/LotsaGlass/GLObject/Data management/regex.c
C | 10689 lines | 8466 code | 922 blank | 1301 comment | 2532 complexity | 7cd64486ca5a35e4d4301e2360a0a241 MD5 | raw file
Large files files are truncated, but you can click here to view the full file
- /* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003, 2005 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
- #ifdef HAVE_CONFIG_H
- #include "config.h"
- #endif
- #include <stdbool.h>
- #define __USE_GNU
- /* Make sure noone compiles this code with a C++ compiler. */
- #ifdef __cplusplus
- # error "This is C code, use a C compiler"
- #endif
- #ifdef _LIBC
- /* We have to keep the namespace clean. */
- # define regfree(preg) __regfree (preg)
- # define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef)
- # define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags)
- # define regerror(errcode, preg, errbuf, errbuf_size) \
- __regerror(errcode, preg, errbuf, errbuf_size)
- # define re_set_registers(bu, re, nu, st, en) \
- __re_set_registers (bu, re, nu, st, en)
- # define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \
- __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop)
- # define re_match(bufp, string, size, pos, regs) \
- __re_match (bufp, string, size, pos, regs)
- # define re_search(bufp, string, size, startpos, range, regs) \
- __re_search (bufp, string, size, startpos, range, regs)
- # define re_compile_pattern(pattern, length, bufp) \
- __re_compile_pattern (pattern, length, bufp)
- # define re_set_syntax(syntax) __re_set_syntax (syntax)
- # define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \
- __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop)
- # define re_compile_fastmap(bufp) __re_compile_fastmap (bufp)
- # include "../locale/localeinfo.h"
- #endif
- /* On some systems, limits.h sets RE_DUP_MAX to a lower value than
- GNU regex allows. Include it before <regex.h>, which correctly
- #undefs RE_DUP_MAX and sets it to the right value. */
- #include <limits.h>
- #include "regex.h"
- /* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003, 2004, 2005, 2007 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
- #ifndef _REGEX_INTERNAL_H
- #define _REGEX_INTERNAL_H 1
- #include <assert.h>
- #include <ctype.h>
- #include <stdio.h>
- #include <stdlib.h>
- #include <string.h>
- #if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC
- # include <langinfo.h>
- #endif
- #if defined HAVE_LOCALE_H || defined _LIBC
- # include <locale.h>
- #endif
- #if defined HAVE_WCHAR_H || defined _LIBC
- # include <wchar.h>
- #endif /* HAVE_WCHAR_H || _LIBC */
- #if defined HAVE_WCTYPE_H || defined _LIBC
- # include <wctype.h>
- #endif /* HAVE_WCTYPE_H || _LIBC */
- #if defined HAVE_STDBOOL_H || defined _LIBC
- # include <stdbool.h>
- #endif /* HAVE_STDBOOL_H || _LIBC */
- #if defined HAVE_STDINT_H || defined _LIBC
- # include <stdint.h>
- #endif /* HAVE_STDINT_H || _LIBC */
- #if defined _LIBC
- # include <bits/libc-lock.h>
- #else
- # define __libc_lock_define(CLASS,NAME)
- # define __libc_lock_init(NAME) do { } while (0)
- # define __libc_lock_lock(NAME) do { } while (0)
- # define __libc_lock_unlock(NAME) do { } while (0)
- #endif
- /* In case that the system doesn't have isblank(). */
- #if !defined _LIBC && !defined HAVE_ISBLANK && !defined isblank
- # define isblank(ch) ((ch) == ' ' || (ch) == '\t')
- #endif
- #ifdef _LIBC
- # ifndef _RE_DEFINE_LOCALE_FUNCTIONS
- # define _RE_DEFINE_LOCALE_FUNCTIONS 1
- # include <locale/localeinfo.h>
- # include <locale/elem-hash.h>
- # include <locale/coll-lookup.h>
- # endif
- #endif
- /* This is for other GNU distributions with internationalized messages. */
- #if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC
- # include <libintl.h>
- # ifdef _LIBC
- # undef gettext
- # define gettext(msgid) \
- INTUSE(__dcgettext) (_libc_intl_domainname, msgid, LC_MESSAGES)
- # endif
- #else
- # define gettext(msgid) (msgid)
- #endif
- #ifndef gettext_noop
- /* This define is so xgettext can find the internationalizable
- strings. */
- # define gettext_noop(String) String
- #endif
- /* For loser systems without the definition. */
- #ifndef SIZE_MAX
- # define SIZE_MAX ((size_t) -1)
- #endif
- #if (defined MB_CUR_MAX && HAVE_LOCALE_H && HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_WCRTOMB && HAVE_MBRTOWC && HAVE_WCSCOLL) || _LIBC
- # define RE_ENABLE_I18N
- #endif
- #if __GNUC__ >= 3
- # define BE(expr, val) __builtin_expect (expr, val)
- #else
- # define BE(expr, val) (expr)
- # define inline
- #endif
- /* Number of single byte character. */
- #define SBC_MAX 256
- #define COLL_ELEM_LEN_MAX 8
- /* The character which represents newline. */
- #define NEWLINE_CHAR '\n'
- #define WIDE_NEWLINE_CHAR L'\n'
- /* Rename to standard API for using out of glibc. */
- #ifndef _LIBC
- # define __wctype wctype
- # define __iswctype iswctype
- # define __btowc btowc
- # define __mempcpy mempcpy
- # define __wcrtomb wcrtomb
- # define __regfree regfree
- # define attribute_hidden
- #endif /* not _LIBC */
- #ifdef __GNUC__
- # define __attribute(arg) __attribute__ (arg)
- #else
- # define __attribute(arg)
- #endif
- extern const char __re_error_msgid[] attribute_hidden;
- extern const size_t __re_error_msgid_idx[] attribute_hidden;
- /* An integer used to represent a set of bits. It must be unsigned,
- and must be at least as wide as unsigned int. */
- typedef unsigned long int bitset_word_t;
- /* All bits set in a bitset_word_t. */
- #define BITSET_WORD_MAX ULONG_MAX
- /* Number of bits in a bitset_word_t. */
- #define BITSET_WORD_BITS (sizeof (bitset_word_t) * CHAR_BIT)
- /* Number of bitset_word_t in a bit_set. */
- #define BITSET_WORDS (SBC_MAX / BITSET_WORD_BITS)
- typedef bitset_word_t bitset_t[BITSET_WORDS];
- typedef bitset_word_t *re_bitset_ptr_t;
- typedef const bitset_word_t *re_const_bitset_ptr_t;
- #define bitset_set(set,i) \
- (set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS)
