1 /* $OpenBSD: flexdef.h,v 1.17 2024/11/09 18:03:44 op 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 #ifdef HAVE_CONFIG_H 40 #include <config.h> 41 #endif 42 43 #ifdef STDC_HEADERS 44 #include <stdio.h> 45 #include <stdlib.h> 46 #include <stdarg.h> 47 #include <setjmp.h> 48 #include <ctype.h> 49 #include <string.h> 50 #include <math.h> 51 #endif 52 #ifdef HAVE_ASSERT_H 53 #include <assert.h> 54 #else 55 #define assert(Pred) 56 #endif 57 58 #ifdef HAVE_LIMITS_H 59 #include <limits.h> 60 #endif 61 #ifdef HAVE_UNISTD_H 62 #include <unistd.h> 63 #endif 64 #ifdef HAVE_NETINET_IN_H 65 #include <netinet/in.h> 66 #endif 67 #ifdef HAVE_SYS_PARAMS_H 68 #include <sys/params.h> 69 #endif 70 #ifdef HAVE_SYS_WAIT_H 71 #include <sys/wait.h> 72 #endif 73 #ifdef HAVE_STDBOOL_H 74 #include <stdbool.h> 75 #else 76 #define bool int 77 #define true 1 78 #define false 0 79 #endif 80 #ifdef HAVE_REGEX_H 81 #include <regex.h> 82 #endif 83 #include "flexint.h" 84 85 #define _(STRING) STRING 86 87 /* Always be prepared to generate an 8-bit scanner. */ 88 #define CSIZE 256 89 90 /* Size of input alphabet - should be size of ASCII set. */ 91 #ifndef DEFAULT_CSIZE 92 #define DEFAULT_CSIZE 128 93 #endif 94 95 #ifndef PROTO 96 #if defined(__STDC__) 97 #define PROTO(proto) proto 98 #else 99 #define PROTO(proto) () 100 #endif 101 #endif 102 103 /* Maximum line length we'll have to deal with. */ 104 #define MAXLINE 2048 105 106 #ifndef MIN 107 #define MIN(x,y) ((x) < (y) ? (x) : (y)) 108 #endif 109 #ifndef MAX 110 #define MAX(x,y) ((x) > (y) ? (x) : (y)) 111 #endif 112 #ifndef ABS 113 #define ABS(x) ((x) < 0 ? -(x) : (x)) 114 #endif 115 116 117 /* ANSI C does not guarantee that isascii() is defined */ 118 #ifndef isascii 119 #define isascii(c) ((c) <= 0177) 120 #endif 121 122 #define unspecified -1 123 124 /* Special chk[] values marking the slots taking by end-of-buffer and action 125 * numbers. 126 */ 127 #define EOB_POSITION -1 128 #define ACTION_POSITION -2 129 130 /* Number of data items per line for -f output. */ 131 #define NUMDATAITEMS 10 132 133 /* Number of lines of data in -f output before inserting a blank line for 134 * readability. 135 */ 136 #define NUMDATALINES 10 137 138 /* transition_struct_out() definitions. */ 139 #define TRANS_STRUCT_PRINT_LENGTH 14 140 141 /* Returns true if an nfa state has an epsilon out-transition slot 142 * that can be used. This definition is currently not used. 143 */ 144 #define FREE_EPSILON(state) \ 145 (transchar[state] == SYM_EPSILON && \ 146 trans2[state] == NO_TRANSITION && \ 147 finalst[state] != state) 148 149 /* Returns true if an nfa state has an epsilon out-transition character 150 * and both slots are free 151 */ 152 #define SUPER_FREE_EPSILON(state) \ 153 (transchar[state] == SYM_EPSILON && \ 154 trans1[state] == NO_TRANSITION) \ 155 156 /* Maximum number of NFA states that can comprise a DFA state. It's real 157 * big because if there's a lot of rules, the initial state will have a 158 * huge epsilon closure. 159 */ 160 #define INITIAL_MAX_DFA_SIZE 750 161 #define MAX_DFA_SIZE_INCREMENT 750 162 163 164 /* A note on the following masks. They are used to mark accepting numbers 165 * as being special. As such, they implicitly limit the number of accepting 166 * numbers (i.e., rules) because if there are too many rules the rule numbers 167 * will overload the mask bits. Fortunately, this limit is \large/ (0x2000 == 168 * 8192) so unlikely to actually cause any problems. A check is made in 169 * new_rule() to ensure that this limit is not reached. 170 */ 171 172 /* Mask to mark a trailing context accepting number. */ 173 #define YY_TRAILING_MASK 0x2000 174 175 /* Mask to mark the accepting number of the "head" of a trailing context 176 * rule. 177 */ 178 #define YY_TRAILING_HEAD_MASK 0x4000 179 180 /* Maximum number of rules, as outlined in the above note. */ 181 #define MAX_RULE (YY_TRAILING_MASK - 1) 182 183 184 /* NIL must be 0. If not, its special meaning when making equivalence classes 185 * (it marks the representative of a given e.c.) will be unidentifiable. 186 */ 187 #define NIL 0 188 189 #define JAM -1 /* to mark a missing DFA transition */ 190 #define NO_TRANSITION NIL 191 #define UNIQUE -1 /* marks a symbol as an e.c. representative */ 192 #define INFINITE_REPEAT -1 /* for x{5,} constructions */ 193 194 #define INITIAL_MAX_CCLS 100 /* max number of unique character classes */ 195 #define MAX_CCLS_INCREMENT 100 196 197 /* Size of table holding members of character classes. */ 198 #define INITIAL_MAX_CCL_TBL_SIZE 500 199 #define MAX_CCL_TBL_SIZE_INCREMENT 250 200 201 #define INITIAL_MAX_RULES 100 /* default maximum number of rules */ 202 #define MAX_RULES_INCREMENT 100 203 204 #define INITIAL_MNS 2000 /* default maximum number of nfa states */ 205 #define MNS_INCREMENT 1000 /* amount to bump above by if it's not enough */ 206 207 #define INITIAL_MAX_DFAS 1000 /* default maximum number of dfa states */ 208 #define MAX_DFAS_INCREMENT 1000 209 210 #define JAMSTATE -32766 /* marks a reference to the state that always jams */ 211 212 /* Maximum number of NFA states. */ 213 #define MAXIMUM_MNS 31999 214 #define MAXIMUM_MNS_LONG 1999999999 215 216 /* Enough so that if it's subtracted from an NFA state number, the result 217 * is guaranteed to be negative. 