1 /* $NetBSD: flexdef.h,v 1.7 2018/12/23 16:27:17 christos Exp $ */ 2 3 /* flexdef - definitions file for flex */ 4 5 /* Copyright (c) 1990 The Regents of the University of California. */ 6 /* All rights reserved. */ 7 8 /* This code is derived from software contributed to Berkeley by */ 9 /* Vern Paxson. */ 10 11 /* The United States Government has rights in this work pursuant */ 12 /* to contract no. DE-AC03-76SF00098 between the United States */ 13 /* Department of Energy and the University of California. */ 14 15 /* This file is part of flex. */ 16 17 /* Redistribution and use in source and binary forms, with or without */ 18 /* modification, are permitted provided that the following conditions */ 19 /* are met: */ 20 21 /* 1. Redistributions of source code must retain the above copyright */ 22 /* notice, this list of conditions and the following disclaimer. */ 23 /* 2. Redistributions in binary form must reproduce the above copyright */ 24 /* notice, this list of conditions and the following disclaimer in the */ 25 /* documentation and/or other materials provided with the distribution. */ 26 27 /* Neither the name of the University nor the names of its contributors */ 28 /* may be used to endorse or promote products derived from this software */ 29 /* without specific prior written permission. */ 30 31 /* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR */ 32 /* IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */ 33 /* WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */ 34 /* PURPOSE. */ 35 36 #ifndef FLEXDEF_H 37 #define FLEXDEF_H 1 38 39 #if HAVE_NBTOOL_CONFIG_H 40 #include "nbtool_config.h" 41 #undef PACKAGE_BUGREPORT 42 #undef PACKAGE_NAME 43 #undef PACKAGE_STRING 44 #undef PACKAGE_TARNAME 45 #undef PACKAGE_URL 46 #undef PACKAGE_VERSION 47 #endif 48 49 #ifdef HAVE_CONFIG_H 50 #include <config.h> 51 #endif 52 53 #include <stdio.h> 54 #include <stdlib.h> 55 #include <stdarg.h> 56 #include <setjmp.h> 57 #include <ctype.h> 58 #include <libgen.h> /* for XPG version of basename(3) */ 59 #include <string.h> 60 #include <math.h> 61 62 #ifdef HAVE_ASSERT_H 63 #include <assert.h> 64 #else 65 #define assert(Pred) 66 #endif 67 68 #ifdef HAVE_LIMITS_H 69 #include <limits.h> 70 #endif 71 /* Required: dup() and dup2() in <unistd.h> */ 72 #include <unistd.h> 73 #ifdef HAVE_NETINET_IN_H 74 #include <netinet/in.h> 75 #endif 76 #ifdef HAVE_SYS_PARAMS_H 77 #include <sys/params.h> 78 #endif 79 /* Required: stat() in <sys/stat.h> */ 80 #include <sys/stat.h> 81 /* Required: wait() in <sys/wait.h> */ 82 #include <sys/wait.h> 83 #include <stdbool.h> 84 #include <stdarg.h> 85 /* Required: regcomp(), regexec() and regerror() in <regex.h> */ 86 #include <regex.h> 87 /* Required: strcasecmp() in <strings.h> */ 88 #include <strings.h> 89 #include "flexint.h" 90 91 /* We use gettext. So, when we write strings which should be translated, we mark them with _() */ 92 #ifdef ENABLE_NLS 93 #ifdef HAVE_LOCALE_H 94 #include <locale.h> 95 #endif /* HAVE_LOCALE_H */ 96 #include "gettext.h" 97 #define _(String) gettext (String) 98 #else 99 #define _(STRING) STRING 100 #endif /* ENABLE_NLS */ 101 102 /* Always be prepared to generate an 8-bit scanner. */ 103 #define CSIZE 256 104 105 /* Size of input alphabet - should be size of ASCII set. */ 106 #ifndef DEFAULT_CSIZE 107 #define DEFAULT_CSIZE 128 108 #endif 109 110 /* Maximum line length we'll have to deal with. */ 111 #define MAXLINE 2048 112 113 #ifndef MIN 114 #define MIN(x,y) ((x) < (y) ? (x) : (y)) 115 #endif 116 #ifndef MAX 117 #define MAX(x,y) ((x) > (y) ? (x) : (y)) 118 #endif 119 #ifndef ABS 120 #define ABS(x) ((x) < 0 ? -(x) : (x)) 121 #endif 122 123 /* Whether an integer is a power of two */ 124 #define is_power_of_2(n) ((n) > 0 && ((n) & ((n) - 1)) == 0) 125 126 #define unspecified -1 127 128 /* Special chk[] values marking the slots taking by end-of-buffer and action 129 * numbers. 130 */ 131 #define EOB_POSITION -1 132 #define ACTION_POSITION -2 133 134 /* Number of data items per line for -f output. */ 135 #define NUMDATAITEMS 10 136 137 /* Number of lines of data in -f output before inserting a blank line for 138 * readability. 139 */ 140 #define NUMDATALINES 10 141 142 /* transition_struct_out() definitions. */ 143 #define TRANS_STRUCT_PRINT_LENGTH 14 144 145 /* Returns true if an nfa state has an epsilon out-transition slot 146 * that can be used. This definition is currently not used. 147 */ 148 #define FREE_EPSILON(state) \ 149 (transchar[state] == SYM_EPSILON && \ 150 trans2[state] == NO_TRANSITION && \ 151 finalst[state] != state) 152 153 /* Returns true if an nfa state has an epsilon out-transition character 154 * and both slots are free 155 */ 156 #define SUPER_FREE_EPSILON(state) \ 157 (transchar[state] == SYM_EPSILON && \ 158 trans1[state] == NO_TRANSITION) \ 159 160 /* Maximum number of NFA states that can comprise a DFA state. It's real 161 * big because if there's a lot of rules, the initial state will have a 162 * huge epsilon closure. 163 */ 164 #define INITIAL_MAX_DFA_SIZE 750 165 #define MAX_DFA_SIZE_INCREMENT 750 166 167 168 /* A note on the following masks. They are used to mark accepting numbers 169 * as being special. As such, they implicitly limit the number of accepting 170 * numbers (i.e., rules) because if there are too many rules the rule numbers 171 * will overload the mask bits. Fortunately, this limit is \large/ (0x2000 == 172 * 8192) so unlikely to actually cause any problems. A check is made in 173 * new_rule() to ensure that this limit is not reached. 