/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1465 - (show annotations)
Fri Mar 14 15:58:19 2014 UTC (5 years, 7 months ago) by ph10
File size: 263940 byte(s)
Final updates for 8.35-RC1.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.35-RC1 14-March-2014
5 ------------------------------
6
7 1. A new flag is set, when property checks are present in an XCLASS.
8 When this flag is not set, PCRE can perform certain optimizations
9 such as studying these XCLASS-es.
10
11 2. The auto-possessification of character sets were improved: a normal
12 and an extended character set can be compared now. Furthermore
13 the JIT compiler optimizes more character set checks.
14
15 3. Got rid of some compiler warnings for potentially uninitialized variables
16 that show up only when compiled with -O2.
17
18 4. A pattern such as (?=ab\K) that uses \K in an assertion can set the start
19 of a match later then the end of the match. The pcretest program was not
20 handling the case sensibly - it was outputting from the start to the next
21 binary zero. It now reports this situation in a message, and outputs the
22 text from the end to the start.
23
24 5. Fast forward search is improved in JIT. Instead of the first three
25 characters, any three characters with fixed position can be searched.
26 Search order: first, last, middle.
27
28 6. Improve character range checks in JIT. Characters are read by an inprecise
29 function now, which returns with an unknown value if the character code is
30 above a certain treshold (e.g: 256). The only limitation is that the value
31 must be bigger than the treshold as well. This function is useful, when
32 the characters above the treshold are handled in the same way.
33
34 7. The macros whose names start with RAWUCHAR are placeholders for a future
35 mode in which only the bottom 21 bits of 32-bit data items are used. To
36 make this more memorable for those maintaining the code, the names have
37 been changed to start with UCHAR21, and an extensive comment has been added
38 to their definition.
39
40 8. Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
41 to the export list in Makefile.am (they were accidentally omitted from the
42 8.34 tarball).
43
44 9. The informational output from pcretest used the phrase "starting byte set"
45 which is inappropriate for the 16-bit and 32-bit libraries. As the output
46 for "first char" and "need char" really means "non-UTF-char", I've changed
47 "byte" to "char", and slightly reworded the output. The documentation about
48 these values has also been (I hope) clarified.
49
50 10. Another JIT related optimization: use table jumps for selecting the correct
51 backtracking path, when more than four alternatives are present inside a
52 bracket.
53
54 11. Empty match is not possible, when the minimum length is greater than zero,
55 and there is no \K in the pattern. JIT should avoid empty match checks in
56 such cases.
57
58 12. In a caseless character class with UCP support, when a character with more
59 than one alternative case was not the first character of a range, not all
60 the alternative cases were added to the class. For example, s and \x{17f}
61 are both alternative cases for S: the class [RST] was handled correctly,
62 but [R-T] was not.
63
64 13. The configure.ac file always checked for pthread support when JIT was
65 enabled. This is not used in Windows, so I have put this test inside a
66 check for the presence of windows.h (which was already tested for).
67
68 14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
69 The algorithm provides a way to skip certain starting offsets, and usually
70 faster than linear prefix searches.
71
72 15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
73 as for 'fr_FR' and 'french'. For some reason, however, it then used the
74 Windows-specific input and output files, which have 'french' screwed in.
75 So this could never have worked. One of the problems with locales is that
76 they aren't always the same. I have now updated RunTest so that it checks
77 the output of the locale test (test 3) against three different output
78 files, and it allows the test to pass if any one of them matches. With luck
79 this should make the test pass on some versions of Solaris where it was
80 failing. Because of the uncertainty, the script did not used to stop if
81 test 3 failed; it now does. If further versions of a French locale ever
82 come to light, they can now easily be added.
83
84 16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
85 there was a message during ./configure, but it did not stop. This now
86 provokes an error. The invalid example in README has been corrected.
87 If a value less than the minimum is given, the minimum value has always
88 been used, but now a warning is given.
89
90 17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
91 was a bug in the test system, which is now fixed. Also, the list of various
92 configurations that are tested for each release did not have one with both
93 16/32 bits and --enable-bar-anycrlf. It now does.
94
95 18. pcretest was missing "-C bsr" for displaying the \R default setting.
96
97 19. Little endian PowerPC systems are supported now by the JIT compiler.
98
99 20. The fast forward newline mechanism could enter to an infinite loop on
100 certain invalid UTF-8 input. Although we don't support these cases
101 this issue can be fixed by a performance optimization.