- #define bitset_clear(set,i) \
- (set[i / BITSET_WORD_BITS] &= ~((bitset_word_t) 1 << i % BITSET_WORD_BITS))
- #define bitset_contain(set,i) \
- (set[i / BITSET_WORD_BITS] & ((bitset_word_t) 1 << i % BITSET_WORD_BITS))
- #define bitset_empty(set) memset (set, '\0', sizeof (bitset_t))
- #define bitset_set_all(set) memset (set, '\xff', sizeof (bitset_t))
- #define bitset_copy(dest,src) memcpy (dest, src, sizeof (bitset_t))
- #define PREV_WORD_CONSTRAINT 0x0001
- #define PREV_NOTWORD_CONSTRAINT 0x0002
- #define NEXT_WORD_CONSTRAINT 0x0004
- #define NEXT_NOTWORD_CONSTRAINT 0x0008
- #define PREV_NEWLINE_CONSTRAINT 0x0010
- #define NEXT_NEWLINE_CONSTRAINT 0x0020
- #define PREV_BEGBUF_CONSTRAINT 0x0040
- #define NEXT_ENDBUF_CONSTRAINT 0x0080
- #define WORD_DELIM_CONSTRAINT 0x0100
- #define NOT_WORD_DELIM_CONSTRAINT 0x0200
- typedef enum
- {
- INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
- WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
- WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
- INSIDE_NOTWORD = PREV_NOTWORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
- LINE_FIRST = PREV_NEWLINE_CONSTRAINT,
- LINE_LAST = NEXT_NEWLINE_CONSTRAINT,
- BUF_FIRST = PREV_BEGBUF_CONSTRAINT,
- BUF_LAST = NEXT_ENDBUF_CONSTRAINT,
- WORD_DELIM = WORD_DELIM_CONSTRAINT,
- NOT_WORD_DELIM = NOT_WORD_DELIM_CONSTRAINT
- } re_context_type;
- typedef struct
- {
- int alloc;
- int nelem;
- int *elems;
- } re_node_set;
- typedef enum
- {
- NON_TYPE = 0,
- /* Node type, These are used by token, node, tree. */
- CHARACTER = 1,
- END_OF_RE = 2,
- SIMPLE_BRACKET = 3,
- OP_BACK_REF = 4,
- OP_PERIOD = 5,
- #ifdef RE_ENABLE_I18N
- COMPLEX_BRACKET = 6,
- OP_UTF8_PERIOD = 7,
- #endif /* RE_ENABLE_I18N */
- /* We define EPSILON_BIT as a macro so that OP_OPEN_SUBEXP is used
- when the debugger shows values of this enum type. */
- #define EPSILON_BIT 8
- OP_OPEN_SUBEXP = EPSILON_BIT | 0,
- OP_CLOSE_SUBEXP = EPSILON_BIT | 1,
- OP_ALT = EPSILON_BIT | 2,
- OP_DUP_ASTERISK = EPSILON_BIT | 3,
- ANCHOR = EPSILON_BIT | 4,
- /* Tree type, these are used only by tree. */
- CONCAT = 16,
- SUBEXP = 17,
- /* Token type, these are used only by token. */
- OP_DUP_PLUS = 18,
- OP_DUP_QUESTION,
- OP_OPEN_BRACKET,
- OP_CLOSE_BRACKET,
- OP_CHARSET_RANGE,
- OP_OPEN_DUP_NUM,
- OP_CLOSE_DUP_NUM,
- OP_NON_MATCH_LIST,
- OP_OPEN_COLL_ELEM,
- OP_CLOSE_COLL_ELEM,
- OP_OPEN_EQUIV_CLASS,
- OP_CLOSE_EQUIV_CLASS,
- OP_OPEN_CHAR_CLASS,
- OP_CLOSE_CHAR_CLASS,
- OP_WORD,
- OP_NOTWORD,
- OP_SPACE,
- OP_NOTSPACE,
- BACK_SLASH
- } re_token_type_t;
- #ifdef RE_ENABLE_I18N
- typedef struct
- {
- /* Multibyte characters. */
- wchar_t *mbchars;
- /* Collating symbols. */
- # ifdef _LIBC
- int32_t *coll_syms;
- # endif
- /* Equivalence classes. */
- # ifdef _LIBC
- int32_t *equiv_classes;
- # endif
- /* Range expressions. */
- # ifdef _LIBC
- uint32_t *range_starts;
- uint32_t *range_ends;
- # else /* not _LIBC */
- wchar_t *range_starts;
- wchar_t *range_ends;
- # endif /* not _LIBC */
- /* Character classes. */
- wctype_t *char_classes;
- /* If this character set is the non-matching list. */
- unsigned int non_match : 1;
- /* # of multibyte characters. */
- int nmbchars;
- /* # of collating symbols. */
- int ncoll_syms;
- /* # of equivalence classes. */
- int nequiv_classes;
- /* # of range expressions. */
- int nranges;
- /* # of character classes. */
- int nchar_classes;
- } re_charset_t;
- #endif /* RE_ENABLE_I18N */
- typedef struct
- {
- union
- {
- unsigned char c; /* for CHARACTER */
- re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */
- #ifdef RE_ENABLE_I18N
- re_charset_t *mbcset; /* for COMPLEX_BRACKET */
- #endif /* RE_ENABLE_I18N */
- int idx; /* for BACK_REF */
- re_context_type ctx_type; /* for ANCHOR */
- } opr;
- #if __GNUC__ >= 2
- re_token_type_t type : 8;
- #else
- re_token_type_t type;
- #endif
- unsigned int constraint : 10; /* context constraint */
- unsigned int duplicated : 1;
- unsigned int opt_subexp : 1;
- #ifdef RE_ENABLE_I18N
- unsigned int accept_mb : 1;
- /* These 2 bits can be moved into the union if needed (e.g. if running out
- of bits; move opr.c to opr.c.c and move the flags to opr.c.flags). */
- unsigned int mb_partial : 1;
- #endif
- unsigned int word_char : 1;
- } re_token_t;
- #define IS_EPSILON_NODE(type) ((type) & EPSILON_BIT)
- struct re_string_t
- {
- /* Indicate the raw buffer which is the original string passed as an
- argument of regexec(), re_search(), etc.. */
- const unsigned char *raw_mbs;
- /* Store the multibyte string. In case of "case insensitive mode" like
- REG_ICASE, upper cases of the string are stored, otherwise MBS points
- the same address that RAW_MBS points. */
- unsigned char *mbs;
- #ifdef RE_ENABLE_I18N
- /* Store the wide character string which is corresponding to MBS. */
- wint_t *wcs;
- int *offsets;
- mbstate_t cur_state;
- #endif
- /* Index in RAW_MBS. Each character mbs[i] corresponds to
- raw_mbs[raw_mbs_idx + i]. */
- int raw_mbs_idx;
- /* The length of the valid characters in the buffers. */
- int valid_len;
- /* The corresponding number of bytes in raw_mbs array. */
- int valid_raw_len;
- /* The length of the buffers MBS and WCS. */
- int bufs_len;
- /* The index in MBS, which is updated by re_string_fetch_byte. */
- int cur_idx;
- /* length of RAW_MBS array. */
- int raw_len;
- /* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */
- int len;
- /* End of the buffer may be shorter than its length in the cases such
- as re_match_2, re_search_2. Then, we use STOP for end of the buffer
- instead of LEN. */
- int raw_stop;
- /* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */
- int stop;
- /* The context of mbs[0]. We store the context independently, since
- the context of mbs[0] may be different from raw_mbs[0], which is
- the beginning of the input string. */
- unsigned int tip_context;
- /* The translation passed as a part of an argument of re_compile_pattern. */
