1 /* $OpenBSD: flexdef.h,v 1.14 2015/11/19 23:40:16 tedu 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 /* AIX requires this to be the first thing in the file. */ 44 #ifndef __GNUC__ 45 # if HAVE_ALLOCA_H 46 # include <alloca.h> 47 # else 48 # ifdef _AIX 49 #pragma alloca 50 # else 51 # ifndef alloca /* predefined by HP cc +Olibcalls */ 52 char *alloca (); 53 # endif 54 # endif 55 # endif 56 #endif 57 58 #ifdef STDC_HEADERS 59 #include <stdio.h> 60 #include <stdlib.h> 61 #include <stdarg.h> 62 #include <setjmp.h> 63 #include <ctype.h> 64 #include <string.h> 65 #include <math.h> 66 #endif 67 #ifdef HAVE_ASSERT_H 68 #include <assert.h> 69 #else 70 #define assert(Pred) 71 #endif 72 73 #ifdef HAVE_LIMITS_H 74 #include <limits.h> 75 #endif 76 #ifdef HAVE_UNISTD_H 77 #include <unistd.h> 78 #endif 79 #ifdef HAVE_NETINET_IN_H 80 #include <netinet/in.h> 81 #endif 82 #ifdef HAVE_SYS_PARAMS_H 83 #include <sys/params.h> 84 #endif 85 #ifdef HAVE_SYS_WAIT_H 86 #include <sys/wait.h> 87 #endif 88 #ifdef HAVE_STDBOOL_H 89 #include <stdbool.h> 90 #else 91 #define bool int 92 #define true 1 93 #define false 0 94 #endif 95 #ifdef HAVE_REGEX_H 96 #include <regex.h> 97 #endif 98 #include "flexint.h" 99 100 #define _(STRING) STRING 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 #ifndef PROTO 111 #if defined(__STDC__) 112 #define PROTO(proto) proto 113 #else 114 #define PROTO(proto) () 115 #endif 116 #endif 117 118 /* Maximum line length we'll have to deal with. */ 119 #define MAXLINE 2048 120 121 #ifndef MIN 122 #define MIN(x,y) ((x) < (y) ? (x) : (y)) 123 #endif 124 #ifndef MAX 125 #define MAX(x,y) ((x) > (y) ? (x) : (y)) 126 #endif 127 #ifndef ABS 128 #define ABS(x) ((x) < 0 ? -(x) : (x)) 129 #endif 130 131 132 /* ANSI C does not guarantee that isascii() is defined */ 133 #ifndef isascii 134 #define isascii(c) ((c) <= 0177) 135 #endif 136 137 #define unspecified -1 138 139 /* Special chk[] values marking the slots taking by end-of-buffer and action 140 * numbers. 141 */ 142 #define EOB_POSITION -1 143 #define ACTION_POSITION -2 144 145 /* Number of data items per line for -f output. */ 146 #define NUMDATAITEMS 10 147 148 /* Number of lines of data in -f output before inserting a blank line for 149 * readability. 150 */ 151 #define NUMDATALINES 10 152 153 /* transition_struct_out() definitions. */ 154 #define TRANS_STRUCT_PRINT_LENGTH 14 155 156 /* Returns true if an nfa state has an epsilon out-transition slot 157 * that can be used. This definition is currently not used. 158 */ 159 #define FREE_EPSILON(state) \ 160 (transchar[state] == SYM_EPSILON && \ 161 trans2[state] == NO_TRANSITION && \ 162 finalst[state] != state) 163 164 /* Returns true if an nfa state has an epsilon out-transition character 165 * and both slots are free 166 */ 167 #define SUPER_FREE_EPSILON(state) \ 168 (transchar[state] == SYM_EPSILON && \ 169 trans1[state] == NO_TRANSITION) \ 170 171 /* Maximum number of NFA states that can comprise a DFA state. It's real 172 * big because if there's a lot of rules, the initial state will have a 173 * huge epsilon closure. 174 */ 175 #define INITIAL_MAX_DFA_SIZE 750 176 #define MAX_DFA_SIZE_INCREMENT 750 177 178 179 /* A note on the following masks. They are used to mark accepting numbers 180 * as being special. As such, they implicitly limit the number of accepting 181 * numbers (i.e., rules) because if there are too many rules the rule numbers 182 * will overload the mask bits. Fortunately, this limit is \large/ (0x2000 == 183 * 8192) so unlikely to actually cause any problems. A check is made in 184 * new_rule() to ensure that this limit is not reached. 185 */ 186 187 /* Mask to mark a trailing context accepting number. */ 188 #define YY_TRAILING_MASK 0x2000 189 190 /* Mask to mark the accepting number of the "head" of a trailing context 191 * rule. 192 */ 193 #define YY_TRAILING_HEAD_MASK 0x4000 194 195 /* Maximum number of rules, as outlined in the above note. */ 196 #define MAX_RULE (YY_TRAILING_MASK - 1) 197 198 199 /* NIL must be 0. If not, its special meaning when making equivalence classes 200 * (it marks the representative of a given e.c.) will be unidentifiable. 201 */ 202 #define NIL 0 203 204 #define JAM -1 /* to mark a missing DFA transition */ 205 #define NO_TRANSITION NIL 206 #define UNIQUE -1 /* marks a symbol as an e.c. representative */ 207 #define INFINITE_REPEAT -1 /* for x{5,} constructions */ 208 209 #define INITIAL_MAX_CCLS 100 /* max number of unique character classes */ 210 #define MAX_CCLS_INCREMENT 100 211 212 /* Size of table holding members of character classes. */ 213 #define INITIAL_MAX_CCL_TBL_SIZE 500 214 #define MAX_CCL_TBL_SIZE_INCREMENT 250 215 216 #define INITIAL_MAX_RULES 100 /* default maximum number of rules */ 217 #define MAX_RULES_INCREMENT 100 218 219 #define INITIAL_MNS 2000 /* default maximum number of nfa states */ 220 #define MNS_INCREMENT 1000 /* amount to bump above by if it's not enough */ 221 222 #define INITIAL_MAX_DFAS 1000 /* default maximum number of dfa states */ 223 #define MAX_DFAS_INCREMENT 1000 224 225 #define JAMSTATE -32766 /* marks a reference to the state that always jams */ 226 227 /* Maximum number of NFA states. */ 228 #define MAXIMUM_MNS 31999 229 #define MAXIMUM_MNS_LONG 1999999999 230 231 /* Enough so that if it's subtracted from an NFA state number, the result 232 * is guaranteed to be negative. 