" inside the nxt/chk table. 276 */ 277 #define INTERIOR_FIT_PERCENTAGE 15 278 279 /* Size of region set aside to cache the complete transition table of 280 * protos on the proto queue to enable quick comparisons. 281 */ 282 #define PROT_SAVE_SIZE 2000 283 284 #define MSP 50 /* maximum number of saved protos (protos on the proto queue) */ 285 286 /* Maximum number of out-transitions a state can have that we'll rummage 287 * around through the interior of the internal fast table looking for a 288 * spot for it. 289 */ 290 #define MAX_XTIONS_FULL_INTERIOR_FIT 4 291 292 /* Maximum number of rules which will be reported as being associated 293 * with a DFA state. 294 */ 295 #define MAX_ASSOC_RULES 100 296 297 /* Number that, if used to subscript an array, has a good chance of producing 298 * an error; should be small enough to fit into a short. 299 */ 300 #define BAD_SUBSCRIPT -32767 301 302 /* Absolute value of largest number that can be stored in a short, with a 303 * bit of slop thrown in for general paranoia. 304 */ 305 #define MAX_SHORT 32700 306 307 308 /* Declarations for global variables. */ 309 310 311 /* Variables for flags: 312 * printstats - if true (-v), dump statistics 313 * syntaxerror - true if a syntax error has been found 314 * eofseen - true if we've seen an eof in the input file 315 * ddebug - if true (-d), make a "debug" scanner 316 * trace - if true (-T), trace processing 317 * nowarn - if true (-w), do not generate warnings 318 * spprdflt - if true (-s), suppress the default rule 319 * interactive - if true (-I), generate an interactive scanner 320 * lex_compat - if true (-l), maximize compatibility with AT&T lex 321 * posix_compat - if true (-X), maximize compatibility with POSIX lex 322 * do_yylineno - if true, generate code to maintain yylineno 323 * useecs - if true (-Ce flag), use equivalence classes 324 * fulltbl - if true (-Cf flag), don't compress the DFA state table 325 * usemecs - if true (-Cm flag), use meta-equivalence classes 326 * fullspd - if true (-F flag), use Jacobson method of table representation 327 * gen_line_dirs - if true (i.e., no -L flag), generate #line directives 328 * performance_report - if > 0 (i.e., -p flag), generate a report relating 329 * to scanner performance; if > 1 (-p -p), report on minor performance 330 * problems, too 331 * backing_up_report - if true (i.e., -b flag), generate "lex.backup" file 332 * listing backing-up states 333 * C_plus_plus - if true (i.e., -+ flag), generate a C++ scanner class; 334 * otherwise, a standard C scanner 335 * reentrant - if true (-R), generate a reentrant C scanner. 336 * bison_bridge_lval - if true (--bison-bridge), bison pure calling convention. 337 * bison_bridge_lloc - if true (--bison-locations), bison yylloc. 338 * long_align - if true (-Ca flag), favor long-word alignment. 339 * use_read - if true (-f, -F, or -Cr) then use read() for scanner input; 340 * otherwise, use fread(). 341 * yytext_is_array - if true (i.e., %array directive), then declare 342 * yytext as a array instead of a character pointer. Nice and inefficient. 343 * do_yywrap - do yywrap() processing on EOF. If false, EOF treated as 344 * "no more files". 345 * csize - size of character set for the scanner we're generating; 346 * 128 for 7-bit chars and 256 for 8-bit 347 * yymore_used - if true, yymore() is used in input rules 348 * reject - if true, generate back-up tables for REJECT macro 349 * real_reject - if true, scanner really uses REJECT (as opposed to just 350 * having "reject" set for variable trailing context) 351 * continued_action - true if this rule's action is to "fall through" to 352 * the next rule's action (i.e., the '|' action) 353 * in_rule - true if we're inside an individual rule, false if not. 354 * yymore_really_used - whether to treat yymore() as really used, regardless 355 * of what we think based on references to it in the user's actions. 