102
103 21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
104 not take account if existing stack usage. There is now a new global
105 variable called pcre_stack_guard that can be set to point to an external
106 function to check stack availability. It is called at the start of
107 processing every parenthesized group.
108
109 22. A typo in the code meant that in ungreedy mode the max/min qualifier
110 behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
111 not match "ab".
112
113 23. When UTF was disabled, the JIT program reported some incorrect compile
114 errors. These messages are silenced now.
115
116 24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
117 compiler.
118
119
120 Version 8.34 15-December-2013
121 -----------------------------
122
123 1. Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
124 executable memory. Patch inspired by Carsten Klein.
125
126 2. ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
127 this macro is never tested and has no effect, because the work to support
128 coverage involves only compiling and linking options and special targets in
129 the Makefile. The comment in config.h implied that defining the macro would
130 enable coverage support, which is totally false. There was also support for
131 setting this macro in the CMake files (my fault, I just copied it from
132 configure). SUPPORT_GCOV has now been removed.
133
134 3. Make a small performance improvement in strlen16() and strlen32() in
135 pcretest.
136
137 4. Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
138 detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
139 these cases). There was also one in pcretest.c.
140
141 5. Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
142
143 6. In UTF mode, the code for checking whether a group could match an empty
144 string (which is used for indefinitely repeated groups to allow for
145 breaking an infinite loop) was broken when the group contained a repeated
146 negated single-character class with a character that occupied more than one
147 data item and had a minimum repetition of zero (for example, [^\x{100}]* in
148 UTF-8 mode). The effect was undefined: the group might or might not be
149 deemed as matching an empty string, or the program might have crashed.
150
151 7. The code for checking whether a group could match an empty string was not
152 recognizing that \h, \H, \v, \V, and \R must match a character.
153
154 8. Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
155 an empty string. If it can, pcretest shows this in its information output.
156
157 9. Fixed two related bugs that applied to Unicode extended grapheme clusters
158 that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
159 matched by pcre_exec() without using JIT:
160
161 (a) If the rest of the pattern did not match after a maximal run of
162 grapheme clusters, the code for backing up to try with fewer of them
163 did not always back up over a full grapheme when characters that do not
164 have the modifier quality were involved, e.g. Hangul syllables.
165
166 (b) If the match point in a subject started with modifier character, and
167 there was no match, the code could incorrectly back up beyond the match
168 point, and potentially beyond the first character in the subject,
169 leading to a segfault or an incorrect match result.
170
171 10. A conditional group with an assertion condition could lead to PCRE
172 recording an incorrect first data item for a match if no other first data
173 item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
174 first data item, and therefore matched "ca" after "c" instead of at the
175 start.
176
177 11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
178 bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
179
180 12. The source of pcregrep now includes z/OS-specific code so that it can be
181 compiled for z/OS as part of the special z/OS distribution.
182
183 13. Added the -T and -TM options to pcretest.
184
185 14. The code in pcre_compile.c for creating the table of named capturing groups
186 has been refactored. Instead of creating the table dynamically during the
187 actual compiling pass, the information is remembered during the pre-compile
188 pass (on the stack unless there are more than 20 named groups, in which
189 case malloc() is used) and the whole table is created before the actual
190 compile happens. This has simplified the code (it is now nearly 150 lines
191 shorter) and prepared the way for better handling of references to groups
192 with duplicate names.
193
194 15. A back reference to a named subpattern when there is more than one of the
195 same name now checks them in the order in which they appear in the pattern.
196 The first one that is set is used for the reference. Previously only the
197 first one was inspected. This change makes PCRE more compatible with Perl.
198
199 16. Unicode character properties were updated from Unicode 6.3.0.
200
201 17. The compile-time code for auto-possessification has been refactored, based
202 on a patch by Zoltan Herczeg. It now happens after instead of during
203 compilation. The code is cleaner, and more cases are handled. The option
204 PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
205 options in pcretest are provided to set it. It can also be set by
206 (*NO_AUTO_POSSESS) at the start of a pattern.
207
208 18. The character VT has been added to the default ("C" locale) set of
209 characters that match \s and are generally treated as white space,
210 following this same change in Perl 5.18. There is now no difference between
211 "Perl space" and "POSIX space". Whether VT is treated as white space in
212 other locales depends on the locale.