174 */ 175 176 /* Mask to mark a trailing context accepting number. */ 177 #define YY_TRAILING_MASK 0x2000 178 179 /* Mask to mark the accepting number of the "head" of a trailing context 180 * rule. 181 */ 182 #define YY_TRAILING_HEAD_MASK 0x4000 183 184 /* Maximum number of rules, as outlined in the above note. */ 185 #define MAX_RULE (YY_TRAILING_MASK - 1) 186 187 188 /* NIL must be 0. If not, its special meaning when making equivalence classes 189 * (it marks the representative of a given e.c.) will be unidentifiable. 190 */ 191 #define NIL 0 192 193 #define JAM -1 /* to mark a missing DFA transition */ 194 #define NO_TRANSITION NIL 195 #define UNIQUE -1 /* marks a symbol as an e.c. representative */ 196 #define INFINITE_REPEAT -1 /* for x{5,} constructions */ 197 198 #define INITIAL_MAX_CCLS 100 /* max number of unique character classes */ 199 #define MAX_CCLS_INCREMENT 100 200 201 /* Size of table holding members of character classes. */ 202 #define INITIAL_MAX_CCL_TBL_SIZE 500 203 #define MAX_CCL_TBL_SIZE_INCREMENT 250 204 205 #define INITIAL_MAX_RULES 100 /* default maximum number of rules */ 206 #define MAX_RULES_INCREMENT 100 207 208 #define INITIAL_MNS 2000 /* default maximum number of nfa states */ 209 #define MNS_INCREMENT 1000 /* amount to bump above by if it's not enough */ 210 211 #define INITIAL_MAX_DFAS 1000 /* default maximum number of dfa states */ 212 #define MAX_DFAS_INCREMENT 1000 213 214 #define JAMSTATE -32766 /* marks a reference to the state that always jams */ 215 216 /* Maximum number of NFA states. */ 217 #define MAXIMUM_MNS 31999 218 #define MAXIMUM_MNS_LONG 1999999999 219 220 /* Enough so that if it's subtracted from an NFA state number, the result 221 * is guaranteed to be negative. 222 */ 223 #define MARKER_DIFFERENCE (maximum_mns+2) 224 225 /* Maximum number of nxt/chk pairs for non-templates. */ 226 #define INITIAL_MAX_XPAIRS 2000 227 #define MAX_XPAIRS_INCREMENT 2000 228 229 /* Maximum number of nxt/chk pairs needed for templates. */ 230 #define INITIAL_MAX_TEMPLATE_XPAIRS 2500 231 #define MAX_TEMPLATE_XPAIRS_INCREMENT 2500 232 233 #define SYM_EPSILON (CSIZE + 1) /* to mark transitions on the symbol epsilon */ 234 235 #define INITIAL_MAX_SCS 40 /* maximum number of start conditions */ 236 #define MAX_SCS_INCREMENT 40 /* amount to bump by if it's not enough */ 237 238 #define ONE_STACK_SIZE 500 /* stack of states with only one out-transition */ 239 #define SAME_TRANS -1 /* transition is the same as "default" entry for state */ 240 241 /* The following percentages are used to tune table compression: 242 243 * The percentage the number of out-transitions a state must be of the 244 * number of equivalence classes in order to be considered for table 245 * compaction by using protos. 246 */ 247 #define PROTO_SIZE_PERCENTAGE 15 248 249 /* The percentage the number of homogeneous out-transitions of a state 250 * must be of the number of total out-transitions of the state in order 251 * that the state's transition table is first compared with a potential 252 * template of the most common out-transition instead of with the first 253 * proto in the proto queue. 254 */ 255 #define CHECK_COM_PERCENTAGE 50 256 257 /* The percentage the number of differences between a state's transition 258 * table and the proto it was first compared with must be of the total 259 * number of out-transitions of the state in order to keep the first 260 * proto as a good match and not search any further. 261 */ 262 #define FIRST_MATCH_DIFF_PERCENTAGE 10 263 264 /* The percentage the number of differences between a state's transition 265 * table and the most similar proto must be of the state's total number 266 * of out-transitions to use the proto as an acceptable close match. 267 */ 268 #define ACCEPTABLE_DIFF_PERCENTAGE 50 269 270 /* The percentage the number of homogeneous out-transitions of a state 271 * must be of the number of total out-transitions of the state in order 272 * to consider making a template from the state. 273 */ 274 #define TEMPLATE_SAME_PERCENTAGE 60 275 276 /* The percentage the number of differences between a state's transition 277 * table and the most similar proto must be of the state's total number 278 * of out-transitions to create a new proto from the state. 279 */ 280 #define NEW_PROTO_DIFF_PERCENTAGE 20 281 282 /* The percentage the total number of out-transitions of a state must be 283 * of the number of equivalence classes in order to consider trying to 284 * fit the transition table into "holes" inside the nxt/chk table. 285 */ 286 #define INTERIOR_FIT_PERCENTAGE 15 287 288 /* Size of region set aside to cache the complete transition table of 289 * protos on the proto queue to enable quick comparisons. 290 */ 291 #define PROT_SAVE_SIZE 2000 292 293 #define MSP 50 /* maximum number of saved protos (protos on the proto queue) */ 294 295 /* Maximum number of out-transitions a state can have that we'll rummage 296 * around through the interior of the internal fast table looking for a 297 * spot for it. 298 */ 299 #define MAX_XTIONS_FULL_INTERIOR_FIT 4 300 301 /* Maximum number of rules which will be reported as being associated 302 * with a DFA state. 303 */ 304 #define MAX_ASSOC_RULES 100 305 306 /* Number that, if used to subscript an array, has a good chance of producing 307 * an error; should be small enough to fit into a short. 308 */ 309 #define BAD_SUBSCRIPT -32767 310 311 /* Absolute value of largest number that can be stored in a short, with a 312 * bit of slop thrown in for general paranoia. 