218 */ 219 #define MARKER_DIFFERENCE (maximum_mns+2) 220 221 /* Maximum number of nxt/chk pairs for non-templates. */ 222 #define INITIAL_MAX_XPAIRS 2000 223 #define MAX_XPAIRS_INCREMENT 2000 224 225 /* Maximum number of nxt/chk pairs needed for templates. */ 226 #define INITIAL_MAX_TEMPLATE_XPAIRS 2500 227 #define MAX_TEMPLATE_XPAIRS_INCREMENT 2500 228 229 #define SYM_EPSILON (CSIZE + 1) /* to mark transitions on the symbol epsilon */ 230 231 #define INITIAL_MAX_SCS 40 /* maximum number of start conditions */ 232 #define MAX_SCS_INCREMENT 40 /* amount to bump by if it's not enough */ 233 234 #define ONE_STACK_SIZE 500 /* stack of states with only one out-transition */ 235 #define SAME_TRANS -1 /* transition is the same as "default" entry for state */ 236 237 /* The following percentages are used to tune table compression: 238 239 * The percentage the number of out-transitions a state must be of the 240 * number of equivalence classes in order to be considered for table 241 * compaction by using protos. 242 */ 243 #define PROTO_SIZE_PERCENTAGE 15 244 245 /* The percentage the number of homogeneous out-transitions of a state 246 * must be of the number of total out-transitions of the state in order 247 * that the state's transition table is first compared with a potential 248 * template of the most common out-transition instead of with the first 249 * proto in the proto queue. 250 */ 251 #define CHECK_COM_PERCENTAGE 50 252 253 /* The percentage the number of differences between a state's transition 254 * table and the proto it was first compared with must be of the total 255 * number of out-transitions of the state in order to keep the first 256 * proto as a good match and not search any further. 257 */ 258 #define FIRST_MATCH_DIFF_PERCENTAGE 10 259 260 /* The percentage the number of differences between a state's transition 261 * table and the most similar proto must be of the state's total number 262 * of out-transitions to use the proto as an acceptable close match. 263 */ 264 #define ACCEPTABLE_DIFF_PERCENTAGE 50 265 266 /* The percentage the number of homogeneous out-transitions of a state 267 * must be of the number of total out-transitions of the state in order 268 * to consider making a template from the state. 269 */ 270 #define TEMPLATE_SAME_PERCENTAGE 60 271 272 /* The percentage the number of differences between a state's transition 273 * table and the most similar proto must be of the state's total number 274 * of out-transitions to create a new proto from the state. 275 */ 276 #define NEW_PROTO_DIFF_PERCENTAGE 20 277 278 /* The percentage the total number of out-transitions of a state must be 279 * of the number of equivalence classes in order to consider trying to 280 * fit the transition table into "holes" inside the nxt/chk table. 281 */ 282 #define INTERIOR_FIT_PERCENTAGE 15 283 284 /* Size of region set aside to cache the complete transition table of 285 * protos on the proto queue to enable quick comparisons. 286 */ 287 #define PROT_SAVE_SIZE 2000 288 289 #define MSP 50 /* maximum number of saved protos (protos on the proto queue) */ 290 291 /* Maximum number of out-transitions a state can have that we'll rummage 292 * around through the interior of the internal fast table looking for a 293 * spot for it. 294 */ 295 #define MAX_XTIONS_FULL_INTERIOR_FIT 4 296 297 /* Maximum number of rules which will be reported as being associated 298 * with a DFA state. 299 */ 300 #define MAX_ASSOC_RULES 100 301 302 /* Number that, if used to subscript an array, has a good chance of producing 303 * an error; should be small enough to fit into a short. 304 */ 305 #define BAD_SUBSCRIPT -32767 306 307 /* Absolute value of largest number that can be stored in a short, with a 308 * bit of slop thrown in for general paranoia. 309 */ 310 #define MAX_SHORT 32700 311 312 313 /* Declarations for global variables. */ 314 315 316 /* Variables for flags: 317 * printstats - if true (-v), dump statistics 318 * syntaxerror - true if a syntax error has been found 319 * eofseen - true if we've seen an eof in the input file 320 * ddebug - if true (-d), make a "debug" scanner 321 * trace - if true (-T), trace processing 322 * nowarn - if true (-w), do not generate warnings 323 * spprdflt - if true (-s), suppress the default rule 324 * interactive - if true (-I), generate an interactive scanner 325 * lex_compat - if true (-l), maximize compatibility with AT&T lex 326 * posix_compat - if true (-X), maximize compatibility with POSIX lex 327 * do_yylineno - if true, generate code to maintain yylineno 328 * useecs - if true (-Ce flag), use equivalence classes 329 * fulltbl - if true (-Cf flag), don't compress the DFA state table 330 * usemecs - if true (-Cm flag), use meta-equivalence classes 331 * fullspd - if true (-F flag), use Jacobson method of table representation 332 * gen_line_dirs - if true (i.e., no -L flag), generate #line directives 333 * performance_report - if > 0 (i.e., -p flag), generate a report relating 334 * to scanner performance; if > 1 (-p -p), report on minor performance 335 * problems, too 336 * backing_up_report - if true (i.e., -b flag), generate "lex.backup" file 337 * listing backing-up states 338 * C_plus_plus - if true (i.e., -+ flag), generate a C++ scanner class; 339 * otherwise, a standard C scanner 340 * reentrant - if true (-R), generate a reentrant C scanner. 341 * bison_bridge_lval - if true (--bison-bridge), bison pure calling convention. 342 * bison_bridge_lloc - if true (--bison-locations), bison yylloc. 343 * long_align - if true (-Ca flag), favor long-word alignment. 344 * use_read - if true (-f, -F, or -Cr) then use read() for scanner input; 345 * otherwise, use fread(). 346 * yytext_is_array - if true (i.e., %array directive), then declare 347 * yytext as a array instead of a character pointer. Nice and inefficient. 