213
214 19. The code for checking named groups as conditions, either for being set or
215 for being recursed, has been refactored (this is related to 14 and 15
216 above). Processing unduplicated named groups should now be as fast at
217 numerical groups, and processing duplicated groups should be faster than
218 before.
219
220 20. Two patches to the CMake build system, by Alexander Barkov:
221
222 (1) Replace the "source" command by "." in CMakeLists.txt because
223 "source" is a bash-ism.
224
225 (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
226 without these the CMake build does not work on Solaris.
227
228 21. Perl has changed its handling of \8 and \9. If there is no previously
229 encountered capturing group of those numbers, they are treated as the
230 literal characters 8 and 9 instead of a binary zero followed by the
231 literals. PCRE now does the same.
232
233 22. Following Perl, added \o{} to specify codepoints in octal, making it
234 possible to specify values greater than 0777 and also making them
235 unambiguous.
236
237 23. Perl now gives an error for missing closing braces after \x{... instead of
238 treating the string as literal. PCRE now does the same.
239
240 24. RunTest used to grumble if an inappropriate test was selected explicitly,
241 but just skip it when running all tests. This make it awkward to run ranges
242 of tests when one of them was inappropriate. Now it just skips any
243 inappropriate tests, as it always did when running all tests.
244
245 25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
246 character types such as \d or \w, too many callouts were inserted, and the
247 data that they returned was rubbish.
248
249 26. In UCP mode, \s was not matching two of the characters that Perl matches,
250 namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
251 were matched by \h. The code has now been refactored so that the lists of
252 the horizontal and vertical whitespace characters used for \h and \v (which
253 are defined only in one place) are now also used for \s.
254
255 27. Add JIT support for the 64 bit TileGX architecture.
256 Patch by Jiong Wang (Tilera Corporation).
257
258 28. Possessive quantifiers for classes (both explicit and automatically
259 generated) now use special opcodes instead of wrapping in ONCE brackets.
260
261 29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
262 (because it's meaningless), this was not happening when PCRE_CASELESS was
263 set. Not wrong, but inefficient.
264
265 30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
266 properties for \w, \d, etc) is present in a test regex. Otherwise if the
267 test contains no characters greater than 255, Perl doesn't realise it
268 should be using Unicode semantics.
269
270 31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
271 [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
272 does in Unicode mode.
273
274 32. Added the "forbid" facility to pcretest so that putting tests into the
275 wrong test files can sometimes be quickly detected.
276
277 33. There is now a limit (default 250) on the depth of nesting of parentheses.
278 This limit is imposed to control the amount of system stack used at compile
279 time. It can be changed at build time by --with-parens-nest-limit=xxx or
280 the equivalent in CMake.
281
282 34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
283 errors. Perl warns for these when in warning mode, but PCRE has no facility
284 for giving warnings.
285
286 35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
287 However, this was not working for (?!) because it is optimized to (*FAIL),
288 for which PCRE does not allow quantifiers. The optimization is now disabled
289 when a quantifier follows (?!). I can't see any use for this, but it makes
290 things uniform.
291
292 36. Perl no longer allows group names to start with digits, so I have made this
293 change also in PCRE. It simplifies the code a bit.
294
295 37. In extended mode, Perl ignores spaces before a + that indicates a
296 possessive quantifier. PCRE allowed a space before the quantifier, but not
297 before the possessive +. It now does.
298
299 38. The use of \K (reset reported match start) within a repeated possessive
300 group such as (a\Kb)*+ was not working.
301
302 40. Document that the same character tables must be used at compile time and
303 run time, and that the facility to pass tables to pcre_exec() and
304 pcre_dfa_exec() is for use only with saved/restored patterns.
305
306 41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
307 features for Builds with MSVC:
308
309 1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
310 builds). The .rc files can be used to set FileDescription and many other
311 attributes.
312
313 2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
314 This allows higher-level build scripts which want .pdb files to avoid
315 hard-coding the exact files needed."
316
317 42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
318 mean "start of word" and "end of word", respectively, as a transition aid.
319
320 43. A minimizing repeat of a class containing codepoints greater than 255 in
321 non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
322 compiled to use the heap for recursion.
323
324 44. Got rid of some compiler warnings for unused variables when UTF but not UCP
325 is configured.
326
327
328 Version 8.33 28-May-2013
329 ------------------------
330
331 1. Added 'U' to some constants that are compared to unsigned integers, to
332 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
333 variables that are added to signed variables, to ensure the result is
334 signed and can be negated.
