ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 91 - (show annotations)
Sat Feb 24 21:41:34 2007 UTC (14 years, 2 months ago) by nigel
File size: 104662 byte(s)
Load pcre-6.7 into code/trunk.
1 ChangeLog for PCRE
2 ------------------
4 Version 6.7 04-Jul-06
5 ---------------------
7 1. In order to handle tests when input lines are enormously long, pcretest has
8 been re-factored so that it automatically extends its buffers when
9 necessary. The code is crude, but this _is_ just a test program. The
10 default size has been increased from 32K to 50K.
12 2. The code in pcre_study() was using the value of the re argument before
13 testing it for NULL. (Of course, in any sensible call of the function, it
14 won't be NULL.)
16 3. The memmove() emulation function in pcre_internal.h, which is used on
17 systems that lack both memmove() and bcopy() - that is, hardly ever -
18 was missing a "static" storage class specifier.
20 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
21 containing an extended class (one that cannot be represented by a bitmap
22 because it contains high-valued characters or Unicode property items, e.g.
23 [\pZ]). Almost always one would set UTF-8 mode when processing such a
24 pattern, but PCRE should not loop if you do not (it no longer does).
25 [Detail: two cases were found: (a) a repeated subpattern containing an
26 extended class; (b) a recursive reference to a subpattern that followed a
27 previous extended class. It wasn't skipping over the extended class
28 correctly when UTF-8 mode was not set.]
30 5. A negated single-character class was not being recognized as fixed-length
31 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
32 compile error "lookbehind assertion is not fixed length".
34 6. The RunPerlTest auxiliary script was showing an unexpected difference
35 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
36 write a Perl script that can interpret lines of an input file either as
37 byte characters or as UTF-8, which is what "perltest" was being required to
38 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
39 can't do is switch easily at run time between having the "use utf8;" pragma
40 or not. In the end, I fudged it by using the RunPerlTest script to insert
41 "use utf8;" explicitly for the UTF-8 tests.
43 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
44 the end of the subject string, contrary to the documentation and to what
45 Perl does. This was true of both matching functions. Now it matches only at
46 the start of the subject and immediately after *internal* newlines.
48 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
49 a pointer to an int instead of a pointer to an unsigned long int. This
50 caused problems on 64-bit systems.
52 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
53 instance of the 'standard' template library not being so standard".
55 10. There was no check on the number of named subpatterns nor the maximum
56 length of a subpattern name. The product of these values is used to compute
57 the size of the memory block for a compiled pattern. By supplying a very
58 long subpattern name and a large number of named subpatterns, the size
59 computation could be caused to overflow. This is now prevented by limiting
60 the length of names to 32 characters, and the number of named subpatterns
61 to 10,000.
63 11. Subpatterns that are repeated with specific counts have to be replicated in
64 the compiled pattern. The size of memory for this was computed from the
65 length of the subpattern and the repeat count. The latter is limited to
66 65535, but there was no limit on the former, meaning that integer overflow
67 could in principle occur. The compiled length of a repeated subpattern is
68 now limited to 30,000 bytes in order to prevent this.
70 12. Added the optional facility to have named substrings with the same name.
72 13. Added the ability to use a named substring as a condition, using the
73 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
74 are numbers (not recommended). Forward references are permitted.
76 14. Added forward references in named backreferences (if you see what I mean).
78 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
79 pattern could run off the end of the subject. For example, the pattern
80 "(?s)(.{1,5})"8 did this with the subject "ab".
82 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
83 PCRE_CASELESS was set when matching characters that were quantified with ?
84 or *.
86 17. A character class other than a single negated character that had a minimum
87 but no maximum quantifier - for example [ab]{6,} - was not handled
88 correctly by pce_dfa_exec(). It would match only one character.
90 18. A valid (though odd) pattern that looked like a POSIX character
91 class but used an invalid character after [ (for example [[,abc,]]) caused
92 pcre_compile() to give the error "Failed: internal error: code overflow" or
93 in some cases to crash with a glibc free() error. This could even happen if
94 the pattern terminated after [[ but there just happened to be a sequence of
95 letters, a binary zero, and a closing ] in the memory that followed.
97 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
98 over the years. Originally (before any Unicode support), just the bottom 8
99 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
100 output from "man perlunicode" includes this:
102 The regular expression compiler produces polymorphic opcodes. That
103 is, the pattern adapts to the data and automatically switches to
104 the Unicode character scheme when presented with Unicode data--or
105 instead uses a traditional byte scheme when presented with byte
106 data.
108 Sadly, a wide octal escape does not cause a switch, and in a string with
109 no other multibyte characters, these octal escapes are treated as before.
110 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
111 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
112 Unicode string.
114 I have not perpetrated such confusion in PCRE. Up till now, it took just
115 the bottom 8 bits, as in old Perl. I have now made octal escapes with
116 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
117 translate to the appropriate multibyte character.
119 29. Applied some refactoring to reduce the number of warnings from Microsoft
120 and Borland compilers. This has included removing the fudge introduced
121 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
122 a warning about an unused variable.
124 21. PCRE has not included VT (character 0x0b) in the set of whitespace
125 characters since release 4.0, because Perl (from release 5.004) does not.
126 [Or at least, is documented not to: some releases seem to be in conflict
127 with the documentation.] However, when a pattern was studied with
128 pcre_study() and all its branches started with \s, PCRE still included VT
129 as a possible starting character. Of course, this did no harm; it just
130 caused an unnecessary match attempt.
132 22. Removed a now-redundant internal flag bit that recorded the fact that case
133 dependency changed within the pattern. This was once needed for "required
134 byte" processing, but is no longer used. This recovers a now-scarce options
135 bit. Also moved the least significant internal flag bit to the most-
136 significant bit of the word, which was not previously used (hangover from
137 the days when it was an int rather than a uint) to free up another bit for
138 the future.
140 23. Added support for CRLF line endings as well as CR and LF. As well as the
141 default being selectable at build time, it can now be changed at runtime
142 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
143 specify that it is scanning data with non-default line endings.
145 24. Changed the definition of CXXLINK to make it agree with the definition of
146 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
148 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
149 recursions. This makes a big different to stack usage for some patterns.
151 26. If a subpattern containing a named recursion or subroutine reference such
152 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
153 the space required for the compiled pattern went wrong and gave too small a
154 value. Depending on the environment, this could lead to "Failed: internal
155 error: code overflow at offset 49" or "glibc detected double free or
156 corruption" errors.
158 27. Applied patches from Google (a) to support the new newline modes and (b) to
159 advance over multibyte UTF-8 characters in GlobalReplace.
161 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
162 difference for some implementation of PCRE in some Windows version.
164 29. Added some extra testing facilities to pcretest:
166 \q<number> in a data line sets the "match limit" value
167 \Q<number> in a data line sets the "match recursion limt" value
168 -S <number> sets the stack size, where <number> is in megabytes
170 The -S option isn't available for Windows.
173 Version 6.6 06-Feb-06
174 ---------------------
176 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
177 in pcreposix.h. I have copied the definition from pcre.h.
179 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
180 because pcre.h is no longer a built file.
182 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
183 not normally included in the compiled code.
186 Version 6.5 01-Feb-06
187 ---------------------
189 1. When using the partial match feature with pcre_dfa_exec(), it was not
190 anchoring the second and subsequent partial matches at the new starting
191 point. This could lead to incorrect results. For example, with the pattern
192 /1234/, partially matching against "123" and then "a4" gave a match.
194 2. Changes to pcregrep:
196 (a) All non-match returns from pcre_exec() were being treated as failures
197 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
198 error message is output. Some extra information is given for the
200 probably the only errors that are likely to be caused by users (by
201 specifying a regex that has nested indefinite repeats, for instance).
202 If there are more than 20 of these errors, pcregrep is abandoned.
204 (b) A binary zero was treated as data while matching, but terminated the
205 output line if it was written out. This has been fixed: binary zeroes
206 are now no different to any other data bytes.
208 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
209 used to set a locale for matching. The --locale=xxxx long option has
210 been added (no short equivalent) to specify a locale explicitly on the
211 pcregrep command, overriding the environment variables.
213 (d) When -B was used with -n, some line numbers in the output were one less
214 than they should have been.
216 (e) Added the -o (--only-matching) option.
218 (f) If -A or -C was used with -c (count only), some lines of context were
219 accidentally printed for the final match.
221 (g) Added the -H (--with-filename) option.
223 (h) The combination of options -rh failed to suppress file names for files
224 that were found from directory arguments.
226 (i) Added the -D (--devices) and -d (--directories) options.
228 (j) Added the -F (--fixed-strings) option.
230 (k) Allow "-" to be used as a file name for -f as well as for a data file.
232 (l) Added the --colo(u)r option.
234 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
235 is not present by default.
237 3. A nasty bug was discovered in the handling of recursive patterns, that is,
238 items such as (?R) or (?1), when the recursion could match a number of
239 alternatives. If it matched one of the alternatives, but subsequently,
240 outside the recursion, there was a failure, the code tried to back up into
241 the recursion. However, because of the way PCRE is implemented, this is not
242 possible, and the result was an incorrect result from the match.