313 */ 314 #define MAX_SHORT 32700 315 316 317 /* Declarations for global variables. */ 318 319 320 /* Variables for flags: 321 * printstats - if true (-v), dump statistics 322 * syntaxerror - true if a syntax error has been found 323 * eofseen - true if we've seen an eof in the input file 324 * ddebug - if true (-d), make a "debug" scanner 325 * trace - if true (-T), trace processing 326 * nowarn - if true (-w), do not generate warnings 327 * spprdflt - if true (-s), suppress the default rule 328 * interactive - if true (-I), generate an interactive scanner 329 * lex_compat - if true (-l), maximize compatibility with AT&T lex 330 * posix_compat - if true (-X), maximize compatibility with POSIX lex 331 * do_yylineno - if true, generate code to maintain yylineno 332 * useecs - if true (-Ce flag), use equivalence classes 333 * fulltbl - if true (-Cf flag), don't compress the DFA state table 334 * usemecs - if true (-Cm flag), use meta-equivalence classes 335 * fullspd - if true (-F flag), use Jacobson method of table representation 336 * gen_line_dirs - if true (i.e., no -L flag), generate #line directives 337 * performance_report - if > 0 (i.e., -p flag), generate a report relating 338 * to scanner performance; if > 1 (-p -p), report on minor performance 339 * problems, too 340 * backing_up_report - if true (i.e., -b flag), generate "lex.backup" file 341 * listing backing-up states 342 * C_plus_plus - if true (i.e., -+ flag), generate a C++ scanner class; 343 * otherwise, a standard C scanner 344 * reentrant - if true (-R), generate a reentrant C scanner. 345 * bison_bridge_lval - if true (--bison-bridge), bison pure calling convention. 346 * bison_bridge_lloc - if true (--bison-locations), bison yylloc. 347 * long_align - if true (-Ca flag), favor long-word alignment. 348 * use_read - if true (-f, -F, or -Cr) then use read() for scanner input; 349 * otherwise, use fread(). 350 * yytext_is_array - if true (i.e., %array directive), then declare 351 * yytext as a array instead of a character pointer. Nice and inefficient. 352 * do_yywrap - do yywrap() processing on EOF. If false, EOF treated as 353 * "no more files". 354 * csize - size of character set for the scanner we're generating; 355 * 128 for 7-bit chars and 256 for 8-bit 356 * yymore_used - if true, yymore() is used in input rules 357 * reject - if true, generate back-up tables for REJECT macro 358 * real_reject - if true, scanner really uses REJECT (as opposed to just 359 * having "reject" set for variable trailing context) 360 * continued_action - true if this rule's action is to "fall through" to 361 * the next rule's action (i.e., the '|' action) 362 * in_rule - true if we're inside an individual rule, false if not. 363 * yymore_really_used - whether to treat yymore() as really used, regardless 364 * of what we think based on references to it in the user's actions. 365 * reject_really_used - same for REJECT 366 * trace_hex - use hexadecimal numbers in trace/debug outputs instead of octals 367 */ 368 369 extern int printstats, syntaxerror, eofseen, ddebug, trace, nowarn, 370 spprdflt; 371 extern int interactive, lex_compat, posix_compat, do_yylineno; 372 extern int useecs, fulltbl, usemecs, fullspd; 373 extern int gen_line_dirs, performance_report, backing_up_report; 374 extern int reentrant, bison_bridge_lval, bison_bridge_lloc; 375 extern int C_plus_plus, long_align, use_read, yytext_is_array, do_yywrap; 376 extern int csize; 377 extern int yymore_used, reject, real_reject, continued_action, in_rule; 378 379 extern int yymore_really_used, reject_really_used; 380 extern int trace_hex; 381 382 /* Variables used in the flex input routines: 383 * datapos - characters on current output line 384 * dataline - number of contiguous lines of data in current data 385 * statement. Used to generate readable -f output 386 * linenum - current input line number 387 * skelfile - the skeleton file 388 * skel - compiled-in skeleton array 389 * skel_ind - index into "skel" array, if skelfile is nil 390 * yyin - input file 391 * backing_up_file - file to summarize backing-up states to 392 * infilename - name of input file 393 * outfilename - name of output file 394 * headerfilename - name of the .h file to generate 395 * did_outfilename - whether outfilename was explicitly set 396 * prefix - the prefix used for externally visible names ("yy" by default) 397 * yyclass - yyFlexLexer subclass to use for YY_DECL 398 * do_stdinit - whether to initialize yyin/yyout to stdin/stdout 399 * use_stdout - the -t flag 400 * input_files - array holding names of input files 401 * num_input_files - size of input_files array 402 * program_name - name with which program was invoked 403 * 404 * action_array - array to hold the rule actions 405 * action_size - size of action_array 406 * defs1_offset - index where the user's section 1 definitions start 407 * in action_array 408 * prolog_offset - index where the prolog starts in action_array 409 * action_offset - index where the non-prolog starts in action_array 410 * action_index - index where the next action should go, with respect 411 * to "action_array" 412 */ 413 414 extern int datapos, dataline, linenum; 415 extern FILE *skelfile, *backing_up_file; 416 extern const char *skel[]; 417 extern int skel_ind; 418 extern char *infilename, *outfilename, *headerfilename; 419 extern int did_outfilename; 420 extern char *prefix, *yyclass, *extra_type; 421 extern int do_stdinit, use_stdout; 422 extern char **input_files; 423 extern int num_input_files; 424 extern char *program_name; 425 426 extern char *action_array; 427 extern int action_size; 428 extern int defs1_offset, prolog_offset, action_offset, action_index; 429 430 431 /* Variables for stack of states having only one out-transition: 432 * onestate - state number 433 * onesym - transition symbol 434 * onenext - target state 435 * onedef - default base