- RE_TRANSLATE_TYPE trans;
- /* Copy of re_dfa_t's word_char. */
- re_const_bitset_ptr_t word_char;
- /* 1 if REG_ICASE. */
- unsigned char icase;
- unsigned char is_utf8;
- unsigned char map_notascii;
- unsigned char mbs_allocated;
- unsigned char offsets_needed;
- unsigned char newline_anchor;
- unsigned char word_ops_used;
- int mb_cur_max;
- };
- typedef struct re_string_t re_string_t;
- struct re_dfa_t;
- typedef struct re_dfa_t re_dfa_t;
- #ifndef _LIBC
- # ifdef __i386__
- # define internal_function __attribute ((regparm (3), stdcall))
- # else
- # define internal_function
- # endif
- #endif
- #ifndef NOT_IN_libc
- static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr,
- int new_buf_len)
- internal_function;
- # ifdef RE_ENABLE_I18N
- static void build_wcs_buffer (re_string_t *pstr) internal_function;
- static int build_wcs_upper_buffer (re_string_t *pstr) internal_function;
- # endif /* RE_ENABLE_I18N */
- static void build_upper_buffer (re_string_t *pstr) internal_function;
- static void re_string_translate_buffer (re_string_t *pstr) internal_function;
- static unsigned int re_string_context_at (const re_string_t *input, int idx,
- int eflags)
- internal_function __attribute ((pure));
- #endif
- #define re_string_peek_byte(pstr, offset) \
- ((pstr)->mbs[(pstr)->cur_idx + offset])
- #define re_string_fetch_byte(pstr) \
- ((pstr)->mbs[(pstr)->cur_idx++])
- #define re_string_first_byte(pstr, idx) \
- ((idx) == (pstr)->valid_len || (pstr)->wcs[idx] != WEOF)
- #define re_string_is_single_byte_char(pstr, idx) \
- ((pstr)->wcs[idx] != WEOF && ((pstr)->valid_len == (idx) + 1 \
- || (pstr)->wcs[(idx) + 1] != WEOF))
- #define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx)
- #define re_string_cur_idx(pstr) ((pstr)->cur_idx)
- #define re_string_get_buffer(pstr) ((pstr)->mbs)
- #define re_string_length(pstr) ((pstr)->len)
- #define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx])
- #define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx))
- #define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx))
- //#include <alloca.h>
- #ifndef _LIBC
- # if HAVE_ALLOCA
- /* The OS usually guarantees only one guard page at the bottom of the stack,
- and a page size can be as small as 4096 bytes. So we cannot safely
- allocate anything larger than 4096 bytes. Also care for the possibility
- of a few compiler-allocated temporary stack slots. */
- # define __libc_use_alloca(n) ((n) < 4032)
- # else
- /* alloca is implemented with malloc, so just use malloc. */
- # define __libc_use_alloca(n) 0
- # endif
- #endif
- #define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t)))
- #define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t)))
- #define re_free(p) free (p)
- struct bin_tree_t
- {
- struct bin_tree_t *parent;
- struct bin_tree_t *left;
- struct bin_tree_t *right;
- struct bin_tree_t *first;
- struct bin_tree_t *next;
- re_token_t token;
- /* `node_idx' is the index in dfa->nodes, if `type' == 0.
- Otherwise `type' indicate the type of this node. */
- int node_idx;
- };
- typedef struct bin_tree_t bin_tree_t;
- #define BIN_TREE_STORAGE_SIZE \
- ((1024 - sizeof (void *)) / sizeof (bin_tree_t))
- struct bin_tree_storage_t
- {
- struct bin_tree_storage_t *next;
- bin_tree_t data[BIN_TREE_STORAGE_SIZE];
- };
- typedef struct bin_tree_storage_t bin_tree_storage_t;
- #define CONTEXT_WORD 1
- #define CONTEXT_NEWLINE (CONTEXT_WORD << 1)
- #define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1)
- #define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1)
- #define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD)
- #define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE)
- #define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF)
- #define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF)
- #define IS_ORDINARY_CONTEXT(c) ((c) == 0)
- #define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_')
- #define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR)
- #define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_')
- #define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR)
- #define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \
- ((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
- || ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
- || ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\
- || ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context)))
- #define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \
- ((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
- || (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
- || (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \
- || (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context)))
- struct re_dfastate_t
- {
- unsigned int hash;
- re_node_set nodes;
- re_node_set non_eps_nodes;
- re_node_set inveclosure;
- re_node_set *entrance_nodes;
- struct re_dfastate_t **trtable, **word_trtable;
- unsigned int context : 4;
- unsigned int halt : 1;
- /* If this state can accept `multi byte'.
- Note that we refer to multibyte characters, and multi character
- collating elements as `multi byte'. */
- unsigned int accept_mb : 1;
- /* If this state has backreference node(s). */
- unsigned int has_backref : 1;
- unsigned int has_constraint : 1;
- };
- typedef struct re_dfastate_t re_dfastate_t;
- struct re_state_table_entry
- {
- int num;
- int alloc;
- re_dfastate_t **array;
- };
- /* Array type used in re_sub_match_last_t and re_sub_match_top_t. */
- typedef struct
- {
- int next_idx;
- int alloc;
- re_dfastate_t **array;
- } state_array_t;
- /* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */
- typedef struct
- {
- int node;
- int str_idx; /* The position NODE match at. */
- state_array_t path;
- } re_sub_match_last_t;
- /* Store information about the node NODE whose type is OP_OPEN_SUBEXP.