348 * do_yywrap - do yywrap() processing on EOF. If false, EOF treated as 349 * "no more files". 350 * csize - size of character set for the scanner we're generating; 351 * 128 for 7-bit chars and 256 for 8-bit 352 * yymore_used - if true, yymore() is used in input rules 353 * reject - if true, generate back-up tables for REJECT macro 354 * real_reject - if true, scanner really uses REJECT (as opposed to just 355 * having "reject" set for variable trailing context) 356 * continued_action - true if this rule's action is to "fall through" to 357 * the next rule's action (i.e., the '|' action) 358 * in_rule - true if we're inside an individual rule, false if not. 359 * yymore_really_used - whether to treat yymore() as really used, regardless 360 * of what we think based on references to it in the user's actions. 361 * reject_really_used - same for REJECT 362 */ 363 364 extern int printstats, syntaxerror, eofseen, ddebug, trace, nowarn, 365 spprdflt; 366 extern int interactive, lex_compat, posix_compat, do_yylineno; 367 extern int useecs, fulltbl, usemecs, fullspd; 368 extern int gen_line_dirs, performance_report, backing_up_report; 369 extern int reentrant, bison_bridge_lval, bison_bridge_lloc; 370 extern bool ansi_func_defs, ansi_func_protos; 371 extern int C_plus_plus, long_align, use_read, yytext_is_array, do_yywrap; 372 extern int csize; 373 extern int yymore_used, reject, real_reject, continued_action, in_rule; 374 375 extern int yymore_really_used, reject_really_used; 376 377 378 /* Variables used in the flex input routines: 379 * datapos - characters on current output line 380 * dataline - number of contiguous lines of data in current data 381 * statement. Used to generate readable -f output 382 * linenum - current input line number 383 * skelfile - the skeleton file 384 * skel - compiled-in skeleton array 385 * skel_ind - index into "skel" array, if skelfile is nil 386 * yyin - input file 387 * backing_up_file - file to summarize backing-up states to 388 * infilename - name of input file 389 * outfilename - name of output file 390 * headerfilename - name of the .h file to generate 391 * did_outfilename - whether outfilename was explicitly set 392 * prefix - the prefix used for externally visible names ("yy" by default) 393 * yyclass - yyFlexLexer subclass to use for YY_DECL 394 * do_stdinit - whether to initialize yyin/yyout to stdin/stdout 395 * use_stdout - the -t flag 396 * input_files - array holding names of input files 397 * num_input_files - size of input_files array 398 * program_name - name with which program was invoked 399 * 400 * action_array - array to hold the rule actions 401 * action_size - size of action_array 402 * defs1_offset - index where the user's section 1 definitions start 403 * in action_array 404 * prolog_offset - index where the prolog starts in action_array 405 * action_offset - index where the non-prolog starts in action_array 406 * action_index - index where the next action should go, with respect 407 * to "action_array" 408 */ 409 410 extern int datapos, dataline, linenum; 411 extern FILE *skelfile, *yyin, *backing_up_file; 412 extern const char *skel[]; 413 extern int skel_ind; 414 extern char *infilename, *outfilename, *headerfilename; 415 extern int did_outfilename; 416 extern char *prefix, *yyclass, *extra_type; 417 extern int do_stdinit, use_stdout; 418 extern char **input_files; 419 extern int num_input_files; 420 extern char *program_name; 421 422 extern char *action_array; 423 extern int action_size; 424 extern int defs1_offset, prolog_offset, action_offset, action_index; 425 426 427 /* Variables for stack of states having only one out-transition: 428 * onestate - state number 429 * onesym - transition symbol 430 * onenext - target state 431 * onedef - default base entry 432 * onesp - stack pointer 433 */ 434 435 extern int onestate[ONE_STACK_SIZE], onesym[ONE_STACK_SIZE]; 436 extern int onenext[ONE_STACK_SIZE], onedef[ONE_STACK_SIZE], onesp; 437 438 439 /* Variables for nfa machine data: 440 * maximum_mns - maximal number of NFA states supported by tables 441 * current_mns - current maximum on number of NFA states 442 * num_rules - number of the last accepting state; also is number of 443 * rules created so far 444 * num_eof_rules - number of <<EOF>> rules 445 * default_rule - number of the default rule 446 * current_max_rules - current maximum number of rules 447 * lastnfa - last nfa state number created 448 * firstst - physically the first state of a fragment 449 * lastst - last physical state of fragment 450 * finalst - last logical state of fragment 451 * transchar - transition character 452 * trans1 - transition state 453 * trans2 - 2nd transition state for epsilons 454 * accptnum - accepting number 455 * assoc_rule - rule associated with this NFA state (or 0 if none) 456 * state_type - a STATE_xxx type identifying whether the state is part 457 * of a normal rule, the leading state in a trailing context 458 * rule (i.e., the state which marks the transition from 459 * recognizing the text-to-be-matched to the beginning of 460 * the trailing context), or a subsequent state in a trailing 461 * context rule 462 * rule_type - a RULE_xxx type identifying whether this a ho-hum 463 * normal rule or one which has variable head & trailing 464 * context 465 * rule_linenum - line number associated with rule 466 * rule_useful - true if we've determined that the rule can be matched 467 * rule_has_nl - true if rule could possibly match a newline 468 * ccl_has_nl - true if current ccl could match a newline 469 * nlch - default eol char 470 */ 471 472 extern int maximum_mns, current_mns, current_max_rules; 473 extern int num_rules, num_eof_rules, default_rule, lastnfa; 474 extern int *firstst, *lastst, *finalst, *transchar, *trans1, *trans2; 475 extern int *accptnum, *assoc_rule, *state_type; 476 extern int *rule_type, *rule_linenum, *rule_useful; 477 extern bool *rule_has_nl, *ccl_has_nl; 478 extern int nlch; 479 480 /* Different types of states; values are useful as masks, as well, for 481 * routines like check_trailing_context(). 