356 * reject_really_used - same for REJECT 357 * trace_hex - use hexadecimal numbers in trace/debug outputs instead of octals 358 */ 359 360 extern int printstats, syntaxerror, eofseen, ddebug, trace, nowarn, 361 spprdflt; 362 extern int interactive, lex_compat, posix_compat, do_yylineno; 363 extern int useecs, fulltbl, usemecs, fullspd; 364 extern int gen_line_dirs, performance_report, backing_up_report; 365 extern int reentrant, bison_bridge_lval, bison_bridge_lloc; 366 extern int C_plus_plus, long_align, use_read, yytext_is_array, do_yywrap; 367 extern int csize; 368 extern int yymore_used, reject, real_reject, continued_action, in_rule; 369 370 extern int yymore_really_used, reject_really_used; 371 extern int trace_hex; 372 373 /* Variables used in the flex input routines: 374 * datapos - characters on current output line 375 * dataline - number of contiguous lines of data in current data 376 * statement. Used to generate readable -f output 377 * linenum - current input line number 378 * skelfile - the skeleton file 379 * skel - compiled-in skeleton array 380 * skel_ind - index into "skel" array, if skelfile is nil 381 * yyin - input file 382 * backing_up_file - file to summarize backing-up states to 383 * infilename - name of input file 384 * outfilename - name of output file 385 * headerfilename - name of the .h file to generate 386 * did_outfilename - whether outfilename was explicitly set 387 * prefix - the prefix used for externally visible names ("yy" by default) 388 * yyclass - yyFlexLexer subclass to use for YY_DECL 389 * do_stdinit - whether to initialize yyin/yyout to stdin/stdout 390 * use_stdout - the -t flag 391 * input_files - array holding names of input files 392 * num_input_files - size of input_files array 393 * program_name - name with which program was invoked 394 * 395 * action_array - array to hold the rule actions 396 * action_size - size of action_array 397 * defs1_offset - index where the user's section 1 definitions start 398 * in action_array 399 * prolog_offset - index where the prolog starts in action_array 400 * action_offset - index where the non-prolog starts in action_array 401 * action_index - index where the next action should go, with respect 402 * to "action_array" 403 */ 404 405 extern int datapos, dataline, linenum; 406 extern FILE *skelfile, *backing_up_file; 407 extern const char *skel[]; 408 extern int skel_ind; 409 extern char *infilename, *outfilename, *headerfilename; 410 extern int did_outfilename; 411 extern char *prefix, *yyclass, *extra_type; 412 extern int do_stdinit, use_stdout; 413 extern char **input_files; 414 extern int num_input_files; 415 extern char *program_name; 416 417 extern char *action_array; 418 extern int action_size; 419 extern int defs1_offset, prolog_offset, action_offset, action_index; 420 421 422 /* Variables for stack of states having only one out-transition: 423 * onestate - state number 424 * onesym - transition symbol 425 * onenext - target state 426 * onedef - default base entry 427 * onesp - stack pointer 428 */ 429 430 extern int onestate[ONE_STACK_SIZE], onesym[ONE_STACK_SIZE]; 431 extern int onenext[ONE_STACK_SIZE], onedef[ONE_STACK_SIZE], onesp; 432 433 434 /* Variables for nfa machine data: 435 * maximum_mns - maximal number of NFA states supported by tables 436 * current_mns - current maximum on number of NFA states 437 * num_rules - number of the last accepting state; also is number of 438 * rules created so far 439 * num_eof_rules - number of <<EOF>> rules 440 * default_rule - number of the default rule 441 * current_max_rules - current maximum number of rules 442 * lastnfa - last nfa state number created 443 * firstst - physically the first state of a fragment 444 * lastst - last physical state of fragment 445 * finalst - last logical state of fragment 446 * transchar - transition character 447 * trans1 - transition state 448 * trans2 - 2nd transition state for epsilons 449 * accptnum - accepting number 450 * assoc_rule - rule associated with this NFA state (or 0 if none) 451 * state_type - a STATE_xxx type identifying whether the state is part 452 * of a normal rule, the leading state in a trailing context 453 * rule (i.e., the state which marks the transition from 454 * recognizing the text-to-be-matched to the beginning of 455 * the trailing context), or a subsequent state in a trailing 456 * context rule 457 * rule_type - a RULE_xxx type identifying whether this a ho-hum 458 * normal rule or one which has variable head & trailing 459 * context 460 * rule_linenum - line number associated with rule 461 * rule_useful - true if we've determined that the rule can be matched 462 * rule_has_nl - true if rule could possibly match a newline 463 * ccl_has_nl - true if current ccl could match a newline 464 * nlch - default eol char 465 */ 466 467 extern int maximum_mns, current_mns, current_max_rules; 468 extern int num_rules, num_eof_rules, default_rule, lastnfa; 469 extern int *firstst, *lastst, *finalst, *transchar, *trans1, *trans2; 470 extern int *accptnum, *assoc_rule, *state_type; 471 extern int *rule_type, *rule_linenum, *rule_useful; 472 extern bool *rule_has_nl, *ccl_has_nl; 473 extern int nlch; 474 475 /* Different types of states; values are useful as masks, as well, for 476 * routines like check_trailing_context(). 