335
336 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
337 CMake config files.
338
339 3. Revise the creation of config.h.generic so that all boolean macros are
340 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
341 overriding via -D on the command line possible.
342
343 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
344 to unsigned int is reported to make a quite noticeable speed difference in
345 a specific Windows environment. Testing on Linux did also appear to show
346 some benefit (and it is clearly not harmful). Also fixed the definition of
347 Xop which should be unsigned.
348
349 5. Related to (4), changing the definition of the intermediate variable cc
350 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
351 improvements.
352
353 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
354 unnecessary spaces.
355
356 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
357 and later.
358
359 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
360
361 9. Optimizing fast_forward_start_bits in JIT.
362
363 10. Adding support for callouts in JIT, and fixing some issues revealed
364 during this work. Namely:
365
366 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
367
368 (b) Minimum length was not checked before the matching is started.
369
370 11. The value of capture_last that is passed to callouts was incorrect in some
371 cases when there was a capture on one path that was subsequently abandoned
372 after a backtrack. Also, the capture_last value is now reset after a
373 recursion, since all captures are also reset in this case.
374
375 12. The interpreter no longer returns the "too many substrings" error in the
376 case when an overflowing capture is in a branch that is subsequently
377 abandoned after a backtrack.
378
379 13. In the pathological case when an offset vector of size 2 is used, pcretest
380 now prints out the matched string after a yield of 0 or 1.
381
382 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
383 Only supported by the JIT compiler at the moment.
384
385 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
386
387 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
388 offset of the starting point of the matching process, provided the offsets
389 vector is large enough.
390
391 17. The \A escape now records a lookbehind value of 1, though its execution
392 does not actually inspect the previous character. This is to ensure that,
393 in partial multi-segment matching, at least one character from the old
394 segment is retained when a new segment is processed. Otherwise, if there
395 are no lookbehinds in the pattern, \A might match incorrectly at the start
396 of a new segment.
397
398 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
399
400 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
401 gives some modest performance improvement in 8-bit mode.
402
403 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
404 be expressed in certain programming languages using Universal Character
405 Names.
406
407 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
408 which points out that "non characters" are not "characters that may not
409 appear in Unicode strings" but rather "characters that are reserved for
410 internal use and have only local meaning".
411
412 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
413 there was a conditional group that depended on an assertion, if the
414 assertion was false, the callout that immediately followed the alternation
415 in the condition was skipped when pcre_exec() was used for matching.
416
417 23. Allow an explicit callout to be inserted before an assertion that is the
418 condition for a conditional group, for compatibility with automatic
419 callouts, which always insert a callout at this point.
420
421 24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
422 confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
423
424 25. (*PRUNE) is now supported by the JIT compiler.
425
426 26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
427
428 27. Fix the case where there are two or more SKIPs with arguments that may be
429 ignored.
430
431 28. (*SKIP) is now supported by the JIT compiler.
432
433 29. (*THEN) is now supported by the JIT compiler.
434
435 30. Update RunTest with additional test selector options.
436
437 31. The way PCRE handles backtracking verbs has been changed in two ways.
438
439 (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
440 SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
441 backtracking. In some cases this makes it more Perl-compatible, but Perl's
442 rather obscure rules do not always do the same thing.
443
444 (2) Previously, backtracking verbs were confined within assertions. This is
445 no longer the case for positive assertions, except for (*ACCEPT). Again,
446 this sometimes improves Perl compatibility, and sometimes does not.
447
448 32. A number of tests that were in test 2 because Perl did things differently
449 have been moved to test 1, because either Perl or PCRE has changed, and
450 these tests are now compatible.
451
452 32. Backtracking control verbs are now handled in the same way in JIT and
453 interpreter.
454
455 33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
456 contained a forward subroutine reference caused a compile error.
457
458 34. Auto-detect and optimize limited repetitions in JIT.
459
460 35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
461 blocking (*UTF) etc.
462
463 36. In the interpreter, maximizing pattern repetitions for characters and
464 character types now use tail recursion, which reduces stack usage.
465
466 37. The value of the max lookbehind was not correctly preserved if a compiled
467 and saved regex was reloaded on a host of different endianness.
468
469 38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
470 of the compiled pattern block, expand the flags field from 16 to 32 bits
471 because it was almost full.
472
473 39. Try madvise first before posix_madvise.