233 */ 234 #define MARKER_DIFFERENCE (maximum_mns+2) 235 236 /* Maximum number of nxt/chk pairs for non-templates. */ 237 #define INITIAL_MAX_XPAIRS 2000 238 #define MAX_XPAIRS_INCREMENT 2000 239 240 /* Maximum number of nxt/chk pairs needed for templates. */ 241 #define INITIAL_MAX_TEMPLATE_XPAIRS 2500 242 #define MAX_TEMPLATE_XPAIRS_INCREMENT 2500 243 244 #define SYM_EPSILON (CSIZE + 1) /* to mark transitions on the symbol epsilon */ 245 246 #define INITIAL_MAX_SCS 40 /* maximum number of start conditions */ 247 #define MAX_SCS_INCREMENT 40 /* amount to bump by if it's not enough */ 248 249 #define ONE_STACK_SIZE 500 /* stack of states with only one out-transition */ 250 #define SAME_TRANS -1 /* transition is the same as "default" entry for state */ 251 252 /* The following percentages are used to tune table compression: 253 254 * The percentage the number of out-transitions a state must be of the 255 * number of equivalence classes in order to be considered for table 256 * compaction by using protos. 257 */ 258 #define PROTO_SIZE_PERCENTAGE 15 259 260 /* The percentage the number of homogeneous out-transitions of a state 261 * must be of the number of total out-transitions of the state in order 262 * that the state's transition table is first compared with a potential 263 * template of the most common out-transition instead of with the first 264 * proto in the proto queue. 265 */ 266 #define CHECK_COM_PERCENTAGE 50 267 268 /* The percentage the number of differences between a state's transition 269 * table and the proto it was first compared with must be of the total 270 * number of out-transitions of the state in order to keep the first 271 * proto as a good match and not search any further. 272 */ 273 #define FIRST_MATCH_DIFF_PERCENTAGE 10 274 275 /* The percentage the number of differences between a state's transition 276 * table and the most similar proto must be of the state's total number 277 * of out-transitions to use the proto as an acceptable close match. 278 */ 279 #define ACCEPTABLE_DIFF_PERCENTAGE 50 280 281 /* The percentage the number of homogeneous out-transitions of a state 282 * must be of the number of total out-transitions of the state in order 283 * to consider making a template from the state. 284 */ 285 #define TEMPLATE_SAME_PERCENTAGE 60 286 287 /* The percentage the number of differences between a state's transition 288 * table and the most similar proto must be of the state's total number 289 * of out-transitions to create a new proto from the state. 290 */ 291 #define NEW_PROTO_DIFF_PERCENTAGE 20 292 293 /* The percentage the total number of out-transitions of a state must be 294 * of the number of equivalence classes in order to consider trying to 295 * fit the transition table into "holes" inside the nxt/chk table. 296 */ 297 #define INTERIOR_FIT_PERCENTAGE 15 298 299 /* Size of region set aside to cache the complete transition table of 300 * protos on the proto queue to enable quick comparisons. 301 */ 302 #define PROT_SAVE_SIZE 2000 303 304 #define MSP 50 /* maximum number of saved protos (protos on the proto queue) */ 305 306 /* Maximum number of out-transitions a state can have that we'll rummage 307 * around through the interior of the internal fast table looking for a 308 * spot for it. 309 */ 310 #define MAX_XTIONS_FULL_INTERIOR_FIT 4 311 312 /* Maximum number of rules which will be reported as being associated 313 * with a DFA state. 314 */ 315 #define MAX_ASSOC_RULES 100 316 317 /* Number that, if used to subscript an array, has a good chance of producing 318 * an error; should be small enough to fit into a short. 319 */ 320 #define BAD_SUBSCRIPT -32767 321 322 /* Absolute value of largest number that can be stored in a short, with a 323 * bit of slop thrown in for general paranoia. 324 */ 325 #define MAX_SHORT 32700 326 327 328 /* Declarations for global variables. */ 329 330 331 /* Variables for flags: 332 * printstats - if true (-v), dump statistics 333 * syntaxerror - true if a syntax error has been found 334 * eofseen - true if we've seen an eof in the input file 335 * ddebug - if true (-d), make a "debug" scanner 336 * trace - if true (-T), trace processing 337 * nowarn - if true (-w), do not generate warnings 338 * spprdflt - if true (-s), suppress the default rule 339 * interactive - if true (-I), generate an interactive scanner 340 * lex_compat - if true (-l), maximize compatibility with AT&T lex 341 * posix_compat - if true (-X), maximize compatibility with POSIX lex 342 * do_yylineno - if true, generate code to maintain yylineno 343 * useecs - if true (-Ce flag), use equivalence classes 344 * fulltbl - if true (-Cf flag), don't compress the DFA state table 345 * usemecs - if true (-Cm flag), use meta-equivalence classes 346 * fullspd - if true (-F flag), use Jacobson method of table representation 347 * gen_line_dirs - if true (i.e., no -L flag), generate #line directives 348 * performance_report - if > 0 (i.e., -p flag), generate a report relating 349 * to scanner performance; if > 1 (-p -p), report on minor performance 350 * problems, too 351 * backing_up_report - if true (i.e., -b flag), generate "lex.backup" file 352 * listing backing-up states 353 * C_plus_plus - if true (i.e., -+ flag), generate a C++ scanner class; 354 * otherwise, a standard C scanner 355 * reentrant - if true (-R), generate a reentrant C scanner. 356 * bison_bridge_lval - if true (--bison-bridge), bison pure calling convention. 357 * bison_bridge_lloc - if true (--bison-locations), bison yylloc. 358 * long_align - if true (-Ca flag), favor long-word alignment. 359 * use_read - if true (-f, -F, or -Cr) then use read() for scanner input; 360 * otherwise, use fread(). 361 * yytext_is_array - if true (i.e., %array directive), then declare 362 * yytext as a array instead of a character pointer. Nice and inefficient. 