- And information about the node, whose type is OP_CLOSE_SUBEXP,
- corresponding to NODE is stored in LASTS. */
- typedef struct
- {
- int str_idx;
- int node;
- state_array_t *path;
- int alasts; /* Allocation size of LASTS. */
- int nlasts; /* The number of LASTS. */
- re_sub_match_last_t **lasts;
- } re_sub_match_top_t;
- struct re_backref_cache_entry
- {
- int node;
- int str_idx;
- int subexp_from;
- int subexp_to;
- char more;
- char unused;
- unsigned short int eps_reachable_subexps_map;
- };
- typedef struct
- {
- /* The string object corresponding to the input string. */
- re_string_t input;
- #if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)
- const re_dfa_t *const dfa;
- #else
- const re_dfa_t *dfa;
- #endif
- /* EFLAGS of the argument of regexec. */
- int eflags;
- /* Where the matching ends. */
- int match_last;
- int last_node;
- /* The state log used by the matcher. */
- re_dfastate_t **state_log;
- int state_log_top;
- /* Back reference cache. */
- int nbkref_ents;
- int abkref_ents;
- struct re_backref_cache_entry *bkref_ents;
- int max_mb_elem_len;
- int nsub_tops;
- int asub_tops;
- re_sub_match_top_t **sub_tops;
- } re_match_context_t;
- typedef struct
- {
- re_dfastate_t **sifted_states;
- re_dfastate_t **limited_states;
- int last_node;
- int last_str_idx;
- re_node_set limits;
- } re_sift_context_t;
- struct re_fail_stack_ent_t
- {
- int idx;
- int node;
- regmatch_t *regs;
- re_node_set eps_via_nodes;
- };
- struct re_fail_stack_t
- {
- int num;
- int alloc;
- struct re_fail_stack_ent_t *stack;
- };
- struct re_dfa_t
- {
- re_token_t *nodes;
- size_t nodes_alloc;
- size_t nodes_len;
- int *nexts;
- int *org_indices;
- re_node_set *edests;
- re_node_set *eclosures;
- re_node_set *inveclosures;
- struct re_state_table_entry *state_table;
- re_dfastate_t *init_state;
- re_dfastate_t *init_state_word;
- re_dfastate_t *init_state_nl;
- re_dfastate_t *init_state_begbuf;
- bin_tree_t *str_tree;
- bin_tree_storage_t *str_tree_storage;
- re_bitset_ptr_t sb_char;
- int str_tree_storage_idx;
- /* number of subexpressions `re_nsub' is in regex_t. */
- unsigned int state_hash_mask;
- int init_node;
- int nbackref; /* The number of backreference in this dfa. */
- /* Bitmap expressing which backreference is used. */
- bitset_word_t used_bkref_map;
- bitset_word_t completed_bkref_map;
- unsigned int has_plural_match : 1;
- /* If this dfa has "multibyte node", which is a backreference or
- a node which can accept multibyte character or multi character
- collating element. */
- unsigned int has_mb_node : 1;
- unsigned int is_utf8 : 1;
- unsigned int map_notascii : 1;
- unsigned int word_ops_used : 1;
- int mb_cur_max;
- bitset_t word_char;
- reg_syntax_t syntax;
- int *subexp_map;
- #ifdef DEBUG
- char* re_str;
- #endif
- __libc_lock_define (, lock)
- };
- #define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set))
- #define re_node_set_remove(set,id) \
- (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1))
- #define re_node_set_empty(p) ((p)->nelem = 0)
- #define re_node_set_free(set) re_free ((set)->elems)
- typedef enum
- {
- SB_CHAR,
- MB_CHAR,
- EQUIV_CLASS,
- COLL_SYM,
- CHAR_CLASS
- } bracket_elem_type;
- typedef struct
- {
- bracket_elem_type type;
- union
- {
- unsigned char ch;
- unsigned char *name;
- wchar_t wch;
- } opr;
- } bracket_elem_t;
- /* Inline functions for bitset operation. */
- static inline void
- bitset_not (bitset_t set)
- {
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
- set[bitset_i] = ~set[bitset_i];
- }
- static inline void
- bitset_merge (bitset_t dest, const bitset_t src)
- {
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
- dest[bitset_i] |= src[bitset_i];
- }
- static inline void
- bitset_mask (bitset_t dest, const bitset_t src)
- {
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
- dest[bitset_i] &= src[bitset_i];
- }
- #ifdef RE_ENABLE_I18N
- /* Inline functions for re_string. */
- static inline int
- internal_function __attribute ((pure))
- re_string_char_size_at (const re_string_t *pstr, int idx)
- {
- int byte_idx;
- if (pstr->mb_cur_max == 1)
- return 1;
- for (byte_idx = 1; idx + byte_idx < pstr->valid_len; ++byte_idx)
- if (pstr->wcs[idx + byte_idx] != WEOF)
- break;
- return byte_idx;
- }
- static inline wint_t
- internal_function __attribute ((pure))
- re_string_wchar_at (const re_string_t *pstr, int idx)
- {
- if (pstr->mb_cur_max == 1)
- return (wint_t) pstr->mbs[idx];
- return (wint_t) pstr->wcs[idx];
- }
- # ifndef NOT_IN_libc
- static int
- internal_function __attribute ((pure))
- re_string_elem_size_at (const re_string_t *pstr, int idx)
- {
- # ifdef _LIBC
- const unsigned char *p, *extra;
- const int32_t *table, *indirect;
- int32_t tmp;
- # include <locale/weight.h>
- uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
- if (nrules != 0)
- {
- table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
- extra = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
- indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_INDIRECTMB);
- p = pstr->mbs + idx;
- tmp = findidx (&p);
- return p - pstr->mbs - idx;
- }
- else
- # endif /* _LIBC */
- return 1;
- }
- # endif
- #endif /* RE_ENABLE_I18N */
- #endif /* _REGEX_INTERNAL_H */
- /* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
- static void re_string_construct_common (const char *str, int len,
- re_string_t *pstr,
- RE_TRANSLATE_TYPE trans, int icase,
- const re_dfa_t *dfa) internal_function;
- static re_dfastate_t *create_ci_newstate (const re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int hash) internal_function;
- static re_dfastate_t *create_cd_newstate (const re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int context,
- unsigned int hash) internal_function;
- /* Functions for string operation. */
- /* This function allocate the buffers. It is necessary to call
- re_string_reconstruct before using the object. */
- static reg_errcode_t
- internal_function
- re_string_allocate (re_string_t *pstr, const char *str, int len, int init_len,