477 */ 478 #define STATE_NORMAL 0x1 479 #define STATE_TRAILING_CONTEXT 0x2 480 481 /* Global holding current type of state we're making. */ 482 483 extern int current_state_type; 484 485 /* Different types of rules. */ 486 #define RULE_NORMAL 0 487 #define RULE_VARIABLE 1 488 489 /* True if the input rules include a rule with both variable-length head 490 * and trailing context, false otherwise. 491 */ 492 extern int variable_trailing_context_rules; 493 494 495 /* Variables for protos: 496 * numtemps - number of templates created 497 * numprots - number of protos created 498 * protprev - backlink to a more-recently used proto 499 * protnext - forward link to a less-recently used proto 500 * prottbl - base/def table entry for proto 501 * protcomst - common state of proto 502 * firstprot - number of the most recently used proto 503 * lastprot - number of the least recently used proto 504 * protsave contains the entire state array for protos 505 */ 506 507 extern int numtemps, numprots, protprev[MSP], protnext[MSP], prottbl[MSP]; 508 extern int protcomst[MSP], firstprot, lastprot, protsave[PROT_SAVE_SIZE]; 509 510 511 /* Variables for managing equivalence classes: 512 * numecs - number of equivalence classes 513 * nextecm - forward link of Equivalence Class members 514 * ecgroup - class number or backward link of EC members 515 * nummecs - number of meta-equivalence classes (used to compress 516 * templates) 517 * tecfwd - forward link of meta-equivalence classes members 518 * tecbck - backward link of MEC's 519 */ 520 521 /* Reserve enough room in the equivalence class arrays so that we 522 * can use the CSIZE'th element to hold equivalence class information 523 * for the NUL character. Later we'll move this information into 524 * the 0th element. 525 */ 526 extern int numecs, nextecm[CSIZE + 1], ecgroup[CSIZE + 1], nummecs; 527 528 /* Meta-equivalence classes are indexed starting at 1, so it's possible 529 * that they will require positions from 1 .. CSIZE, i.e., CSIZE + 1 530 * slots total (since the arrays are 0-based). nextecm[] and ecgroup[] 531 * don't require the extra position since they're indexed from 1 .. CSIZE - 1. 532 */ 533 extern int tecfwd[CSIZE + 1], tecbck[CSIZE + 1]; 534 535 536 /* Variables for start conditions: 537 * lastsc - last start condition created 538 * current_max_scs - current limit on number of start conditions 539 * scset - set of rules active in start condition 540 * scbol - set of rules active only at the beginning of line in a s.c. 541 * scxclu - true if start condition is exclusive 542 * sceof - true if start condition has EOF rule 543 * scname - start condition name 544 */ 545 546 extern int lastsc, *scset, *scbol, *scxclu, *sceof; 547 extern int current_max_scs; 548 extern char **scname; 549 550 551 /* Variables for dfa machine data: 552 * current_max_dfa_size - current maximum number of NFA states in DFA 553 * current_max_xpairs - current maximum number of non-template xtion pairs 554 * current_max_template_xpairs - current maximum number of template pairs 555 * current_max_dfas - current maximum number DFA states 556 * lastdfa - last dfa state number created 557 * nxt - state to enter upon reading character 558 * chk - check value to see if "nxt" applies 559 * tnxt - internal nxt table for templates 560 * base - offset into "nxt" for given state 561 * def - where to go if "chk" disallows "nxt" entry 562 * nultrans - NUL transition for each state 563 * NUL_ec - equivalence class of the NUL character 564 * tblend - last "nxt/chk" table entry being used 565 * firstfree - first empty entry in "nxt/chk" table 566 * dss - nfa state set for each dfa 567 * dfasiz - size of nfa state set for each dfa 568 * dfaacc - accepting set for each dfa state (if using REJECT), or accepting 569 * number, if not 570 * accsiz - size of accepting set for each dfa state 571 * dhash - dfa state hash value 572 * numas - number of DFA accepting states created; note that