363 * do_yywrap - do yywrap() processing on EOF. If false, EOF treated as 364 * "no more files". 365 * csize - size of character set for the scanner we're generating; 366 * 128 for 7-bit chars and 256 for 8-bit 367 * yymore_used - if true, yymore() is used in input rules 368 * reject - if true, generate back-up tables for REJECT macro 369 * real_reject - if true, scanner really uses REJECT (as opposed to just 370 * having "reject" set for variable trailing context) 371 * continued_action - true if this rule's action is to "fall through" to 372 * the next rule's action (i.e., the '|' action) 373 * in_rule - true if we're inside an individual rule, false if not. 374 * yymore_really_used - whether to treat yymore() as really used, regardless 375 * of what we think based on references to it in the user's actions. 376 * reject_really_used - same for REJECT 377 */ 378 379 extern int printstats, syntaxerror, eofseen, ddebug, trace, nowarn, 380 spprdflt; 381 extern int interactive, lex_compat, posix_compat, do_yylineno; 382 extern int useecs, fulltbl, usemecs, fullspd; 383 extern int gen_line_dirs, performance_report, backing_up_report; 384 extern int reentrant, bison_bridge_lval, bison_bridge_lloc; 385 extern bool ansi_func_defs, ansi_func_protos; 386 extern int C_plus_plus, long_align, use_read, yytext_is_array, do_yywrap; 387 extern int csize; 388 extern int yymore_used, reject, real_reject, continued_action, in_rule; 389 390 extern int yymore_really_used, reject_really_used; 391 392 393 /* Variables used in the flex input routines: 394 * datapos - characters on current output line 395 * dataline - number of contiguous lines of data in current data 396 * statement. Used to generate readable -f output 397 * linenum - current input line number 398 * skelfile - the skeleton file 399 * skel - compiled-in skeleton array 400 * skel_ind - index into "skel" array, if skelfile is nil 401 * yyin - input file 402 * backing_up_file - file to summarize backing-up states to 403 * infilename - name of input file 404 * outfilename - name of output file 405 * headerfilename - name of the .h file to generate 406 * did_outfilename - whether outfilename was explicitly set 407 * prefix - the prefix used for externally visible names ("yy" by default) 408 * yyclass - yyFlexLexer subclass to use for YY_DECL 409 * do_stdinit - whether to initialize yyin/yyout to stdin/stdout 410 * use_stdout - the -t flag 411 * input_files - array holding names of input files 412 * num_input_files - size of input_files array 413 * program_name - name with which program was invoked 414 * 415 * action_array - array to hold the rule actions 416 * action_size - size of action_array 417 * defs1_offset - index where the user's section 1 definitions start 418 * in action_array 419 * prolog_offset - index where the prolog starts in action_array 420 * action_offset - index where the non-prolog starts in action_array 421 * action_index - index where the next action should go, with respect 422 * to "action_array" 423 */ 424 425 extern int datapos, dataline, linenum; 426 extern FILE *skelfile, *yyin, *backing_up_file; 427 extern const char *skel[]; 428 extern int skel_ind; 429 extern char *infilename, *outfilename, *headerfilename; 430 extern int did_outfilename; 431 extern char *prefix, *yyclass, *extra_type; 432 extern int do_stdinit, use_stdout; 433 extern char **input_files; 434 extern int num_input_files; 435 extern char *program_name; 436 437 extern char *action_array; 438 extern int action_size; 439 extern int defs1_offset, prolog_offset, action_offset, action_index; 440 441 442 /* Variables for stack of states having only one out-transition: 443 * onestate - state number 444 * onesym - transition symbol 445 * onenext - target state 446 * onedef - default base entry 447 * onesp - stack pointer 448 */ 449 450 extern int onestate[ONE_STACK_SIZE], onesym[ONE_STACK_SIZE]; 451 extern int onenext[ONE_STACK_SIZE], onedef[ONE_STACK_SIZE], onesp; 452 453 454 /* Variables for nfa machine data: 455 * maximum_mns - maximal number of NFA states supported by tables 456 * current_mns - current maximum on number of NFA states 457 * num_rules - number of the last accepting state; also is number of 458 * rules created so far 459 * num_eof_rules - number of <<EOF>> rules 460 * default_rule - number of the default rule 461 * current_max_rules - current maximum number of rules 462 * lastnfa - last nfa state number created 463 * firstst - physically the first state of a fragment 464 * lastst - last physical state of fragment 465 * finalst - last logical state of fragment 466 * transchar - transition character 467 * trans1 - transition state 468 * trans2 - 2nd transition state for epsilons 469 * accptnum - accepting number 470 * assoc_rule - rule associated with this NFA state (or 0 if none) 471 * state_type - a STATE_xxx type identifying whether the state is part 472 * of a normal rule, the leading state in a trailing context 473 * rule (i.e., the state which marks the transition from 474 * recognizing the text-to-be-matched to the beginning of 475 * the trailing context), or a subsequent state in a trailing 476 * context rule 477 * rule_type - a RULE_xxx type identifying whether this a ho-hum 478 * normal rule or one which has variable head & trailing 479 * context 480 * rule_linenum - line number associated with rule 481 * rule_useful - true if we've determined that the rule can be matched 482 * rule_has_nl - true if rule could possibly match a newline 483 * ccl_has_nl - true if current ccl could match a newline 484 * nlch - default eol char 485 */ 486 487 extern int maximum_mns, current_mns, current_max_rules; 488 extern int num_rules, num_eof_rules, default_rule, lastnfa; 489 extern int *firstst, *lastst, *finalst, *transchar, *trans1, *trans2; 490 extern int *accptnum, *assoc_rule, *state_type; 491 extern int *rule_type, *rule_linenum, *rule_useful; 492 extern bool *rule_has_nl, *ccl_has_nl; 493 extern int nlch; 494 495 /* Different types of states; values are useful as masks, as well, for 496 * routines like check_trailing_context(). 