482 */ 483 #define STATE_NORMAL 0x1 484 #define STATE_TRAILING_CONTEXT 0x2 485 486 /* Global holding current type of state we're making. */ 487 488 extern int current_state_type; 489 490 /* Different types of rules. */ 491 #define RULE_NORMAL 0 492 #define RULE_VARIABLE 1 493 494 /* True if the input rules include a rule with both variable-length head 495 * and trailing context, false otherwise. 496 */ 497 extern int variable_trailing_context_rules; 498 499 500 /* Variables for protos: 501 * numtemps - number of templates created 502 * numprots - number of protos created 503 * protprev - backlink to a more-recently used proto 504 * protnext - forward link to a less-recently used proto 505 * prottbl - base/def table entry for proto 506 * protcomst - common state of proto 507 * firstprot - number of the most recently used proto 508 * lastprot - number of the least recently used proto 509 * protsave contains the entire state array for protos 510 */ 511 512 extern int numtemps, numprots, protprev[MSP], protnext[MSP], prottbl[MSP]; 513 extern int protcomst[MSP], firstprot, lastprot, protsave[PROT_SAVE_SIZE]; 514 515 516 /* Variables for managing equivalence classes: 517 * numecs - number of equivalence classes 518 * nextecm - forward link of Equivalence Class members 519 * ecgroup - class number or backward link of EC members 520 * nummecs - number of meta-equivalence classes (used to compress 521 * templates) 522 * tecfwd - forward link of meta-equivalence classes members 523 * tecbck - backward link of MEC's 524 */ 525 526 /* Reserve enough room in the equivalence class arrays so that we 527 * can use the CSIZE'th element to hold equivalence class information 528 * for the NUL character. Later we'll move this information into 529 * the 0th element. 530 */ 531 extern int numecs, nextecm[CSIZE + 1], ecgroup[CSIZE + 1], nummecs; 532 533 /* Meta-equivalence classes are indexed starting at 1, so it's possible 534 * that they will require positions from 1 .. CSIZE, i.e., CSIZE + 1 535 * slots total (since the arrays are 0-based). nextecm[] and ecgroup[] 536 * don't require the extra position since they're indexed from 1 .. CSIZE - 1. 537 */ 538 extern int tecfwd[CSIZE + 1], tecbck[CSIZE + 1]; 539 540 541 /* Variables for start conditions: 542 * lastsc - last start condition created 543 * current_max_scs - current limit on number of start conditions 544 * scset - set of rules active in start condition 545 * scbol - set of rules active only at the beginning of line in a s.c. 546 * scxclu - true if start condition is exclusive 547 * sceof - true if start condition has EOF rule 548 * scname - start condition name 549 */ 550 551 extern int lastsc, *scset, *scbol, *scxclu, *sceof; 552 extern int current_max_scs; 553 extern char **scname; 554 555 556 /* Variables for dfa machine data: 557 * current_max_dfa_size - current maximum number of NFA states in DFA 558 * current_max_xpairs - current maximum number of non-template xtion pairs 559 * current_max_template_xpairs - current maximum number of template pairs 560 * current_max_dfas - current maximum number DFA states 561 * lastdfa - last dfa state number created 562 * nxt - state to enter upon reading character 563 * chk - check value to see if "nxt" applies 564 * tnxt - internal nxt table for templates 565 * base - offset into "nxt" for given state 566 * def - where to go if "chk" disallows "nxt" entry 567 * nultrans - NUL transition for each state 568 * NUL_ec - equivalence class of the NUL character 569 * tblend - last "nxt/chk" table entry being used 570 * firstfree - first empty entry in "nxt/chk" table 571 * dss - nfa state set for each dfa 572 * dfasiz - size of nfa state set for each dfa 573 * dfaacc - accepting set for each dfa state (if using REJECT), or accepting 574 * number, if not 575 * accsiz - size of accepting set for each dfa state 576 * dhash - dfa state hash value 577 * numas - number of DFA accepting states created; note that this 578 * is not necessarily the same value as num_rules, which is the analogous 579 * value for the NFA 580 * numsnpairs - number of state/nextstate transition pairs 581 * jambase - position in base/def where the default jam table starts 582 * jamstate - state number corresponding to "jam" state 583 * end_of_buffer_state - end-of-buffer dfa state number 584 */ 585 586 extern int current_max_dfa_size, current_max_xpairs; 587 extern int current_max_template_xpairs, current_max_dfas; 588 extern int lastdfa, *nxt, *chk, *tnxt; 589 extern int *base, *def, *nultrans, NUL_ec, tblend, firstfree, **dss, 590 *dfasiz; 591 extern union dfaacc_union { 592 int *dfaacc_set; 593 int dfaacc_state; 594 } *dfaacc; 595 extern int *accsiz, *dhash, numas; 596 extern int numsnpairs, jambase, jamstate; 597 extern int end_of_buffer_state; 598 599 /* Variables for ccl information: 600 * lastccl - ccl index of the last created ccl 601 * current_maxccls - current limit on the maximum number of