this 573 * is not necessarily the same value as num_rules, which is the analogous 574 * value for the NFA 575 * numsnpairs - number of state/nextstate transition pairs 576 * jambase - position in base/def where the default jam table starts 577 * jamstate - state number corresponding to "jam" state 578 * end_of_buffer_state - end-of-buffer dfa state number 579 */ 580 581 extern int current_max_dfa_size, current_max_xpairs; 582 extern int current_max_template_xpairs, current_max_dfas; 583 extern int lastdfa, *nxt, *chk, *tnxt; 584 extern int *base, *def, *nultrans, NUL_ec, tblend, firstfree, **dss, 585 *dfasiz; 586 extern union dfaacc_union { 587 int *dfaacc_set; 588 int dfaacc_state; 589 } *dfaacc; 590 extern int *accsiz, *dhash, numas; 591 extern int numsnpairs, jambase, jamstate; 592 extern int end_of_buffer_state; 593 594 /* Variables for ccl information: 595 * lastccl - ccl index of the last created ccl 596 * current_maxccls - current limit on the maximum number of unique ccl's 597 * cclmap - maps a ccl index to its set pointer 598 * ccllen - gives the length of a ccl 599 * cclng - true for a given ccl if the ccl is negated 600 * cclreuse - counts how many times a ccl is re-used 601 * current_max_ccl_tbl_size - current limit on number of characters needed 602 * to represent the unique ccl's 603 * ccltbl - holds the characters in each ccl - indexed by cclmap 604 */ 605 606 extern int lastccl, *cclmap, *ccllen, *cclng, cclreuse; 607 extern int current_maxccls, current_max_ccl_tbl_size; 608 extern unsigned char *ccltbl; 609 610 611 /* Variables for miscellaneous information: 612 * nmstr - last NAME scanned by the scanner 613 * sectnum - section number currently being parsed 614 * nummt - number of empty nxt/chk table entries 615 * hshcol - number of hash collisions detected by snstods 616 * dfaeql - number of times a newly created dfa was equal to an old one 617 * numeps - number of epsilon NFA states created 618 * eps2 - number of epsilon states which have 2 out-transitions 619 * num_reallocs - number of times it was necessary to realloc() a group 620 * of arrays 621 * tmpuses - number of DFA states that chain to templates 622 * totnst - total number of NFA states used to make DFA states 623 * peakpairs - peak number of transition pairs we had to store internally 624 * numuniq - number of unique transitions 625 * numdup - number of duplicate transitions 626 * hshsave - number of hash collisions saved by checking number of states 627 * num_backing_up - number of DFA states requiring backing up 628 * bol_needed - whether scanner needs beginning-of-line recognition 629 */ 630 631 extern char nmstr[MAXLINE]; 632 extern int sectnum, nummt, hshcol, dfaeql, numeps, eps2, num_reallocs; 633 extern int tmpuses, totnst, peakpairs, numuniq, numdup, hshsave; 634 extern int num_backing_up, bol_needed; 635 636 #ifndef HAVE_REALLOCARRAY 637 void *reallocarray(void *, size_t, size_t); 638 #endif 639 640 void *allocate_array(int, size_t); 641 void *reallocate_array(void *, int, size_t); 642 643 #define allocate_integer_array(size) \ 644 allocate_array(size, sizeof(int)) 645 646 #define reallocate_integer_array(array,size) \ 647 reallocate_array((void *) array, size, sizeof(int)) 648 649 #define allocate_bool_array(size) \ 650 allocate_array(size, sizeof(bool)) 651 652 #define reallocate_bool_array(array,size) \ 653 reallocate_array((void *) array, size, sizeof(bool)) 654 655 #define allocate_int_ptr_array(size) \ 656 allocate_array(size, sizeof(int *)) 657 658 #define allocate_char_ptr_array(size) \ 659 allocate_array(size, sizeof(char *)) 660 661 #define allocate_dfaacc_union(size) \ 662 allocate_array(size, sizeof(union dfaacc_union)) 663 664 #define reallocate_int_ptr_array(array,size) \ 665 reallocate_array((void *) array, size, sizeof(int *)) 666 667 #define reallocate_char_ptr_array(array,size) \ 668 reallocate_array((void *) array, size, sizeof(char *)) 669 670 #define reallocate_dfaacc_union(array, size) \ 671 reallocate_array((void *) array, size, sizeof(union dfaacc_union)) 672 673 #define allocate_character_array(size) \ 674 allocate_array( size, sizeof(char)) 675 676 #define reallocate_character_array(array,size) \ 677 reallocate_array((void *) array, size, sizeof(char)) 678 679 #define allocate_Character_array(size) \ 680 allocate_array(size, sizeof(unsigned char)) 681 682 #define reallocate_Character_array(array,size) \ 