497 */ 498 #define STATE_NORMAL 0x1 499 #define STATE_TRAILING_CONTEXT 0x2 500 501 /* Global holding current type of state we're making. */ 502 503 extern int current_state_type; 504 505 /* Different types of rules. */ 506 #define RULE_NORMAL 0 507 #define RULE_VARIABLE 1 508 509 /* True if the input rules include a rule with both variable-length head 510 * and trailing context, false otherwise. 511 */ 512 extern int variable_trailing_context_rules; 513 514 515 /* Variables for protos: 516 * numtemps - number of templates created 517 * numprots - number of protos created 518 * protprev - backlink to a more-recently used proto 519 * protnext - forward link to a less-recently used proto 520 * prottbl - base/def table entry for proto 521 * protcomst - common state of proto 522 * firstprot - number of the most recently used proto 523 * lastprot - number of the least recently used proto 524 * protsave contains the entire state array for protos 525 */ 526 527 extern int numtemps, numprots, protprev[MSP], protnext[MSP], prottbl[MSP]; 528 extern int protcomst[MSP], firstprot, lastprot, protsave[PROT_SAVE_SIZE]; 529 530 531 /* Variables for managing equivalence classes: 532 * numecs - number of equivalence classes 533 * nextecm - forward link of Equivalence Class members 534 * ecgroup - class number or backward link of EC members 535 * nummecs - number of meta-equivalence classes (used to compress 536 * templates) 537 * tecfwd - forward link of meta-equivalence classes members 538 * tecbck - backward link of MEC's 539 */ 540 541 /* Reserve enough room in the equivalence class arrays so that we 542 * can use the CSIZE'th element to hold equivalence class information 543 * for the NUL character. Later we'll move this information into 544 * the 0th element. 545 */ 546 extern int numecs, nextecm[CSIZE + 1], ecgroup[CSIZE + 1], nummecs; 547 548 /* Meta-equivalence classes are indexed starting at 1, so it's possible 549 * that they will require positions from 1 .. CSIZE, i.e., CSIZE + 1 550 * slots total (since the arrays are 0-based). nextecm[] and ecgroup[] 551 * don't require the extra position since they're indexed from 1 .. CSIZE - 1. 552 */ 553 extern int tecfwd[CSIZE + 1], tecbck[CSIZE + 1]; 554 555 556 /* Variables for start conditions: 557 * lastsc - last start condition created 558 * current_max_scs - current limit on number of start conditions 559 * scset - set of rules active in start condition 560 * scbol - set of rules active only at the beginning of line in a s.c. 561 * scxclu - true if start condition is exclusive 562 * sceof - true if start condition has EOF rule 563 * scname - start condition name 564 */ 565 566 extern int lastsc, *scset, *scbol, *scxclu, *sceof; 567 extern int current_max_scs; 568 extern char **scname; 569 570 571 /* Variables for dfa machine data: 572 * current_max_dfa_size - current maximum number of NFA states in DFA 573 * current_max_xpairs - current maximum number of non-template xtion pairs 574 * current_max_template_xpairs - current maximum number of template pairs 575 * current_max_dfas - current maximum number DFA states 576 * lastdfa - last dfa state number created 577 * nxt - state to enter upon reading character 578 * chk - check value to see if "nxt" applies 579 * tnxt - internal nxt table for templates 580 * base - offset into "nxt" for given state 581 * def - where to go if "chk" disallows "nxt" entry 582 * nultrans - NUL transition for each state 583 * NUL_ec - equivalence class of the NUL character 584 * tblend - last "nxt/chk" table entry being used 585 * firstfree - first empty entry in "nxt/chk" table 586 * dss - nfa state set for each dfa 587 * dfasiz - size of nfa state set for each dfa 588 * dfaacc - accepting set for each dfa state (if using REJECT), or accepting 589 * number, if not 590 * accsiz - size of accepting set for each dfa state 591 * dhash - dfa state hash value 592 * numas - number of DFA accepting states created; note that this 593 * is not necessarily the same value as num_rules, which is the analogous 594 * value for the NFA 595 * numsnpairs - number of state/nextstate transition pairs 596 * jambase - position in base/def where the default jam table starts 597 * jamstate - state number corresponding to "jam" state 598 * end_of_buffer_state - end-of-buffer dfa state number 599 */ 600 601 extern int current_max_dfa_size, current_max_xpairs; 602 extern int current_max_template_xpairs, current_max_dfas; 603 extern int lastdfa, *nxt, *chk, *tnxt; 604 extern int *base, *def, *nultrans, NUL_ec, tblend, firstfree, **dss, 605 *dfasiz; 606 extern union dfaacc_union { 607 int *dfaacc_set; 608 int dfaacc_state; 609 } *dfaacc; 610 extern int *accsiz, *dhash, numas; 611 extern int numsnpairs, jambase, jamstate; 612 extern int end_of_buffer_state; 613 614 /* Variables for ccl information: 615 * lastccl - ccl index of the last created ccl 616 * current_maxccls - current limit on the maximum number of unique ccl's 617 * cclmap - maps a ccl index to its set pointer 618 * ccllen - gives the length of a ccl 619 * cclng - true for a given ccl if the ccl is negated 620 * cclreuse - counts how many times a ccl is re-used 