- RE_TRANSLATE_TYPE trans, int icase, const re_dfa_t *dfa)
- {
- reg_errcode_t ret;
- int init_buf_len;
- /* Ensure at least one character fits into the buffers. */
- if (init_len < dfa->mb_cur_max)
- init_len = dfa->mb_cur_max;
- init_buf_len = (len + 1 < init_len) ? len + 1: init_len;
- re_string_construct_common (str, len, pstr, trans, icase, dfa);
- ret = re_string_realloc_buffers (pstr, init_buf_len);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- pstr->word_char = dfa->word_char;
- pstr->word_ops_used = dfa->word_ops_used;
- pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str;
- pstr->valid_len = (pstr->mbs_allocated || dfa->mb_cur_max > 1) ? 0 : len;
- pstr->valid_raw_len = pstr->valid_len;
- return REG_NOERROR;
- }
- /* This function allocate the buffers, and initialize them. */
- static reg_errcode_t
- internal_function
- re_string_construct (re_string_t *pstr, const char *str, int len,
- RE_TRANSLATE_TYPE trans, int icase, const re_dfa_t *dfa)
- {
- reg_errcode_t ret;
- memset (pstr, '\0', sizeof (re_string_t));
- re_string_construct_common (str, len, pstr, trans, icase, dfa);
- if (len > 0)
- {
- ret = re_string_realloc_buffers (pstr, len + 1);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- }
- pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str;
- if (icase)
- {
- #ifdef RE_ENABLE_I18N
- if (dfa->mb_cur_max > 1)
- {
- while (1)
- {
- ret = build_wcs_upper_buffer (pstr);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- if (pstr->valid_raw_len >= len)
- break;
- if (pstr->bufs_len > pstr->valid_len + dfa->mb_cur_max)
- break;
- ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- }
- }
- else
- #endif /* RE_ENABLE_I18N */
- build_upper_buffer (pstr);
- }
- else
- {
- #ifdef RE_ENABLE_I18N
- if (dfa->mb_cur_max > 1)
- build_wcs_buffer (pstr);
- else
- #endif /* RE_ENABLE_I18N */
- {
- if (trans != NULL)
- re_string_translate_buffer (pstr);
- else
- {
- pstr->valid_len = pstr->bufs_len;
- pstr->valid_raw_len = pstr->bufs_len;
- }
- }
- }
- return REG_NOERROR;
- }
- /* Helper functions for re_string_allocate, and re_string_construct. */
- static reg_errcode_t
- internal_function
- re_string_realloc_buffers (re_string_t *pstr, int new_buf_len)
- {
- #ifdef RE_ENABLE_I18N
- if (pstr->mb_cur_max > 1)
- {
- wint_t *new_wcs = re_realloc (pstr->wcs, wint_t, new_buf_len);
- if (BE (new_wcs == NULL, 0))
- return REG_ESPACE;
- pstr->wcs = new_wcs;
- if (pstr->offsets != NULL)
- {
- int *new_offsets = re_realloc (pstr->offsets, int, new_buf_len);
- if (BE (new_offsets == NULL, 0))
- return REG_ESPACE;
- pstr->offsets = new_offsets;
- }
- }
- #endif /* RE_ENABLE_I18N */
- if (pstr->mbs_allocated)
- {
- unsigned char *new_mbs = re_realloc (pstr->mbs, unsigned char,
- new_buf_len);
- if (BE (new_mbs == NULL, 0))
- return REG_ESPACE;
- pstr->mbs = new_mbs;
- }
- pstr->bufs_len = new_buf_len;
- return REG_NOERROR;
- }
- static void
- internal_function
- re_string_construct_common (const char *str, int len, re_string_t *pstr,
- RE_TRANSLATE_TYPE trans, int icase,
- const re_dfa_t *dfa)
- {
- pstr->raw_mbs = (const unsigned char *) str;
- pstr->len = len;
- pstr->raw_len = len;
- pstr->trans = trans;
- pstr->icase = icase ? 1 : 0;
- pstr->mbs_allocated = (trans != NULL || icase);
- pstr->mb_cur_max = dfa->mb_cur_max;
- pstr->is_utf8 = dfa->is_utf8;
- pstr->map_notascii = dfa->map_notascii;
- pstr->stop = pstr->len;
- pstr->raw_stop = pstr->stop;
- }
- #ifdef RE_ENABLE_I18N
- /* Build wide character buffer PSTR->WCS.
- If the byte sequence of the string are:
- <mb1>(0), <mb1>(1), <mb2>(0), <mb2>(1), <sb3>
- Then wide character buffer will be:
- <wc1> , WEOF , <wc2> , WEOF , <wc3>
- We use WEOF for padding, they indicate that the position isn't
- a first byte of a multibyte character.
- Note that this function assumes PSTR->VALID_LEN elements are already
- built and starts from PSTR->VALID_LEN. */
- static void
- internal_function
- build_wcs_buffer (re_string_t *pstr)
- {
- #ifdef _LIBC
- unsigned char buf[MB_LEN_MAX];
- assert (MB_LEN_MAX >= pstr->mb_cur_max);
- #else
- unsigned char buf[64];
- #endif
- mbstate_t prev_st;
- int byte_idx, end_idx, remain_len;
- size_t mbclen;
- /* Build the buffers from pstr->valid_len to either pstr->len or
- pstr->bufs_len. */
- end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
- for (byte_idx = pstr->valid_len; byte_idx < end_idx;)
- {
- wchar_t wc;
- const char *p;
- remain_len = end_idx - byte_idx;
- prev_st = pstr->cur_state;
- /* Apply the translation if we need. */
- if (BE (pstr->trans != NULL, 0))
- {
- int i, ch;
- for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i)
- {
- ch = pstr->raw_mbs [pstr->raw_mbs_idx + byte_idx + i];
- buf[i] = pstr->mbs[byte_idx + i] = pstr->trans[ch];
- }
- p = (const char *) buf;
- }
- else
- p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx;
- mbclen = mbrtowc (&wc, p, remain_len, &pstr->cur_state);
- if (BE (mbclen == (size_t) -2, 0))
- {
- /* The buffer doesn't have enough space, finish to build. */
- pstr->cur_state = prev_st;
- break;
- }
- else if (BE (mbclen == (size_t) -1 || mbclen == 0, 0))
- {
- /* We treat these cases as a singlebyte character. */
- mbclen = 1;
- wc = (wchar_t) pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
- if (BE (pstr->trans != NULL, 0))
- wc = pstr->trans[wc];
- pstr->cur_state = prev_st;
- }
- /* Write wide character and padding. */
- pstr->wcs[byte_idx++] = wc;
- /* Write paddings. */
- for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
- pstr->wcs[byte_idx++] = WEOF;
- }
- pstr->valid_len = byte_idx;
- pstr->valid_raw_len = byte_idx;
- }
- /* Build wide character buffer PSTR->WCS like build_wcs_buffer,
- but for REG_ICASE. */
- static reg_errcode_t
- internal_function
- build_wcs_upper_buffer (re_string_t *pstr)
- {
- mbstate_t prev_st;
- int src_idx, byte_idx, end_idx, remain_len;
- size_t mbclen;
- #ifdef _LIBC
- char buf[MB_LEN_MAX];
- assert (MB_LEN_MAX >= pstr->mb_cur_max);
- #else
- char buf[64];
- #endif
- byte_idx = pstr->valid_len;