683 reallocate_array((void *) array, size, sizeof(unsigned char)) 684 685 686 /* External functions that are cross-referenced among the flex source files. */ 687 688 689 /* from file ccl.c */ 690 691 extern void ccladd(int, int); /* add a single character to a ccl */ 692 extern int cclinit(void); /* make an empty ccl */ 693 extern void cclnegate(int); /* negate a ccl */ 694 extern int ccl_set_diff (int a, int b); /* set difference of two ccls. */ 695 extern int ccl_set_union (int a, int b); /* set union of two ccls. */ 696 697 /* List the members of a set of characters in CCL form. */ 698 extern void list_character_set(FILE *, int[]); 699 700 701 /* from file dfa.c */ 702 703 /* Check a DFA state for backing up. */ 704 extern void check_for_backing_up(int, int[]); 705 706 /* Check to see if NFA state set constitutes "dangerous" trailing context. */ 707 extern void check_trailing_context(int *, int, int *, int); 708 709 /* Construct the epsilon closure of a set of ndfa states. */ 710 extern int *epsclosure(int *, int *, int[], int *, int *); 711 712 /* Increase the maximum number of dfas. */ 713 extern void increase_max_dfas(void); 714 715 extern void ntod(void); /* convert a ndfa to a dfa */ 716 717 /* Converts a set of ndfa states into a dfa state. */ 718 extern int snstods(int[], int, int[], int, int, int *); 719 720 721 /* from file ecs.c */ 722 723 /* Convert character classes to set of equivalence classes. */ 724 extern void ccl2ecl(void); 725 726 /* Associate equivalence class numbers with class members. */ 727 extern int cre8ecs(int[], int[], int); 728 729 /* Update equivalence classes based on character class transitions. */ 730 extern void mkeccl(unsigned char[], int, int[], int[], int, int); 731 732 /* Create equivalence class for single character. */ 733 extern void mkechar(int, int[], int[]); 734 735 736 /* from file gen.c */ 737 738 extern void do_indent(void); /* indent to the current level */ 739 740 /* Generate the code to keep backing-up information. */ 741 extern void gen_backing_up(void); 742 743 /* Generate the code to perform the backing up. */ 744 extern void gen_bu_action(void); 745 746 /* Generate full speed compressed transition table. */ 747 extern void genctbl(void); 748 749 /* Generate the code to find the action number. */ 750 extern void gen_find_action(void); 751 752 extern void genftbl(void); /* generate full transition table */ 753 754 /* Generate the code to find the next compressed-table state. */ 755 extern void gen_next_compressed_state(char *); 756 757 /* Generate the code to find the next match. */ 758 extern void gen_next_match(void); 759 760 /* Generate the code to find the next state. */ 761 extern void gen_next_state(int); 762 763 /* Generate the code to make a NUL transition. */ 764 extern void gen_NUL_trans(void); 765 766 /* Generate the code to find the start state. */ 767 extern void gen_start_state(void); 768 769 /* Generate data statements for the transition tables. */ 770 extern void gentabs(void); 771 772 /* Write out a formatted string at the current indentation level. */ 773 extern void indent_put2s(const char *, const char *); 774 775 /* Write out a string + newline at the current indentation level. */ 776 extern void indent_puts(const char *); 777 778 extern void make_tables(void); /* generate transition tables */ 779 780 781 /* from file main.c */ 782 783 extern void check_options(void); 784 extern void flexend(int); 785 extern void usage(void); 786 787 788 /* from file misc.c */ 789 790 /* Add a #define to the action file. */ 791 extern void action_define(const char *defname, int value); 792 793 /* Add the given text to the stored actions. */ 794 extern void add_action(const char *new_text); 795 796 /* True if a string is all lower case. */ 797 extern int all_lower(char *); 798 799 /* True if a string is all upper case. */ 800 extern int all_upper(char *); 801 802 /* Compare two integers for use by qsort. */ 803 extern int intcmp(const void *, const void *); 804 805 /* Check a character to make sure it's in the expected range. */ 806 extern void check_char(int c); 807 808 /* Replace upper-case letter to lower-case. */ 809 extern unsigned char clower(int); 810 811 /* strdup() that fails fatally on allocation failures. */ 812 extern char *xstrdup(const char *); 813 814 /* Compare two characters for use by qsort with '\0' sorting last. */ 815 