unique ccl's 602 * cclmap - maps a ccl index to its set pointer 603 * ccllen - gives the length of a ccl 604 * cclng - true for a given ccl if the ccl is negated 605 * cclreuse - counts how many times a ccl is re-used 606 * current_max_ccl_tbl_size - current limit on number of characters needed 607 * to represent the unique ccl's 608 * ccltbl - holds the characters in each ccl - indexed by cclmap 609 */ 610 611 extern int lastccl, *cclmap, *ccllen, *cclng, cclreuse; 612 extern int current_maxccls, current_max_ccl_tbl_size; 613 extern u_char *ccltbl; 614 615 616 /* Variables for miscellaneous information: 617 * nmstr - last NAME scanned by the scanner 618 * sectnum - section number currently being parsed 619 * nummt - number of empty nxt/chk table entries 620 * hshcol - number of hash collisions detected by snstods 621 * dfaeql - number of times a newly created dfa was equal to an old one 622 * numeps - number of epsilon NFA states created 623 * eps2 - number of epsilon states which have 2 out-transitions 624 * num_reallocs - number of times it was necessary to realloc() a group 625 * of arrays 626 * tmpuses - number of DFA states that chain to templates 627 * totnst - total number of NFA states used to make DFA states 628 * peakpairs - peak number of transition pairs we had to store internally 629 * numuniq - number of unique transitions 630 * numdup - number of duplicate transitions 631 * hshsave - number of hash collisions saved by checking number of states 632 * num_backing_up - number of DFA states requiring backing up 633 * bol_needed - whether scanner needs beginning-of-line recognition 634 */ 635 636 extern char nmstr[MAXLINE]; 637 extern int sectnum, nummt, hshcol, dfaeql, numeps, eps2, num_reallocs; 638 extern int tmpuses, totnst, peakpairs, numuniq, numdup, hshsave; 639 extern int num_backing_up, bol_needed; 640 641 void *allocate_array PROTO ((int, size_t)); 642 void *reallocate_array PROTO ((void *, int, size_t)); 643 644 #define allocate_integer_array(size) \ 645 (int *) allocate_array( size, sizeof( int ) ) 646 647 #define reallocate_integer_array(array,size) \ 648 (int *) reallocate_array( (void *) array, size, sizeof( int ) ) 649 650 #define allocate_bool_array(size) \ 651 (bool *) allocate_array( size, sizeof( bool ) ) 652 653 #define reallocate_bool_array(array,size) \ 654 (bool *) reallocate_array( (void *) array, size, sizeof( bool ) ) 655 656 #define allocate_int_ptr_array(size) \ 657 (int **) allocate_array( size, sizeof( int * ) ) 658 659 #define allocate_char_ptr_array(size) \ 660 (char **) allocate_array( size, sizeof( char * ) ) 661 662 #define allocate_dfaacc_union(size) \ 663 (union dfaacc_union *) \ 664 allocate_array( size, sizeof( union dfaacc_union ) ) 665 666 #define reallocate_int_ptr_array(array,size) \ 667 (int **) reallocate_array( (void *) array, size, sizeof( int * ) ) 668 669 #define reallocate_char_ptr_array(array,size) \ 670 (char **) reallocate_array( (void *) array, size, sizeof( char * ) ) 671 672 #define reallocate_dfaacc_union(array, size) \ 673 (union dfaacc_union *) \ 674 reallocate_array( (void *) array, size, sizeof( union dfaacc_union ) ) 675 676 #define allocate_character_array(size) \ 677 (char *) allocate_array( size, sizeof( char ) ) 678 679 #define reallocate_character_array(array,size) \ 680 (char *) reallocate_array( (void *) array, size, sizeof( char ) ) 681 682 #define allocate_Character_array(size) \ 683 (u_char *) allocate_array( size, sizeof( u_char ) ) 684 685 #define reallocate_Character_array(array,size) \ 686 (u_char *) reallocate_array( (void *) array, size, sizeof( u_char ) ) 687 688 689 /* Used to communicate between scanner and parser. The type should really 690 * be YYSTYPE, but we can't easily get our hands on it. 691 */ 692 extern int yylval; 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 PROTO ((int, int)); /* add a single character to a ccl */ 701 extern int cclinit PROTO ((void)); /* make an empty ccl */ 702 extern void cclnegate PROTO ((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 PROTO ((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 PROTO ((int, int[])); 714 715 /* Check to see if NFA state set constitutes "dangerous" trailing context. */ 716 extern void check_trailing_context PROTO ((int *, int, int *, int)); 717 718 /* Construct the epsilon closure of a set of ndfa states. */ 719 extern int *epsclosure PROTO ((int *, int *, int[], int *, int *)); 720 721 /* Increase the maximum number of dfas. */ 722 extern void increase_max_dfas PROTO ((void)); 723 724 extern void ntod PROTO ((void)); /* convert a ndfa to a dfa */ 725 726 /* Converts a set of ndfa states into a dfa state. */ 727 extern int snstods PROTO ((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 PROTO ((void)); 734 735 /* Associate equivalence class numbers with class members. */ 736 extern int cre8ecs PROTO ((int[], int[], int)); 737 738 /* Update equivalence classes based on character class transitions. */ 739 extern void mkeccl PROTO ((u_char[], int, int[], int[], int, int)); 740 741 /* Create equivalence class for single character. */ 742 extern void mkechar PROTO ((int, int[], int[])); 743 744 745 /* from file gen.c */ 746 747 extern void do_indent PROTO ((void)); /* indent to the current level */ 748 749 /* Generate the code to keep backing-up information. */ 750 extern void gen_backing_up PROTO ((void)); 751 752 /* Generate the code to perform the backing up. */ 753 extern void gen_bu_action PROTO ((void)); 754 755 /* Generate full speed compressed transition