entry 436 * onesp - stack pointer 437 */ 438 439 extern int onestate[ONE_STACK_SIZE], onesym[ONE_STACK_SIZE]; 440 extern int onenext[ONE_STACK_SIZE], onedef[ONE_STACK_SIZE], onesp; 441 442 443 /* Variables for nfa machine data: 444 * maximum_mns - maximal number of NFA states supported by tables 445 * current_mns - current maximum on number of NFA states 446 * num_rules - number of the last accepting state; also is number of 447 * rules created so far 448 * num_eof_rules - number of <<EOF>> rules 449 * default_rule - number of the default rule 450 * current_max_rules - current maximum number of rules 451 * lastnfa - last nfa state number created 452 * firstst - physically the first state of a fragment 453 * lastst - last physical state of fragment 454 * finalst - last logical state of fragment 455 * transchar - transition character 456 * trans1 - transition state 457 * trans2 - 2nd transition state for epsilons 458 * accptnum - accepting number 459 * assoc_rule - rule associated with this NFA state (or 0 if none) 460 * state_type - a STATE_xxx type identifying whether the state is part 461 * of a normal rule, the leading state in a trailing context 462 * rule (i.e., the state which marks the transition from 463 * recognizing the text-to-be-matched to the beginning of 464 * the trailing context), or a subsequent state in a trailing 465 * context rule 466 * rule_type - a RULE_xxx type identifying whether this a ho-hum 467 * normal rule or one which has variable head & trailing 468 * context 469 * rule_linenum - line number associated with rule 470 * rule_useful - true if we've determined that the rule can be matched 471 * rule_has_nl - true if rule could possibly match a newline 472 * ccl_has_nl - true if current ccl could match a newline 473 * nlch - default eol char 474 */ 475 476 extern int maximum_mns, current_mns, current_max_rules; 477 extern int num_rules, num_eof_rules, default_rule, lastnfa; 478 extern int *firstst, *lastst, *finalst, *transchar, *trans1, *trans2; 479 extern int *accptnum, *assoc_rule, *state_type; 480 extern int *rule_type, *rule_linenum, *rule_useful; 481 extern bool *rule_has_nl, *ccl_has_nl; 482 extern int nlch; 483 484 /* Different types of states; values are useful as masks, as well, for 485 * routines like check_trailing_context(). 486 */ 487 #define STATE_NORMAL 0x1 488 #define STATE_TRAILING_CONTEXT 0x2 489 490 /* Global holding current type of state we're making. */ 491 492 extern int current_state_type; 493 494 /* Different types of rules. */ 495 #define RULE_NORMAL 0 496 #define RULE_VARIABLE 1 497 498 /* True if the input rules include a rule with both variable-length head 499 * and trailing context, false otherwise. 500 */ 501 extern int variable_trailing_context_rules; 502 503 504 /* Variables for protos: 505 * numtemps - number of templates created 506 * numprots - number of protos created 507 * protprev - backlink to a more-recently used proto 508 * protnext - forward link to a less-recently used proto 509 * prottbl - base/def table entry for proto 510 * protcomst - common state of proto 511 * firstprot - number of the most recently used proto 512 * lastprot - number of the least recently used proto 513 * protsave contains the entire state array for protos 514 */ 515 516 extern int numtemps, numprots, protprev[MSP], protnext[MSP], prottbl[MSP]; 517 extern int protcomst[MSP], firstprot, lastprot, protsave[PROT_SAVE_SIZE]; 518 519 520 /* Variables for managing equivalence classes: 521 * numecs - number of equivalence classes 522 * nextecm - forward link of Equivalence Class members 523 * ecgroup - class number or backward link of EC members 524 * nummecs - number of meta-equivalence classes (used to compress 525 * templates) 526 * tecfwd - forward link of meta-equivalence classes members 527 * tecbck - backward link of MEC's 528 */ 529 530 /* Reserve enough room in the equivalence class arrays so that we 531 * can use the CSIZE'th element to hold equivalence class information 532 * for the NUL character. Later we'll move this information into 533 * the 0th element. 534 */ 535 extern int numecs, nextecm[CSIZE + 1], ecgroup[CSIZE + 1], nummecs; 536 537 /* Meta-equivalence classes are indexed starting at 1, so it's possible 538 * that they will require positions from 1 .. CSIZE, i.e., CSIZE + 1 539 * slots total (since the arrays are 0-based). nextecm[] and ecgroup[] 540 * don't require the extra position since they're indexed from 1 .. CSIZE - 1. 541 */ 542 extern int tecfwd[CSIZE + 1], tecbck[CSIZE + 1]; 543 544 545 /* Variables for start conditions: 546 * lastsc - last start condition created 547 * current_max_scs - current limit on number of start conditions 548 * scset - set of rules active in start condition 549 * scbol - set of rules active only at the beginning of line in a s.c. 550 * scxclu - true if start condition is exclusive 551 * sceof - true if start condition has EOF rule 552 * scname - start condition name 553 */ 554 555 extern int lastsc, *scset, *scbol, *scxclu, *sceof; 556 extern int current_max_scs; 557 extern char **scname; 558 559 560 /* Variables for dfa machine data: 561 * current_max_dfa_size - current maximum number of NFA states in DFA 562 * current_max_xpairs - current maximum number of non-template xtion pairs 563 * current_max_template_xpairs - current maximum number of template pairs 564 * current_max_dfas - current maximum number DFA states 565 * lastdfa - last dfa state number created 566 * nxt - state to enter upon reading character 567 * chk - check value to see if "nxt" applies 568 * tnxt - internal nxt table for templates 569 * base - offset into "nxt" for given state 570 * def - where to go if "chk" disallows "nxt" entry 571 * nultrans - NUL transition for each state 572 * NUL_ec - equivalence class of the NUL character 573 * tblend - last "nxt/chk" table entry being used 574 * firstfree - first empty entry in "nxt/chk" table 575 * dss - nfa state set for each dfa 576 * dfasiz - size of nfa state set for each dfa 577 * dfaacc - accepting set for each dfa state (if using REJECT), or accepting 578 * number, if not 579 * accsiz - size of accepting set for each dfa state 580 * dhash - dfa state hash value 581 * numas - number of DFA accepting states created; note that this 582 * is not necessarily the same value as num_rules, which is the analogous 583 * value for the NFA 584 * numsnpairs - number of state/nextstate transition pairs 585 * jambase - position in base/def where the default jam table starts 586 * jamstate - state number corresponding to "jam" state 587 * end_of_buffer_state - end-of-buffer dfa state number 588 */ 589 590 extern int current_max_dfa_size, current_max_xpairs; 591 extern int current_max_template_xpairs, current_max_dfas; 592 extern int lastdfa, *nxt, *chk, *tnxt; 593 extern int *base, *def, *nultrans, NUL_ec, tblend, firstfree, **dss, 594 *dfasiz; 595 extern union dfaacc_union { 596 int *dfaacc_set; 597 int dfaacc_state; 598 } *dfaacc; 599 extern int *accsiz, *dhash, numas; 600 extern int numsnpairs, jambase, jamstate; 601 extern int end_of_buffer_state; 602 603 /* Variables for ccl information: 604 * lastccl - ccl index of the last created ccl 605 * current_maxccls - current limit on the maximum number of unique ccl's 606 * cclmap - maps a ccl index to its set pointer 607 * ccllen - gives the length of a ccl 608 * cclng - true for a given ccl if the ccl is negated 609 * cclreuse - counts how many times a ccl is re-used 610 * current_max_ccl_tbl_size - current limit on number of characters needed 611 * to represent the unique ccl's 612 * ccltbl - holds the characters in each ccl - indexed by cclmap 613 */ 614 615 extern int lastccl, *cclmap, *ccllen, *cclng, cclreuse; 616 extern int current_maxccls, current_max_ccl_tbl_size; 617 extern unsigned char *ccltbl; 618 619 620 /* Variables for miscellaneous information: 621 * nmstr - last NAME scanned by the scanner 622 * sectnum - section number currently being parsed 623 * nummt - number of empty nxt/chk table entries 624 * hshcol - number of hash collisions detected by snstods 625 * dfaeql - number of times a newly created dfa was equal to an old one 626 * numeps - number of epsilon NFA states created 627 * eps2 - number of epsilon states which have 2 out-transitions 628 * num_reallocs - number of times it was necessary to realloc() a group 629 * of arrays 630 * tmpuses - number of DFA states that chain to templates 631 * totnst - total number of NFA states used to make DFA states 632 * peakpairs - peak number of transition pairs we had to store internally 633 * numuniq - number of unique transitions 634 * numdup - number of duplicate transitions 635 * hshsave - number of hash collisions saved by checking number of states 636 * num_backing_up - number of DFA states requiring backing up 637 * bol_needed - whether scanner needs beginning-of-line recognition 638 */ 639 640 extern char nmstr[MAXLINE]; 641 extern int sectnum, nummt, hshcol, dfaeql, numeps, eps2, num_reallocs; 642 extern int tmpuses, totnst, peakpairs, numuniq, numdup, hshsave; 643 extern int num_backing_up, bol_needed; 644 645 #ifndef HAVE_REALLOCARRAY 646 void *reallocarray(void *, size_t, size_t); 647 #endif 648 649 void *allocate_array(int, size_t); 650 void *reallocate_array(void *, int, size_t); 651 652 #define allocate_integer_array(size) \ 653 allocate_array(size, sizeof(int)) 654 655 #define reallocate_integer_array(array,size) \ 656 reallocate_array((void *) array, size, sizeof(int)) 657 658 #define allocate_bool_array(size) \ 659 allocate_array(size, sizeof(bool)) 660 661 #define reallocate_bool_array(array,size) \ 662 reallocate_array((void *) array, size, sizeof(bool)) 663 664 #define allocate_int_ptr_array(size) \ 665 allocate_array(size, sizeof(int *)) 666 667 #define allocate_char_ptr_array(size) \ 668 allocate_array(size, sizeof(char *)) 669 670 #define allocate_dfaacc_union(size) \ 671 allocate_array(size, sizeof(union dfaacc_union)) 672 673 #define reallocate_int_ptr_array(array,size) \ 674 reallocate_array((void *) array, size, sizeof(int *)) 675 676 #define reallocate_char_ptr_array(array,size) \ 677 reallocate_array((void *) array, size, sizeof(char *)) 678 679 #define reallocate_dfaacc_union(array, size) \ 680 reallocate_array((void *) array, size, sizeof(union dfaacc_union)) 681 682 #define allocate_character_array(size) \ 683 allocate_array( size, sizeof(char)) 684 685 #define reallocate_character_array(array,size) \ 686 reallocate_array((void *) array, size, sizeof(char)) 687 688 #define allocate_Character_array(size) \ 689 allocate_array(size, sizeof(unsigned char)) 690 691 #define reallocate_Character_array(array,size) \ 692 reallocate_array((void *) array, size, sizeof(unsigned char)) 693 694 695 /* External functions that are cross-referenced among the flex source files. */ 696 697 698 /* from file ccl.c */ 699 700 extern void ccladd(int, int); /* add a single character to a ccl */ 701 extern int cclinit(void); /* make an empty ccl */ 702 extern void cclnegate(int); /* negate a ccl */ 703 extern int ccl_set_diff (int a, int b); /* set difference of two ccls. */ 704 extern int ccl_set_union (int a, int b); /* set union of two ccls. */ 705 706 /* List the members of a set of characters in CCL form. */ 707 extern void list_character_set(FILE *, int[]); 708 709 