474
475 40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
476 with a pattern such as ^$. It has taken 4 years for anybody to notice! The
477 original change locked out all matches of empty strings. This has been
478 changed so that one match of an empty string per line is recognized.
479 Subsequent searches on the same line (for colouring or for --only-matching,
480 for example) do not recognize empty strings.
481
482 41. Applied a user patch to fix a number of spelling mistakes in comments.
483
484 42. Data lines longer than 65536 caused pcretest to crash.
485
486 43. Clarified the data type for length and startoffset arguments for pcre_exec
487 and pcre_dfa_exec in the function-specific man pages, where they were
488 explicitly stated to be in bytes, never having been updated. I also added
489 some clarification to the pcreapi man page.
490
491 44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
492 a segmentation fault.
493
494
495 Version 8.32 30-November-2012
496 -----------------------------
497
498 1. Improved JIT compiler optimizations for first character search and single
499 character iterators.
500
501 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
502 Patch by Daniel Richard G.
503
504 3. Single character iterator optimizations in the JIT compiler.
505
506 4. Improved JIT compiler optimizations for character ranges.
507
508 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
509 Reported by Giuseppe D'Angelo.
510
511 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
512 of a line, was being set incorrectly in cases where .* appeared inside
513 atomic brackets at the start of a pattern, or where there was a subsequent
514 *PRUNE or *SKIP.
515
516 7. Improved instruction cache flush for POWER/PowerPC.
517 Patch by Daniel Richard G.
518
519 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
520 grep:
521
522 (a) There is now no limit to the number of patterns to be matched.
523
524 (b) An error is given if a pattern is too long.
525
526 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
527 are now supported.
528
529 (d) --exclude-from and --include-from (multiple use) have been added.
530
531 (e) Exclusions and inclusions now apply to all files and directories, not
532 just to those obtained from scanning a directory recursively.
533
534 (f) Multiple uses of -f and --file-list are now supported.
535
536 (g) In a Windows environment, the default for -d has been changed from
537 "read" (the GNU grep default) to "skip", because otherwise the presence
538 of a directory in the file list provokes an error.
539
540 (h) The documentation has been revised and clarified in places.
541
542 9. Improve the matching speed of capturing brackets.
543
544 10. Changed the meaning of \X so that it now matches a Unicode extended
545 grapheme cluster.
546
547 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
548 out POSIX threads when JIT support is configured.
549
550 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
551
552 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
553 structure could go wrong in environments where size_t is not the same size
554 as int.
555
556 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
557 set.
558
559 15. The EBCDIC support had decayed; later updates to the code had included
560 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
561 general tidy up of EBCDIC-related issues, and the documentation was also
562 not quite right. There is now a test that can be run on ASCII systems to
563 check some of the EBCDIC-related things (but is it not a full test).
564
565 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
566 in a small tidy to the code.
567
568 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
569
570 18. If the --only-matching (-o) option in pcregrep is specified multiple
571 times, each one causes appropriate output. For example, -o1 -o2 outputs the
572 substrings matched by the 1st and 2nd capturing parentheses. A separating
573 string can be specified by --om-separator (default empty).
574
575 19. Improving the first n character searches.
576
577 20. Turn case lists for horizontal and vertical white space into macros so that
578 they are defined only once.
579
580 21. This set of changes together give more compatible Unicode case-folding
581 behaviour for characters that have more than one other case when UCP
582 support is available.
583
584 (a) The Unicode property table now has offsets into a new table of sets of
585 three or more characters that are case-equivalent. The MultiStage2.py
586 script that generates these tables (the pcre_ucd.c file) now scans
587 CaseFolding.txt instead of UnicodeData.txt for character case
588 information.
589
590 (b) The code for adding characters or ranges of characters to a character
591 class has been abstracted into a generalized function that also handles
592 case-independence. In UTF-mode with UCP support, this uses the new data
593 to handle characters with more than one other case.
594
595 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
596 whose other case is greater than 256 are now correctly matched
597 caselessly. Previously, the high codepoint matched the low one, but not
598 vice versa.
599
600 (d) The processing of \h, \H, \v, and \ in character classes now makes use
601 of the new class addition function, using character lists defined as
602 macros alongside the case definitions of 20 above.
603
604 (e) Caseless back references now work with characters that have more than
605 one other case.
606
607 (f) General caseless matching of characters with more than one other case
608 is supported.