table. */ 756 extern void genctbl PROTO ((void)); 757 758 /* Generate the code to find the action number. */ 759 extern void gen_find_action PROTO ((void)); 760 761 extern void genftbl PROTO ((void)); /* generate full transition table */ 762 763 /* Generate the code to find the next compressed-table state. */ 764 extern void gen_next_compressed_state PROTO ((char *)); 765 766 /* Generate the code to find the next match. */ 767 extern void gen_next_match PROTO ((void)); 768 769 /* Generate the code to find the next state. */ 770 extern void gen_next_state PROTO ((int)); 771 772 /* Generate the code to make a NUL transition. */ 773 extern void gen_NUL_trans PROTO ((void)); 774 775 /* Generate the code to find the start state. */ 776 extern void gen_start_state PROTO ((void)); 777 778 /* Generate data statements for the transition tables. */ 779 extern void gentabs PROTO ((void)); 780 781 /* Write out a formatted string at the current indentation level. */ 782 extern void indent_put2s PROTO ((const char *, const char *)); 783 784 /* Write out a string + newline at the current indentation level. */ 785 extern void indent_puts PROTO ((const char *)); 786 787 extern void make_tables PROTO ((void)); /* generate transition tables */ 788 789 790 /* from file main.c */ 791 792 extern void check_options PROTO ((void)); 793 extern void flexend PROTO ((int)); 794 extern void usage PROTO ((void)); 795 796 797 /* from file misc.c */ 798 799 /* Add a #define to the action file. */ 800 extern void action_define PROTO ((const char *defname, int value)); 801 802 /* Add the given text to the stored actions. */ 803 extern void add_action PROTO ((const char *new_text)); 804 805 /* True if a string is all lower case. */ 806 extern int all_lower PROTO ((char *)); 807 808 /* True if a string is all upper case. */ 809 extern int all_upper PROTO ((char *)); 810 811 /* Compare two integers for use by qsort. */ 812 extern int intcmp PROTO ((const void *, const void *)); 813 814 /* Check a character to make sure it's in the expected range. */ 815 extern void check_char PROTO ((int c)); 816 817 /* Replace upper-case letter to lower-case. */ 818 extern u_char clower PROTO ((int)); 819 820 /* Returns a dynamically allocated copy of a string. */ 821 extern char *copy_string PROTO ((const char *)); 822 823 /* Returns a dynamically allocated copy of a (potentially) unsigned string. */ 824 extern u_char *copy_unsigned_string PROTO ((u_char *)); 825 826 /* Compare two characters for use by qsort with '\0' sorting last. */ 827 extern int cclcmp PROTO ((const void *, const void *)); 828 829 /* Finish up a block of data declarations. */ 830 extern void dataend PROTO ((void)); 831 832 /* Flush generated data statements. */ 833 extern void dataflush PROTO ((void)); 834 835 /* Report an error message and terminate. */ 836 extern void flexerror PROTO ((const char *)); 837 838 /* Report a fatal error message and terminate. */ 839 extern void flexfatal PROTO ((const char *)); 840 841 /* Report a fatal error with a pinpoint, and terminate */ 842 #if HAVE_DECL___FUNC__ 843 #define flex_die(msg) \ 844 do{ \ 845 fprintf (stderr,\ 846 _("%s: fatal internal error at %s:%d (%s): %s\n"),\ 847 program_name, __FILE__, (int)__LINE__,\ 848 __func__,msg);\ 849 FLEX_EXIT(1);\ 850 }while(0) 851 #else /* ! HAVE_DECL___FUNC__ */ 852 #define flex_die(msg) \ 853 do{ \ 854 fprintf (stderr,\ 855 _("%s: fatal internal error at %s:%d %s\n"),\ 856 program_name, __FILE__, (int)__LINE__,\ 857 msg);\ 858 FLEX_EXIT(1);\ 859 }while(0) 860 #endif /* ! HAVE_DECL___func__ */ 861 862 /* Convert a hexadecimal digit string to an integer value. */ 863 extern int htoi PROTO ((u_char[])); 864 865 /* Report an error message formatted with one integer argument. */ 866 extern void lerrif PROTO ((const char *, int)); 867 868 /* Report an error message formatted with one string argument. */ 869 extern void lerrsf PROTO ((const char *, const char *)); 870 871 /* Like lerrsf, but also exit after displaying message. */ 872 extern void lerrsf_fatal PROTO ((const char *, const char *)); 873 874 /* Spit out a "#line" statement. */ 875 extern void line_directive_out PROTO ((FILE *, int)); 876 877 /* Mark the current position in the action array as the end of the section 1 878 * user defs. 879 */ 880 extern void mark_defs1 PROTO ((void)); 881 882 /* Mark the current position in the action array as the end of the prolog. */ 883 extern void mark_prolog PROTO ((void)); 884 885 /* Generate a data statement for a two-dimensional array. */ 886 extern void mk2data PROTO ((int)); 887 888 extern void mkdata PROTO ((int)); /* generate a data statement */ 889 890 /* Return the integer represented by a string of digits. */ 891 extern int myctoi PROTO ((const char *)); 892 893 /* Return character corresponding to escape sequence. */ 894 extern u_char myesc PROTO ((u_char[])); 895 896 /* Convert an octal digit string to an integer value. */ 897 extern int otoi PROTO ((u_char[])); 898 899 /* Output a (possibly-formatted) string to the generated scanner. */ 900 extern void out PROTO ((const char *)); 901 extern void out_dec PROTO ((const char *, int)); 902 extern void out_dec2 PROTO ((const char *, int, int)); 903 extern void out_hex PROTO ((const char *, unsigned int)); 904 extern void out_str PROTO ((const char *, const char *)); 905 extern void out_str3 906 PROTO ((const char *, const char *, const char *, const char *)); 907 extern void out_str_dec PROTO ((const char *, const char *, int)); 908 extern void outc PROTO ((int)); 909 extern void