Update.

2003-11-24  Jakub Jelinek  <jakub@redhat.com>

	* posix/regex_internal.h (re_token_t): Add word_char bit.  Add
	comment.
	(re_dfa_t): Add sb_char field.
	(bitset_mask): New function.
	* posix/regcomp.c (free_dfa_content): Free sb_char.
	(init_dfa): Don't initialize word_char unnecessarily.
	Initialize sb_char.
	(duplicate_node): Don't duplicate !word_char CHARACTERs with
	NEXT_WORD_CONSTRAINT constraint or word_char CHARACTERs with
	NEXT_NOTWORD_CONSTRAINT.  Return -1 in *new_idx instead.
	(duplicate_node_closure): Handle clone_dest == -1 from
	duplicate_node.
	(peek_token): Initialize word_char bit.
	(parse_expression, parse_dup_op): Add comments.
	(parse_bracket_exp): Don't set bitmask bits for multi-byte char
	starting bytes here at the beginning.  Mask off the bits right
	before creating SIMPLE_BRACKET.
	(build_charclass_op): Likewise.
	* posix/regexec.c (group_nodes_into_DFAstates) <case OP_PERIOD>: Only
	set accept bits for single-byte characters.
	(group_nodes_into_DFAstates): Don't rely on characters 0 .. 127
	being single byte encoded and the rest multi-byte.
	* posix/bug-regex19.c (tests): Add new tests.
	(do_mb_tests): Initialize t to *test.
	(main): Fail even on do_mb_tests errors.

0 files changed, 0 insertions, 0 deletions