Jim Meyering [Mon, 3 Apr 2000 06:24:20 +0000 (06:24 +0000)]
(jm_FUNC_GNU_STRFTIME): Set TZ environment
variable in the shell rather than using putenv, which isn't
portable. This avoids the configure-time inter-test dependency
on the potentially-renamed putenv function.
Stefan Monnier [Sun, 2 Apr 2000 23:56:45 +0000 (23:56 +0000)]
* regex.c (PTR_TO_OFFSET) [!emacs]: Remove.
(RE_MULTIBYTE_P, RE_STRING_CHAR_AND_LENGTH): New macros.
(GET_CHAR_BEFORE_2): Moved from charset.h plus fixed minor bug when
we are between str1 and str2.
(MAX_MULTIBYTE_LENGTH, CHAR_STRING) [!emacs]: Provide trivial default.
(PATFETCH): Use `TRANSLATE'.
(PATFETCH_RAW): Fetch multibyte char if applicable.
(PATUNFETCH): Remove.
(regex_compile): Rely on PATFETCH to do most of the multibyte magic.
When writing a char, write it directly into the pattern buffer rather
than going needlessly through a temp char-array.
(re_match_2_internal): Similarly, rely on RE_STRING_CHAR to do the
multibyte magic and remove the useless `#ifdef emacs'.
(bcmp_translate): Don't compare as multibyte chars when in a unibyte
buffer.
* regex.h (struct re_pattern_buffer): Make field `multibyte'
conditional on `emacs'.
* charset.h (GET_CHAR_BEFORE_2): Moved to regex.c.
Stefan Monnier [Wed, 29 Mar 2000 04:01:25 +0000 (04:01 +0000)]
(analyse_first): New function obtained by ripping out most
of re_compile_fastmap and generalizing it a little bit so that it
can also just return whether a given (sub)pattern can match the empty
string or not.
(regex_compile): Use `analyse_first' to decide whether the loop-check
needs to be done or not for *, +, *? and +? (the loop check is costly
for non-greedy repetition).
(re_compile_fastmap): Delegate the actual work to `analyse_first'.
Stefan Monnier [Mon, 27 Mar 2000 22:26:37 +0000 (22:26 +0000)]
(REGEX_FREE_STACK, RESET_FAIL_STACK): Make them usable as an expression.
(enum re_opcode_t): Update description of succeed_n.
(PATFETCH): Always define.
(regex_compile): Use lookahead rather than PATUNFETCH (for repetition
operators, char classes, shy-groups and intervals).
Optimize special cases of intervals so as to only use succeed_n and
jump_n when really needed.
(re_compile_fastmap): Simplify handling of jump_n and succeed_n now
that we don't have to handle the special cases any more.
Simplify on_failure_jump handling as well.
Stefan Monnier [Sun, 26 Mar 2000 23:05:51 +0000 (23:05 +0000)]
(enum re_opcode_t): New opcode on_failure_jump_nastyloop.
(print_partial_compiled_pattern, re_compile_fastmap): Handle new opcode.
(regex_compile): Use on_failure_jump_nastyloop for non-greedy loops.
(re_match_2_internal): Add code for on_failure_jump_nastyloop when
executing it as well as when popping it off the stack to find infinite
loops in non-greedy repetition operators.
Jim Meyering [Fri, 24 Mar 2000 05:54:46 +0000 (05:54 +0000)]
(EXTRA_DIST): Add ref-add.sed.in and ref-del.sed.in.
(charset_alias): Define.
(install-exec-local): Factor out common code.
(uninstall-local): Split lines longer than 80.
(ref-add.sed, ref-del.sed): Remove rules... (do the following instead)
(SUFFIXES): Define.
(.sed.in.sed): New rule. Don't redirect directly to $@.
(CLEANFILES): Add ref-add.sed and ref-del.sed.
Stefan Monnier [Wed, 22 Mar 2000 04:17:32 +0000 (04:17 +0000)]
(CHAR_CHARSET, CHARSET_LEADING_CODE_BASE): Add default
definitions for non-Emacs compilation.
(enum re_opcode_t): Remove (not)wordchar and move (not)syntaxspec
outside of `#ifdef emacs'.
(print_partial_compiled_pattern): Update.
(regex_compile): Use (not)syntaxspec(Sword) instead of (not)wordchar.