244 In order to prevent this happening, the specification of recursion has
245 been changed so that all such subpatterns are automatically treated as
246 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
248 4. I had overlooked the fact that, in some locales, there are characters for
249 which isalpha() is true but neither isupper() nor islower() are true. In
250 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
251 and ordfeminine) are like this. This affected the treatment of \w and \W
252 when they appeared in character classes, but not when they appeared outside
253 a character class. The bit map for "word" characters is now created
254 separately from the results of isalnum() instead of just taking it from the
255 upper, lower, and digit maps. (Plus the underscore character, of course.)
257 5. The above bug also affected the handling of POSIX character classes such as
258 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
259 permanent tables. Instead, the bit maps for such a class were previously
260 created as the appropriate unions of the upper, lower, and digit bitmaps.
261 Now they are created by subtraction from the [[:word:]] class, which has
262 its own bitmap.
264 6. The [[:blank:]] character class matches horizontal, but not vertical space.
265 It is created by subtracting the vertical space characters (\x09, \x0a,
266 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
267 subtraction was done in the overall bitmap for a character class, meaning
268 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
269 be recognized. This bug has been fixed.
271 7. Patches from the folks at Google:
273 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
274 real life, but is still worth protecting against".
276 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
277 regular expressions".
279 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
280 have it.
282 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
283 "configure" and the latter not, in order to fix a problem somebody had
284 with compiling the Arg class on HP-UX.
286 (e) Improve the error-handling of the C++ wrapper a little bit.
288 (f) New tests for checking recursion limiting.
290 8. The pcre_memmove() function, which is used only if the environment does not
291 have a standard memmove() function (and is therefore rarely compiled),
292 contained two bugs: (a) use of int instead of size_t, and (b) it was not
293 returning a result (though PCRE never actually uses the result).
295 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
296 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
297 returned instead of calling malloc() with an overflowing number that would
298 most likely cause subsequent chaos.
300 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
302 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
303 with this option is matched, the nmatch and pmatch options of regexec() are
304 ignored.
306 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
307 provided in case anyone wants to the the POSIX interface with UTF-8
308 strings.
310 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
311 C++ linking (needed for some HP-UX environments).
313 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
314 (unused parameter) and in the pcre_printint() function (omitted "default"
315 switch label when the default is to do nothing).
317 15. Added some code to make it possible, when PCRE is compiled as a C++
318 library, to replace subject pointers for pcre_exec() with a smart pointer
319 class, thus making it possible to process discontinuous strings.
321 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
322 much the same function. They were added by different people who were trying
323 to make PCRE easy to compile on non-Unix systems. It has been suggested
324 that PCRE_EXPORT be abolished now that there is more automatic apparatus
325 for compiling on Windows systems. I have therefore replaced it with
326 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
327 defaults to "extern" for C or "extern C" for C++, which works fine on
328 Unix-like systems. It is now possible to override the value of PCRE_DATA_
329 SCOPE with something explicit in config.h. In addition:
331 (a) pcreposix.h still had just "extern" instead of either of these macros;
332 I have replaced it with PCRE_DATA_SCOPE.
334 (b) Functions such as _pcre_xclass(), which are internal to the library,
335 but external in the C sense, all had PCRE_EXPORT in their definitions.
336 This is apparently wrong for the Windows case, so I have removed it.
337 (It makes no difference on Unix-like systems.)
339 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
340 of recursive calls to match(). This is different to MATCH_LIMIT because
341 that limits the total number of calls to match(), not all of which increase
342 the depth of recursion. Limiting the recursion depth limits the amount of
343 stack (or heap if NO_RECURSE is set) that is used. The default can be set
344 when PCRE is compiled, and changed at run time. A patch from Google adds
345 this functionality to the C++ interface.
347 18. Changes to the handling of Unicode character properties:
349 (a) Updated the table to Unicode 4.1.0.
351 (b) Recognize characters that are not in the table as "Cn" (undefined).
353 (c) I revised the way the table is implemented to a much improved format
354 which includes recognition of ranges. It now supports the ranges that
355 are defined in UnicodeData.txt, and it also amalgamates other
356 characters into ranges. This has reduced the number of entries in the
357 table from around 16,000 to around 3,000, thus reducing its size
358 considerably. I realized I did not need to use a tree structure after
359 all - a binary chop search is just as efficient. Having reduced the
360 number of entries, I extended their size from 6 bytes to 8 bytes to
361 allow for more data.
363 (d) Added support for Unicode script names via properties such as \p{Han}.
365 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
366 matching that character.
368 20. When matching a repeated Unicode property with a minimum greater than zero,
369 (for example \pL{2,}), PCRE could look past the end of the subject if it
370 reached it while seeking the minimum number of characters. This could
371 happen only if some of the characters were more than one byte long, because
372 there is a check for at least the minimum number of bytes.
374 21. Refactored the implementation of \p and \P so as to be more general, to
375 allow for more different types of property in future. This has changed the
376 compiled form incompatibly. Anybody with saved compiled patterns that use
377 \p or \P will have to recompile them.
379 22. Added "Any" and "L&" to the supported property types.
381 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
382 but give a compile time error if the value is greater than 0xff.
384 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
385 accidentally not being installed or uninstalled.
387 25. The pcre.h file was built from pcre.h.in, but the only changes that were
388 made were to insert the current release number. This seemed silly, because
389 it made things harder for people building PCRE on systems that don't run
390 "configure". I have turned pcre.h into a distributed file, no longer built
391 by "configure", with the version identification directly included. There is
392 no longer a pcre.h.in file.
394 However, this change necessitated a change to the pcre-config script as
395 well. It is built from pcre-config.in, and one of the substitutions was the
396 release number. I have updated configure.ac so that ./configure now finds
397 the release number by grepping pcre.h.
399 26. Added the ability to run the tests under valgrind.
402 Version 6.4 05-Sep-05
403 ---------------------
405 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
406 "--" to be printed when multiple files were scanned, even when none of the
407 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
408 consider it to be a bug, and have restored the previous behaviour.
410 2. A couple of code tidies to get rid of compiler warnings.
412 3. The pcretest program used to cheat by referring to symbols in the library
413 whose names begin with _pcre_. These are internal symbols that are not
414 really supposed to be visible externally, and in some environments it is
415 possible to suppress them. The cheating is now confined to including
416 certain files from the library's source, which is a bit cleaner.
418 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
419 file's purpose clearer.
421 5. Reorganized pcre_ucp_findchar().
424 Version 6.3 15-Aug-05
425 ---------------------
427 1. The file libpcre.pc.in did not have general read permission in the tarball.
429 2. There were some problems when building without C++ support:
431 (a) If C++ support was not built, "make install" and "make test" still
432 tried to test it.
434 (b) There were problems when the value of CXX was explicitly set. Some
435 changes have been made to try to fix these, and ...
437 (c) --disable-cpp can now be used to explicitly disable C++ support.
439 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
440 backslash in a target when C++ was disabled. This confuses some
441 versions of "make", apparently. Using an intermediate variable solves
442 this. (Same for CPP_LOBJ.)
444 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
445 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
446 necessary on certain architectures.
448 4. Added a setting of -export-symbols-regex to the link command to remove
449 those symbols that are exported in the C sense, but actually are local
450 within the library, and not documented. Their names all begin with
451 "_pcre_". This is not a perfect job, because (a) we have to except some
452 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
453 available (and never for static libraries). I have made a note to try to
454 find a way round (a) in the future.
457 Version 6.2 01-Aug-05
458 ---------------------
460 1. There was no test for integer overflow of quantifier values. A construction
461 such as {1111111111111111} would give undefined results. What is worse, if
462 a minimum quantifier for a parenthesized subpattern overflowed and became
463 negative, the calculation of the memory size went wrong. This could have
464 led to memory overwriting.
466 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
468 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
469 operating environments where this matters.
471 4. Applied Giuseppe Maxia's patch to add additional features for controlling
472 PCRE options from within the C++ wrapper.
474 5. Named capturing subpatterns were not being correctly counted when a pattern
475 was compiled. This caused two problems: (a) If there were more than 100
476 such subpatterns, the calculation of the memory needed for the whole
477 compiled pattern went wrong, leading to an overflow error. (b) Numerical
478 back references of the form \12, where the number was greater than 9, were
479 not recognized as back references, even though there were sufficient
480 previous subpatterns.
482 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
483 versions of gcc, e.g. 2.95.4.
486 Version 6.1 21-Jun-05
487 ---------------------
489 1. There was one reference to the variable "posix" in pcretest.c that was not
490 surrounded by "#if !defined NOPOSIX".
492 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
493 the cross-check on the old pcre_info() function, for the benefit of the
494 cut-down version of PCRE that is currently imported into Exim.
496 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
497 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
498 patterns that start (?i) but I don't think that matters. The use of (?i) is
499 just an example; this all applies to the other options as well.
501 4. Since libtool seems to echo the compile commands it is issuing, the output
502 from "make" can be reduced a bit by putting "@" in front of each libtool
503 compile command.
505 5. Patch from the folks at Google for configure.in to be a bit more thorough
506 in checking for a suitable C++ installation before trying to compile the
507 C++ stuff. This should fix a reported problem when a compiler was present,
508 but no suitable headers.
510 6. The man pages all had just "PCRE" as their title. I have changed them to
511 be the relevant file name. I have also arranged that these names are
512 retained in the file doc/pcre.txt, which is a concatenation in text format
513 of all the man pages except the little individual ones for each function.
515 7. The NON-UNIX-USE file had not been updated for the different set of source
516 files that come with release 6. I also added a few comments about the C++
517 wrapper.