621 * current_max_ccl_tbl_size - current limit on number of characters needed 622 * to represent the unique ccl's 623 * ccltbl - holds the characters in each ccl - indexed by cclmap 624 */ 625 626 extern int lastccl, *cclmap, *ccllen, *cclng, cclreuse; 627 extern int current_maxccls, current_max_ccl_tbl_size; 628 extern u_char *ccltbl; 629 630 631 /* Variables for miscellaneous information: 632 * nmstr - last NAME scanned by the scanner 633 * sectnum - section number currently being parsed 634 * nummt - number of empty nxt/chk table entries 635 * hshcol - number of hash collisions detected by snstods 636 * dfaeql - number of times a newly created dfa was equal to an old one 637 * numeps - number of epsilon NFA states created 638 * eps2 - number of epsilon states which have 2 out-transitions 639 * num_reallocs - number of times it was necessary to realloc() a group 640 * of arrays 641 * tmpuses - number of DFA states that chain to templates 642 * totnst - total number of NFA states used to make DFA states 643 * peakpairs - peak number of transition pairs we had to store internally 644 * numuniq - number of unique transitions 645 * numdup - number of duplicate transitions 646 * hshsave - number of hash collisions saved by checking number of states 647 * num_backing_up - number of DFA states requiring backing up 648 * bol_needed - whether scanner needs beginning-of-line recognition 649 */ 650 651 extern char nmstr[MAXLINE]; 652 extern int sectnum, nummt, hshcol, dfaeql, numeps, eps2, num_reallocs; 653 extern int tmpuses, totnst, peakpairs, numuniq, numdup, hshsave; 654 extern int num_backing_up, bol_needed; 655 656 void *allocate_array PROTO ((int, size_t)); 657 void *reallocate_array PROTO ((void *, int, size_t)); 658 659 #define allocate_integer_array(size) \ 660 (int *) allocate_array( size, sizeof( int ) ) 661 662 #define reallocate_integer_array(array,size) \ 663 (int *) reallocate_array( (void *) array, size, sizeof( int ) ) 664 665 #define allocate_bool_array(size) \ 666 (bool *) allocate_array( size, sizeof( bool ) ) 667 668 #define reallocate_bool_array(array,size) \ 669 (bool *) reallocate_array( (void *) array, size, sizeof( bool ) ) 670 671 #define allocate_int_ptr_array(size) \ 672 (int **) allocate_array( size, sizeof( int * ) ) 673 674 #define allocate_char_ptr_array(size) \ 675 (char **) allocate_array( size, sizeof( char * ) ) 676 677 #define allocate_dfaacc_union(size) \ 678 (union dfaacc_union *) \ 679 allocate_array( size, sizeof( union dfaacc_union ) ) 680 681 #define reallocate_int_ptr_array(array,size) \ 682 (int **) reallocate_array( (void *) array, size, sizeof( int * ) ) 683 684 #define reallocate_char_ptr_array(array,size) \ 685 (char **) reallocate_array( (void *) array, size, sizeof( char * ) ) 686 687 #define reallocate_dfaacc_union(array, size) \ 688 (union dfaacc_union *) \ 689 reallocate_array( (void *) array, size, sizeof( union dfaacc_union ) ) 690 691 #define allocate_character_array(size) \ 692 (char *) allocate_array( size, sizeof( char ) ) 693 694 #define reallocate_character_array(array,size) \ 695 (char *) reallocate_array( (void *) array, size, sizeof( char ) ) 696 697 #define allocate_Character_array(size) \ 698 (u_char *) allocate_array( size, sizeof( u_char ) ) 699 700 #define reallocate_Character_array(array,size) \ 701 (u_char *) reallocate_array( (void *) array, size, sizeof( u_char ) ) 702 703 704 /* Used to communicate between scanner and parser. The type should really 705 * be YYSTYPE, but we can't easily get our hands on it. 706 */ 707 extern int yylval; 708 709 710 /* External functions that are cross-referenced among the flex source files. */ 711 712 713 /* from file ccl.c */ 714 715 extern void ccladd PROTO ((int, int)); /* add a single character to a ccl */ 716 extern int cclinit PROTO ((void)); /* make an empty ccl */ 717 extern void cclnegate PROTO ((int)); /* negate a ccl */ 718 extern int ccl_set_diff (int a, int b); /* set difference of two ccls. */ 719 extern int ccl_set_union (int a, int b); /* set union of two ccls. */ 720 721 /* List the members of a set of characters in CCL form. */ 722 extern void list_character_set PROTO ((FILE *, int[])); 723 724 725 /* from file dfa.c */ 726 727 /* Check a DFA state for backing up. */ 728 extern void check_for_backing_up PROTO ((int, int[])); 729 730 /* Check to see if NFA state set constitutes "dangerous" trailing context. */ 731 extern void check_trailing_context PROTO ((int *, int, int *, int)); 732 733 /* Construct the epsilon closure of a set of ndfa states. */ 734 extern int *epsclosure PROTO ((int *, int *, int[], int *, int *)); 735 736 /* Increase the maximum number of dfas. */ 737 extern void increase_max_dfas PROTO ((void)); 738 739 extern void ntod PROTO ((void)); /* convert a ndfa to a dfa */ 740 741 /* Converts a set of ndfa states into a dfa state. */ 742 extern int snstods PROTO ((int[], int, int[], int, int, int *)); 743 744 745 /* from file ecs.c */ 746 747 /* Convert character classes to set of equivalence classes. */ 748 extern void ccl2ecl PROTO ((void)); 749 750 /* Associate equivalence class numbers with class members. */ 751 extern int cre8ecs PROTO ((int[], int[], int)); 752 753 /* Update equivalence classes based on character class transitions. */ 754 extern void mkeccl PROTO ((u_char[], int, int[], int[], int, int)); 755 756 /* Create equivalence class for single character. */ 757 extern void mkechar PROTO ((int, int[], int[])); 758 759 760 /* from file gen.c */ 761 762 extern void do_indent PROTO ((void)); /* indent to the current level */ 763 764 /* Generate the code to keep backing-up