diff options
Diffstat (limited to 'posix')
| -rw-r--r-- | posix/regex_internal.c | 29 | ||||
| -rw-r--r-- | posix/regex_internal.h | 26 | ||||
| -rw-r--r-- | posix/regexec.c | 36 |
3 files changed, 45 insertions, 46 deletions
diff --git a/posix/regex_internal.c b/posix/regex_internal.c index 25e7b7e..c4400a8 100644 --- a/posix/regex_internal.c +++ b/posix/regex_internal.c @@ -57,8 +57,8 @@ #include "regex.h" #include "regex_internal.h" -static void re_string_construct_common (const unsigned char *str, - int len, re_string_t *pstr, +static void re_string_construct_common (const char *str, int len, + re_string_t *pstr, RE_TRANSLATE_TYPE trans, int icase); #ifdef RE_ENABLE_I18N static int re_string_skip_chars (re_string_t *pstr, int new_raw_idx); @@ -86,7 +86,7 @@ static unsigned int inline calc_state_hash (const re_node_set *nodes, static reg_errcode_t re_string_allocate (pstr, str, len, init_len, trans, icase) re_string_t *pstr; - const unsigned char *str; + const char *str; int len, init_len, icase; RE_TRANSLATE_TYPE trans; { @@ -99,8 +99,7 @@ re_string_allocate (pstr, str, len, init_len, trans, icase) if (BE (ret != REG_NOERROR, 0)) return ret; - pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case - : (unsigned char *)str); + pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case : (char *) str); pstr->mbs = MBS_ALLOCATED (pstr) ? pstr->mbs : pstr->mbs_case; pstr->valid_len = (MBS_CASE_ALLOCATED (pstr) || MBS_ALLOCATED (pstr) || MB_CUR_MAX > 1) ? pstr->valid_len : len; @@ -112,7 +111,7 @@ re_string_allocate (pstr, str, len, init_len, trans, icase) static reg_errcode_t re_string_construct (pstr, str, len, trans, icase) re_string_t *pstr; - const unsigned char *str; + const char *str; int len, icase; RE_TRANSLATE_TYPE trans; { @@ -128,8 +127,7 @@ re_string_construct (pstr, str, len, trans, icase) if (BE (ret != REG_NOERROR, 0)) return ret; } - pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case - : (unsigned char *)str); + pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case : (char *) str); pstr->mbs = MBS_ALLOCATED (pstr) ? pstr->mbs : pstr->mbs_case; if (icase) @@ -178,13 +176,13 @@ re_string_realloc_buffers (pstr, new_buf_len) #endif /* RE_ENABLE_I18N */ if (MBS_ALLOCATED (pstr)) { - pstr->mbs = re_realloc (pstr->mbs, unsigned char, new_buf_len); + pstr->mbs = re_realloc (pstr->mbs, char, new_buf_len); if (BE (pstr->mbs == NULL, 0)) return REG_ESPACE; } if (MBS_CASE_ALLOCATED (pstr)) { - pstr->mbs_case = re_realloc (pstr->mbs_case, unsigned char, new_buf_len); + pstr->mbs_case = re_realloc (pstr->mbs_case, char, new_buf_len); if (BE (pstr->mbs_case == NULL, 0)) return REG_ESPACE; if (!MBS_ALLOCATED (pstr)) @@ -197,7 +195,7 @@ re_string_realloc_buffers (pstr, new_buf_len) static void re_string_construct_common (str, len, pstr, trans, icase) - const unsigned char *str; + const char *str; int len; re_string_t *pstr; RE_TRANSLATE_TYPE trans; @@ -256,8 +254,9 @@ build_wcs_buffer (pstr) /* Apply the translateion if we need. */ if (pstr->trans != NULL && mbclen == 1) { - int ch = pstr->trans[pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]]; - pstr->mbs_case[byte_idx] = ch; + int ch = *((unsigned char *) pstr->raw_mbs + pstr->raw_mbs_idx + + byte_idx); + pstr->mbs_case[byte_idx] = pstr->trans[ch]; } /* Write wide character and padding. */ pstr->wcs[byte_idx++] = wc; @@ -421,9 +420,9 @@ re_string_reconstruct (pstr, idx, eflags, newline) pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF : CONTEXT_NEWLINE | CONTEXT_BEGBUF); if (!MBS_CASE_ALLOCATED (pstr)) - pstr->mbs_case = (unsigned char *)pstr->raw_mbs; + pstr->mbs_case = (char *) pstr->raw_mbs; if (!MBS_ALLOCATED (pstr) && !MBS_CASE_ALLOCATED (pstr)) - pstr->mbs = (unsigned char *)pstr->raw_mbs; + pstr->mbs = (char *) pstr->raw_mbs; offset = idx; } diff --git a/posix/regex_internal.h b/posix/regex_internal.h index a7f6042..574bf65 100644 --- a/posix/regex_internal.h +++ b/posix/regex_internal.h @@ -228,15 +228,15 @@ 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; + const 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; + char *mbs; /* Store the case sensitive multibyte string. In case of "case insensitive mode", the original string are stored, otherwise MBS_CASE points the same address that MBS points. */ - unsigned char *mbs_case; + char *mbs_case; #ifdef RE_ENABLE_I18N /* Store the wide character string which is corresponding to MBS. */ wint_t *wcs; @@ -275,13 +275,12 @@ typedef struct re_string_t re_string_t; #define MBS_CASE_ALLOCATED(pstr) (pstr->trans != NULL) -static reg_errcode_t re_string_allocate (re_string_t *pstr, - const unsigned char *str, int len, - int init_len, +static reg_errcode_t re_string_allocate (re_string_t *pstr, const char *str, + int len, int init_len, RE_TRANSLATE_TYPE trans, int icase); -static reg_errcode_t re_string_construct (re_string_t *pstr, - const unsigned char *str, int