520 Version 6.0 07-Jun-05
521 ---------------------
523 1. Some minor internal re-organization to help with my DFA experiments.
525 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
526 didn't matter for the library itself when fully configured, but did matter
527 when compiling without UCP support, or within Exim, where the ucp files are
528 not imported.
530 3. Refactoring of the library code to split up the various functions into
531 different source modules. The addition of the new DFA matching code (see
532 below) to a single monolithic source would have made it really too
533 unwieldy, quite apart from causing all the code to be include in a
534 statically linked application, when only some functions are used. This is
535 relevant even without the DFA addition now that patterns can be compiled in
536 one application and matched in another.
538 The downside of splitting up is that there have to be some external
539 functions and data tables that are used internally in different modules of
540 the library but which are not part of the API. These have all had their
541 names changed to start with "_pcre_" so that they are unlikely to clash
542 with other external names.
544 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
545 a different (DFA) algorithm. Although it is slower than the original
546 function, it does have some advantages for certain types of matching
547 problem.
549 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
550 including restarting after a partial match.
552 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
553 defined when compiling for Windows was sent to me. I have put it into the
554 code, though I have no means of testing or verifying it.
556 7. Added the pcre_refcount() auxiliary function.
558 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
559 match before or at the first newline in the subject string. In pcretest,
560 the /f option on a pattern can be used to set this.
562 9. A repeated \w when used in UTF-8 mode with characters greater than 256
563 would behave wrongly. This has been present in PCRE since release 4.0.
565 10. A number of changes to the pcregrep command:
567 (a) Refactored how -x works; insert ^(...)$ instead of setting
568 PCRE_ANCHORED and checking the length, in preparation for adding
569 something similar for -w.
571 (b) Added the -w (match as a word) option.
573 (c) Refactored the way lines are read and buffered so as to have more
574 than one at a time available.
576 (d) Implemented a pcregrep test script.
578 (e) Added the -M (multiline match) option. This allows patterns to match
579 over several lines of the subject. The buffering ensures that at least
580 8K, or the rest of the document (whichever is the shorter) is available
581 for matching (and similarly the previous 8K for lookbehind assertions).
583 (f) Changed the --help output so that it now says
585 -w, --word-regex(p)
587 instead of two lines, one with "regex" and the other with "regexp"
588 because that confused at least one person since the short forms are the
589 same. (This required a bit of code, as the output is generated
590 automatically from a table. It wasn't just a text change.)
592 (g) -- can be used to terminate pcregrep options if the next thing isn't an
593 option but starts with a hyphen. Could be a pattern or a path name
594 starting with a hyphen, for instance.
596 (h) "-" can be given as a file name to represent stdin.
598 (i) When file names are being printed, "(standard input)" is used for
599 the standard input, for compatibility with GNU grep. Previously
600 "<stdin>" was used.
602 (j) The option --label=xxx can be used to supply a name to be used for
603 stdin when file names are being printed. There is no short form.
605 (k) Re-factored the options decoding logic because we are going to add
606 two more options that take data. Such options can now be given in four
607 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
609 (l) Added the -A, -B, and -C options for requesting that lines of context
610 around matches be printed.
612 (m) Added the -L option to print the names of files that do not contain
613 any matching lines, that is, the complement of -l.
615 (n) The return code is 2 if any file cannot be opened, but pcregrep does
616 continue to scan other files.
618 (o) The -s option was incorrectly implemented. For compatibility with other
619 greps, it now suppresses the error message for a non-existent or non-
620 accessible file (but not the return code). There is a new option called
621 -q that suppresses the output of matching lines, which was what -s was
622 previously doing.
624 (p) Added --include and --exclude options to specify files for inclusion
625 and exclusion when recursing.
627 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
628 Hopefully, it now does.
630 12. Missing cast in pcre_study().
632 13. Added an "uninstall" target to the makefile.
634 14. Replaced "extern" in the function prototypes in Makefile.in with
635 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
636 world, but is set differently for Windows.
638 15. Added a second compiling function called pcre_compile2(). The only
639 difference is that it has an extra argument, which is a pointer to an
640 integer error code. When there is a compile-time failure, this is set
641 non-zero, in addition to the error test pointer being set to point to an
642 error message. The new argument may be NULL if no error number is required
643 (but then you may as well call pcre_compile(), which is now just a
644 wrapper). This facility is provided because some applications need a
645 numeric error indication, but it has also enabled me to tidy up the way
646 compile-time errors are handled in the POSIX wrapper.
648 16. Added VPATH=.libs to the makefile; this should help when building with one
649 prefix path and installing with another. (Or so I'm told by someone who
650 knows more about this stuff than I do.)
652 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
653 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
654 match everything, including newlines. This is not POSIX-compatible, but
655 somebody wanted the feature. From pcretest it can be activated by using
656 both the P and the s flags.
658 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
660 19. libpcre.pc was being incorrectly installed as executable.
662 20. A couple of places in pcretest check for end-of-line by looking for '\n';
663 it now also looks for '\r' so that it will work unmodified on Windows.
665 21. Added Google's contributed C++ wrapper to the distribution.
667 22. Added some untidy missing memory free() calls in pcretest, to keep
668 Electric Fence happy when testing.
672 Version 5.0 13-Sep-04
673 ---------------------
675 1. Internal change: literal characters are no longer packed up into items
676 containing multiple characters in a single byte-string. Each character
677 is now matched using a separate opcode. However, there may be more than one
678 byte in the character in UTF-8 mode.
680 2. The pcre_callout_block structure has two new fields: pattern_position and
681 next_item_length. These contain the offset in the pattern to the next match
682 item, and its length, respectively.
684 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
685 insertion of callouts before each pattern item. Added the /C option to
686 pcretest to make use of this.
688 4. On the advice of a Windows user, the lines
690 #if defined(_WIN32) || defined(WIN32)
691 _setmode( _fileno( stdout ), 0x8000 );
692 #endif /* defined(_WIN32) || defined(WIN32) */
694 have been added to the source of pcretest. This apparently does useful
695 magic in relation to line terminators.
697 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
698 for the benefit of those environments where the "b" makes a difference.
700 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
701 to know about it. I have put a hack into configure.in that adds in code
702 to set GCC=yes if CC=icc. This seems to end up at a point in the
703 generated configure script that is early enough to affect the setting of
704 compiler options, which is what is needed, but I have no means of testing
705 whether it really works. (The user who reported this had patched the
706 generated configure script, which of course I cannot do.)
708 LATER: After change 22 below (new libtool files), the configure script
709 seems to know about icc (and also ecc). Therefore, I have commented out
710 this hack in configure.in.
712 7. Added support for pkg-config (2 patches were sent in).
714 8. Negated POSIX character classes that used a combination of internal tables
715 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
716 [[:^ascii]]. Typically, they would match almost any characters. The other
717 POSIX classes were not broken in this way.
719 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
720 to find the match, as PCRE was deluded into thinking that the match had to
721 start at the start point or following a newline. The same bug applied to
722 patterns with negative forward assertions or any backward assertions
723 preceding ".*" at the start, unless the pattern required a fixed first
724 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
726 10. In UTF-8 mode, when moving forwards in the subject after a failed match
727 starting at the last subject character, bytes beyond the end of the subject
728 string were read.
730 11. Renamed the variable "class" as "classbits" to make life easier for C++
731 users. (Previously there was a macro definition, but it apparently wasn't
732 enough.)
734 12. Added the new field "tables" to the extra data so that tables can be passed
735 in at exec time, or the internal tables can be re-selected. This allows
736 a compiled regex to be saved and re-used at a later time by a different
737 program that might have everything at different addresses.
739 13. Modified the pcre-config script so that, when run on Solaris, it shows a
740 -R library as well as a -L library.
742 14. The debugging options of pcretest (-d on the command line or D on a
743 pattern) showed incorrect output for anything following an extended class
744 that contained multibyte characters and which was followed by a quantifier.
746 15. Added optional support for general category Unicode character properties
747 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
748 support. It adds about 90K to the size of the library. The meanings of the
749 inbuilt class escapes such as \d and \s have NOT been changed.
751 16. Updated pcredemo.c to include calls to free() to release the memory for the
752 compiled pattern.
754 17. The generated file chartables.c was being created in the source directory
755 instead of in the building directory. This caused the build to fail if the
756 source directory was different from the building directory, and was
757 read-only.
759 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
760 file. No doubt somebody will tell me if they don't make sense... Also added
761 Dan Mooney's comments about building on OpenVMS.
763 19. Added support for partial matching via the PCRE_PARTIAL option for
764 pcre_exec() and the \P data escape in pcretest.
766 20. Extended pcretest with 3 new pattern features:
768 (i) A pattern option of the form ">rest-of-line" causes pcretest to
769 write the compiled pattern to the file whose name is "rest-of-line".
770 This is a straight binary dump of the data, with the saved pointer to
771 the character tables forced to be NULL. The study data, if any, is
772 written too. After writing, pcretest reads a new pattern.
774 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
775 compiled pattern from the given file. There must not be any
776 occurrences of "<" in the file name (pretty unlikely); if there are,
777 pcretest will instead treat the initial "<" as a pattern delimiter.
778 After reading in the pattern, pcretest goes on to read data lines as
779 usual.
781 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
782 and 16-bit fields in a compiled pattern, to simulate a pattern that
783 was compiled on a host of opposite endianness.
