1 /* Extended regular expression matching and search library.
2 Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
6 This program is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2, or (at your option)
11 This program is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License along
17 with this program; if not, write to the Free Software Foundation,
18 Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
20 #ifndef _REGEX_INTERNAL_H
21 #define _REGEX_INTERNAL_H 1
34 #if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC
35 # include <langinfo.h>
37 #if defined HAVE_LOCALE_H || defined _LIBC
40 #if defined HAVE_WCHAR_H || defined _LIBC
42 #endif /* HAVE_WCHAR_H || _LIBC */
43 #if defined HAVE_WCTYPE_H || defined _LIBC
45 #endif /* HAVE_WCTYPE_H || _LIBC */
48 # include <bits/libc-lock.h>
50 # define __libc_lock_init(NAME) do { } while (0)
51 # define __libc_lock_lock(NAME) do { } while (0)
52 # define __libc_lock_unlock(NAME) do { } while (0)
55 /* In case that the system doesn't have isblank(). */
56 #if !defined _LIBC && !HAVE_DECL_ISBLANK && !defined isblank
57 # define isblank(ch) ((ch) == ' ' || (ch) == '\t')
61 # ifndef _RE_DEFINE_LOCALE_FUNCTIONS
62 # define _RE_DEFINE_LOCALE_FUNCTIONS 1
63 # include <locale/localeinfo.h>
64 # include <locale/elem-hash.h>
65 # include <locale/coll-lookup.h>
69 /* This is for other GNU distributions with internationalized messages. */
70 #if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC
74 # define gettext(msgid) \
75 INTUSE(__dcgettext) (_libc_intl_domainname, msgid, LC_MESSAGES)
78 # define gettext(msgid) (msgid)
82 /* This define is so xgettext can find the internationalizable
84 # define gettext_noop(String) String
87 /* For loser systems without the definition. */
89 # define SIZE_MAX ((size_t) -1)
92 #if (defined MB_CUR_MAX && HAVE_LOCALE_H && HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_WCRTOMB && HAVE_MBRTOWC && HAVE_WCSCOLL) || _LIBC
93 # define RE_ENABLE_I18N
97 # define BE(expr, val) __builtin_expect (expr, val)
99 # define BE(expr, val) (expr)
105 /* Number of ASCII characters. */
106 #define ASCII_CHARS 0x80
108 /* Number of single byte characters. */
109 #define SBC_MAX (UCHAR_MAX + 1)
111 #define COLL_ELEM_LEN_MAX 8
113 /* The character which represents newline. */
114 #define NEWLINE_CHAR '\n'
115 #define WIDE_NEWLINE_CHAR L'\n'
117 /* Rename to standard API for using out of glibc. */
119 # define __wctype wctype
120 # define __iswctype iswctype
121 # define __btowc btowc
123 # define __mempcpy mempcpy
125 # define __wcrtomb wcrtomb
126 # define __regfree regfree
127 # define attribute_hidden
128 #endif /* not _LIBC */
130 #if __GNUC__ >= 4 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)
131 # define __attribute(arg) __attribute__ (arg)
133 # define __attribute(arg)
136 typedef __re_idx_t Idx;
138 /* Special return value for failure to match. */
139 #define REG_MISSING ((Idx) -1)
141 /* Special return value for internal error. */
142 #define REG_ERROR ((Idx) -2)
144 /* Test whether N is a valid index, and is not one of the above. */
145 #ifdef _REGEX_LARGE_OFFSETS
146 # define REG_VALID_INDEX(n) ((Idx) (n) < REG_ERROR)
148 # define REG_VALID_INDEX(n) (0 <= (n))
151 /* Test whether N is a valid nonzero index. */
152 #ifdef _REGEX_LARGE_OFFSETS
153 # define REG_VALID_NONZERO_INDEX(n) ((Idx) ((n) - 1) < (Idx) (REG_ERROR - 1))
155 # define REG_VALID_NONZERO_INDEX(n) (0 < (n))
158 /* A hash value, suitable for computing hash tables. */
159 typedef __re_size_t re_hashval_t;
161 /* An integer used to represent a set of bits. It must be unsigned,
162 and must be at least as wide as unsigned int. */
163 typedef unsigned long int bitset_word_t;
164 /* All bits set in a bitset_word_t. */
165 #define BITSET_WORD_MAX ULONG_MAX
167 /* Number of bits in a bitset_word_t. For portability to hosts with