609
610 22. Unicode character properties were updated from Unicode 6.2.0
611
612 23. Improved CMake support under Windows. Patch by Daniel Richard G.
613
614 24. Add support for 32-bit character strings, and UTF-32
615
616 25. Major JIT compiler update (code refactoring and bugfixing).
617 Experimental Sparc 32 support is added.
618
619 26. Applied a modified version of Daniel Richard G's patch to create
620 pcre.h.generic and config.h.generic by "make" instead of in the
621 PrepareRelease script.
622
623 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
624 pcre_compile.c when checking for a zero character.
625
626 28. Introducing a native interface for JIT. Through this interface, the compiled
627 machine code can be directly executed. The purpose of this interface is to
628 provide fast pattern matching, so several sanity checks are not performed.
629 However, feature tests are still performed. The new interface provides
630 1.4x speedup compared to the old one.
631
632 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
633 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
634 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
635
636 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
637 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
638 the "old" RFC 2279). Instead, it ended up passing a negative length to
639 pcre_exec().
640
641 31. Add support for GCC's visibility feature to hide internal functions.
642
643 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
644 "unknown -C option" after outputting 0 or 1.
645
646 33. There is now support for generating a code coverage report for the test
647 suite in environments where gcc is the compiler and lcov is installed. This
648 is mainly for the benefit of the developers.
649
650 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
651 unaddressable using valgrind annotations, allowing valgrind to detect
652 invalid memory accesses. This is mainly for the benefit of the developers.
653
654 25. (*UTF) can now be used to start a pattern in any of the three libraries.
655
656 26. Give configure error if --enable-cpp but no C++ compiler found.
657
658
659 Version 8.31 06-July-2012
660 -------------------------
661
662 1. Fixing a wrong JIT test case and some compiler warnings.
663
664 2. Removed a bashism from the RunTest script.
665
666 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
667 to unsigned type, result still unsigned" that was given by an MS compiler
668 on encountering the code "-sizeof(xxx)".
669
670 4. Partial matching support is added to the JIT compiler.
671
672 5. Fixed several bugs concerned with partial matching of items that consist
673 of more than one character:
674
675 (a) /^(..)\1/ did not partially match "aba" because checking references was
676 done on an "all or nothing" basis. This also applied to repeated
677 references.
678
679 (b) \R did not give a hard partial match if \r was found at the end of the
680 subject.
681
682 (c) \X did not give a hard partial match after matching one or more
683 characters at the end of the subject.
684
685 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
686 a partial match for the string "\r".
687
688 (e) When newline was set to CRLF, the metacharacter "." did not recognize
689 a partial match for a CR character at the end of the subject string.
690
691 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
692 running pcretest, the text "(JIT)" added to the output whenever JIT is
693 actually used to run the match.
694
695 7. Individual JIT compile options can be set in pcretest by following -s+[+]
696 or /S+[+] with a digit between 1 and 7.
697
698 8. OP_NOT now supports any UTF character not just single-byte ones.
699
700 9. (*MARK) control verb is now supported by the JIT compiler.
701
702 10. The command "./RunTest list" lists the available tests without actually
703 running any of them. (Because I keep forgetting what they all are.)
704
705 11. Add PCRE_INFO_MAXLOOKBEHIND.
706
707 12. Applied a (slightly modified) user-supplied patch that improves performance
708 when the heap is used for recursion (compiled with --disable-stack-for-
709 recursion). Instead of malloc and free for each heap frame each time a
710 logical recursion happens, frames are retained on a chain and re-used where
711 possible. This sometimes gives as much as 30% improvement.
712
713 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
714 call.
715
716 14. As documented, (*COMMIT) is now confined to within a positive assertion.
717
718 15. It is now possible to link pcretest with libedit as an alternative to
719 libreadline.
720
721 16. (*COMMIT) control verb is now supported by the JIT compiler.
722
723 17. The Unicode data tables have been updated to Unicode 6.1.0.
724
725 18. Added --file-list option to pcregrep.
726
727 19. Added binary file support to pcregrep, including the -a, --binary-files,
728 -I, and --text options.
729
730 20. The madvise function is renamed for posix_madvise for QNX compatibility
731 reasons. Fixed by Giuseppe D'Angelo.
732
733 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
734
735 22. Changed the option for creating symbolic links for 16-bit man pages from
736 -s to -sf so that re-installing does not cause issues.