extern int cclcmp(const void *, const void *); 816 817 /* Finish up a block of data declarations. */ 818 extern void dataend(void); 819 820 /* Flush generated data statements. */ 821 extern void dataflush(void); 822 823 /* Report an error message and terminate. */ 824 extern void flexerror(const char *); 825 826 /* Report a fatal error message and terminate. */ 827 extern void flexfatal(const char *); 828 829 /* Report a fatal error with a pinpoint, and terminate */ 830 #if HAVE_DECL___FUNC__ 831 #define flex_die(msg) \ 832 do{ \ 833 fprintf (stderr,\ 834 _("%s: fatal internal error at %s:%d (%s): %s\n"),\ 835 program_name, __FILE__, (int)__LINE__,\ 836 __func__,msg);\ 837 FLEX_EXIT(1);\ 838 }while(0) 839 #else /* ! HAVE_DECL___FUNC__ */ 840 #define flex_die(msg) \ 841 do{ \ 842 fprintf (stderr,\ 843 _("%s: fatal internal error at %s:%d %s\n"),\ 844 program_name, __FILE__, (int)__LINE__,\ 845 msg);\ 846 FLEX_EXIT(1);\ 847 }while(0) 848 #endif /* ! HAVE_DECL___func__ */ 849 850 /* Report an error message formatted */ 851 extern void lerr(const char *, ...) 852 #if defined(__GNUC__) && __GNUC__ >= 3 853 __attribute__((__format__(__printf__, 1, 2))) 854 #endif 855 ; 856 857 /* Like lerr, but also exit after displaying message. */ 858 extern void lerr_fatal(const char *, ...) 859 #if defined(__GNUC__) && __GNUC__ >= 3 860 __attribute__((__format__(__printf__, 1, 2))) 861 #endif 862 ; 863 864 /* Spit out a "#line" statement. */ 865 extern void line_directive_out(FILE *, int); 866 867 /* Mark the current position in the action array as the end of the section 1 868 * user defs. 869 */ 870 extern void mark_defs1(void); 871 872 /* Mark the current position in the action array as the end of the prolog. */ 873 extern void mark_prolog(void); 874 875 /* Generate a data statment for a two-dimensional array. */ 876 extern void mk2data(int); 877 878 extern void mkdata(int); /* generate a data statement */ 879 880 /* Return the integer represented by a string of digits. */ 881 extern int myctoi(const char *); 882 883 /* Return character corresponding to escape sequence. */ 884 extern unsigned char myesc(unsigned char[]); 885 886 /* Output a (possibly-formatted) string to the generated scanner. */ 887 extern void out(const char *); 888 extern void out_dec(const char *, int); 889 extern void out_dec2(const char *, int, int); 890 extern void out_hex(const char *, unsigned int); 891 extern void out_str(const char *, const char *); 892 extern void out_str3(const char *, const char *, const char *, const char *); 893 extern void out_str_dec(const char *, const char *, int); 894 extern void outc(int); 895 extern void outn(const char *); 896 extern void out_m4_define(const char* def, const char* val); 897 898 /* Return a printable version of the given character, which might be 899 * 8-bit. 900 */ 901 extern char *readable_form(int); 902 903 /* Write out one section of the skeleton file. */ 904 extern void skelout(void); 905 906 /* Output a yy_trans_info structure. */ 907 extern void transition_struct_out(int, int); 908 909 /* Only needed when using certain broken versions of bison to build parse.c. */ 910 extern void *yy_flex_xmalloc(int); 911 912 913 /* from file nfa.c */ 914 915 /* Add an accepting state to a machine. */ 916 extern void add_accept(int, int); 917 918 /* Make a given number of copies of a singleton machine. */ 919 extern int copysingl(int, int); 920 921 /* Debugging routine to write out an nfa. */ 922 extern void dumpnfa(int); 923 924 /* Finish up the processing for a rule. */ 925 extern void finish_rule(int, int, int, int, int); 926 927 /* Connect two machines together. */ 928 extern int link_machines(int, int); 929 930 /* Mark each "beginning" state in a machine as being a "normal" (i.e., 931 * not trailing context associated) state. 