(re_compile_fastmap): Merge handling of charset and charset_not (for
emacs and non-emacs compilation as well).
Similarly for (not)categoryspec and (not)syntaxspec.
Don't use the fastmap when reaching `anychar' since the added
complexity is not justified.
(re_match_2_internal): Merge (not)wordchar (emacs and non-emacs)
and (not)syntaxspec. Merge (not)categoryspec.
Stefan Monnier [Sun, 19 Mar 2000 23:22:06 +0000 (23:22 +0000)]
(RE_STRING_CHAR): New macro.
(GET_CHAR_AFER_2): Remove.
(RE_TRANSLATE, RE_TRANSLATE_P): New macros moved from regex.h.
(enum re_opcode_t): Remove on_failure_jump_exclusive.
(print_partial_compiled_pattern, re_compile_fastmap)
(re_match_2_internal): Remove on_failure_jump_exclusive.
(regex_compile): Turn optimizable P+ loops into PP*, so that the
optimization only need to work for * (ie. can use of_keep_string_jump).
Remove the special case for .*\n since it is now covered by the general
optimization.
(re_search_2): Don't bother with `room'.
(skip_one_char): New function.
(skip_noops): Simplify since `memory' is not needed any more.
(mutually_exclusive_p): Restructure slightly to use `switch' and
add handling for "all" remaining cases.
(re_match_2_internal): Change on_failure_jump_smart to use
on_failure_keep_string_jump (and redirect the end-of-loop jump)
rather than on_failure_jump_exclusive.
Jim Meyering [Fri, 17 Mar 2000 19:32:27 +0000 (19:32 +0000)]
(utf8_wctomb): New function.
(print_unicode_char): Pass the Unicode character to iconv in UTF-8
format instead of in UCS-4 with platform dependent endianness.
Stefan Monnier [Thu, 16 Mar 2000 02:53:38 +0000 (02:53 +0000)]
(re_match_2): Fix string shortening (to fit `stop') to make sure
POINTER_TO_OFFSET gives the same value before and after PREFETCH.
Use `dfail' to guarantee "atomic" matching.
(PTR_TO_OFFSET): Use POINTER_TO_OFFSET.
(debug): Now only active if > 0 rather than if != 0.
(DEBUG_*): Update for the new meaning of `debug'.
(print_partial_compiled_pattern): Add missing `succeed' case.
Use CHARSET_* macros in the charset(_not) branch.
Fix off-by-two bugs in `succeed_n', `jump_n' and `set_number_at'.
(store_op1, store_op2, insert_op1, insert_op2)
(at_begline_loc_p, at_endline_loc_p): Add prototype.
(group_in_compile_stack): Move to after its arg's types are declared
and add a prototype.
(PATFETCH): Define in terms of PATFETCH_RAW.
(GET_UNSIGNED_NUMBER): Add the usual `do { ... } while(0)' wrapper.
(QUIT): Redefine as a nop except for NTemacs.
(regex_compile): Handle intervals {,M} as if it was {0,M}.
Fix indentation of the greedy-op and shy-group code.
(at_(beg|end)line_loc_p): Fix argument's types.
(re_compile_fastmap): Ifdef out failure_stack_ptr to shut up gcc.
(re_search_2): Use POS_AS_IN_BUFFER. Simplify `room' computation.
(MATCHING_IN_FIRST_STRING): Remove.
(re_match_2): Use POS_AS_IN_BUFFER.
Ifdef out failure_stack_ptr to shut up gcc.
Use FIRST_STRING_P and POINTER_TO_OFFSET.
Use QUIT unconditionally.
Stefan Monnier [Tue, 14 Mar 2000 00:27:57 +0000 (00:27 +0000)]
* regex.c: Declare a new type `re_char' used throughout the code for the
string char type. It's `const unsigned char' to match the rest of Emacs.
Consistently make sure all pointers to strings use it and make sure all
pointers into the pattern use `unsigned char'.
(re_match_2_internal): Use `PREFETCH+STRING_CHAR' instead of
GET_CHAR_AFTER_2.
Also merge wordbound and notwordbound to reduce code duplication.
* charset.h (GET_CHAR_AFTER_2): Remove.
(GET_CHAR_BEFORE_2): Use unsigned chars, like everywhere else.
Stefan Monnier [Wed, 8 Mar 2000 23:25:41 +0000 (23:25 +0000)]
This is a big redesign of failure-stack and register handling, prompted
by bugs revealed when trying to add shy-groups. Overall, what happened
is that loops are now structured a little differently, groups can be
shy and the code is a little simpler.
