* posix/regcomp.c (peek_token): Remove recent changes for anchor
	handling again.
	(parse_reg_exp): Likewise.
	* posix/regex.h: Remove RE_CARET_ANCHORS_HERE.

	(peek_token): Accept \s and \S as OP_SPACE and OP_NOTSPACE.
	(parse_expression): Replace build_word_op with
This commit is contained in:
Ulrich Drepper 2003-09-25 20:48:59 +00:00
parent 63461e7544
commit a0788ec48c
3 changed files with 13 additions and 12 deletions

View File

@ -1,5 +1,10 @@
2003-09-25 Ulrich Drepper <drepper@redhat.com>
* posix/regcomp.c (peek_token): Remove recent changes for anchor
handling again.
(parse_reg_exp): Likewise.
* posix/regex.h: Remove RE_CARET_ANCHORS_HERE.
* csu/Makefile ($(objpfx)version-info.h): Fix sed script to handle
banners.
@ -239,8 +244,8 @@
* posix/regcomp.c (build_word_op): Rename like...
(build_charclass_op): ...this. Accept two extra parameters,
CLASS_NAME and EXTRA. Add EXTRA to the result, not only _.
(peek_token): accept \s and \S as OP_SPACE and OP_NOTSPACE.
(parse_expression): replace build_word_op with
(peek_token): Accept \s and \S as OP_SPACE and OP_NOTSPACE.
(parse_expression): Replace build_word_op with
build_charclass_op, add new arguments, accept OP_SPACE
and OP_NOTSPACE.
* posix/regex_internal.h (re_token_type_t): Add OP_SPACE

View File

@ -119,7 +119,7 @@ static reg_errcode_t build_charclass (RE_TRANSLATE_TYPE trans,
reg_syntax_t syntax);
#endif /* not RE_ENABLE_I18N */
static bin_tree_t *build_charclass_op (re_dfa_t *dfa, RE_TRANSLATE_TYPE trans,
const unsigned char *class_name,
const unsigned char *class_name,
const unsigned char *extra, int not,
reg_errcode_t *err);
static void free_bin_tree (bin_tree_t *tree);
@ -1660,11 +1660,12 @@ peek_token (token, input, syntax)
token->type = OP_PERIOD;
break;
case '^':
if (!(syntax & (RE_CONTEXT_INDEP_ANCHORS | RE_CARET_ANCHORS_HERE)) &&
if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) &&
re_string_cur_idx (input) != 0)
{
char prev = re_string_peek_byte (input, -1);
if (!(syntax & RE_NEWLINE_ALT) || prev != '\n')
if (prev != '|' && prev != '(' &&
(!(syntax & RE_NEWLINE_ALT) || prev != '\n'))
break;
}
token->type = ANCHOR;
@ -1799,7 +1800,7 @@ 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 | RE_CARET_ANCHORS_HERE);
current_token = fetch_token (regexp, syntax);
tree = parse_reg_exp (regexp, preg, &current_token, syntax, 0, err);
if (BE (*err != REG_NOERROR && tree == NULL, 0))
return NULL;
@ -1846,7 +1847,7 @@ 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 | RE_CARET_ANCHORS_HERE);
*token = fetch_token (regexp, syntax);
if (token->type != OP_ALT && token->type != END_OF_RE
&& (nest == 0 || token->type != OP_CLOSE_SUBEXP))
{

View File

@ -170,11 +170,6 @@ 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