785 21. The pcre-exec() function can now cope with patterns that were compiled on
786 hosts of opposite endianness, with this restriction:
788 As for any compiled expression that is saved and used later, the tables
789 pointer field cannot be preserved; the extra_data field in the arguments
790 to pcre_exec() should be used to pass in a tables address if a value
791 other than the default internal tables were used at compile time.
793 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
794 now diagnosed as an error. Previously, most of the time, a negative number
795 would have been treated as zero, but if in addition "ovector" was passed as
796 NULL, a crash could occur.
798 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
799 new versions from the libtool 1.5 distribution (the last one is a copy of
800 a file called libtool.m4). This seems to have fixed the need to patch
801 "configure" to support Darwin 1.3 (which I used to do). However, I still
802 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
803 workstation).
805 24. Changed the PCRE licence to be the more standard "BSD" licence.
808 Version 4.5 01-Dec-03
809 ---------------------
811 1. There has been some re-arrangement of the code for the match() function so
812 that it can be compiled in a version that does not call itself recursively.
813 Instead, it keeps those local variables that need separate instances for
814 each "recursion" in a frame on the heap, and gets/frees frames whenever it
815 needs to "recurse". Keeping track of where control must go is done by means
816 of setjmp/longjmp. The whole thing is implemented by a set of macros that
817 hide most of the details from the main code, and operates only if
818 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
819 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
820 operating.
822 To make it easier for callers to provide specially tailored get/free
823 functions for this usage, two new functions, pcre_stack_malloc, and
824 pcre_stack_free, are used. They are always called in strict stacking order,
825 and the size of block requested is always the same.
827 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
828 PCRE has been compiled to use the stack or the heap for recursion. The
829 -C option of pcretest uses this to show which version is compiled.
831 A new data escape \S, is added to pcretest; it causes the amounts of store
832 obtained and freed by both kinds of malloc/free at match time to be added
833 to the output.
835 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
836 what's available on my current Linux desktop machine.
838 3. When matching a UTF-8 string, the test for a valid string at the start has
839 been extended. If start_offset is not zero, PCRE now checks that it points
840 to a byte that is the start of a UTF-8 character. If not, it returns
841 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
842 this is necessary because there may be backward assertions in the pattern.
843 When matching the same subject several times, it may save resources to use
844 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
846 4. The code for checking the validity of UTF-8 strings has been tightened so
847 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
848 containing "overlong sequences".
850 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
851 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
852 should have been "&", but it just so happened that all the cases this let
853 through by mistake were picked up later in the function.
855 6. I had used a variable called "isblank" - this is a C99 function, causing
856 some compilers to warn. To avoid this, I renamed it (as "blankclass").
858 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
859 prompting; (b) run "./pcretest /dev/null" at the start of the test script
860 so the version is shown; (c) stop "make test" echoing "./RunTest".
862 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
864 9. The prototype for memmove() for systems that don't have it was using
865 size_t, but the inclusion of the header that defines size_t was later. I've
866 moved the #includes for the C headers earlier to avoid this.
868 10. Added some adjustments to the code to make it easier to compiler on certain
869 special systems:
871 (a) Some "const" qualifiers were missing.
872 (b) Added the macro EXPORT before all exported functions; by default this
873 is defined to be empty.
874 (c) Changed the dftables auxiliary program (that builds chartables.c) so
875 that it reads its output file name as an argument instead of writing
876 to the standard output and assuming this can be redirected.
878 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
879 class containing characters with values greater than 255, PCRE compilation
880 went into a loop.
882 12. A recursive reference to a subpattern that was within another subpattern
883 that had a minimum quantifier of zero caused PCRE to crash. For example,
884 (x(y(?2))z)? provoked this bug with a subject that got as far as the
885 recursion. If the recursively-called subpattern itself had a zero repeat,
886 that was OK.
888 13. In pcretest, the buffer for reading a data line was set at 30K, but the
889 buffer into which it was copied (for escape processing) was still set at
890 1024, so long lines caused crashes.
892 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
893 "internal error: code overflow...". This applied to any character class
894 that was followed by a possessive quantifier.
896 15. Modified the Makefile to add libpcre.la as a prerequisite for
897 libpcreposix.la because I was told this is needed for a parallel build to
898 work.
900 16. If a pattern that contained .* following optional items at the start was
901 studied, the wrong optimizing data was generated, leading to matching
902 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
903 matching string must start with a or b or c. The correct conclusion for
904 this pattern is that a match can start with any character.
907 Version 4.4 13-Aug-03
908 ---------------------
910 1. In UTF-8 mode, a character class containing characters with values between
911 127 and 255 was not handled correctly if the compiled pattern was studied.
912 In fixing this, I have also improved the studying algorithm for such
913 classes (slightly).
915 2. Three internal functions had redundant arguments passed to them. Removal
916 might give a very teeny performance improvement.
918 3. Documentation bug: the value of the capture_top field in a callout is *one
919 more than* the number of the hightest numbered captured substring.
921 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
922 in incorrectly linking with a previously installed version. They now link
923 explicitly with libpcre.la.
925 5. configure.in no longer needs to recognize Cygwin specially.
927 6. A problem in pcre.in for Windows platforms is fixed.
929 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
930 pcretest, it used to include the size of the study block as part of its
931 output. Unfortunately, the structure contains a field that has a different
932 size on different hardware architectures. This meant that the tests that
933 showed this size failed. As the block is currently always of a fixed size,
934 this information isn't actually particularly useful in pcretest output, so
935 I have just removed it.
937 8. Three pre-processor statements accidentally did not start in column 1.
938 Sadly, there are *still* compilers around that complain, even though
939 standard C has not required this for well over a decade. Sigh.
941 9. In pcretest, the code for checking callouts passed small integers in the
942 callout_data field, which is a void * field. However, some picky compilers
943 complained about the casts involved for this on 64-bit systems. Now
944 pcretest passes the address of the small integer instead, which should get
945 rid of the warnings.
947 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
948 both compile and run time, and gives an error if an invalid UTF-8 sequence
949 is found. There is a option for disabling this check in cases where the
950 string is known to be correct and/or the maximum performance is wanted.
952 11. In response to a bug report, I changed one line in Makefile.in from
954 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
955 to
956 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
958 to look similar to other lines, but I have no way of telling whether this
959 is the right thing to do, as I do not use Windows. No doubt I'll get told
960 if it's wrong...
963 Version 4.3 21-May-03
964 ---------------------
966 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
967 Makefile.
969 2. Some refactoring to improve the quality of the code:
971 (i) The utf8_table... variables are now declared "const".
973 (ii) The code for \cx, which used the "case flipping" table to upper case
974 lower case letters, now just substracts 32. This is ASCII-specific,
975 but the whole concept of \cx is ASCII-specific, so it seems
976 reasonable.
978 (iii) PCRE was using its character types table to recognize decimal and
979 hexadecimal digits in the pattern. This is silly, because it handles
980 only 0-9, a-f, and A-F, but the character types table is locale-
981 specific, which means strange things might happen. A private
982 table is now used for this - though it costs 256 bytes, a table is
983 much faster than multiple explicit tests. Of course, the standard
984 character types table is still used for matching digits in subject
985 strings against \d.
987 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
988 ending in _t are). So I've renamed it as ESC_tee.
990 3. The first argument for regexec() in the POSIX wrapper should have been
991 defined as "const".
993 4. Changed pcretest to use malloc() for its buffers so that they can be
994 Electric Fenced for debugging.
996 5. There were several places in the code where, in UTF-8 mode, PCRE would try
997 to read one or more bytes before the start of the subject string. Often this
998 had no effect on PCRE's behaviour, but in some circumstances it could
999 provoke a segmentation fault.
1001 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1002 to try to read one or more bytes before the start of the subject string.
1004 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1005 UTF-8 support could misbehave in various ways if the subject string
1006 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1007 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1008 back over UTF-8 characters.)
1011 Version 4.2 14-Apr-03
1012 ---------------------
1014 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1016 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1017 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1018 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1019 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1020 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1022 Note: automatic setting of the BUILD variables is not yet working
1023 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1024 compile-time but not at link-time
1025 [LINK]: use for linking executables only
1026 make different versions for Windows and non-Windows
1027 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1028 libraries
1029 [LINK_FOR_BUILD]: new variable
1030 [OBJEXT]: use throughout
1031 [EXEEXT]: use throughout
1032 <winshared>: new target
1033 <wininstall>: new target
1034 <dftables.o>: use native compiler
1035 <dftables>: use native linker
1036 <install>: handle Windows platform correctly
1037 <clean>: ditto
1038 <check>: ditto
1039 copy DLL to top builddir before testing
1041 As part of these changes, -no-undefined was removed again. This was reported
1042 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1043 in any case.
1045 3. Some tidies to get rid of compiler warnings:
1047 . In the match_data structure, match_limit was an unsigned long int, whereas
1048 match_call_count was an int. I've made them both unsigned long ints.
1050 . In pcretest the fact that a const uschar * doesn't automatically cast to
1051 a void * provoked a warning.
1053 . Turning on some more compiler warnings threw up some "shadow" variables
1054 and a few more missing casts.
1056 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1057 option, a class that contained a single character with a value between 128
1058 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1060 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1061 option, a class that contained several characters, but with at least one
1062 whose value was between 128 and 255 caused PCRE to crash.