len, - RE_TRANSLATE_TYPE trans, int icase); +static reg_errcode_t re_string_construct (re_string_t *pstr, const char *str, + int len, RE_TRANSLATE_TYPE trans, + int icase); static reg_errcode_t re_string_reconstruct (re_string_t *pstr, int idx, int eflags, int newline); static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr, @@ -513,7 +512,7 @@ typedef struct union { unsigned char ch; - unsigned char *name; + char *name; wchar_t wch; } opr; } bracket_elem_t; @@ -581,8 +580,7 @@ re_string_elem_size_at (pstr, idx) int idx; { #ifdef _LIBC - const unsigned char *p; - const char *extra; + const char *extra, *p; const int32_t *table, *indirect; int32_t tmp; # include <locale/weight.h> @@ -595,8 +593,8 @@ re_string_elem_size_at (pstr, idx) indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); p = pstr->mbs + idx; - tmp = findidx (&p); - return p - (const unsigned char *) pstr->mbs - idx; + tmp = findidx ((const unsigned char **) &p); + return p - pstr->mbs - idx; } else #endif /* _LIBC */ diff --git a/posix/regexec.c b/posix/regexec.c index f9b162b..4a9c64a 100644 --- a/posix/regexec.c +++ b/posix/regexec.c @@ -123,7 +123,7 @@ static re_dfastate_t **build_trtable (const regex_t *dfa, static int check_node_accept_bytes (const regex_t *preg, int node_idx, const re_string_t *input, int idx); # ifdef _LIBC -static unsigned int find_collation_sequence_value (const unsigned char *mbs, +static unsigned int find_collation_sequence_value (const char *mbs, size_t name_len); # endif /* _LIBC */ #endif /* RE_ENABLE_I18N */ @@ -153,8 +153,8 @@ static reg_errcode_t extend_buffers (re_match_context_t *mctx); int regexec (preg, string, nmatch, pmatch, eflags) - const regex_t *preg; - const char *string; + const regex_t *__restrict preg; + const char *__restrict string; size_t nmatch; regmatch_t pmatch[]; int eflags; @@ -1621,7 +1621,7 @@ transit_state_bkref_loop (preg, nodes, work_state_log, mctx) for (i = 0; i < nodes->nelem; ++i) { - unsigned char *buf; + char *buf; int dest_str_idx, subexp_idx, prev_nelem, subexp_len; int node_idx = nodes->elems[i]; unsigned int context; @@ -2069,7 +2069,7 @@ check_node_accept_bytes (preg, node_idx, input, str_idx) { const re_charset_t *cset = node->opr.mbcset; # ifdef _LIBC - const unsigned char *pin = re_string_get_buffer (input) + str_idx; + const char *pin = re_string_get_buffer (input) + str_idx; # endif /* _LIBC */ int match_len = 0; wchar_t wc = ((cset->nranges || cset->nchar_classes || cset->nmbchars) @@ -2098,18 +2098,17 @@ check_node_accept_bytes (preg, node_idx, input, str_idx) { unsigned int in_collseq = 0; const int32_t *table, *indirect; - const unsigned char *weights, *extra, *collseqwc; + const char *weights, *extra, *collseqwc; int32_t idx; /* This #include defines a local function! */ # include <locale/weight.h> /* match with collating_symbol? */ if (cset->ncoll_syms) - extra = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + extra = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); for (i = 0; i < cset->ncoll_syms; ++i) { - const unsigned char *coll_sym = extra + cset->coll_syms[i]; + const char *coll_sym = extra + cset->coll_syms[i]; /* Compare the length of input collating element and the length of current collating element. */ if (*coll_sym != elem_len) @@ -2148,13 +2147,11 @@ check_node_accept_bytes (preg, node_idx, input, str_idx) /* match with equivalence_class? */ if (cset->nequiv_classes) { - const unsigned char *cp = pin; + const unsigned char *cp = (const unsigned char *) pin; table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); - weights = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); - extra = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + weights = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); idx = findidx (&cp); @@ -2183,7 +2180,12 @@ check_node_accept_bytes (preg, node_idx, input, str_idx) # endif /* _LIBC */ { /* match with range expression? */ - wchar_t cmp_buf[6] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'}; +#if __GNUC__ >= 2 + wchar_t cmp_buf[] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'}; +#else + wchar_t cmp_buf[] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'}; + cmp_buf[2] = wc; +#endif for (i = 0; i < cset->nranges; ++i) { cmp_buf[0] = cset->range_starts[i]; @@ -2213,7 +2215,7 @@ check_node_accept_bytes (preg, node_idx, input, str_idx) # ifdef _LIBC static unsigned int find_collation_sequence_value (mbs, mbs_len) - const unsigned char *mbs; + const char *mbs; size_t mbs_len; { uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); @@ -2224,7 +2226,7 @@ find_collation_sequence_value (mbs, mbs_len) /* No valid character. Match it as a single byte character. */ const unsigned char *collseq = (const unsigned char *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB); - return collseq[mbs[0]]; + return collseq[*(unsigned char *) mbs]; } return UINT_MAX; } |