information. */ 765 extern void gen_backing_up PROTO ((void)); 766 767 /* Generate the code to perform the backing up. */ 768 extern void gen_bu_action PROTO ((void)); 769 770 /* Generate full speed compressed transition table. */ 771 extern void genctbl PROTO ((void)); 772 773 /* Generate the code to find the action number. */ 774 extern void gen_find_action PROTO ((void)); 775 776 extern void genftbl PROTO ((void)); /* generate full transition table */ 777 778 /* Generate the code to find the next compressed-table state. */ 779 extern void gen_next_compressed_state PROTO ((char *)); 780 781 /* Generate the code to find the next match. */ 782 extern void gen_next_match PROTO ((void)); 783 784 /* Generate the code to find the next state. */ 785 extern void gen_next_state PROTO ((int)); 786 787 /* Generate the code to make a NUL transition. */ 788 extern void gen_NUL_trans PROTO ((void)); 789 790 /* Generate the code to find the start state. */ 791 extern void gen_start_state PROTO ((void)); 792 793 /* Generate data statements for the transition tables. */ 794 extern void gentabs PROTO ((void)); 795 796 /* Write out a formatted string at the current indentation level. */ 797 extern void indent_put2s PROTO ((const char *, const char *)); 798 799 /* Write out a string + newline at the current indentation level. */ 800 extern void indent_puts PROTO ((const char *)); 801 802 extern void make_tables PROTO ((void)); /* generate transition tables */ 803 804 805 /* from file main.c */ 806 807 extern void check_options PROTO ((void)); 808 extern void flexend PROTO ((int)); 809 extern void usage PROTO ((void)); 810 811 812 /* from file misc.c */ 813 814 /* Add a #define to the action file. */ 815 extern void action_define PROTO ((const char *defname, int value)); 816 817 /* Add the given text to the stored actions. */ 818 extern void add_action PROTO ((const char *new_text)); 819 820 /* True if a string is all lower case. */ 821 extern int all_lower PROTO ((char *)); 822 823 /* True if a string is all upper case. */ 824 extern int all_upper PROTO ((char *)); 825 826 /* Compare two integers for use by qsort. */ 827 extern int intcmp PROTO ((const void *, const void *)); 828 829 /* Check a character to make sure it's in the expected range. */ 830 extern void check_char PROTO ((int c)); 831 832 /* Replace upper-case letter to lower-case. */ 833 extern u_char clower PROTO ((int)); 834 835 /* Returns a dynamically allocated copy of a string. */ 836 extern char *copy_string PROTO ((const char *)); 837 838 /* Returns a dynamically allocated copy of a (potentially) unsigned string. */ 839 extern u_char *copy_unsigned_string PROTO ((u_char *)); 840 841 /* Compare two characters for use by qsort with '\0' sorting last. */ 842 extern int cclcmp PROTO ((const void *, const void *)); 843 844 /* Finish up a block of data declarations. */ 845 extern void dataend PROTO ((void)); 846 847 /* Flush generated data statements. */ 848 extern void dataflush PROTO ((void)); 849 850 /* Report an error message and terminate. */ 851 extern void flexerror PROTO ((const char *)); 852 853 /* Report a fatal error message and terminate. */ 854 extern void flexfatal PROTO ((const char *)); 855 856 /* Report a fatal error with a pinpoint, and terminate */ 857 #if HAVE_DECL___FUNC__ 858 #define flex_die(msg) \ 859 do{ \ 860 fprintf (stderr,\ 861 _("%s: fatal internal error at %s:%d (%s): %s\n"),\ 862 program_name, __FILE__, (int)__LINE__,\ 863 __func__,msg);\ 864 FLEX_EXIT(1);\ 865 }while(0) 866 #else /* ! HAVE_DECL___FUNC__ */ 867 #define flex_die(msg) \ 868 do{ \ 869 fprintf (stderr,\ 870 _("%s: fatal internal error at %s:%d %s\n"),\ 871 program_name, __FILE__, (int)__LINE__,\ 872 msg);\ 873 FLEX_EXIT(1);\ 874 }while(0) 875 #endif /* ! HAVE_DECL___func__ */ 876 877 /* Convert a hexadecimal digit string to an integer value. */ 878 extern int htoi PROTO ((u_char[])); 879 880 /* Report an error message formatted with one integer argument. */ 881 extern void lerrif PROTO ((const char *, int)); 882 883 /* Report an error message formatted with one string argument. */ 884 extern void lerrsf PROTO ((const char *, const char *)); 885 886 /* Like lerrsf, but also exit after displaying message. */ 887 extern void lerrsf_fatal PROTO ((const char *, const char *)); 888 889 /* Spit out a "#line" statement. */ 890 extern void line_directive_out PROTO ((FILE *, int)); 891 892 /* Mark the current position in the action array as the end of the section 1 893 * user defs. 894 */ 895 extern void mark_defs1 PROTO ((void)); 896 897 /* Mark the current position in the action array as the end of the prolog. */ 898 extern void mark_prolog PROTO ((void)); 899 900 /* Generate a data statment for a two-dimensional array. */ 901 extern void mk2data PROTO ((int)); 902 903 extern void mkdata PROTO ((int)); /* generate a data statement */ 904 905 /* Return the integer represented by a string of digits. */ 906 extern int myctoi PROTO ((const char *)); 907 908 /* Return character corresponding to escape sequence. */ 909 extern u_char myesc PROTO ((u_char[])); 910 911 /* Convert an octal digit string to an integer value. */ 912 extern int otoi PROTO ((u_char[])); 913 914 /* Output a (possibly-formatted) string to the generated scanner. */ 915 extern void out PROTO ((const char *)); 916 extern void out_dec PROTO ((const char *, int)); 917 extern void out_dec2 PROTO ((const char *, int, int)); 918 extern void out_hex PROTO ((const char *, unsigned int)); 919 extern void out_str PROTO ((const char *, const char *)); 920 extern void out_str3 921 PROTO ((const char *, const char *, const