737
738 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
739
740 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
741 matches in certain environments (the workspace was not being correctly
742 retained). Also added to pcre_dfa_exec() a simple plausibility check on
743 some of the workspace data at the beginning of a restart.
744
745 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
746 was not doing so when it should - probably a typo introduced by SVN 528
747 (change 8.10/14).
748
749 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
750 \w+ when the character tables indicated that \x{c4} was a word character.
751 There were several related cases, all because the tests for doing a table
752 lookup were testing for characters less than 127 instead of 255.
753
754 27. If a pattern contains capturing parentheses that are not used in a match,
755 their slots in the ovector are set to -1. For those that are higher than
756 any matched groups, this happens at the end of processing. In the case when
757 there were back references that the ovector was too small to contain
758 (causing temporary malloc'd memory to be used during matching), and the
759 highest capturing number was not used, memory off the end of the ovector
760 was incorrectly being set to -1. (It was using the size of the temporary
761 memory instead of the true size.)
762
763 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
764 ovector size, it uses memory at the end of the block that it has got.
765
766 29. Check for an overlong MARK name and give an error at compile time. The
767 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
768
769 30. JIT compiler update.
770
771 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
772 Rill for the patch.
773
774 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
775
776 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
777
778 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
779 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
780 was enabled.
781
782 35. Improve JIT code generation for greedy plus quantifier.
783
784 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
785 1 to "aa" instead of to an empty string. The bug affected repeated groups
786 that could potentially match an empty string.
787
788 37. Optimizing single character iterators in JIT.
789
790 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
791 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
792 codepoints that are too big for the mode are faulted, and in a UTF mode,
793 disallowed codepoints are also faulted.
794
795 39. If PCRE was compiled with UTF support, in three places in the DFA
796 matcher there was code that should only have been obeyed in UTF mode, but
797 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
798 processing when bytes with values greater than 127 were present. In 16-bit
799 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
800 both cases the values are those that cannot be the first data item in a UTF
801 character. The three items that might have provoked this were recursions,
802 possessively repeated groups, and atomic groups.
803
804 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
805 and pcregrep, because some OS require shared objects to be explicitly
806 passed to ld, causing the link step to fail if they are not.
807
808 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
809 mode, patterns that started with \h* or \R* might be incorrectly matched.
810
811
812 Version 8.30 04-February-2012
813 -----------------------------
814
815 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
816 name is defined in ctype.h.
817
818 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
819 only in quite long subpatterns.
820
821 3. Removed the function pcre_info(), which has been obsolete and deprecated
822 since it was replaced by pcre_fullinfo() in February 2000.
823
824 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
825 match a (*MARK), and the match failed at the start of the subject, a
826 reference to memory before the start of the subject could occur. This bug
827 was introduced by fix 17 of release 8.21.
828
829 5. A reference to an unset group with zero minimum repetition was giving
830 totally wrong answers (in non-JavaScript-compatibility mode). For example,
831 /(another)?(\1?)test/ matched against "hello world test". This bug was
832 introduced in release 8.13.
833
834 6. Add support for 16-bit character strings (a large amount of work involving
835 many changes and refactorings).
836
837 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
838 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
839 from a file.
840
841 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
842 rounding is not applied in this particular case).
843
844 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
845 if they appear, or are escaped, in patterns.
846
847 10. Get rid of a number of -Wunused-but-set-variable warnings.
848
849 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
850 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
851 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
852 also returns the mark "x". This bug applied to capturing parentheses,
853 non-capturing parentheses, and atomic parentheses. It also applied to some
854 assertions.
855
856 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
857 information out of configure.ac instead of relying on pcre.h.generic, which
858 is not stored in the repository.
859
860 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
861 -lreadline.
862
863 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
864
865 15. Applied Graycode's patch to put the top-level frame on the stack rather
866 than the heap when not using the stack for recursion. This gives a
867 performance improvement in many cases when recursion is not deep.
868
869 16. Experimental code added to "pcretest -C" to output the stack frame size.
870
871
872 Version 8.21 12-Dec-2011
873 ------------------------
874
875 1. Updating the JIT compiler.
876
877 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
878 are added as well.
879
880 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
881 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
882 calling _pcre_jit_exec. Some extra comments are added.
883
884 4. (*MARK) settings inside atomic groups that do not contain any capturing
885 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
886 was introduced by change 18 for 8.20.