168 padding bits, do not use '(sizeof (bitset_word_t) * CHAR_BIT)';
169 instead, deduce it directly from BITSET_WORD_MAX. Avoid
170 greater-than-32-bit integers and unconditional shifts by more than
171 31 bits, as they're not portable. */
172 #if BITSET_WORD_MAX == 0xffffffff
173 # define BITSET_WORD_BITS 32
174 #elif BITSET_WORD_MAX >> 31 >> 5 == 1
175 # define BITSET_WORD_BITS 36
176 #elif BITSET_WORD_MAX >> 31 >> 16 == 1
177 # define BITSET_WORD_BITS 48
178 #elif BITSET_WORD_MAX >> 31 >> 28 == 1
179 # define BITSET_WORD_BITS 60
180 #elif BITSET_WORD_MAX >> 31 >> 31 >> 1 == 1
181 # define BITSET_WORD_BITS 64
182 #elif BITSET_WORD_MAX >> 31 >> 31 >> 9 == 1
183 # define BITSET_WORD_BITS 72
184 #elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 3 == 1
185 # define BITSET_WORD_BITS 128
186 #elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 == 1
187 # define BITSET_WORD_BITS 256
188 #elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 > 1
189 # define BITSET_WORD_BITS 257 /* any value > SBC_MAX will do here */
190 # if BITSET_WORD_BITS <= SBC_MAX
191 # error "Invalid SBC_MAX"
193 #elif BITSET_WORD_MAX == (0xffffffff + 2) * 0xffffffff
194 /* Work around a bug in 64-bit PGC (before version 6.1-2), where the
195 preprocessor mishandles large unsigned values as if they were signed. */
196 # define BITSET_WORD_BITS 64
198 # error "Add case for new bitset_word_t size"
201 /* Number of bitset_word_t values in a bitset_t. */
202 #define BITSET_WORDS ((SBC_MAX + BITSET_WORD_BITS - 1) / BITSET_WORD_BITS)
204 typedef bitset_word_t bitset_t[BITSET_WORDS];
205 typedef bitset_word_t *re_bitset_ptr_t;
206 typedef const bitset_word_t *re_const_bitset_ptr_t;
208 #define PREV_WORD_CONSTRAINT 0x0001
209 #define PREV_NOTWORD_CONSTRAINT 0x0002
210 #define NEXT_WORD_CONSTRAINT 0x0004
211 #define NEXT_NOTWORD_CONSTRAINT 0x0008
212 #define PREV_NEWLINE_CONSTRAINT 0x0010
213 #define NEXT_NEWLINE_CONSTRAINT 0x0020
214 #define PREV_BEGBUF_CONSTRAINT 0x0040
215 #define NEXT_ENDBUF_CONSTRAINT 0x0080
216 #define WORD_DELIM_CONSTRAINT 0x0100
217 #define NOT_WORD_DELIM_CONSTRAINT 0x0200
221 INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
222 WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
223 WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
224 INSIDE_NOTWORD = PREV_NOTWORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
225 LINE_FIRST = PREV_NEWLINE_CONSTRAINT,
226 LINE_LAST = NEXT_NEWLINE_CONSTRAINT,
227 BUF_FIRST = PREV_BEGBUF_CONSTRAINT,
228 BUF_LAST = NEXT_ENDBUF_CONSTRAINT,
229 WORD_DELIM = WORD_DELIM_CONSTRAINT,
230 NOT_WORD_DELIM = NOT_WORD_DELIM_CONSTRAINT
244 /* Node type, These are used by token, node, tree. */
250 #ifdef RE_ENABLE_I18N
253 #endif /* RE_ENABLE_I18N */
255 /* We define EPSILON_BIT as a macro so that OP_OPEN_SUBEXP is used
256 when the debugger shows values of this enum type. */
257 #define EPSILON_BIT 8
258 OP_OPEN_SUBEXP = EPSILON_BIT | 0,
259 OP_CLOSE_SUBEXP = EPSILON_BIT | 1,
260 OP_ALT = EPSILON_BIT | 2,
261 OP_DUP_ASTERISK = EPSILON_BIT | 3,
262 ANCHOR = EPSILON_BIT | 4,
264 /* Tree type, these are used only by tree. */
268 /* Token type, these are used only by token. */
280 OP_CLOSE_EQUIV_CLASS,
291 #ifdef RE_ENABLE_I18N
294 /* Multibyte characters. */
297 /* Collating symbols. */
302 /* Equivalence classes. */
304 int32_t *equiv_classes;
307 /* Range expressions. */
309 uint32_t *range_starts;
310 uint32_t *range_ends;
311 # else /* not _LIBC */
312 wchar_t *range_starts;
314 # endif /* not _LIBC */
316 /* Character classes. */
317 wctype_t *char_classes;
319 /* If this character set is the non-matching list. */
320 unsigned int non_match : 1;
322 /* # of multibyte characters. */
325 /* # of collating symbols. */
328 /* # of equivalence classes. */
331 /* # of range expressions. */
334 /* # of character classes. */
337 #endif /* RE_ENABLE_I18N */