char *, const char *)); 922 extern void out_str_dec PROTO ((const char *, const char *, int)); 923 extern void outc PROTO ((int)); 924 extern void outn PROTO ((const char *)); 925 extern void out_m4_define (const char* def, const char* val); 926 927 /* Return a printable version of the given character, which might be 928 * 8-bit. 929 */ 930 extern char *readable_form PROTO ((int)); 931 932 /* Write out one section of the skeleton file. */ 933 extern void skelout PROTO ((void)); 934 935 /* Output a yy_trans_info structure. */ 936 extern void transition_struct_out PROTO ((int, int)); 937 938 /* Only needed when using certain broken versions of bison to build parse.c. */ 939 extern void *yy_flex_xmalloc PROTO ((int)); 940 941 /* from file nfa.c */ 942 943 /* Add an accepting state to a machine. */ 944 extern void add_accept PROTO ((int, int)); 945 946 /* Make a given number of copies of a singleton machine. */ 947 extern int copysingl PROTO ((int, int)); 948 949 /* Debugging routine to write out an nfa. */ 950 extern void dumpnfa PROTO ((int)); 951 952 /* Finish up the processing for a rule. */ 953 extern void finish_rule PROTO ((int, int, int, int, int)); 954 955 /* Connect two machines together. */ 956 extern int link_machines PROTO ((int, int)); 957 958 /* Mark each "beginning" state in a machine as being a "normal" (i.e., 959 * not trailing context associated) state. 960 */ 961 extern void mark_beginning_as_normal PROTO ((int)); 962 963 /* Make a machine that branches to two machines. */ 964 extern int mkbranch PROTO ((int, int)); 965 966 extern int mkclos PROTO ((int)); /* convert a machine into a closure */ 967 extern int mkopt PROTO ((int)); /* make a machine optional */ 968 969 /* Make a machine that matches either one of two machines. */ 970 extern int mkor PROTO ((int, int)); 971 972 /* Convert a machine into a positive closure. */ 973 extern int mkposcl PROTO ((int)); 974 975 extern int mkrep PROTO ((int, int, int)); /* make a replicated machine */ 976 977 /* Create a state with a transition on a given symbol. */ 978 extern int mkstate PROTO ((int)); 979 980 extern void new_rule PROTO ((void)); /* initialize for a new rule */ 981 982 983 /* from file parse.y */ 984 985 /* Build the "<<EOF>>" action for the active start conditions. */ 986 extern void build_eof_action PROTO ((void)); 987 988 /* Write out a message formatted with one string, pinpointing its location. */ 989 extern void format_pinpoint_message PROTO ((const char *, const char *)); 990 991 /* Write out a message, pinpointing its location. */ 992 extern void pinpoint_message PROTO ((const char *)); 993 994 /* Write out a warning, pinpointing it at the given line. */ 995 extern void line_warning PROTO ((const char *, int)); 996 997 /* Write out a message, pinpointing it at the given line. */ 998 extern void line_pinpoint PROTO ((const char *, int)); 999 1000 /* Report a formatted syntax error. */ 1001 extern void format_synerr PROTO ((const char *, const char *)); 1002 extern void synerr PROTO ((const char *)); /* report a syntax error */ 1003 extern void format_warn PROTO ((const char *, const char *)); 1004 extern void warn PROTO ((const char *)); /* report a warning */ 1005 extern void yyerror PROTO ((const char *)); /* report a parse error */ 1006 extern int yyparse PROTO ((void)); /* the YACC parser */ 1007 1008 1009 /* from file scan.l */ 1010 1011 /* The Flex-generated scanner for flex. */ 1012 extern int flexscan PROTO ((void)); 1013 1014 /* Open the given file (if NULL, stdin) for scanning. */ 1015 extern void set_input_file PROTO ((char *)); 1016 1017 /* Wrapup a file in the lexical analyzer. */ 1018 extern int yywrap PROTO ((void)); 1019 1020 1021 /* from file sym.c */ 1022 1023 /* Save the text of a character class. */ 1024 extern void cclinstal PROTO ((u_char[], int)); 1025 1026 /* Lookup the number associated with character class. */ 1027 extern int ccllookup PROTO ((u_char[])); 1028 1029 extern void ndinstal PROTO ((const char *, u_char[])); /* install a name definition */ 1030 extern u_char *ndlookup PROTO ((const char *)); /* lookup a name definition */ 1031 1032 /* Increase maximum number of SC's. */ 1033 extern void scextend PROTO ((void)); 1034 extern void scinstal PROTO ((const char *, int)); /* make a start condition */ 1035 1036 /* Lookup the number associated with a start condition. */ 1037 extern int sclookup PROTO ((const char *)); 1038 1039 1040 /* from file tblcmp.c */ 1041 1042 /* Build table entries for dfa state. */ 1043 extern void bldtbl PROTO ((int[], int, int, int, int)); 1044 1045 extern void cmptmps PROTO ((void)); /* compress template table entries */ 1046 extern void expand_nxt_chk PROTO ((void)); /* increase nxt/chk arrays */ 1047 1048 /* Finds a space in the table for a state to be placed. */ 1049 extern int find_table_space PROTO ((int *, int)); 1050 extern void inittbl PROTO ((void)); /* initialize transition tables */ 1051 1052 /* Make the default, "jam" table entries. */ 1053 extern void mkdeftbl PROTO ((void)); 1054 1055 /* Create table entries for a state (or state fragment) which has 1056 * only one out-transition. 