1065 Version 4.1 12-Mar-03
1066 ---------------------
1068 1. Compiling with gcc -pedantic found a couple of places where casts were
1069 needed, and a string in dftables.c that was longer than standard compilers are
1070 required to support.
1072 2. Compiling with Sun's compiler found a few more places where the code could
1073 be tidied up in order to avoid warnings.
1075 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1076 first of these names is deprecated in the latest Autoconf in favour of the name
1077 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1078 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1079 analogy I have changed it to CFLAGS_FOR_BUILD.
1081 4. Added -no-undefined to the linking command in the Makefile, because this is
1082 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1083 linking step for the pcreposix library.
1085 5. PCRE was failing to diagnose the case of two named groups with the same
1086 name.
1088 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1089 literal character that is needed in the subject for a match, and scans along to
1090 ensure that it is present before embarking on the full matching process. This
1091 saves time in cases of nested unlimited repeats that are never going to match.
1092 Problem: the scan can take a lot of time if the subject is very long (e.g.
1093 megabytes), thus penalizing straightforward matches. It is now done only if the
1094 amount of subject to be scanned is less than 1000 bytes.
1096 7. A lesser problem with the same optimization is that it was recording the
1097 first character of an anchored pattern as "needed", thus provoking a search
1098 right along the subject, even when the first match of the pattern was going to
1099 fail. The "needed" character is now not set for anchored patterns, unless it
1100 follows something in the pattern that is of non-fixed length. Thus, it still
1101 fulfils its original purpose of finding quick non-matches in cases of nested
1102 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1105 Version 4.0 17-Feb-03
1106 ---------------------
1108 1. If a comment in an extended regex that started immediately after a meta-item
1109 extended to the end of string, PCRE compiled incorrect data. This could lead to
1110 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1112 2. Moved to autoconf 2.53 and libtool 1.4.2.
1114 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1115 the special perltest8 script is no longer needed - all the tests can be run
1116 from a single perltest script.
1118 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1119 by \s. It has now been removed in PCRE. This means it isn't recognized as
1120 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1121 class [:space:] *does* include VT, thereby creating a mess.
1123 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1124 space and tab.
1126 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1127 its new features into the main test script, reducing the number of scripts.
1129 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1130 were backward compatible, and made the (?i) apply to the whole pattern, as if
1131 /i were given. Now it behaves more logically, and applies the option setting
1132 only to what follows. PCRE has been changed to follow suit. However, if it
1133 finds options settings right at the start of the pattern, it extracts them into
1134 the global options, as before. Thus, they show up in the info data.
1136 8. Added support for the \Q...\E escape sequence. Characters in between are
1137 treated as literals. This is slightly different from Perl in that $ and @ are
1138 also handled as literals inside the quotes. In Perl, they will cause variable
1139 interpolation. Note the following examples:
1141 Pattern PCRE matches Perl matches
1143 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1144 \Qabc\$xyz\E abc\$xyz abc\$xyz
1145 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1147 For compatibility with Perl, \Q...\E sequences are recognized inside character
1148 classes as well as outside them.
1150 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1151 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1152 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1153 signed/unsigned warnings.
1155 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1156 option for pcretest, so I've replaced it by a simple function that does just
1157 that job.
1159 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1160 "pcregrep -".
1162 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1163 Java package. This provides some syntactic sugar for simple cases of what my
1164 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1165 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1166 item, you can use this simplified notation. Note that only makes sense with
1167 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1168 greediness, whatever the setting of the PCRE_UNGREEDY option.
1170 13. A change of greediness default within a pattern was not taking effect at
1171 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1172 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1173 was abstracted outside.
1175 14. PCRE now supports the \G assertion. It is true when the current matching
1176 position is at the start point of the match. This differs from \A when the
1177 starting offset is non-zero. Used with the /g option of pcretest (or similar
1178 code), it works in the same way as it does for Perl's /g option. If all
1179 alternatives of a regex begin with \G, the expression is anchored to the start
1180 match position, and the "anchored" flag is set in the compiled expression.
1182 15. Some bugs concerning the handling of certain option changes within patterns
1183 have been fixed. These applied to options other than (?ims). For example,
1184 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1185 been the other way round. Some of this was related to change 7 above.
1187 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1188 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1189 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1190 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1192 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1193 mode. Unlike ".", it always matches newline, whatever the setting of
1194 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1195 assertions. Perl allows it, but it doesn't (in general) work because it can't
1196 calculate the length of the lookbehind. At least, that's the case for Perl
1197 5.8.0 - I've been told they are going to document that it doesn't work in
1198 future.
1200 18. Added an error diagnosis for escapes that PCRE does not support: these are
1201 \L, \l, \N, \P, \p, \U, \u, and \X.
1203 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1204 reading past the end of the pattern in cases such as /[abcd/.
1206 20. PCRE was getting more memory than necessary for patterns with classes that
1207 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1209 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1210 compiling PCRE for use with Virtual Pascal.
1212 22. Small fix to the Makefile to make it work properly if the build is done
1213 outside the source tree.
1215 23. Added a new extension: a condition to go with recursion. If a conditional
1216 subpattern starts with (?(R) the "true" branch is used if recursion has
1217 happened, whereas the "false" branch is used only at the top level.
1219 24. When there was a very long string of literal characters (over 255 bytes
1220 without UTF support, over 250 bytes with UTF support), the computation of how
1221 much memory was required could be incorrect, leading to segfaults or other
1222 strange effects.
1224 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1225 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1226 there was a subsequent back reference to those brackets. This meant that, for
1227 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1228 possible to check for precisely this case. All we can do is abandon the
1229 optimization if .* occurs inside capturing brackets when there are any back
1230 references whatsoever. (See below for a better fix that came later.)
1232 26. The handling of the optimization for finding the first character of a
1233 non-anchored pattern, and for finding a character that is required later in the
1234 match were failing in some cases. This didn't break the matching; it just
1235 failed to optimize when it could. The way this is done has been re-implemented.
1237 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1239 28. Added a new feature that provides some of the functionality that Perl
1240 provides with (?{...}). The facility is termed a "callout". The way it is done
1241 in PCRE is for the caller to provide an optional function, by setting
1242 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1243 global variable. By default it is unset, which disables all calling out. To get
1244 the function called, the regex must include (?C) at appropriate points. This
1245 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1246 This provides a means of identifying different callout points. When PCRE
1247 reaches such a point in the regex, if pcre_callout has been set, the external
1248 function is called. It is provided with data in a structure called
1249 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1250 matching continues; if it returns a non-zero value, the match at the current
1251 point fails. However, backtracking will occur if possible. [This was changed
1252 later and other features added - see item 49 below.]
1254 29. pcretest is upgraded to test the callout functionality. It provides a
1255 callout function that displays information. By default, it shows the start of
1256 the match and the current position in the text. There are some new data escapes
1257 to vary what happens:
1259 \C+ in addition, show current contents of captured substrings
1260 \C- do not supply a callout function
1261 \C!n return 1 when callout number n is reached
1262 \C!n!m return 1 when callout number n is reached for the mth time
1264 30. If pcregrep was called with the -l option and just a single file name, it
1265 output "<stdin>" if a match was found, instead of the file name.
1267 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1268 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1269 pcre_exec(). This saves a malloc/free per call. The default value of
1270 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1271 when configuring.
1273 32. The default maximum size of a compiled pattern is 64K. There have been a
1274 few cases of people hitting this limit. The code now uses macros to handle the
1275 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1276 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1277 configuring. Tests 2 and 5 work only with 2-byte links because they output
1278 debugging information about compiled patterns.
1280 33. Internal code re-arrangements:
1282 (a) Moved the debugging function for printing out a compiled regex into
1283 its own source file (printint.c) and used #include to pull it into
1284 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1285 separate copies.
1287 (b) Defined the list of op-code names for debugging as a macro in
1288 internal.h so that it is next to the definition of the opcodes.
1290 (c) Defined a table of op-code lengths for simpler skipping along compiled
1291 code. This is again a macro in internal.h so that it is next to the
1292 definition of the opcodes.
1294 34. Added support for recursive calls to individual subpatterns, along the
1295 lines of Robin Houston's patch (but implemented somewhat differently).
1297 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1298 allow it to read and process whole directories in Win32. This code was
1299 contributed by Lionel Fourquaux; it has not been tested by me.
1301 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1302 used to name a group. Names consist of alphanumerics and underscores, and must
1303 be unique. Back references use the syntax (?P=name) and recursive calls use
1304 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1305 numbers. The function pcre_fullinfo() can be used after compilation to extract
1306 a name/number map. There are three relevant calls:
1308 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1309 PCRE_INFO_NAMECOUNT yields the number of entries
1310 PCRE_INFO_NAMETABLE yields a pointer to the map.
1312 The map is a vector of fixed-size entries. The size of each entry depends on
1313 the length of the longest name used. The first two bytes of each entry are the
1314 group number, most significant byte first. There follows the corresponding
1315 name, zero terminated. The names are in alphabetical order.
1317 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1318 case instead of 255. Making it the same both with and without UTF-8 support
1319 means that the same test output works with both.
1321 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1322 calling malloc() with a zero argument.
1324 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1325 optimization. I've improved things by keeping a bitmap of backreferences with
1326 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1327 fact referenced, the optimization can be applied. It is unlikely that a
1328 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1329 the match to follow \n) will appear inside brackets with a number greater than
1330 31, but if it does, any back reference > 31 suppresses the optimization.