932 */ 933 extern void mark_beginning_as_normal(int); 934 935 /* Make a machine that branches to two machines. */ 936 extern int mkbranch(int, int); 937 938 extern int mkclos(int); /* convert a machine into a closure */ 939 extern int mkopt(int); /* make a machine optional */ 940 941 /* Make a machine that matches either one of two machines. */ 942 extern int mkor(int, int); 943 944 /* Convert a machine into a positive closure. */ 945 extern int mkposcl(int); 946 947 extern int mkrep(int, int, int); /* make a replicated machine */ 948 949 /* Create a state with a transition on a given symbol. */ 950 extern int mkstate(int); 951 952 extern void new_rule(void); /* initialize for a new rule */ 953 954 955 /* from file parse.y */ 956 957 /* Build the "<<EOF>>" action for the active start conditions. */ 958 extern void build_eof_action(void); 959 960 /* Write out a message formatted with one string, pinpointing its location. */ 961 extern void format_pinpoint_message(const char *, const char *); 962 963 /* Write out a message, pinpointing its location. */ 964 extern void pinpoint_message(const char *); 965 966 /* Write out a warning, pinpointing it at the given line. */ 967 extern void line_warning(const char *, int); 968 969 /* Write out a message, pinpointing it at the given line. */ 970 extern void line_pinpoint(const char *, int); 971 972 /* Report a formatted syntax error. */ 973 extern void format_synerr(const char *, const char *); 974 extern void synerr(const char *); /* report a syntax error */ 975 extern void format_warn(const char *, const char *); 976 extern void lwarn(const char *); /* report a warning */ 977 extern void yyerror(const char *); /* report a parse error */ 978 extern int yyparse(void); /* the YACC parser */ 979 980 981 /* from file scan.l */ 982 983 /* The Flex-generated scanner for flex. */ 984 extern int flexscan(void); 985 986 /* Open the given file (if NULL, stdin) for scanning. */ 987 extern void set_input_file(char *); 988 989 990 /* from file sym.c */ 991 992 /* Save the text of a character class. */ 993 extern void cclinstal(char[], int); 994 995 /* Lookup the number associated with character class. */ 996 extern int ccllookup(char[]); 997 998 extern void ndinstal(const char *, char[]); /* install a name definition */ 999 extern char *ndlookup(const char *); /* lookup a name definition */ 1000 1001 /* Increase maximum number of SC's. */ 1002 extern void scextend(void); 1003 extern void scinstal(const char *, int); /* make a start condition */ 1004 1005 /* Lookup the number associated with a start condition. */ 1006 extern int sclookup(const char *); 1007 1008 1009 /* from file tblcmp.c */ 1010 1011 /* Build table entries for dfa state. */ 1012 extern void bldtbl(int[], int, int, int, int); 1013 1014 extern void cmptmps(void); /* compress template table entries */ 1015 extern void expand_nxt_chk(void); /* increase nxt/chk arrays */ 1016 1017 /* Finds a space in the table for a state to be placed. */ 1018 extern int find_table_space(int *, int); 1019 extern void inittbl(void); /* initialize transition tables */ 1020 1021 /* Make the default, "jam" table entries. */ 1022 extern void mkdeftbl(void); 1023 1024 /* Create table entries for a state (or state fragment) which has 1025 * only one out-transition. 1026 */ 1027 extern void mk1tbl(int, int, int, int); 1028 1029 /* Place a state into full speed transition table. */ 1030 extern void place_state(int *, int, int); 1031 1032 /* Save states with only one out-transition to be processed later. */ 1033 extern void stack1(int, int, int, int); 1034 1035 1036 /* from file yylex.c */ 1037 1038 extern int yylex(void); 1039 1040 /* A growable array. See buf.c. */ 1041 struct Buf { 1042 void *elts; /* elements. */ 1043 int nelts; /* number of elements. */ 1044 size_t elt_size; /* in bytes. */ 1045 int nmax; /* max capacity of elements. */ 1046 }; 1047 1048 extern void buf_init(struct Buf * buf, size_t elem_size); 1049 extern void buf_destroy(struct Buf * buf); 1050 extern struct Buf *buf_append(struct Buf * buf, const void *ptr, int n_elem); 1051 extern struct Buf *buf_concat(struct Buf* dest, const struct Buf* src); 1052 extern struct Buf *buf_strappend(struct Buf *, const char *str); 1053 extern struct Buf *buf_strnappend(struct Buf *, const char *str, int nchars); 1054 extern struct Buf *buf_strdefine(struct Buf * buf, const char *str, const char *def); 1055 extern struct Buf *buf_prints(struct Buf *buf, const char *fmt, const char* s); 1056 extern struct Buf *buf_m4_define(struct Buf *buf, const char* def, const char* val); 1057 extern struct Buf *buf_m4_undefine(struct Buf *buf, const char* def); 1058 extern struct Buf *buf_print_strings(struct Buf * buf, FILE* out); 1059 extern struct Buf *buf_linedir(struct Buf *buf, const char* filename, int lineno); 1060 1061 extern struct Buf userdef_buf; /* a string buffer for #define's generated by user-options on cmd line. */ 1062 extern struct Buf defs_buf; /* a char* buffer