- end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
- /* The following optimization assumes that ASCII characters can be
- mapped to wide characters with a simple cast. */
- if (! pstr->map_notascii && pstr->trans == NULL && !pstr->offsets_needed)
- {
- while (byte_idx < end_idx)
- {
- wchar_t wc;
- if (isascii (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx])
- && mbsinit (&pstr->cur_state))
- {
- /* In case of a singlebyte character. */
- pstr->mbs[byte_idx]
- = toupper (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]);
- /* The next step uses the assumption that wchar_t is encoded
- ASCII-safe: all ASCII values can be converted like this. */
- pstr->wcs[byte_idx] = (wchar_t) pstr->mbs[byte_idx];
- ++byte_idx;
- continue;
- }
- remain_len = end_idx - byte_idx;
- prev_st = pstr->cur_state;
- mbclen = mbrtowc (&wc,
- ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx
- + byte_idx), remain_len, &pstr->cur_state);
- if (BE (mbclen + 2 > 2, 1))
- {
- wchar_t wcu = wc;
- if (iswlower (wc))
- {
- size_t mbcdlen;
- wcu = towupper (wc);
- mbcdlen = wcrtomb (buf, wcu, &prev_st);
- if (BE (mbclen == mbcdlen, 1))
- memcpy (pstr->mbs + byte_idx, buf, mbclen);
- else
- {
- src_idx = byte_idx;
- goto offsets_needed;
- }
- }
- else
- memcpy (pstr->mbs + byte_idx,
- pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx, mbclen);
- pstr->wcs[byte_idx++] = wcu;
- /* Write paddings. */
- for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
- pstr->wcs[byte_idx++] = WEOF;
- }
- else if (mbclen == (size_t) -1 || mbclen == 0)
- {
- /* It is an invalid character or '\0'. Just use the byte. */
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
- pstr->mbs[byte_idx] = ch;
- /* And also cast it to wide char. */
- pstr->wcs[byte_idx++] = (wchar_t) ch;
- if (BE (mbclen == (size_t) -1, 0))
- pstr->cur_state = prev_st;
- }
- else
- {
- /* The buffer doesn't have enough space, finish to build. */
- pstr->cur_state = prev_st;
- break;
- }
- }
- pstr->valid_len = byte_idx;
- pstr->valid_raw_len = byte_idx;
- return REG_NOERROR;
- }
- else
- for (src_idx = pstr->valid_raw_len; byte_idx < end_idx;)
- {
- wchar_t wc;
- const char *p;
- offsets_needed:
- remain_len = end_idx - byte_idx;
- prev_st = pstr->cur_state;
- if (BE (pstr->trans != NULL, 0))
- {
- int i, ch;
- for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i)
- {
- ch = pstr->raw_mbs [pstr->raw_mbs_idx + src_idx + i];
- buf[i] = pstr->trans[ch];
- }
- p = (const char *) buf;
- }
- else
- p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + src_idx;
- mbclen = mbrtowc (&wc, p, remain_len, &pstr->cur_state);
- if (BE (mbclen + 2 > 2, 1))
- {
- wchar_t wcu = wc;
- if (iswlower (wc))
- {
- size_t mbcdlen;
- wcu = towupper (wc);
- mbcdlen = wcrtomb ((char *) buf, wcu, &prev_st);
- if (BE (mbclen == mbcdlen, 1))
- memcpy (pstr->mbs + byte_idx, buf, mbclen);
- else if (mbcdlen != (size_t) -1)
- {
- size_t i;
- if (byte_idx + mbcdlen > pstr->bufs_len)
- {
- pstr->cur_state = prev_st;
- break;
- }
- if (pstr->offsets == NULL)
- {
- pstr->offsets = re_malloc (int, pstr->bufs_len);
- if (pstr->offsets == NULL)
- return REG_ESPACE;
- }
- if (!pstr->offsets_needed)
- {
- for (i = 0; i < (size_t) byte_idx; ++i)
- pstr->offsets[i] = i;
- pstr->offsets_needed = 1;
- }
- memcpy (pstr->mbs + byte_idx, buf, mbcdlen);
- pstr->wcs[byte_idx] = wcu;
- pstr->offsets[byte_idx] = src_idx;
- for (i = 1; i < mbcdlen; ++i)
- {
- pstr->offsets[byte_idx + i]
- = src_idx + (i < mbclen ? i : mbclen - 1);
- pstr->wcs[byte_idx + i] = WEOF;
- }
- pstr->len += mbcdlen - mbclen;
- if (pstr->raw_stop > src_idx)
- pstr->stop += mbcdlen - mbclen;
- end_idx = (pstr->bufs_len > pstr->len)
- ? pstr->len : pstr->bufs_len;
- byte_idx += mbcdlen;
- src_idx += mbclen;
- continue;
- }
- else
- memcpy (pstr->mbs + byte_idx, p, mbclen);
- }
- else
- memcpy (pstr->mbs + byte_idx, p, mbclen);
- if (BE (pstr->offsets_needed != 0, 0))
- {
- size_t i;
- for (i = 0; i < mbclen; ++i)
- pstr->offsets[byte_idx + i] = src_idx + i;
- }
- src_idx += mbclen;
- pstr->wcs[byte_idx++] = wcu;
- /* Write paddings. */
- for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
- pstr->wcs[byte_idx++] = WEOF;
- }
- else if (mbclen == (size_t) -1 || mbclen == 0)
- {
- /* It is an invalid character or '\0'. Just use the byte. */
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + src_idx];
- if (BE (pstr->trans != NULL, 0))
- ch = pstr->trans [ch];
- pstr->mbs[byte_idx] = ch;
- if (BE (pstr->offsets_needed != 0, 0))
- pstr->offsets[byte_idx] = src_idx;
- ++src_idx;
- /* And also cast it to wide char. */
- pstr->wcs[byte_idx++] = (wchar_t) ch;
- if (BE (mbclen == (size_t) -1, 0))
- pstr->cur_state = prev_st;
- }
- else
- {
- /* The buffer doesn't have enough space, finish to build. */
- pstr->cur_state = prev_st;
- break;
- }
- }
- pstr->valid_len = byte_idx;
- pstr->valid_raw_len = src_idx;
- return REG_NOERROR;
- }
- /* Skip characters until the index becomes greater than NEW_RAW_IDX.
- Return the index. */
- static int
- internal_function
- re_string_skip_chars (re_string_t *pstr, int new_raw_idx, wint_t *last_wc)
- {
- mbstate_t prev_st;
- int rawbuf_idx;
- size_t mbclen;
- wchar_t wc = WEOF;
- /* Skip the characters which are not necessary to check. */
- for (rawbuf_idx = pstr->raw_mbs_idx + pstr->valid_raw_len;
- rawbuf_idx < new_raw_idx;)
- {
- int remain_len;
- remain_len = pstr->len - rawbuf_idx;
- prev_st = pstr->cur_state;
- mbclen = mbrtowc (&wc, (const char *) pstr->raw_mbs + rawbuf_idx,
- remain_len, &pstr->cur_state);
- if (BE (mbclen == (size_t) -2 || mbclen == (size_t) -1 || mbclen == 0, 0))
- {
- /* We treat these cases as a single byte character. */
- if (mbclen == 0 || remain_len == 0)
- wc = L'\0';
- else
- wc = *(unsigned char *) (pstr->raw_mbs + rawbuf_idx);
- mbclen = 1;
- pstr->cur_state = prev_st;
- }
- /* Then proceed the next character. */
- rawbuf_idx += mbclen;
- }
- *last_wc = (wint_t) wc;
- return rawbuf_idx;
- }
- #endif /* RE_ENABLE_I18N */
- /* Build the buffer PSTR->MBS, and apply the translation if we need.