1332 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1333 of disabling numbered capturing parentheses. Any opening parenthesis that is
1334 not followed by ? behaves as if it were followed by ?: but named parentheses
1335 can still be used for capturing (and they will acquire numbers in the usual
1336 way).
1338 41. Redesigned the return codes from the match() function into yes/no/error so
1339 that errors can be passed back from deep inside the nested calls. A malloc
1340 failure while inside a recursive subpattern call now causes the
1341 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1343 42. It is now possible to set a limit on the number of times the match()
1344 function is called in a call to pcre_exec(). This facility makes it possible to
1345 limit the amount of recursion and backtracking, though not in a directly
1346 obvious way, because the match() function is used in a number of different
1347 circumstances. The count starts from zero for each position in the subject
1348 string (for non-anchored patterns). The default limit is, for compatibility, a
1349 large number, namely 10 000 000. You can change this in two ways:
1351 (a) When configuring PCRE before making, you can use --with-match-limit=n
1352 to set a default value for the compiled library.
1354 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1355 a different value is set. See 45 below.
1357 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1359 43. Added a new function pcre_config(int, void *) to enable run-time extraction
1360 of things that can be changed at compile time. The first argument specifies
1361 what is wanted and the second points to where the information is to be placed.
1362 The current list of available information is:
1366 The output is an integer that is set to one if UTF-8 support is available;
1367 otherwise it is set to zero.
1371 The output is an integer that it set to the value of the code that is used for
1372 newline. It is either LF (10) or CR (13).
1376 The output is an integer that contains the number of bytes used for internal
1377 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1381 The output is an integer that contains the threshold above which the POSIX
1382 interface uses malloc() for output vectors. See item 31 above.
1386 The output is an unsigned integer that contains the default limit of the number
1387 of match() calls in a pcre_exec() execution. See 42 above.
1389 44. pcretest has been upgraded by the addition of the -C option. This causes it
1390 to extract all the available output from the new pcre_config() function, and to
1391 output it. The program then exits immediately.
1393 45. A need has arisen to pass over additional data with calls to pcre_exec() in
1394 order to support additional features. One way would have been to define
1395 pcre_exec2() (for example) with extra arguments, but this would not have been
1396 extensible, and would also have required all calls to the original function to
1397 be mapped to the new one. Instead, I have chosen to extend the mechanism that
1398 is used for passing in "extra" data from pcre_study().
1400 The pcre_extra structure is now exposed and defined in pcre.h. It currently
1401 contains the following fields:
1403 flags a bitmap indicating which of the following fields are set
1404 study_data opaque data from pcre_study()
1405 match_limit a way of specifying a limit on match() calls for a specific
1406 call to pcre_exec()
1407 callout_data data for callouts (see 49 below)
1409 The flag bits are also defined in pcre.h, and are
1415 The pcre_study() function now returns one of these new pcre_extra blocks, with
1416 the actual study data pointed to by the study_data field, and the
1417 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1418 before. That is, this change is entirely upwards-compatible and requires no
1419 change to existing code.
1421 If you want to pass in additional data to pcre_exec(), you can either place it
1422 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1423 block.
1425 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1426 data string contains the escape sequence \M, pcretest calls pcre_exec() several
1427 times with different match limits, until it finds the minimum value needed for
1428 pcre_exec() to complete. The value is then output. This can be instructive; for
1429 most simple matches the number is quite small, but for pathological cases it
1430 gets very large very quickly.
1432 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1433 returns the size of the data block pointed to by the study_data field in a
1434 pcre_extra block, that is, the value that was passed as the argument to
1435 pcre_malloc() when PCRE was getting memory in which to place the information
1436 created by pcre_study(). The fourth argument should point to a size_t variable.
1437 pcretest has been extended so that this information is shown after a successful
1438 pcre_study() call when information about the compiled regex is being displayed.
1440 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1441 because what follows is always an absolute path. (Later: it turns out that this
1442 is more than cosmetic for MinGW, because it doesn't like empty path
1443 components.)
1445 49. Some changes have been made to the callout feature (see 28 above):
1447 (i) A callout function now has three choices for what it returns:
1449 0 => success, carry on matching
1450 > 0 => failure at this point, but backtrack if possible
1451 < 0 => serious error, return this value from pcre_exec()
1453 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1454 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1455 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1456 use by callout functions. It will never be used by PCRE itself.
1458 (ii) The pcre_extra structure (see 45 above) has a void * field called
1459 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1460 pcre_callout_block structure has a field of the same name. The contents of
1461 the field passed in the pcre_extra structure are passed to the callout
1462 function in the corresponding field in the callout block. This makes it
1463 easier to use the same callout-containing regex from multiple threads. For
1464 testing, the pcretest program has a new data escape
1466 \C*n pass the number n (may be negative) as callout_data
1468 If the callout function in pcretest receives a non-zero value as
1469 callout_data, it returns that value.
1471 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1472 there were some redundant $(CFLAGS) in commands that are now specified as
1473 $(LINK), which already includes $(CFLAGS).
1475 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1476 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1477 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1478 one-byte characters throughout. Note that case-insensitive matching applies
1479 only to characters whose values are less than 256. PCRE doesn't support the
1480 notion of cases for higher-valued characters.
1482 (i) A character class whose characters are all within 0-255 is handled as
1483 a bit map, and the map is inverted for negative classes. Previously, a
1484 character > 255 always failed to match such a class; however it should
1485 match if the class was a negative one (e.g. [^ab]). This has been fixed.
1487 (ii) A negated character class with a single character < 255 is coded as
1488 "not this character" (OP_NOT). This wasn't working properly when the test
1489 character was multibyte, either singly or repeated.
1491 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1492 mode, for example: \x{100}{2,3}.
1494 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1495 singly or repeated) now correctly test multibyte characters. However,
1496 PCRE doesn't recognize any characters with values greater than 255 as
1497 digits, spaces, or word characters. Such characters always match \D, \S,
1498 and \W, and never match \d, \s, or \w.
1500 (v) Classes may now contain characters and character ranges with values
1501 greater than 255. For example: [ab\x{100}-\x{400}].
1503 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1504 PCRE in UTF-8 mode.
1506 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1507 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1508 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1509 value.)
1511 53. The single man page has become too large. I have therefore split it up into
1512 a number of separate man pages. These also give rise to individual HTML pages;
1513 these are now put in a separate directory, and there is an index.html page that
1514 lists them all. Some hyperlinking between the pages has been installed.
1516 54. Added convenience functions for handling named capturing parentheses.
1518 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1519 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1520 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1521 are faulted.
1523 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1524 calling configure. These values are used when compiling the dftables.c program
1525 which is run to generate the source of the default character tables. They
1526 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1527 you will need to set these values.
1529 57. Updated the building process for Windows DLL, as provided by Fred Cox.
1532 Version 3.9 02-Jan-02
1533 ---------------------
1535 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1537 2. If --disable-static was given, the building process failed when trying to
1538 build pcretest and pcregrep. (For some reason it was using libtool to compile
1539 them, which is not right, as they aren't part of the library.)
1542 Version 3.8 18-Dec-01
1543 ---------------------
1545 1. The experimental UTF-8 code was completely screwed up. It was packing the
1546 bytes in the wrong order. How dumb can you get?
1549 Version 3.7 29-Oct-01
1550 ---------------------
1552 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1553 This caused pcretest, when used on the test data, to segfault. Unfortunately,
1554 this didn't happen under Solaris 8, where I normally test things.
1556 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1557 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1558 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1559 this fix an hour or so after the initial 3.7 release.)
1562 Version 3.6 23-Oct-01
1563 ---------------------
1565 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1566 offsets passed as NULL with zero offset count.
1568 2. The config.guess and config.sub files had not been updated when I moved to
1569 the latest autoconf.
1572 Version 3.5 15-Aug-01
1573 ---------------------
1575 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1576 had been forgotten.
1578 2. By using declared but undefined structures, we can avoid using "void"
1579 definitions in pcre.h while keeping the internal definitions of the structures
1580 private.
1582 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1583 user point of view, this means that both static and shared libraries are built
1584 by default, but this can be individually controlled. More of the work of
1585 handling this static/shared cases is now inside libtool instead of PCRE's make
1586 file.
1588 4. The pcretest utility is now installed along with pcregrep because it is
1589 useful for users (to test regexs) and by doing this, it automatically gets
1590 relinked by libtool. The documentation has been turned into a man page, so
1591 there are now .1, .txt, and .html versions in /doc.
1593 5. Upgrades to pcregrep:
1594 (i) Added long-form option names like gnu grep.
1595 (ii) Added --help to list all options with an explanatory phrase.
1596 (iii) Added -r, --recursive to recurse into sub-directories.
1597 (iv) Added -f, --file to read patterns from a file.
1599 6. pcre_exec() was referring to its "code" argument before testing that
1600 argument for NULL (and giving an error if it was NULL).
1602 7. Upgraded Makefile.in to allow for compiling in a different directory from
1603 the source directory.
1605 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1606 options bits, the pointer it was passed was to an int instead of to an unsigned
1607 long int. This mattered only on 64-bit systems.
1609 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1610 generated) instead of pcre.in, which it its source. Also made the same change
1611 in several of the .c files.
1613 10. A new release of gcc defines printf() as a macro, which broke pcretest
1614 because it had an ifdef in the middle of a string argument for printf(). Fixed
1615 by using separate calls to printf().
