diff options
| -rw-r--r-- | ChangeLog | 12 | ||||
| -rw-r--r-- | posix/regcomp.c | 9 | ||||
| -rw-r--r-- | posix/regex.h | 5 | ||||
| -rw-r--r-- | posix/regex_internal.h | 6 | ||||
| -rw-r--r-- | posix/regexec.c | 1 |
5 files changed, 28 insertions, 5 deletions
@@ -1,3 +1,15 @@ +2003-09-20 Paolo Bonzini <bonzini@gnu.org> + + * posix/regcomp.c (peek_token): Don't look back for ( or | + to check whether to treat a caret as special. It fails + for the (extended) regex \(^. + (parse, parse_reg_exp): Pass RE_CARET_ANCHORS_HERE to fetch_token. + * posix/regex.h: Define RE_CARET_ANCHORS_HERE. + + * posix/regexec.c: Check out of bounds value before shifting. + + * posix/regex_internal.h: Define __attribute for non-gcc. + 2003-09-22 Jakub Jelinek <jakub@redhat.com> * include/atomic.h (atomic_compare_and_exchange_val_rel, diff --git a/posix/regcomp.c b/posix/regcomp.c index d9212de3b9..4682ca6c1b 100644 --- a/posix/regcomp.c +++ b/posix/regcomp.c @@ -1650,12 +1650,11 @@ peek_token (token, input, syntax) token->type = OP_PERIOD; break; case '^': - if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) && + if (!(syntax & (RE_CONTEXT_INDEP_ANCHORS | RE_CARET_ANCHORS_HERE)) && re_string_cur_idx (input) != 0) { char prev = re_string_peek_byte (input, -1); - if (prev != '|' && prev != '(' && - (!(syntax & RE_NEWLINE_ALT) || prev != '\n')) + if (!(syntax & RE_NEWLINE_ALT) || prev != '\n') break; } token->type = ANCHOR; @@ -1790,7 +1789,7 @@ parse (regexp, preg, syntax, err) bin_tree_t *tree, *eor, *root; re_token_t current_token; int new_idx; - current_token = fetch_token (regexp, syntax); + current_token = fetch_token (regexp, syntax | RE_CARET_ANCHORS_HERE); tree = parse_reg_exp (regexp, preg, ¤t_token, syntax, 0, err); if (BE (*err != REG_NOERROR && tree == NULL, 0)) return NULL; @@ -1837,7 +1836,7 @@ parse_reg_exp (regexp, preg, token, syntax, nest, err) { re_token_t alt_token = *token; new_idx = re_dfa_add_node (dfa, alt_token, 0); - *token = fetch_token (regexp, syntax); + *token = fetch_token (regexp, syntax | RE_CARET_ANCHORS_HERE); if (token->type != OP_ALT && token->type != END_OF_RE && (nest == 0 || token->type != OP_CLOSE_SUBEXP)) { diff --git a/posix/regex.h b/posix/regex.h index e251962cef..1ab3e243a7 100644 --- a/posix/regex.h +++ b/posix/regex.h @@ -170,6 +170,11 @@ typedef unsigned long int reg_syntax_t; If not set, then case is significant. */ #define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) +/* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only + for ^, because it is difficult to scan the regex backwards to find + whether ^ should be special. */ +#define RE_CARET_ANCHORS_HERE (RE_ICASE << 1) + /* This global variable defines the particular regexp syntax to use (for some interfaces). When a regexp is compiled, the syntax used is stored in the pattern buffer, so changing this does not affect diff --git a/posix/regex_internal.h b/posix/regex_internal.h index bf84ad6270..9dd3fe55b3 100644 --- a/posix/regex_internal.h +++ b/posix/regex_internal.h @@ -106,6 +106,12 @@ # define attribute_hidden #endif /* not _LIBC */ +#ifdef __GNUC__ +# define __attribute __attribute__ +#else +# define __attribute +#endif + extern const char __re_error_msgid[] attribute_hidden; extern const size_t __re_error_msgid_idx[] attribute_hidden; diff --git a/posix/regexec.c b/posix/regexec.c index 277c935231..39a27d2fed 100644 --- a/posix/regexec.c +++ b/posix/regexec.c @@ -2193,6 +2193,7 @@ check_subexp_matching_top (dfa, mctx, cur_nodes, str_idx) { int node = cur_nodes->elems[node_idx]; if (dfa->nodes[node].type == OP_OPEN_SUBEXP + && dfa->nodes[node].opr.idx < (8 * sizeof (dfa->used_bkref_map)) && dfa->used_bkref_map & (1 << dfa->nodes[node].opr.idx)) { err = match_ctx_add_subtop (mctx, node, str_idx); |