outn PROTO ((const char *)); 910 extern void out_m4_define (const char* def, const char* val); 911 912 /* Return a printable version of the given character, which might be 913 * 8-bit. 914 */ 915 extern char *readable_form PROTO ((int)); 916 917 /* Write out one section of the skeleton file. */ 918 extern void skelout PROTO ((void)); 919 920 /* Output a yy_trans_info structure. */ 921 extern void transition_struct_out PROTO ((int, int)); 922 923 /* Only needed when using certain broken versions of bison to build parse.c. */ 924 extern void *yy_flex_xmalloc PROTO ((int)); 925 926 /* from file nfa.c */ 927 928 /* Add an accepting state to a machine. */ 929 extern void add_accept PROTO ((int, int)); 930 931 /* Make a given number of copies of a singleton machine. */ 932 extern int copysingl PROTO ((int, int)); 933 934 /* Debugging routine to write out an nfa. */ 935 extern void dumpnfa PROTO ((int)); 936 937 /* Finish up the processing for a rule. */ 938 extern void finish_rule PROTO ((int, int, int, int, int)); 939 940 /* Connect two machines together. */ 941 extern int link_machines PROTO ((int, int)); 942 943 /* Mark each "beginning" state in a machine as being a "normal" (i.e., 944 * not trailing context associated) state. 945 */ 946 extern void mark_beginning_as_normal PROTO ((int)); 947 948 /* Make a machine that branches to two machines. */ 949 extern int mkbranch PROTO ((int, int)); 950 951 extern int mkclos PROTO ((int)); /* convert a machine into a closure */ 952 extern int mkopt PROTO ((int)); /* make a machine optional */ 953 954 /* Make a machine that matches either one of two machines. */ 955 extern int mkor PROTO ((int, int)); 956 957 /* Convert a machine into a positive closure. */ 958 extern int mkposcl PROTO ((int)); 959 960 extern int mkrep PROTO ((int, int, int)); /* make a replicated machine */ 961 962 /* Create a state with a transition on a given symbol. */ 963 extern int mkstate PROTO ((int)); 964 965 extern void new_rule PROTO ((void)); /* initialize for a new rule */ 966 967 968 /* from file parse.y */ 969 970 /* Build the "<<EOF>>" action for the active start conditions. */ 971 extern void build_eof_action PROTO ((void)); 972 973 /* Write out a message formatted with one string, pinpointing its location. */ 974 extern void format_pinpoint_message PROTO ((const char *, const char *)); 975 976 /* Write out a message, pinpointing its location. */ 977 extern void pinpoint_message PROTO ((const char *)); 978 979 /* Write out a warning, pinpointing it at the given line. */ 980 extern void line_warning PROTO ((const char *, int)); 981 982 /* Write out a message, pinpointing it at the given line. */ 983 extern void line_pinpoint PROTO ((const char *, int)); 984 985 /* Report a formatted syntax error. */ 986 extern void format_synerr PROTO ((const char *, const char *)); 987 extern void synerr PROTO ((const char *)); /* report a syntax error */ 988 extern void format_warn PROTO ((const char *, const char *)); 989 extern void warn PROTO ((const char *)); /* report a warning */ 990 extern void yyerror PROTO ((const char *)); /* report a parse error */ 991 extern int yyparse PROTO ((void)); /* the YACC parser */ 992 993 994 /* from file scan.l */ 995 996 /* The Flex-generated scanner for flex. */ 997 extern int flexscan PROTO ((void)); 998 999 /* Open the given file (if NULL, stdin) for scanning. */ 1000 extern void set_input_file PROTO ((const char *)); 1001 1002 /* Wrapup a file in the lexical analyzer. */ 1003 extern int yywrap PROTO ((void)); 1004 1005 1006 /* from file sym.c */ 1007 1008 /* Save the text of a character class. */ 1009 extern void cclinstal PROTO ((u_char[], int)); 1010 1011 /* Lookup the number associated with character class. */ 1012 extern int ccllookup PROTO ((u_char[])); 1013 1014 extern void ndinstal PROTO ((const char *, u_char[])); /* install a name definition */ 1015 extern u_char *ndlookup PROTO ((const char *)); /* lookup a name definition */ 1016 1017 /* Increase maximum number of SC's. */ 1018 extern void scextend PROTO ((void)); 1019 extern void scinstal PROTO ((const char *, int)); /* make a start condition */ 1020 1021 /* Lookup the number associated with a start condition. */ 1022 extern int sclookup PROTO ((const char *)); 1023 1024 1025 /* from file tblcmp.c */ 1026 1027 /* Build table entries for dfa state. */ 1028 extern void bldtbl PROTO ((int[], int, int, int, int)); 1029 1030 extern void cmptmps PROTO ((void)); /* compress template table entries */ 1031 extern void expand_nxt_chk PROTO ((void)); /* increase nxt/chk arrays */ 1032 1033 /* Finds a space in the table for a state to be placed. */ 1034 extern int find_table_space PROTO ((int *, int)); 1035 extern void inittbl PROTO ((void)); /* initialize transition tables */ 1036 1037 /* Make the default, "jam" table entries. */ 1038 extern void mkdeftbl PROTO ((void)); 1039 1040 /* Create table entries for a state (or state fragment) which has 1041 * only one out-transition. 