1617 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1618 script, to force use of CR or LF instead of \n in the source. On non-Unix
1619 systems, the value can be set in config.h.
1621 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1622 absolute limit. Changed the text of the error message to make this clear, and
1623 likewise updated the man page.
1625 13. The limit of 99 on the number of capturing subpatterns has been removed.
1626 The new limit is 65535, which I hope will not be a "real" limit.
1629 Version 3.4 22-Aug-00
1630 ---------------------
1632 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1634 2. Diagnose condition (?(0) as an error instead of crashing on matching.
1637 Version 3.3 01-Aug-00
1638 ---------------------
1640 1. If an octal character was given, but the value was greater than \377, it
1641 was not getting masked to the least significant bits, as documented. This could
1642 lead to crashes in some systems.
1644 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1645 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1647 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1648 These just pass their arguments on to (pcre_free)(), but they are provided
1649 because some uses of PCRE bind it to non-C systems that can call its functions,
1650 but cannot call free() or pcre_free() directly.
1652 4. Add "make test" as a synonym for "make check". Corrected some comments in
1653 the Makefile.
1655 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1656 Makefile.
1658 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1659 command called pgrep for grepping around the active processes.
1661 7. Added the beginnings of support for UTF-8 character strings.
1663 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1664 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1665 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1666 out for the ar command.)
1669 Version 3.2 12-May-00
1670 ---------------------
1672 This is purely a bug fixing release.
1674 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1675 of ZA. This was just one example of several cases that could provoke this bug,
1676 which was introduced by change 9 of version 2.00. The code for breaking
1677 infinite loops after an iteration that matches an empty string was't working
1678 correctly.
1680 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1681 when matched against abbab (for example). After matching an empty string, it
1682 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1683 caused it to match further down the string than it should.
1685 3. The code contained an inclusion of sys/types.h. It isn't clear why this
1686 was there because it doesn't seem to be needed, and it causes trouble on some
1687 systems, as it is not a Standard C header. It has been removed.
1689 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1690 were reported on the Macintosh. The changes were from
1692 while ((c = *(++ptr)) != 0 && c != '\n');
1693 to
1694 while ((c = *(++ptr)) != 0 && c != '\n') ;
1696 Totally extraordinary, but if that's what it takes...
1698 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1699 available. Added HAVE_BCOPY and an autoconf test for it; if neither
1700 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1701 assumes the way PCRE uses memmove() (always moving upwards).
1703 6. PCRE is being used in one environment where strchr() is not available. There
1704 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1705 faster code anyway.
1708 Version 3.1 09-Feb-00
1709 ---------------------
1711 The only change in this release is the fixing of some bugs in Makefile.in for
1712 the "install" target:
1714 (1) It was failing to install pcreposix.h.
1716 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
1719 Version 3.0 01-Feb-00
1720 ---------------------
1722 1. Add support for the /+ modifier to perltest (to output $` like it does in
1723 pcretest).
1725 2. Add support for the /g modifier to perltest.
1727 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
1728 matches null strings.
1730 4. Fix perltest so that it doesn't do unwanted things when fed an empty
1731 pattern. Perl treats empty patterns specially - it reuses the most recent
1732 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
1733 effect.
1735 5. The POSIX interface was broken in that it was just handing over the POSIX
1736 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
1737 required a bigger vector, with some working space on the end. This means that
1738 the POSIX wrapper now has to get and free some memory, and copy the results.
1740 6. Added some simple autoconf support, placing the test data and the
1741 documentation in separate directories, re-organizing some of the
1742 information files, and making it build pcre-config (a GNU standard). Also added
1743 libtool support for building PCRE as a shared library, which is now the
1744 default.
1746 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
1747 09 are not valid octal constants. Single digits will be used for minor values
1748 less than 10.
1750 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
1751 existing programs that set these in the POSIX interface can use PCRE without
1752 modification.
1754 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
1755 return all that pcre_info() returns, plus additional data. The pcre_info()
1756 function is retained for compatibility, but is considered to be obsolete.
1758 10. Added experimental recursion feature (?R) to handle one common case that
1759 Perl 5.6 will be able to do with (?p{...}).
1761 11. Added support for POSIX character classes like [:alpha:], which Perl is
1762 adopting.
1765 Version 2.08 31-Aug-99
1766 ----------------------
1768 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
1769 trying to match at the startoffset position, but instead was moving forward to
1770 the next newline as if a previous match had failed.
1772 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
1773 and could get into a loop if a null string was matched other than at the start
1774 of the subject.
1776 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
1777 be distinguished at compile time, and for completeness also added PCRE_DATE.
1779 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
1780 in GnuWin32 environments.
1783 Version 2.07 29-Jul-99
1784 ----------------------
1786 1. The documentation is now supplied in plain text form and HTML as well as in
1787 the form of man page sources.
1789 2. C++ compilers don't like assigning (void *) values to other pointer types.
1790 In particular this affects malloc(). Although there is no problem in Standard
1791 C, I've put in casts to keep C++ compilers happy.
1793 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
1794 should be (const char *).
1796 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
1797 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
1798 However, I haven't made this a standard facility. The documentation doesn't
1799 mention it, and the Makefile doesn't support it.
1801 5. The Makefile now contains an "install" target, with editable destinations at
1802 the top of the file. The pcretest program is not installed.
1804 6. pgrep -V now gives the PCRE version number and date.
1806 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1807 causing the entire string to be ignored, instead of just the last character.
1809 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1810 non-matching string, it can take a very, very long time, even for strings of
1811 quite modest length, because of the nested recursion. PCRE now does better in
1812 some of these cases. It does this by remembering the last required literal
1813 character in the pattern, and pre-searching the subject to ensure it is present
1814 before running the real match. In other words, it applies a heuristic to detect
1815 some types of certain failure quickly, and in the above example, if presented
1816 with a string that has no trailing " it gives "no match" very quickly.
1818 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1819 other alternatives are tried instead.
1822 Version 2.06 09-Jun-99
1823 ----------------------
1825 1. Change pcretest's output for amount of store used to show just the code
1826 space, because the remainder (the data block) varies in size between 32-bit and
1827 64-bit systems.
1829 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1830 start matching at. This allows lookbehinds to work when searching for multiple
1831 occurrences in a string.
1833 3. Added additional options to pcretest for testing multiple occurrences:
1835 /+ outputs the rest of the string that follows a match
1836 /g loops for multiple occurrences, using the new startoffset argument
1837 /G loops for multiple occurrences by passing an incremented pointer
1839 4. PCRE wasn't doing the "first character" optimization for patterns starting
1840 with \b or \B, though it was doing it for other lookbehind assertions. That is,
1841 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1842 the letter 'x'. On long subject strings, this gives a significant speed-up.
1845 Version 2.05 21-Apr-99
1846 ----------------------
1848 1. Changed the type of magic_number from int to long int so that it works
1849 properly on 16-bit systems.
1851 2. Fixed a bug which caused patterns starting with .* not to work correctly
1852 when the subject string contained newline characters. PCRE was assuming
1853 anchoring for such patterns in all cases, which is not correct because .* will
1854 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1855 DOTALL is set at top level; otherwise it knows that patterns starting with .*
1856 must be retried after every newline in the subject.
1859 Version 2.04 18-Feb-99
1860 ----------------------
1862 1. For parenthesized subpatterns with repeats whose minimum was zero, the
1863 computation of the store needed to hold the pattern was incorrect (too large).
1864 If such patterns were nested a few deep, this could multiply and become a real
1865 problem.
1867 2. Added /M option to pcretest to show the memory requirement of a specific
1868 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1870 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1871 compiled in such a way that the backtracking after subsequent failure was
1872 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1873 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1876 Version 2.03 02-Feb-99
1877 ----------------------
1879 1. Fixed typo and small mistake in man page.
1881 2. Added 4th condition (GPL supersedes if conflict) and created separate
1882 LICENCE file containing the conditions.
1884 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1885 Perl, that is the internal \ allows the delimiter to be included in the
1886 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1887 the final delimiter, add \ to the end of the pattern (to test the error).
1889 4. Added the convenience functions for extracting substrings after a successful
1890 match. Updated pcretest to make it able to test these functions.
1893 Version 2.02 14-Jan-99
1894 ----------------------
1896 1. Initialized the working variables associated with each extraction so that
1897 their saving and restoring doesn't refer to uninitialized store.
1899 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1900 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1901 fix the problem.
1903 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1904 calls, and wasn't printing the correct value for compiling calls. Increased the
1905 default value of LOOPREPEAT, and the number of significant figures in the
1906 times.
1908 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1910 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1911 a building problem on Windows NT with a FAT file system.
1914 Version 2.01 21-Oct-98
1915 ----------------------
1917 1. Changed the API for pcre_compile() to allow for the provision of a pointer
1918 to character tables built by pcre_maketables() in the current locale. If NULL
1919 is passed, the default tables are used.
1922 Version 2.00 24-Sep-98
1923 ----------------------
1925 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1926 it any more.
1928 2. Allow quantification of (?>) groups, and make it work correctly.
1930 3. The first character computation wasn't working for (?>) groups.
1932 4. Correct the implementation of \Z (it is permitted to match on the \n at the
1933 end of the subject) and add 5.005's \z, which really does match only at the
1934 very end of the subject.