- This function is used in case of REG_ICASE. */
- static void
- internal_function
- build_upper_buffer (re_string_t *pstr)
- {
- int char_idx, end_idx;
- end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
- for (char_idx = pstr->valid_len; char_idx < end_idx; ++char_idx)
- {
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + char_idx];
- if (BE (pstr->trans != NULL, 0))
- ch = pstr->trans[ch];
- if (islower (ch))
- pstr->mbs[char_idx] = toupper (ch);
- else
- pstr->mbs[char_idx] = ch;
- }
- pstr->valid_len = char_idx;
- pstr->valid_raw_len = char_idx;
- }
- /* Apply TRANS to the buffer in PSTR. */
- static void
- internal_function
- re_string_translate_buffer (re_string_t *pstr)
- {
- int buf_idx, end_idx;
- end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
- for (buf_idx = pstr->valid_len; buf_idx < end_idx; ++buf_idx)
- {
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + buf_idx];
- pstr->mbs[buf_idx] = pstr->trans[ch];
- }
- pstr->valid_len = buf_idx;
- pstr->valid_raw_len = buf_idx;
- }
- /* This function re-construct the buffers.
- Concretely, convert to wide character in case of pstr->mb_cur_max > 1,
- convert to upper case in case of REG_ICASE, apply translation. */
- static reg_errcode_t
- internal_function
- re_string_reconstruct (re_string_t *pstr, int idx, int eflags)
- {
- int offset = idx - pstr->raw_mbs_idx;
- if (BE (offset < 0, 0))
- {
- /* Reset buffer. */
- #ifdef RE_ENABLE_I18N
- if (pstr->mb_cur_max > 1)
- memset (&pstr->cur_state, '\0', sizeof (mbstate_t));
- #endif /* RE_ENABLE_I18N */
- pstr->len = pstr->raw_len;
- pstr->stop = pstr->raw_stop;
- pstr->valid_len = 0;
- pstr->raw_mbs_idx = 0;
- pstr->valid_raw_len = 0;
- pstr->offsets_needed = 0;
- pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF
- : CONTEXT_NEWLINE | CONTEXT_BEGBUF);
- if (!pstr->mbs_allocated)
- pstr->mbs = (unsigned char *) pstr->raw_mbs;
- offset = idx;
- }
- if (BE (offset != 0, 1))
- {
- /* Should the already checked characters be kept? */
- if (BE (offset < pstr->valid_raw_len, 1))
- {
- /* Yes, move them to the front of the buffer. */
- #ifdef RE_ENABLE_I18N
- if (BE (pstr->offsets_needed, 0))
- {
- int low = 0, high = pstr->valid_len, mid;
- do
- {
- mid = (high + low) / 2;
- if (pstr->offsets[mid] > offset)
- high = mid;
- else if (pstr->offsets[mid] < offset)
- low = mid + 1;
- else
- break;
- }
- while (low < high);
- if (pstr->offsets[mid] < offset)
- ++mid;
- pstr->tip_context = re_string_context_at (pstr, mid - 1,
- eflags);
- /* This can be quite complicated, so handle specially
- only the common and easy case where the character with
- different length representation of lower and upper
- case is present at or after offset. */
- if (pstr->valid_len > offset
- && mid == offset && pstr->offsets[mid] == offset)
- {
- memmove (pstr->wcs, pstr->wcs + offset,
- (pstr->valid_len - offset) * sizeof (wint_t));
- memmove (pstr->mbs, pstr->mbs + offset, pstr->valid_len - offset);
- pstr->valid_len -= offset;
- pstr->valid_raw_len -= offset;
- for (low = 0; low < pstr->valid_len; low++)
- pstr->offsets[low] = pstr->offsets[low + offset] - offset;
- }
- else
- {
- /* Otherwise, just find out how long the partial multibyte
- character at offset is and fill it with WEOF/255. */
- pstr->len = pstr->raw_len - idx + offset;
- pstr->stop = pstr->raw_stop - idx + offset;
- pstr->offsets_needed = 0;
- while (mid > 0 && pstr->offsets[mid - 1] == offset)
- --mid;
- while (mid < pstr->valid_len)
- if (pstr->wcs[mid] != WEOF)
- break;
- else
- ++mid;
- if (mid == pstr->valid_len)
- pstr->valid_len = 0;
- else
- {
- pstr->valid_len = pstr->offsets[mid] - offset;
- if (pstr->valid_len)
- {
- for (low = 0; low < pstr->valid_len; ++low)
- pstr->wcs[low] = WEOF;
- memset (pstr->mbs, 255, pstr->valid_len);
- }
- }
- pstr->valid_raw_len = pstr->valid_len;
- }
- }
- else
- #endif
- {
- pstr->tip_context = re_string_context_at (pstr, offset - 1,
- eflags);
- #ifdef RE_ENABLE_I18N
- if (pstr->mb_cur_max > 1)
- memmove (pstr->wcs, pstr->wcs + offset,
- (pstr->valid_len - offset) * sizeof (wint_t));
- #endif /* RE_ENABLE_I18N */
- if (BE (pstr->mbs_allocated, 0))
- memmove (pstr->mbs, pstr->mbs + offset,
- pstr->valid_len - offset);
- pstr->valid_len -= offset;
- pstr->valid_raw_len -= offset;
- #if DEBUG
- assert (pstr->valid_len > 0);
- #endif
- }
- }
- else
- {
- /* No, skip all characters until IDX. */
- int prev_valid_len = pstr->valid_len;
- #ifdef RE_ENABLE_I18N
- if (BE (pstr->offsets_needed, 0))
- {
- pstr->len = pstr->raw_len - idx + offset;
- pstr->stop = pstr->raw_stop - idx + offset;
- pstr->offsets_needed = 0;
- }
- #endif
- pstr->valid_len = 0;
- #ifdef RE_ENABLE_I18N
- if (pstr->mb_cur_max > 1)
- {
- int wcs_idx;
- wint_t wc = WEOF;
- if (pstr->is_utf8)
- {
- const unsigned char *raw, *p, *q, *end;
- /* Special case UTF-8. Multi-byte chars start with any
- byte other than 0x80 - 0xbf. */
- raw = pstr->raw_mbs + pstr->raw_mbs_idx;
- end = raw + (offset - pstr->mb_cur_max);
- if (end < pstr->raw_mbs)
- end = pstr->raw_mbs;
- p = raw + offset - 1;
- #ifdef _LIBC
- /* We know the wchar_t encoding is UCS4, so for the simple
- case, ASCII characters, skip the conversion step. */