to save #define'd some symbols generated by flex. */ 1063 extern struct Buf yydmap_buf; /* a string buffer to hold yydmap elements */ 1064 extern struct Buf m4defs_buf; /* Holds m4 definitions. */ 1065 extern struct Buf top_buf; /* contains %top code. String buffer. */ 1066 extern bool no_section3_escape; /* True if the undocumented option --unsafe-no-m4-sect3-escape was passed */ 1067 1068 /* For blocking out code from the header file. */ 1069 #define OUT_BEGIN_CODE() outn("m4_ifdef( [[M4_YY_IN_HEADER]],,[[m4_dnl") 1070 #define OUT_END_CODE() outn("]])") 1071 1072 /* For setjmp/longjmp (instead of calling exit(2)). Linkage in main.c */ 1073 extern jmp_buf flex_main_jmp_buf; 1074 1075 #define FLEX_EXIT(status) longjmp(flex_main_jmp_buf,(status)+1) 1076 1077 /* Removes all \n and \r chars from tail of str. returns str. */ 1078 extern char *chomp (char *str); 1079 1080 /* ctype functions forced to return boolean */ 1081 #define b_isalnum(c) (isalnum(c)?true:false) 1082 #define b_isalpha(c) (isalpha(c)?true:false) 1083 #define b_isascii(c) (isascii(c)?true:false) 1084 #define b_isblank(c) (isblank(c)?true:false) 1085 #define b_iscntrl(c) (iscntrl(c)?true:false) 1086 #define b_isdigit(c) (isdigit(c)?true:false) 1087 #define b_isgraph(c) (isgraph(c)?true:false) 1088 #define b_islower(c) (islower(c)?true:false) 1089 #define b_isprint(c) (isprint(c)?true:false) 1090 #define b_ispunct(c) (ispunct(c)?true:false) 1091 #define b_isspace(c) (isspace(c)?true:false) 1092 #define b_isupper(c) (isupper(c)?true:false) 1093 #define b_isxdigit(c) (isxdigit(c)?true:false) 1094 1095 /* return true if char is uppercase or lowercase. */ 1096 bool has_case(int c); 1097 1098 /* Change case of character if possible. */ 1099 int reverse_case(int c); 1100 1101 /* return false if [c1-c2] is ambiguous for a caseless scanner. */ 1102 bool range_covers_case (int c1, int c2); 1103 1104 /* 1105 * From "filter.c" 1106 */ 1107 1108 /** A single stdio filter to execute. 1109 * The filter may be external, such as "sed", or it 1110 * may be internal, as a function call. 1111 */ 1112 struct filter { 1113 int (*filter_func)(struct filter*); /**< internal filter function */ 1114 void * extra; /**< extra data passed to filter_func */ 1115 int argc; /**< arg count */ 1116 const char ** argv; /**< arg vector, \0-terminated */ 1117 struct filter * next; /**< next filter or NULL */ 1118 }; 1119 1120 /* output filter chain */ 1121 extern struct filter * output_chain; 1122 extern struct filter *filter_create_ext (struct filter * chain, const char *cmd, ...); 1123 struct filter *filter_create_int(struct filter *chain, 1124 int (*filter_func) (struct filter *), 1125 void *extra); 1126 extern bool filter_apply_chain(struct filter * chain); 1127 extern int filter_truncate(struct filter * chain, int max_len); 1128 extern int filter_tee_header(struct filter *chain); 1129 extern int filter_fix_linedirs(struct filter *chain); 1130 1131 1132 /* 1133 * From "regex.c" 1134 */ 1135 1136 extern regex_t regex_linedir, regex_blank_line; 1137 bool flex_init_regex(void); 1138 void flex_regcomp(regex_t *preg, const char *regex, int cflags); 1139 char *regmatch_dup (regmatch_t * m, const char *src); 1140 char *regmatch_cpy (regmatch_t * m, char *dest, const char *src); 1141 int regmatch_len (regmatch_t * m); 1142 int regmatch_strtol (regmatch_t * m, const char *src, char **endptr, int base); 1143 bool regmatch_empty (regmatch_t * m); 1144 1145 /* From "scanflags.h" */ 1146 typedef unsigned int scanflags_t; 1147 extern scanflags_t* _sf_stk; 1148 extern size_t _sf_top_ix, _sf_max; /**< stack of scanner flags. */ 1149 #define _SF_CASE_INS ((scanflags_t) 0x0001) 1150 #define _SF_DOT_ALL ((scanflags_t) 0x0002) 1151 #define _SF_SKIP_WS ((scanflags_t) 0x0004) 1152 #define sf_top() (_sf_stk[_sf_top_ix]) 1153 #define sf_case_ins() (sf_top() & _SF_CASE_INS) 1154 #define sf_dot_all() (sf_top() & _SF_DOT_ALL) 1155 #define sf_skip_ws() (sf_top() & _SF_SKIP_WS) 1156 #define sf_set_case_ins(X) ((X) ? (sf_top() |= _SF_CASE_INS) : (sf_top() &= ~_SF_CASE_INS)) 1157 #define sf_set_dot_all(X) ((X) ? (sf_top() |= _SF_DOT_ALL) : (sf_top() &= ~_SF_DOT_ALL)) 1158 #define sf_set_skip_ws(X) ((X) ? (sf_top() |= _SF_SKIP_WS) : (sf_top() &= ~_SF_SKIP_WS)) 1159 extern void sf_init(void); 1160 extern void sf_push(void); 1161 extern void sf_pop(void); 1162 1163 1164 #endif /* not defined FLEXDEF_H */ 1165