1936 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1938 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1939 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1940 localized options. All options to pcre_study() were also removed.
1942 7. Add other new features from 5.005:
1944 $(?<= positive lookbehind
1945 $(?<! negative lookbehind
1946 (?imsx-imsx) added the unsetting capability
1947 such a setting is global if at outer level; local otherwise
1948 (?imsx-imsx:) non-capturing groups with option setting
1949 (?(cond)re|re) conditional pattern matching
1951 A backreference to itself in a repeated group matches the previous
1952 captured string.
1954 8. General tidying up of studying (both automatic and via "study")
1955 consequential on the addition of new assertions.
1957 9. As in 5.005, unlimited repeated groups that could match an empty substring
1958 are no longer faulted at compile time. Instead, the loop is forcibly broken at
1959 runtime if any iteration does actually match an empty substring.
1961 10. Include the RunTest script in the distribution.
1963 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1964 discrepancies, some of which were old and were also with respect to 5.004. They
1965 have now been fixed.
1968 Version 1.09 28-Apr-98
1969 ----------------------
1971 1. A negated single character class followed by a quantifier with a minimum
1972 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
1973 program crashes, or just wrong answers. This did not apply to negated classes
1974 containing more than one character, or to minima other than one.
1977 Version 1.08 27-Mar-98
1978 ----------------------
1980 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
1982 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
1983 latter must appear before anything that relies on it in the pattern.
1986 Version 1.07 16-Feb-98
1987 ----------------------
1989 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
1990 repeat of a potentially empty string).
1993 Version 1.06 23-Jan-98
1994 ----------------------
1996 1. Added Markus Oberhumer's little patches for C++.
1998 2. Literal strings longer than 255 characters were broken.
2001 Version 1.05 23-Dec-97
2002 ----------------------
2004 1. Negated character classes containing more than one character were failing if
2005 PCRE_CASELESS was set at run time.
2008 Version 1.04 19-Dec-97
2009 ----------------------
2011 1. Corrected the man page, where some "const" qualifiers had been omitted.
2013 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2014 input syntax.
2016 3. Fixed memory leak which occurred when a regex with back references was
2017 matched with an offsets vector that wasn't big enough. The temporary memory
2018 that is used in this case wasn't being freed if the match failed.
2020 4. Tidied pcretest to ensure it frees memory that it gets.
2022 5. Temporary memory was being obtained in the case where the passed offsets
2023 vector was exactly big enough.
2025 6. Corrected definition of offsetof() from change 5 below.
2027 7. I had screwed up change 6 below and broken the rules for the use of
2028 setjmp(). Now fixed.
2031 Version 1.03 18-Dec-97
2032 ----------------------
2034 1. A erroneous regex with a missing opening parenthesis was correctly
2035 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2036 on some systems.
2038 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2039 it was reported that one broken compiler failed on the former because "code" is
2040 also an independent variable.
2042 3. The erroneous regex a[]b caused an array overrun reference.
2044 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2045 fail on data ending with that character. (It was going on too far, and checking
2046 the next character, typically a binary zero.) This was specific to the
2047 optimized code for single-character negative classes.
2049 5. Added a contributed patch from the TIN world which does the following:
2051 + Add an undef for memmove, in case the the system defines a macro for it.
2053 + Add a definition of offsetof(), in case there isn't one. (I don't know
2054 the reason behind this - offsetof() is part of the ANSI standard - but
2055 it does no harm).
2057 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2058 most of the places where whitespace preceded '#'. I have given up and
2059 allowed the remaining 2 cases to be at the margin.
2061 + Rename some variables in pcre to eliminate shadowing. This seems very
2062 pedantic, but does no harm, of course.
2064 6. Moved the call to setjmp() into its own function, to get rid of warnings
2065 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2067 7. Constructs such as \d{8,} were compiling into the equivalent of
2068 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2069 outcome, but in this particular case used more store than had been allocated,
2070 which caused the bug to be discovered because it threw up an internal error.
2072 8. The debugging code in both pcre and pcretest for outputting the compiled
2073 form of a regex was going wrong in the case of back references followed by
2074 curly-bracketed repeats.
2077 Version 1.02 12-Dec-97
2078 ----------------------
2080 1. Typos in pcre.3 and comments in the source fixed.
2082 2. Applied a contributed patch to get rid of places where it used to remove
2083 'const' from variables, and fixed some signed/unsigned and uninitialized
2084 variable warnings.
2086 3. Added the "runtest" target to Makefile.
2088 4. Set default compiler flag to -O2 rather than just -O.
2091 Version 1.01 19-Nov-97
2092 ----------------------
2094 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2095 like /([ab]*)*/, that is, for classes with more than one character in them.
2097 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2098 as /((?>a*))*/ (a PCRE_EXTRA facility).
2101 Version 1.00 18-Nov-97
2102 ----------------------
2104 1. Added compile-time macros to support systems such as SunOS4 which don't have
2105 memmove() or strerror() but have other things that can be used instead.
2107 2. Arranged that "make clean" removes the executables.
2110 Version 0.99 27-Oct-97
2111 ----------------------
2113 1. Fixed bug in code for optimizing classes with only one character. It was
2114 initializing a 32-byte map regardless, which could cause it to run off the end
2115 of the memory it had got.
2117 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2120 Version 0.98 22-Oct-97
2121 ----------------------
2123 1. Fixed bug in code for handling temporary memory usage when there are more
2124 back references than supplied space in the ovector. This could cause segfaults.
2127 Version 0.97 21-Oct-97
2128 ----------------------
2130 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2132 2. Optimized negated single characters not to use a bit map.
2134 3. Brought error texts together as macro definitions; clarified some of them;
2135 fixed one that was wrong - it said "range out of order" when it meant "invalid
2136 escape sequence".
2138 4. Changed some char * arguments to const char *.
2140 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2142 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2143 pcretest.
2146 Version 0.96 16-Oct-97
2147 ----------------------
2149 1. Added a simple "pgrep" utility to the distribution.
2151 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2152 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2153 where "ddd" means "one or more decimal digits".
2155 3. Fixed serious bug. If a pattern had a back reference, but the call to
2156 pcre_exec() didn't supply a large enough ovector to record the related
2157 identifying subpattern, the match always failed. PCRE now remembers the number
2158 of the largest back reference, and gets some temporary memory in which to save
2159 the offsets during matching if necessary, in order to ensure that
2160 backreferences always work.
2162 4. Increased the compatibility with Perl in a number of ways:
2164 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2165 to request this handling. The option can be set at compile or exec time.
2167 (b) $ matches before a terminating newline by default; an option
2168 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2169 mode). The option can be set at compile or exec time.
2171 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2172 the same as Perl's. If the decimal number it represents is less than 10
2173 or there aren't that many previous left capturing parentheses, an octal
2174 escape is read. Inside a character class, it's always an octal escape,
2175 even if it is a single digit.
2177 (d) An escaped but undefined alphabetic character is taken as a literal,
2178 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2179 escapes.
2181 (e) {0} is now permitted. (The previous item is removed from the compiled
2182 pattern).
2184 5. Changed all the names of code files so that the basic parts are no longer
2185 than 10 characters, and abolished the teeny "globals.c" file.
2187 6. Changed the handling of character classes; they are now done with a 32-byte
2188 bit map always.
2190 7. Added the -d and /D options to pcretest to make it possible to look at the
2191 internals of compilation without having to recompile pcre.
2194 Version 0.95 23-Sep-97
2195 ----------------------
2197 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2198 \x20 at the start of a run of normal characters. These were being treated as
2199 real characters, instead of the source characters being re-checked.
2202 Version 0.94 18-Sep-97
2203 ----------------------
2205 1. The functions are now thread-safe, with the caveat that the global variables
2206 containing pointers to malloc() and free() or alternative functions are the
2207 same for all threads.
2209 2. Get pcre_study() to generate a bitmap of initial characters for non-
2210 anchored patterns when this is possible, and use it if passed to pcre_exec().
2213 Version 0.93 15-Sep-97
2214 ----------------------
2216 1. /(b)|(:+)/ was computing an incorrect first character.
2218 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2219 but not actually doing anything yet.
2221 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2222 as Perl does (e.g. [-az] or [az-]).
2224 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2225 all possible positions.
2227 5. Split up into different modules to avoid including unneeded functions in a
2228 compiled binary. However, compile and exec are still in one module. The "study"
2229 function is split off.
2231 6. The character tables are now in a separate module whose source is generated
2232 by an auxiliary program - but can then be edited by hand if required. There are
2233 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2234 toupper() in the code.
2236 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2237 make them global. Abolish the function for setting them, as the caller can now
2238 set them directly.
2241 Version 0.92 11-Sep-97
2242 ----------------------
2244 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2245 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2247 2. Caseless matching was not working in character classes if the characters in
2248 the pattern were in upper case.
2250 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2252 4. Make PCRE_ANCHORED public and accept as a compile option.
2254 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2255 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2256 pass them.
2258 6. Give an error if bad option bits passed at compile or run time.
2260 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2261 pcretest to cause it to pass that flag.
2263 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2264 options, and the first character, if set.
2266 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2269 Version 0.91 10-Sep-97
2270 ----------------------
2272 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2273 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2275 2. PCRE was looping on encountering an indefinitely repeated back reference to
2276 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2277 Perl does - treats the match as successful.
2279 ****

  ViewVC Help
Powered by ViewVC 1.1.5