1042 */ 1043 extern void mk1tbl PROTO ((int, int, int, int)); 1044 1045 /* Place a state into full speed transition table. */ 1046 extern void place_state PROTO ((int *, int, int)); 1047 1048 /* Save states with only one out-transition to be processed later. */ 1049 extern void stack1 PROTO ((int, int, int, int)); 1050 1051 1052 /* from file yylex.c */ 1053 1054 extern int yylex PROTO ((void)); 1055 1056 /* A growable array. See buf.c. */ 1057 struct Buf { 1058 void *elts; /* elements. */ 1059 int nelts; /* number of elements. */ 1060 size_t elt_size; /* in bytes. */ 1061 int nmax; /* max capacity of elements. */ 1062 }; 1063 1064 extern void buf_init PROTO ((struct Buf * buf, size_t elem_size)); 1065 extern void buf_destroy PROTO ((struct Buf * buf)); 1066 extern struct Buf *buf_append 1067 PROTO ((struct Buf * buf, const void *ptr, int n_elem)); 1068 extern struct Buf *buf_concat PROTO((struct Buf* dest, const struct Buf* src)); 1069 extern struct Buf *buf_strappend PROTO ((struct Buf *, const char *str)); 1070 extern struct Buf *buf_strnappend 1071 PROTO ((struct Buf *, const char *str, int nchars)); 1072 extern struct Buf *buf_strdefine 1073 PROTO ((struct Buf * buf, const char *str, const char *def)); 1074 extern struct Buf *buf_prints PROTO((struct Buf *buf, const char *fmt, const char* s)); 1075 extern struct Buf *buf_m4_define PROTO((struct Buf *buf, const char* def, const char* val)); 1076 extern struct Buf *buf_m4_undefine PROTO((struct Buf *buf, const char* def)); 1077 extern struct Buf *buf_print_strings PROTO((struct Buf * buf, FILE* out)); 1078 extern struct Buf *buf_linedir PROTO((struct Buf *buf, const char* filename, int lineno)); 1079 1080 extern struct Buf userdef_buf; /* a string buffer for #define's generated by user-options on cmd line. */ 1081 extern struct Buf defs_buf; /* a char* buffer to save #define'd some symbols generated by flex. */ 1082 extern struct Buf yydmap_buf; /* a string buffer to hold yydmap elements */ 1083 extern struct Buf m4defs_buf; /* Holds m4 definitions. */ 1084 extern struct Buf top_buf; /* contains %top code. String buffer. */ 1085 1086 /* For blocking out code from the header file. */ 1087 #define OUT_BEGIN_CODE() outn("m4_ifdef( [[M4_YY_IN_HEADER]],,[[") 1088 #define OUT_END_CODE() outn("]])") 1089 1090 /* For setjmp/longjmp (instead of calling exit(2)). Linkage in main.c */ 1091 extern jmp_buf flex_main_jmp_buf; 1092 1093 #define FLEX_EXIT(status) longjmp(flex_main_jmp_buf,(status)+1) 1094 1095 /* Removes all \n and \r chars from tail of str. returns str. */ 1096 extern char *chomp (char *str); 1097 1098 /* ctype functions forced to return boolean */ 1099 #define b_isalnum(c) (isalnum(c)?true:false) 1100 #define b_isalpha(c) (isalpha(c)?true:false) 1101 #define b_isascii(c) (isascii(c)?true:false) 1102 #define b_isblank(c) (isblank(c)?true:false) 1103 #define b_iscntrl(c) (iscntrl(c)?true:false) 1104 #define b_isdigit(c) (isdigit(c)?true:false) 1105 #define b_isgraph(c) (isgraph(c)?true:false) 1106 #define b_islower(c) (islower(c)?true:false) 1107 #define b_isprint(c) (isprint(c)?true:false) 1108 #define b_ispunct(c) (ispunct(c)?true:false) 1109 #define b_isspace(c) (isspace(c)?true:false) 1110 #define b_isupper(c) (isupper(c)?true:false) 1111 #define b_isxdigit(c) (isxdigit(c)?true:false) 1112 1113 /* return true if char is uppercase or lowercase. */ 1114 bool has_case(int c); 1115 1116 /* Change case of character if possible. */ 1117 int reverse_case(int c); 1118 1119 /* return false if [c1-c2] is ambiguous for a caseless scanner. */ 1120 bool range_covers_case (int c1, int c2); 1121 1122 /* 1123 * From "filter.c" 1124 */ 1125 1126 /** A single stdio filter to execute. 1127 * The filter may be external, such as "sed", or it 1128 * may be internal, as a function call. 1129 */ 1130 struct filter { 1131 int (*filter_func)(struct filter*); /**< internal filter function */ 1132 void * extra; /**< extra data passed to filter_func */ 1133 int argc; /**< arg count */ 1134 const char ** argv; /**< arg vector, \0-terminated */ 1135 struct filter * next; /**< next filter or NULL */ 1136 }; 1137 1138 /* output filter chain */ 1139 extern struct filter * output_chain; 1140 extern struct filter *filter_create_ext PROTO((struct filter * chain, const char *cmd, ...)); 1141 struct filter *filter_create_int PROTO((struct filter *chain, 1142 int (*filter_func) (struct filter *), 1143 void *extra)); 1144 extern bool filter_apply_chain PROTO((struct filter * chain)); 1145 extern int filter_truncate (struct filter * chain, int max_len); 1146 extern int filter_tee_header PROTO((struct filter *chain)); 1147 extern int filter_fix_linedirs PROTO((struct filter *chain)); 1148 1149 1150 /* 1151 * From "regex.c" 1152 */ 1153 1154 extern regex_t regex_linedir, regex_blank_line; 1155 bool flex_init_regex(void); 1156 void flex_regcomp(regex_t *preg, const char *regex, int cflags); 1157 char *regmatch_dup (regmatch_t * m, const char *src); 1158 char *regmatch_cpy (regmatch_t * m, char *dest, const char *src); 1159 int regmatch_len (regmatch_t * m); 1160 int regmatch_strtol (regmatch_t * m, const char *src, char **endptr, int base); 1161 bool regmatch_empty (regmatch_t * m); 1162 1163 /* From "scanflags.h" */ 1164 typedef unsigned int scanflags_t; 1165 extern scanflags_t* _sf_stk; 1166 extern size_t _sf_top_ix, _sf_max; /**< stack of scanner flags. */ 1167 #define _SF_CASE_INS 0x0001 1168 #define _SF_DOT_ALL 0x0002 1169 #define _SF_SKIP_WS 0x0004 1170 #define sf_top() (_sf_stk[_sf_top_ix]) 1171 #define sf_case_ins() (sf_top() & _SF_CASE_INS) 1172 #define sf_dot_all() (sf_top() & _SF_DOT_ALL) 1173 #define sf_skip_ws() (sf_top() & _SF_SKIP_WS) 1174 #define sf_set_case_ins(X) ((X) ? (sf_top() |= _SF_CASE_INS) : (sf_top() &= ~_SF_CASE_INS)) 1175 #define sf_set_dot_all(X) ((X) ? (sf_top() |= _SF_DOT_ALL) : (sf_top() &= ~_SF_DOT_ALL)) 1176 #define sf_set_skip_ws(X) ((X) ? (sf_top() |= _SF_SKIP_WS) : (sf_top() &= ~_SF_SKIP_WS)) 1177 extern void sf_init(void); 1178 extern void sf_push(void); 1179 extern void sf_pop(void); 1180 1181 1182 #endif /* not defined FLEXDEF_H */ 1183