1057 */ 1058 extern void mk1tbl PROTO ((int, int, int, int)); 1059 1060 /* Place a state into full speed transition table. */ 1061 extern void place_state PROTO ((int *, int, int)); 1062 1063 /* Save states with only one out-transition to be processed later. */ 1064 extern void stack1 PROTO ((int, int, int, int)); 1065 1066 1067 /* from file yylex.c */ 1068 1069 extern int yylex PROTO ((void)); 1070 1071 /* A growable array. See buf.c. */ 1072 struct Buf { 1073 void *elts; /* elements. */ 1074 int nelts; /* number of elements. */ 1075 size_t elt_size; /* in bytes. */ 1076 int nmax; /* max capacity of elements. */ 1077 }; 1078 1079 extern void buf_init PROTO ((struct Buf * buf, size_t elem_size)); 1080 extern void buf_destroy PROTO ((struct Buf * buf)); 1081 extern struct Buf *buf_append 1082 PROTO ((struct Buf * buf, const void *ptr, int n_elem)); 1083 extern struct Buf *buf_concat PROTO((struct Buf* dest, const struct Buf* src)); 1084 extern struct Buf *buf_strappend PROTO ((struct Buf *, const char *str)); 1085 extern struct Buf *buf_strnappend 1086 PROTO ((struct Buf *, const char *str, int nchars)); 1087 extern struct Buf *buf_strdefine 1088 PROTO ((struct Buf * buf, const char *str, const char *def)); 1089 extern struct Buf *buf_prints PROTO((struct Buf *buf, const char *fmt, const char* s)); 1090 extern struct Buf *buf_m4_define PROTO((struct Buf *buf, const char* def, const char* val)); 1091 extern struct Buf *buf_m4_undefine PROTO((struct Buf *buf, const char* def)); 1092 extern struct Buf *buf_print_strings PROTO((struct Buf * buf, FILE* out)); 1093 extern struct Buf *buf_linedir PROTO((struct Buf *buf, const char* filename, int lineno)); 1094 1095 extern struct Buf userdef_buf; /* a string buffer for #define's generated by user-options on cmd line. */ 1096 extern struct Buf defs_buf; /* a char* buffer to save #define'd some symbols generated by flex. */ 1097 extern struct Buf yydmap_buf; /* a string buffer to hold yydmap elements */ 1098 extern struct Buf m4defs_buf; /* Holds m4 definitions. */ 1099 extern struct Buf top_buf; /* contains %top code. String buffer. */ 1100 1101 /* For blocking out code from the header file. */ 1102 #define OUT_BEGIN_CODE() outn("m4_ifdef( [[M4_YY_IN_HEADER]],,[[") 1103 #define OUT_END_CODE() outn("]])") 1104 1105 /* For setjmp/longjmp (instead of calling exit(2)). Linkage in main.c */ 1106 extern jmp_buf flex_main_jmp_buf; 1107 1108 #define FLEX_EXIT(status) longjmp(flex_main_jmp_buf,(status)+1) 1109 1110 /* Removes all \n and \r chars from tail of str. returns str. */ 1111 extern char *chomp (char *str); 1112 1113 /* ctype functions forced to return boolean */ 1114 #define b_isalnum(c) (isalnum(c)?true:false) 1115 #define b_isalpha(c) (isalpha(c)?true:false) 1116 #define b_isascii(c) (isascii(c)?true:false) 1117 #define b_isblank(c) (isblank(c)?true:false) 1118 #define b_iscntrl(c) (iscntrl(c)?true:false) 1119 #define b_isdigit(c) (isdigit(c)?true:false) 1120 #define b_isgraph(c) (isgraph(c)?true:false) 1121 #define b_islower(c) (islower(c)?true:false) 1122 #define b_isprint(c) (isprint(c)?true:false) 1123 #define b_ispunct(c) (ispunct(c)?true:false) 1124 #define b_isspace(c) (isspace(c)?true:false) 1125 #define b_isupper(c) (isupper(c)?true:false) 1126 #define b_isxdigit(c) (isxdigit(c)?true:false) 1127 1128 /* return true if char is uppercase or lowercase. */ 1129 bool has_case(int c); 1130 1131 /* Change case of character if possible. */ 1132 int reverse_case(int c); 1133 1134 /* return false if [c1-c2] is ambiguous for a caseless scanner. */ 1135 bool range_covers_case (int c1, int c2); 1136 1137 /* 1138 * From "filter.c" 1139 */ 1140 1141 /** A single stdio filter to execute. 1142 * The filter may be external, such as "sed", or it 1143 * may be internal, as a function call. 1144 */ 1145 struct filter { 1146 int (*filter_func)(struct filter*); /**< internal filter function */ 1147 void * extra; /**< extra data passed to filter_func */ 1148 int argc; /**< arg count */ 1149 const char ** argv; /**< arg vector, \0-terminated */ 1150 struct filter * next; /**< next filter or NULL */ 1151 }; 1152 1153 /* output filter chain */ 1154 extern struct filter * output_chain; 1155 extern struct filter *filter_create_ext PROTO((struct filter * chain, const char *cmd, ...)); 1156 struct filter *filter_create_int PROTO((struct filter *chain, 1157 int (*filter_func) (struct filter *), 1158 void *extra)); 1159 extern bool filter_apply_chain PROTO((struct filter * chain)); 1160 extern int filter_truncate (struct filter * chain, int max_len); 1161 extern int filter_tee_header PROTO((struct filter *chain)); 1162 extern int filter_fix_linedirs PROTO((struct filter *chain)); 1163 1164 1165 /* 1166 * From "regex.c" 1167 */ 1168 1169 extern regex_t regex_linedir, regex_blank_line; 1170 bool flex_init_regex(void); 1171 void flex_regcomp(regex_t *preg, const char *regex, int cflags); 1172 char *regmatch_dup (regmatch_t * m, const char *src); 1173 char *regmatch_cpy (regmatch_t * m, char *dest, const char *src); 1174 int regmatch_len (regmatch_t * m); 1175 int regmatch_strtol (regmatch_t * m, const char *src, char **endptr, int base); 1176 bool regmatch_empty (regmatch_t * m); 1177 1178 /* From "scanflags.h" */ 1179 typedef unsigned int scanflags_t; 1180 extern scanflags_t* _sf_stk; 1181 extern size_t _sf_top_ix, _sf_max; /**< stack of scanner flags. */ 1182 #define _SF_CASE_INS 0x0001 1183 #define _SF_DOT_ALL 0x0002 1184 #define _SF_SKIP_WS 0x0004 1185 #define sf_top() (_sf_stk[_sf_top_ix]) 1186 #define sf_case_ins() (sf_top() & _SF_CASE_INS) 1187 #define sf_dot_all() (sf_top() & _SF_DOT_ALL) 1188 #define sf_skip_ws() (sf_top() & _SF_SKIP_WS) 1189 #define sf_set_case_ins(X) ((X) ? (sf_top() |= _SF_CASE_INS) : (sf_top() &= ~_SF_CASE_INS)) 1190 #define sf_set_dot_all(X) ((X) ? (sf_top() |= _SF_DOT_ALL) : (sf_top() &= ~_SF_DOT_ALL)) 1191 #define sf_set_skip_ws(X) ((X) ? (sf_top() |= _SF_SKIP_WS) : (sf_top() &= ~_SF_SKIP_WS)) 1192 extern void sf_init(void); 1193 extern void sf_push(void); 1194 extern void sf_pop(void); 1195 1196 1197 #endif /* not defined FLEXDEF_H */ 1198