(enum re_opcode_t): Remove jump_past_alt, maybe_pop_jump,
push_dummy_failure and dumy_failure_jump.
Add on_failure_jump_(exclusive, loop and smart).
Also fix the comment for (start|stop)_memory since they now only take
one argument (the second has becomes unnecessary).
(print_partial_compiled_pattern): Adjust for changes in re_opcode_t.
(print_compiled_pattern): Use %ld to printf long ints and flush to make
debugging a little easier.
(union fail_stack_elt): Make the integer unsigned.
(struct fail_stack_type): Add a `frame' element.
(INIT_FAIL_STACK): Init `frame' as well.
(POP_PATTERN_OP): New macro for re_compile_fastmap.
(DEBUG_PUSH, DEBUG_POP): Remove.
(NUM_REG_ITEMS): Remove.
(NUM_NONREG_ITEMS): Adjust.
(FAILURE_PAT, FAILURE_STR, NEXT_FAILURE_HANDLE, TOP_FAILURE_HANDLE):
New macros for the cycle detection.
(ENSURE_FAIL_STACK): New macro for PUSH_FAILURE_(REG|POINT).
(PUSH_FAILURE_REG, POP_FAILURE_REG, CHECK_INFINITE_LOOP): New macros.
(PUSH_FAILURE_POINT): Don't push registers any more.
The pattern address pushed is not the destination of the jump
but the source of it instead.
(NUM_FAILURE_ITEMS): Remove.
(POP_FAILURE_POINT): Adapt to the new stack structure (i.e. pop
registers before the actual failure point).
Don't hardcode any meaning for str==NULL anymore.
(union register_info_type, REG_MATCH_NULL_STRING_P, IS_ACTIVE)
(MATCHED_SOMETHING, EVER_MATCHED_SOMETHING, SET_REGS_MATCHED): Remove.
(REG_UNSET_VALUE): Use NULL (why not?).
(compile_range): Remove declaration since it doesn't exist.
(struct compile_stack_elt_t): Remove inner_group_offset.
(old_reg(start|end), reg_info, reg_dummy, reg_info_dummy): Remove.
(regex_grow_registers): Remove dead code.
(FIXUP_ALT_JUMP): New macro.
(regex_compile): Add shy-groups
Change loops to use on_failure_jump_smart&jump instead of
on_failure_jump&maybe_pop_jump.
Change + loops to eliminate the initial (dummy_failure_)jump.
Remove c1_base (looks like unused variable to me).
Use `jump' instead of `jump_past_alt' and don't bother with
push_dummy_failure in alternatives since it is now unnecessary.
Use FIXUP_ALT_JUMP.
Eliminate a useless `#ifdef emacs' for (re)allocating the stack.
(re_compile_fastmap): Remove dead variables i and num_regs.
Exit from loop when bufp->can_be_null rather than jumping to `done'.
Avoid jumping backwards so as to ensure termination.
Use PATTERN_STACK_EMPTY and POP_PATTERN_OP.
Improved handling of backreferences.
Remove dead code in handling of `anychar'.
(skip_noops, mutually_exclusive_p): New functions taken from the
handling of `maybe_pop_jump' in re_match_2_internal.
Slightly improve mutually_exclusive_p to handle ".+\n".
((lowest|highest)_active_reg, NO_(LOWEST|HIGHEST)_ACTIVE_REG)
Remove.
(re_match_2_internal): Use %p instead of 0x%x when printf'ing ptrs.
Don't SET_REGS_MATCHED anymore. Remove many dead variables.
Push register (in `start_memory') on the stack rather than storing it
in old_reg(start|end).
Remove the cycle detection from `stop_memory', replaced by the use
of on_failure_jump_loop for greedy loops.
Add code for the new on_failure_jump_<foo>.
Remove ad-hoc code in `on_failure_jump' to push more registers
in the case of a loop.
Take out code from `maybe_pop_jump' into separate functions and
adapt it to the semantics of `on_failure_jump_smart'.
Remove jump_past_alt, dummy_failure_jump and push_dummy_failure.
Remove dummy_failure handling and handling of `failures to jump
to on_failure_jump' (this last one was already dead code, it seems).
((group|alt|common_op)_match_null_string_p): Remove.