710 /* from file dfa.c */ 711 712 /* Check a DFA state for backing up. */ 713 extern void check_for_backing_up(int, int[]); 714 715 /* Check to see if NFA state set constitutes "dangerous" trailing context. */ 716 extern void check_trailing_context(int *, int, int *, int); 717 718 /* Construct the epsilon closure of a set of ndfa states. */ 719 extern int *epsclosure(int *, int *, int[], int *, int *); 720 721 /* Increase the maximum number of dfas. */ 722 extern void increase_max_dfas(void); 723 724 extern void ntod(void); /* convert a ndfa to a dfa */ 725 726 /* Converts a set of ndfa states into a dfa state. */ 727 extern int snstods(int[], int, int[], int, int, int *); 728 729 730 /* from file ecs.c */ 731 732 /* Convert character classes to set of equivalence classes. */ 733 extern void ccl2ecl(void); 734 735 /* Associate equivalence class numbers with class members. */ 736 extern int cre8ecs(int[], int[], int); 737 738 /* Update equivalence classes based on character class transitions. */ 739 extern void mkeccl(unsigned char[], int, int[], int[], int, int); 740 741 /* Create equivalence class for single character. */ 742 extern void mkechar(int, int[], int[]); 743 744 745 /* from file gen.c */ 746 747 extern void do_indent(void); /* indent to the current level */ 748 749 /* Generate the code to keep backing-up information. */ 750 extern void gen_backing_up(void); 751 752 /* Generate the code to perform the backing up. */ 753 extern void gen_bu_action(void); 754 755 /* Generate full speed compressed transition table. */ 756 extern void genctbl(void); 757 758 /* Generate the code to find the action number. */ 759 extern void gen_find_action(void); 760 761 extern void genftbl(void); /* generate full transition table */ 762 763 /* Generate the code to find the next compressed-table state. */ 764 extern void gen_next_compressed_state(char *); 765 766 /* Generate the code to find the next match. */ 767 extern void gen_next_match(void); 768 769 /* Generate the code to find the next state. */ 770 extern void gen_next_state(int); 771 772 /* Generate the code to make a NUL transition. */ 773 extern void gen_NUL_trans(void); 774 775 /* Generate the code to find the start state. */ 776 extern void gen_start_state(void); 777 778 /* Generate data statements for the transition tables. */ 779 extern void gentabs(void); 780 781 /* Write out a formatted string at the current indentation level. */ 782 extern void indent_put2s(const char *, const char *); 783 784 /* Write out a string + newline at the current indentation level. */ 785 extern void indent_puts(const char *); 786 787 extern void make_tables(void); /* generate transition tables */ 788 789 790 /* from file main.c */ 791 792 extern void check_options(void); 793 extern void flexend(int); 794 extern void usage(void); 795 796 797 /* from file misc.c */ 798 799 /* Add a #define to the action file. */ 800 extern void action_define(const char *defname, int value); 801 802 /* Add the given text to the stored actions. */ 803 extern void add_action(const char *new_text); 804 805 /* True if a string is all lower case. */ 806 extern int all_lower(char *); 807 808 /* True if a string is all upper case. */ 809 extern int all_upper(char *); 810 811 /* Compare two integers for use by qsort. */ 812 extern int intcmp(const void *, const void *); 813 814 /* Check a character to make sure it's in the expected range. */ 815 extern void check_char(int c); 816 817 /* Replace upper-case letter to lower-case. */ 818 extern unsigned char clower(int); 819 820 /* strdup() that fails fatally on allocation failures. */ 821 extern char *xstrdup(const char *); 822 823 /* Compare two characters for use by qsort with '\0' sorting last. */ 824 extern int cclcmp(const void *, const void *); 825 826 /* Finish up a block of data declarations. */ 827 extern void dataend(void); 828 829 /* Flush generated data statements. */ 830 extern void dataflush(void); 831 832 /* Report an error message and terminate. */ 833 extern void flexerror(const char *); 834 835 /* Report a fatal error message and terminate. */ 836 extern void flexfatal(const char *); 837 838 /* Report a fatal error with a pinpoint, and terminate */ 839 #if HAVE_DECL___FUNC__ 840 #define flex_die(msg) \ 841 do{ \ 842 fprintf (stderr,\ 843 _("%s: fatal internal error at %s:%d (%s): %s\n"),\ 844 program_name, __FILE__, (int)__LINE__,\ 845 __func__,msg);\ 846 FLEX_EXIT(1);\ 847 }while(0) 848 #else /* ! HAVE_DECL___FUNC__ */ 849 #define flex_die(msg) \ 850 do{ \ 851 fprintf (stderr,\ 852 _("%s: fatal internal error at %s:%d %s\n"),\ 853 program_name, __FILE__, (int)__LINE__,\ 854 msg);\ 855 FLEX_EXIT(1);\ 856 }while(0) 857 #endif /* ! HAVE_DECL___func__ */ 858 859 /* Report an error message formatted */ 860 extern void lerr(const char *, ...) 861 #if defined(__GNUC__) && __GNUC__ >= 3 862 __attribute__((__format__(__printf__, 1, 2))) 863 #endif 864 ; 865 866 /* Like lerr, but also exit after displaying message. */ 867 extern void lerr_fatal(const char *, ...) 