343 unsigned char c; /* for CHARACTER */
344 re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */
345 #ifdef RE_ENABLE_I18N
346 re_charset_t *mbcset; /* for COMPLEX_BRACKET */
347 #endif /* RE_ENABLE_I18N */
348 Idx idx; /* for BACK_REF */
349 re_context_type ctx_type; /* for ANCHOR */
352 re_token_type_t type : 8;
354 re_token_type_t type;
356 unsigned int constraint : 10; /* context constraint */
357 unsigned int duplicated : 1;
358 unsigned int opt_subexp : 1;
359 #ifdef RE_ENABLE_I18N
360 unsigned int accept_mb : 1;
361 /* These 2 bits can be moved into the union if needed (e.g. if running out
362 of bits; move opr.c to opr.c.c and move the flags to opr.c.flags). */
363 unsigned int mb_partial : 1;
365 unsigned int word_char : 1;
368 #define IS_EPSILON_NODE(type) ((type) & EPSILON_BIT)
372 /* Indicate the raw buffer which is the original string passed as an
373 argument of regexec(), re_search(), etc.. */
374 const unsigned char *raw_mbs;
375 /* Store the multibyte string. In case of "case insensitive mode" like
376 REG_ICASE, upper cases of the string are stored, otherwise MBS points
377 the same address that RAW_MBS points. */
379 #ifdef RE_ENABLE_I18N
380 /* Store the wide character string which is corresponding to MBS. */
385 /* Index in RAW_MBS. Each character mbs[i] corresponds to
386 raw_mbs[raw_mbs_idx + i]. */
388 /* The length of the valid characters in the buffers. */
390 /* The corresponding number of bytes in raw_mbs array. */
392 /* The length of the buffers MBS and WCS. */
394 /* The index in MBS, which is updated by re_string_fetch_byte. */
396 /* length of RAW_MBS array. */
398 /* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */
400 /* End of the buffer may be shorter than its length in the cases such
401 as re_match_2, re_search_2. Then, we use STOP for end of the buffer
404 /* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */
407 /* The context of mbs[0]. We store the context independently, since
408 the context of mbs[0] may be different from raw_mbs[0], which is
409 the beginning of the input string. */
410 unsigned int tip_context;
411 /* The translation passed as a part of an argument of re_compile_pattern. */
412 RE_TRANSLATE_TYPE trans;
413 /* Copy of re_dfa_t's word_char. */
414 re_const_bitset_ptr_t word_char;
415 /* true if REG_ICASE. */
417 unsigned char is_utf8;
418 unsigned char map_notascii;
419 unsigned char mbs_allocated;
420 unsigned char offsets_needed;
421 unsigned char newline_anchor;
422 unsigned char word_ops_used;
425 typedef struct re_string_t re_string_t;
429 typedef struct re_dfa_t re_dfa_t;
433 # define internal_function __attribute ((regparm (3), stdcall))
435 # define internal_function
439 static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr,
442 #ifdef RE_ENABLE_I18N
443 static void build_wcs_buffer (re_string_t *pstr) internal_function;
444 static reg_errcode_t build_wcs_upper_buffer (re_string_t *pstr)
446 #endif /* RE_ENABLE_I18N */
447 static void build_upper_buffer (re_string_t *pstr) internal_function;
448 static void re_string_translate_buffer (re_string_t *pstr) internal_function;
449 static unsigned int re_string_context_at (const re_string_t *input, Idx idx,
451 internal_function __attribute ((pure));
452 #define re_string_peek_byte(pstr, offset) \
453 ((pstr)->mbs[(pstr)->cur_idx + offset])
454 #define re_string_fetch_byte(pstr) \
455 ((pstr)->mbs[(pstr)->cur_idx++])
456 #define re_string_first_byte(pstr, idx) \
457 ((idx) == (pstr)->valid_len || (pstr)->wcs[idx] != WEOF)
458 #define re_string_is_single_byte_char(pstr, idx) \
459 ((pstr)->wcs[idx] != WEOF && ((pstr)->valid_len == (idx) + 1 \
460 || (pstr)->wcs[(idx) + 1] != WEOF))
461 #define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx)
462 #define re_string_cur_idx(pstr) ((pstr)->cur_idx)
463 #define re_string_get_buffer(pstr) ((pstr)->mbs)