- if (isascii (*p) && BE (pstr->trans == NULL, 1))
- {
- memset (&pstr->cur_state, '\0', sizeof (mbstate_t));
- /* pstr->valid_len = 0; */
- wc = (wchar_t) *p;
- }
- else
- #endif
- for (; p >= end; --p)
- if ((*p & 0xc0) != 0x80)
- {
- mbstate_t cur_state;
- wchar_t wc2;
- int mlen = raw + pstr->len - p;
- unsigned char buf[6];
- size_t mbclen;
- q = p;
- if (BE (pstr->trans != NULL, 0))
- {
- int i = mlen < 6 ? mlen : 6;
- while (--i >= 0)
- buf[i] = pstr->trans[p[i]];
- q = buf;
- }
- /* XXX Don't use mbrtowc, we know which conversion
- to use (UTF-8 -> UCS4). */
- memset (&cur_state, 0, sizeof (cur_state));
- mbclen = mbrtowc (&wc2, (const char *) p, mlen,
- &cur_state);
- if (raw + offset - p <= mbclen
- && mbclen < (size_t) -2)
- {
- memset (&pstr->cur_state, '\0',
- sizeof (mbstate_t));
- pstr->valid_len = mbclen - (raw + offset - p);
- wc = wc2;
- }
- break;
- }
- }
- if (wc == WEOF)
- pstr->valid_len = re_string_skip_chars (pstr, idx, &wc) - idx;
- if (wc == WEOF)
- pstr->tip_context
- = re_string_context_at (pstr, prev_valid_len - 1, eflags);
- else
- pstr->tip_context = ((BE (pstr->word_ops_used != 0, 0)
- && IS_WIDE_WORD_CHAR (wc))
- ? CONTEXT_WORD
- : ((IS_WIDE_NEWLINE (wc)
- && pstr->newline_anchor)
- ? CONTEXT_NEWLINE : 0));
- if (BE (pstr->valid_len, 0))
- {
- for (wcs_idx = 0; wcs_idx < pstr->valid_len; ++wcs_idx)
- pstr->wcs[wcs_idx] = WEOF;
- if (pstr->mbs_allocated)
- memset (pstr->mbs, 255, pstr->valid_len);
- }
- pstr->valid_raw_len = pstr->valid_len;
- }
- else
- #endif /* RE_ENABLE_I18N */
- {
- int c = pstr->raw_mbs[pstr->raw_mbs_idx + offset - 1];
- pstr->valid_raw_len = 0;
- if (pstr->trans)
- c = pstr->trans[c];
- pstr->tip_context = (bitset_contain (pstr->word_char, c)
- ? CONTEXT_WORD
- : ((IS_NEWLINE (c) && pstr->newline_anchor)
- ? CONTEXT_NEWLINE : 0));
- }
- }
- if (!BE (pstr->mbs_allocated, 0))
- pstr->mbs += offset;
- }
- pstr->raw_mbs_idx = idx;
- pstr->len -= offset;
- pstr->stop -= offset;
- /* Then build the buffers. */
- #ifdef RE_ENABLE_I18N
- if (pstr->mb_cur_max > 1)
- {
- if (pstr->icase)
- {
- reg_errcode_t ret = build_wcs_upper_buffer (pstr);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- }
- else
- build_wcs_buffer (pstr);
- }
- else
- #endif /* RE_ENABLE_I18N */
- if (BE (pstr->mbs_allocated, 0))
- {
- if (pstr->icase)
- build_upper_buffer (pstr);
- else if (pstr->trans != NULL)
- re_string_translate_buffer (pstr);
- }
- else
- pstr->valid_len = pstr->len;
- pstr->cur_idx = 0;
- return REG_NOERROR;
- }
- static unsigned char
- internal_function __attribute ((pure))
- re_string_peek_byte_case (const re_string_t *pstr, int idx)
- {
- int ch, off;
- /* Handle the common (easiest) cases first. */
- if (BE (!pstr->mbs_allocated, 1))
- return re_string_peek_byte (pstr, idx);
- #ifdef RE_ENABLE_I18N
- if (pstr->mb_cur_max > 1
- && ! re_string_is_single_byte_char (pstr, pstr->cur_idx + idx))
- return re_string_peek_byte (pstr, idx);
- #endif
- off = pstr->cur_idx + idx;
- #ifdef RE_ENABLE_I18N
- if (pstr->offsets_needed)
- off = pstr->offsets[off];
- #endif
- ch = pstr->raw_mbs[pstr->raw_mbs_idx + off];
- #ifdef RE_ENABLE_I18N
- /* Ensure that e.g. for tr_TR.UTF-8 BACKSLASH DOTLESS SMALL LETTER I
- this function returns CAPITAL LETTER I instead of first byte of
- DOTLESS SMALL LETTER I. The latter would confuse the parser,
- since peek_byte_case doesn't advance cur_idx in any way. */
- if (pstr->offsets_needed && !isascii (ch))
- return re_string_peek_byte (pstr, idx);
- #endif
- return ch;
- }
- static unsigned char
- internal_function __attribute ((pure))
- re_string_fetch_byte_case (re_string_t *pstr)
- {
- if (BE (!pstr->mbs_allocated, 1))
- return re_string_fetch_byte (pstr);
- #ifdef RE_ENABLE_I18N
- if (pstr->offsets_needed)
- {
- int off, ch;
- /* For tr_TR.UTF-8 [[:islower:]] there is
- [[: CAPITAL LETTER I WITH DOT lower:]] in mbs. Skip
- in that case the whole multi-byte character and return
- the original letter. On the other side, with
- [[: DOTLESS SMALL LETTER I return [[:I, as doing
- anything else would complicate things too much. */
- if (!re_string_first_byte (pstr, pstr->cur_idx))
- return re_string_fetch_byte (pstr);
- off = pstr->offsets[pstr->cur_idx];
- ch = pstr->raw_mbs[pstr->raw_mbs_idx + off];
- if (! isascii (ch))
- return re_string_fetch_byte (pstr);
- re_string_skip_bytes (pstr,
- re_string_char_size_at (pstr, pstr->cur_idx));
- return ch;
- }
- #endif
- return pstr->raw_mbs[pstr->raw_mbs_idx + pstr->cur_idx++];
- }
- static void
- internal_function
- re_string_destruct (re_string_t *pstr)
- {
- #ifdef RE_ENABLE_I18N
- re_free (pstr->wcs);
- re_free (pstr->offsets);
- #endif /* RE_ENABLE_I18N */
- if (pstr->mbs_allocated)
- re_free (pstr->mbs);
- }
- /* Return the context at IDX in INPUT. */
- static unsigned int
- internal_function
- re_string_context_at (const re_string_t *input, int idx, int eflags)
- {
- int c;
- if (BE (idx < 0, 0))
- /* In this case, we use the value stored in input->tip_context,
- since we can't know the character in input->mbs[-1] here. */
- return input->tip_context;
- if (BE (idx == input->len, 0))
- return ((eflags & REG_NOTEOL) ? CONTEXT_ENDBUF
- …
Large files files are truncated, but you can click here to view the full file