868 #if defined(__GNUC__) && __GNUC__ >= 3 869 __attribute__((__format__(__printf__, 1, 2))) 870 #endif 871 ; 872 873 /* Spit out a "#line" statement. */ 874 extern void line_directive_out(FILE *, int); 875 876 /* Mark the current position in the action array as the end of the section 1 877 * user defs. 878 */ 879 extern void mark_defs1(void); 880 881 /* Mark the current position in the action array as the end of the prolog. */ 882 extern void mark_prolog(void); 883 884 /* Generate a data statment for a two-dimensional array. */ 885 extern void mk2data(int); 886 887 extern void mkdata(int); /* generate a data statement */ 888 889 /* Return the integer represented by a string of digits. */ 890 extern int myctoi(const char *); 891 892 /* Return character corresponding to escape sequence. */ 893 extern unsigned char myesc(unsigned char[]); 894 895 /* Output a (possibly-formatted) string to the generated scanner. */ 896 extern void out(const char *); 897 extern void out_dec(const char *, int); 898 extern void out_dec2(const char *, int, int); 899 extern void out_hex(const char *, unsigned int); 900 extern void out_str(const char *, const char *); 901 extern void out_str3(const char *, const char *, const char *, const char *); 902 extern void out_str_dec(const char *, const char *, int); 903 extern void outc(int); 904 extern void outn(const char *); 905 extern void out_m4_define(const char* def, const char* val); 906 907 /* Return a printable version of the given character, which might be 908 * 8-bit. 909 */ 910 extern char *readable_form(int); 911 912 /* Write out one section of the skeleton file. */ 913 extern void skelout(void); 914 915 /* Output a yy_trans_info structure. */ 916 extern void transition_struct_out(int, int); 917 918 /* Only needed when using certain broken versions of bison to build parse.c. */ 919 extern void *yy_flex_xmalloc(int); 920 921 922 /* from file nfa.c */ 923 924 /* Add an accepting state to a machine. */ 925 extern void add_accept(int, int); 926 927 /* Make a given number of copies of a singleton machine. */ 928 extern int copysingl(int, int); 929 930 /* Debugging routine to write out an nfa. */ 931 extern void dumpnfa(int); 932 933 /* Finish up the processing for a rule. */ 934 extern void finish_rule(int, int, int, int, int); 935 936 /* Connect two machines together. */ 937 extern int link_machines(int, int); 938 939 /* Mark each "beginning" state in a machine as being a "normal" (i.e., 940 * not trailing context associated) state. 941 */ 942 extern void mark_beginning_as_normal(int); 943 944 /* Make a machine that branches to two machines. */ 945 extern int mkbranch(int, int); 946 947 extern int mkclos(int); /* convert a machine into a closure */ 948 extern int mkopt(int); /* make a machine optional */ 949 950 /* Make a machine that matches either one of two machines. */ 951 extern int mkor(int, int); 952 953 /* Convert a machine into a positive closure. */ 954 extern int mkposcl(int); 955 956 extern int mkrep(int, int, int); /* make a replicated machine */ 957 958 /* Create a state with a transition on a given symbol. */ 959 extern int mkstate(int); 960 961 extern void new_rule(void); /* initialize for a new rule */ 962 963 964 /* from file parse.y */ 965 966 /* Build the "<<EOF>>" action for the active start conditions. */ 967 extern void build_eof_action(void); 968 969 /* Write out a message formatted with one string, pinpointing its location. */ 970 extern void format_pinpoint_message(const char *, const char *); 971 972 /* Write out a message, pinpointing its location. */ 973 extern void pinpoint_message(const char *); 974 975 /* Write out a warning, pinpointing it at the given line. */ 976 extern void line_warning(const char *, int); 977 978 /* Write out a message, pinpointing it at the given line. */ 979 extern void line_pinpoint(const char *, int); 980 981 /* Report a formatted syntax error. */ 982 extern void format_synerr(const char *, const char *); 983 extern void synerr(const char *); /* report a syntax error */ 984 extern void format_warn(const char *, const char *); 985 extern void lwarn(const char *); /* report a warning */ 986 extern void yyerror(const char *); /* report a parse error */ 987 extern int yyparse(void); /* the YACC parser */ 988 989 990 /* from file scan.l */ 991 992 /* The Flex-generated scanner for flex. */ 993 extern int flexscan(void); 994 995 /* Open the given file (if NULL, stdin) for scanning. */ 996 extern void set_input_file(char *); 997 998 999 /* from file sym.c */ 1000 1001 /* Save the text of a character class. */ 1002 extern void cclinstal(char[], int); 1003 1004 /* Lookup the number associated with character class. */ 1005 extern int ccllookup(char[]); 1006 1007 extern void ndinstal(const char *, char[]); /* install a name definition */ 1008 extern char *ndlookup(const char *); /* lookup a name definition */ 1009 1010 /* Increase maximum number of SC's. */ 1011 extern void scextend(void); 1012 extern void scinstal(const char *, int); /* make a start condition */ 1013 1014 /* Lookup the number associated with a start condition. */ 1015 extern int sclookup(const char *); 1016 1017 1018 /* from file tblcmp.c */ 1019 1020 /* Build table entries for dfa state. */ 1021 extern void bldtbl(int[], int, int, int, int); 1022 1023 extern void cmptmps(void); /* compress template table entries */ 1024 extern void expand_nxt_chk(void); /* increase nxt/chk arrays */ 1025 1026 /* Finds a space in the table for a state to be placed. */ 1027 extern int find_table_space(int *, int); 1028 extern void inittbl(void); /* initialize transition tables */ 1029 1030 /* Make the default, "jam" table entries. */ 1031 extern void mkdeftbl(void); 1032 1033 /* Create table entries for a state (or state fragment) which has 1034 * only one out-transition. 