464 #define re_string_length(pstr) ((pstr)->len)
465 #define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx])
466 #define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx))
467 #define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx))
473 /* The OS usually guarantees only one guard page at the bottom of the stack,
474 and a page size can be as small as 4096 bytes. So we cannot safely
475 allocate anything larger than 4096 bytes. Also care for the possibility
476 of a few compiler-allocated temporary stack slots. */
477 # define __libc_use_alloca(n) ((n) < 4032)
479 /* alloca is implemented with malloc, so just use malloc. */
480 # define __libc_use_alloca(n) 0
485 # define MAX(a,b) ((a) < (b) ? (b) : (a))
488 #define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t)))
489 #define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t)))
490 #define re_free(p) free (p)
494 struct bin_tree_t *parent;
495 struct bin_tree_t *left;
496 struct bin_tree_t *right;
497 struct bin_tree_t *first;
498 struct bin_tree_t *next;
502 /* `node_idx' is the index in dfa->nodes, if `type' == 0.
503 Otherwise `type' indicate the type of this node. */
506 typedef struct bin_tree_t bin_tree_t;
508 #define BIN_TREE_STORAGE_SIZE \
509 ((1024 - sizeof (void *)) / sizeof (bin_tree_t))
511 struct bin_tree_storage_t
513 struct bin_tree_storage_t *next;
514 bin_tree_t data[BIN_TREE_STORAGE_SIZE];
516 typedef struct bin_tree_storage_t bin_tree_storage_t;
518 #define CONTEXT_WORD 1
519 #define CONTEXT_NEWLINE (CONTEXT_WORD << 1)
520 #define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1)
521 #define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1)
523 #define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD)
524 #define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE)
525 #define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF)
526 #define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF)
527 #define IS_ORDINARY_CONTEXT(c) ((c) == 0)
529 #define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_')
530 #define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR)
531 #define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_')
532 #define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR)
534 #define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \
535 ((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
536 || ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
537 || ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\
538 || ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context)))
540 #define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \
541 ((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
542 || (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
543 || (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \
544 || (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context)))
550 re_node_set non_eps_nodes;
551 re_node_set inveclosure;
552 re_node_set *entrance_nodes;
553 struct re_dfastate_t **trtable, **word_trtable;
554 unsigned int context : 4;
555 unsigned int halt : 1;
556 /* If this state can accept `multi byte'.
557 Note that we refer to multibyte characters, and multi character
558 collating elements as `multi byte'. */
559 unsigned int accept_mb : 1;
560 /* If this state has backreference node(s). */
561 unsigned int has_backref : 1;
562 unsigned int has_constraint : 1;
564 typedef struct re_dfastate_t re_dfastate_t;
566 struct re_state_table_entry
570 re_dfastate_t **array;
573 /* Array type used in re_sub_match_last_t and re_sub_match_top_t. */
579 re_dfastate_t **array;
582 /* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */
587 Idx str_idx; /* The position NODE match at. */
589 } re_sub_match_last_t;
591 /* Store information about the node NODE whose type is OP_OPEN_SUBEXP.