1035 */ 1036 extern void mk1tbl(int, int, int, int); 1037 1038 /* Place a state into full speed transition table. */ 1039 extern void place_state(int *, int, int); 1040 1041 /* Save states with only one out-transition to be processed later. */ 1042 extern void stack1(int, int, int, int); 1043 1044 1045 /* from file yylex.c */ 1046 1047 extern int yylex(void); 1048 1049 /* A growable array. See buf.c. */ 1050 struct Buf { 1051 void *elts; /* elements. */ 1052 int nelts; /* number of elements. */ 1053 size_t elt_size; /* in bytes. */ 1054 int nmax; /* max capacity of elements. */ 1055 }; 1056 1057 extern void buf_init(struct Buf * buf, size_t elem_size); 1058 extern void buf_destroy(struct Buf * buf); 1059 extern struct Buf *buf_append(struct Buf * buf, const void *ptr, int n_elem); 1060 extern struct Buf *buf_concat(struct Buf* dest, const struct Buf* src); 1061 extern struct Buf *buf_strappend(struct Buf *, const char *str); 1062 extern struct Buf *buf_strnappend(struct Buf *, const char *str, int nchars); 1063 extern struct Buf *buf_strdefine(struct Buf * buf, const char *str, const char *def); 1064 extern struct Buf *buf_prints(struct Buf *buf, const char *fmt, const char* s); 1065 extern struct Buf *buf_m4_define(struct Buf *buf, const char* def, const char* val); 1066 extern struct Buf *buf_m4_undefine(struct Buf *buf, const char* def); 1067 extern struct Buf *buf_print_strings(struct Buf * buf, FILE* out); 1068 extern struct Buf *buf_linedir(struct Buf *buf, const char* filename, int lineno); 1069 1070 extern struct Buf userdef_buf; /* a string buffer for #define's generated by user-options on cmd line. */ 1071 extern struct Buf defs_buf; /* a char* buffer to save #define'd some symbols generated by flex. */ 1072 extern struct Buf yydmap_buf; /* a string buffer to hold yydmap elements */ 1073 extern struct Buf m4defs_buf; /* Holds m4 definitions. */ 1074 extern struct Buf top_buf; /* contains %top code. String buffer. */ 1075 extern bool no_section3_escape; /* True if the undocumented option --unsafe-no-m4-sect3-escape was passed */ 1076 1077 /* For blocking out code from the header file. */ 1078 #define OUT_BEGIN_CODE() outn("m4_ifdef( [[M4_YY_IN_HEADER]],,[[m4_dnl") 1079 #define OUT_END_CODE() outn("]])") 1080 1081 /* For setjmp/longjmp (instead of calling exit(2)). Linkage in main.c */ 1082 extern jmp_buf flex_main_jmp_buf; 1083 1084 #define FLEX_EXIT(status) longjmp(flex_main_jmp_buf,(status)+1) 1085 1086 /* Removes all \n and \r chars from tail of str. returns str. */ 1087 extern char *chomp (char *str); 1088 1089 /* ctype functions forced to return boolean */ 1090 #define b_isalnum(c) (isalnum(c)?true:false) 1091 #define b_isalpha(c) (isalpha(c)?true:false) 1092 #define b_isascii(c) (isascii(c)?true:false) 1093 #define b_isblank(c) (isblank(c)?true:false) 1094 #define b_iscntrl(c) (iscntrl(c)?true:false) 1095 #define b_isdigit(c) (isdigit(c)?true:false) 1096 #define b_isgraph(c) (isgraph(c)?true:false) 1097 #define b_islower(c) (islower(c)?true:false) 1098 #define b_isprint(c) (isprint(c)?true:false) 1099 #define b_ispunct(c) (ispunct(c)?true:false) 1100 #define b_isspace(c) (isspace(c)?true:false) 1101 #define b_isupper(c) (isupper(c)?true:false) 1102 #define b_isxdigit(c) (isxdigit(c)?true:false) 1103 1104 /* return true if char is uppercase or lowercase. */ 1105 bool has_case(int c); 1106 1107 /* Change case of character if possible. */ 1108 int reverse_case(int c); 1109 1110 /* return false if [c1-c2] is ambiguous for a caseless scanner. */ 1111 bool range_covers_case (int c1, int c2); 1112 1113 /* 1114 * From "filter.c" 1115 */ 1116 1117 /** A single stdio filter to execute. 1118 * The filter may be external, such as "sed", or it 1119 * may be internal, as a function call. 1120 */ 1121 struct filter { 1122 int (*filter_func)(struct filter*); /**< internal filter function */ 1123 void * extra; /**< extra data passed to filter_func */ 1124 int argc; /**< arg count */ 1125 const char ** argv; /**< arg vector, \0-terminated */ 1126 struct filter * next; /**< next filter or NULL */ 1127 }; 1128 1129 /* output filter chain */ 1130 extern struct filter * output_chain; 1131 extern struct filter *filter_create_ext (struct filter * chain, const char *cmd, ...); 1132 struct filter *filter_create_int(struct filter *chain, 1133 int (*filter_func) (struct filter *), 1134 void *extra); 1135 extern bool filter_apply_chain(struct filter * chain); 1136 extern int filter_truncate(struct filter * chain, int max_len); 1137 extern int filter_tee_header(struct filter *chain); 1138 extern int filter_fix_linedirs(struct filter *chain); 1139 1140 1141 /* 1142 * From "regex.c" 1143 */ 1144 1145 extern regex_t regex_linedir, regex_blank_line; 1146 bool flex_init_regex(void); 1147 void flex_regcomp(regex_t *preg, const char *regex, int cflags); 1148 char *regmatch_dup (regmatch_t * m, const char *src); 1149 char *regmatch_cpy (regmatch_t * m, char *dest, const char *src); 1150 int regmatch_len (regmatch_t * m); 1151 int regmatch_strtol (regmatch_t * m, const char *src, char **endptr, int base); 1152 bool regmatch_empty (regmatch_t * m); 1153 1154 /* From "scanflags.h" */ 1155 typedef unsigned int scanflags_t; 1156 extern scanflags_t* _sf_stk; 1157 extern size_t _sf_top_ix, _sf_max; /**< stack of scanner flags. */ 1158 #define _SF_CASE_INS ((scanflags_t) 0x0001) 1159 #define _SF_DOT_ALL ((scanflags_t) 0x0002) 1160 #define _SF_SKIP_WS ((scanflags_t) 0x0004) 1161 #define sf_top() (_sf_stk[_sf_top_ix]) 1162 #define sf_case_ins() (sf_top() & _SF_CASE_INS) 1163 #define sf_dot_all() (sf_top() & _SF_DOT_ALL) 1164 #define sf_skip_ws() (sf_top() & _SF_SKIP_WS) 1165 #define sf_set_case_ins(X) ((X) ? (sf_top() |= _SF_CASE_INS) : (sf_top() &= ~_SF_CASE_INS)) 1166 #define sf_set_dot_all(X) ((X) ? (sf_top() |= _SF_DOT_ALL) : (sf_top() &= ~_SF_DOT_ALL)) 1167 #define sf_set_skip_ws(X) ((X) ? (sf_top() |= _SF_SKIP_WS) : (sf_top() &= ~_SF_SKIP_WS)) 1168 extern void sf_init(void); 1169 extern void sf_push(void); 1170 extern void sf_pop(void); 1171 1172 #ifndef __RCSID 1173 #define __RCSID(a) 1174 #endif 1175 1176 #endif /* not defined FLEXDEF_H */ 1177