592 And information about the node, whose type is OP_CLOSE_SUBEXP,
593 corresponding to NODE is stored in LASTS. */
600 Idx alasts; /* Allocation size of LASTS. */
601 Idx nlasts; /* The number of LASTS. */
602 re_sub_match_last_t **lasts;
603 } re_sub_match_top_t;
605 struct re_backref_cache_entry
613 unsigned short int eps_reachable_subexps_map;
618 /* The string object corresponding to the input string. */
620 #if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)
621 const re_dfa_t *const dfa;
625 /* EFLAGS of the argument of regexec. */
627 /* Where the matching ends. */
630 /* The state log used by the matcher. */
631 re_dfastate_t **state_log;
633 /* Back reference cache. */
636 struct re_backref_cache_entry *bkref_ents;
640 re_sub_match_top_t **sub_tops;
641 } re_match_context_t;
645 re_dfastate_t **sifted_states;
646 re_dfastate_t **limited_states;
652 struct re_fail_stack_ent_t
657 re_node_set eps_via_nodes;
660 struct re_fail_stack_t
664 struct re_fail_stack_ent_t *stack;
675 re_node_set *eclosures;
676 re_node_set *inveclosures;
677 struct re_state_table_entry *state_table;
678 re_dfastate_t *init_state;
679 re_dfastate_t *init_state_word;
680 re_dfastate_t *init_state_nl;
681 re_dfastate_t *init_state_begbuf;
682 bin_tree_t *str_tree;
683 bin_tree_storage_t *str_tree_storage;
684 re_bitset_ptr_t sb_char;
685 int str_tree_storage_idx;
687 /* number of subexpressions `re_nsub' is in regex_t. */
688 re_hashval_t state_hash_mask;
690 Idx nbackref; /* The number of backreference in this dfa. */
692 /* Bitmap expressing which backreference is used. */
693 bitset_word_t used_bkref_map;
694 bitset_word_t completed_bkref_map;
696 unsigned int has_plural_match : 1;
697 /* If this dfa has "multibyte node", which is a backreference or
698 a node which can accept multibyte character or multi character
699 collating element. */
700 unsigned int has_mb_node : 1;
701 unsigned int is_utf8 : 1;
702 unsigned int map_notascii : 1;
703 unsigned int word_ops_used : 1;
712 __libc_lock_define (, lock)
716 #define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set))
717 #define re_node_set_remove(set,id) \
718 (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1))
719 #define re_node_set_empty(p) ((p)->nelem = 0)
720 #define re_node_set_free(set) re_free ((set)->elems)
734 bracket_elem_type type;
744 /* Inline functions for bitset_t operation. */
747 bitset_set (bitset_t set, Idx i)
749 set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS;
753 bitset_clear (bitset_t set, Idx i)
755 set[i / BITSET_WORD_BITS] &= ~ ((bitset_word_t) 1 << i % BITSET_WORD_BITS);
759 bitset_contain (const bitset_t set, Idx i)
761 return (set[i / BITSET_WORD_BITS] >> i % BITSET_WORD_BITS) & 1;
765 bitset_empty (bitset_t set)
767 memset (set, '\0', sizeof (bitset_t));
771 bitset_set_all (bitset_t set)
773 memset (set, -1, sizeof (bitset_word_t) * (SBC_MAX / BITSET_WORD_BITS));
774 if (SBC_MAX % BITSET_WORD_BITS != 0)
775 set[BITSET_WORDS - 1] =
776 ((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1;
780 bitset_copy (bitset_t dest, const bitset_t src)
782 memcpy (dest, src, sizeof (bitset_t));
786 bitset_not (bitset_t set)
789 for (bitset_i = 0; bitset_i < SBC_MAX / BITSET_WORD_BITS; ++bitset_i)
790 set[bitset_i] = ~set[bitset_i];
791 if (SBC_MAX % BITSET_WORD_BITS != 0)
792 set[BITSET_WORDS - 1] =
793 ((((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1)
794 & ~set[BITSET_WORDS - 1]);
798 bitset_merge (bitset_t dest, const bitset_t src)
801 for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
802 dest[bitset_i] |= src[bitset_i];
806 bitset_mask (bitset_t dest, const bitset_t src)
809 for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
810 dest[bitset_i] &= src[bitset_i];
813 #ifdef RE_ENABLE_I18N
814 /* Inline functions for re_string. */
816 internal_function __attribute ((pure))
817 re_string_char_size_at (const re_string_t *pstr, Idx idx)
820 if (pstr->mb_cur_max == 1)
822 for (byte_idx = 1; idx + byte_idx < pstr->valid_len; ++byte_idx)
823 if (pstr->wcs[idx + byte_idx] != WEOF)
829 internal_function __attribute ((pure))
830 re_string_wchar_at (const re_string_t *pstr, Idx idx)
832 if (pstr->mb_cur_max == 1)
833 return (wint_t) pstr->mbs[idx];
834 return (wint_t) pstr->wcs[idx];
838 internal_function __attribute ((pure))
839 re_string_elem_size_at (const re_string_t *pstr, Idx idx)
842 const unsigned char *p, *extra;
843 const int32_t *table, *indirect;
845 # include <locale/weight.h>
846 uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
850 table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
851 extra = (const unsigned char *)
852 _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
853 indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
854 _NL_COLLATE_INDIRECTMB);
857 return p - pstr->mbs - idx;
863 #endif /* RE_ENABLE_I18N */
865 #endif /* _REGEX_INTERNAL_H */