887
888 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
889 ECMA-262 standard.
890
891 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
892 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
893 This bug was probably introduced by change 9 of 8.13.
894
895 7. While fixing 6 above, I noticed that a number of other items were being
896 incorrectly rejected as "not fixed length". This arose partly because newer
897 opcodes had not been added to the fixed-length checking code. I have (a)
898 corrected the bug and added tests for these items, and (b) arranged for an
899 error to occur if an unknown opcode is encountered while checking for fixed
900 length instead of just assuming "not fixed length". The items that were
901 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
902 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
903 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
904
905 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
906 being incorrectly compiled and would have given unpredicatble results.
907
908 9. A possessively repeated subpattern with minimum repeat count greater than
909 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
910 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
911 the first (A) could occur when it should not.
912
913 10. Add a cast and remove a redundant test from the code.
914
915 11. JIT should use pcre_malloc/pcre_free for allocation.
916
917 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
918 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
919 is anything other than /usr, -L is still shown).
920
921 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
922
923 14. Perl does not support \N without a following name in a [] class; PCRE now
924 also gives an error.
925
926 15. If a forward reference was repeated with an upper limit of around 2000,
927 it caused the error "internal error: overran compiling workspace". The
928 maximum number of forward references (including repeats) was limited by the
929 internal workspace, and dependent on the LINK_SIZE. The code has been
930 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
931 the default depends on LINK_SIZE. There is a new upper limit (for safety)
932 of around 200,000 forward references. While doing this, I also speeded up
933 the filling in of repeated forward references.
934
935 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
936 incorrectly expecting the subject to contain another "a" after the start.
937
938 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
939 in the match, the SKIP should be ignored. This was not happening; instead
940 the SKIP was being treated as NOMATCH. For patterns such as
941 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
942 tested.
943
944 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
945 now much more compatible with Perl, in particular in cases where the result
946 is a non-match for a non-anchored pattern. For example, if
947 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
948 "m", where previously it did not return a name. A side effect of this
949 change is that for partial matches, the last encountered mark name is
950 returned, as for non matches. A number of tests that were previously not
951 Perl-compatible have been moved into the Perl-compatible test files. The
952 refactoring has had the pleasing side effect of removing one argument from
953 the match() function, thus reducing its stack requirements.
954
955 19. If the /S+ option was used in pcretest to study a pattern using JIT,
956 subsequent uses of /S (without +) incorrectly behaved like /S+.
957
958 21. Retrieve executable code size support for the JIT compiler and fixing
959 some warnings.
960
961 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
962 not work when the shorter character appeared right at the end of the
963 subject string.
964
965 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
966 systems.
967
968 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
969 output it when the /M option is used in pcretest.
970
971 25. The CheckMan script was not being included in the distribution. Also, added
972 an explicit "perl" to run Perl scripts from the PrepareRelease script
973 because this is reportedly needed in Windows.
974
975 26. If study data was being save in a file and studying had not found a set of
976 "starts with" bytes for the pattern, the data written to the file (though
977 never used) was taken from uninitialized memory and so caused valgrind to
978 complain.
979
980 27. Updated RunTest.bat as provided by Sheri Pierce.
981
982 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
983
984 29. Computation of memory usage for the table of capturing group names was
985 giving an unnecessarily large value.
986
987
988 Version 8.20 21-Oct-2011
989 ------------------------
990
991 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
992 a POSIX class. After further experiments with Perl, which convinced me that
993 Perl has bugs and confusions, a closing square bracket is no longer allowed
994 in a POSIX name. This bug also affected patterns with classes that started
995 with full stops.
996
997 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
998 captured substring, but while checking the failing first alternative,
999 substring 1 is temporarily captured. If the output vector supplied to
1000 pcre_exec() was not big enough for this capture, the yield of the function
1001 was still zero ("insufficient space for captured substrings"). This cannot
1002 be totally fixed without adding another stack variable, which seems a lot
1003 of expense for a edge case. However, I have improved the situation in cases
1004 such as /(a)(b)x|abc/ matched against "abc", where the return code
1005 indicates that fewer than the maximum number of slots in the ovector have
1006 been set.
1007
1008 3. Related to (2) above: when there are more back references in a pattern than
1009 slots in the output vector, pcre_exec() uses temporary memory during
1010 matching, and copies in the captures as far as possible afterwards. It was
1011 using the entire output vector, but this conflicts with the specification
1012 that only 2/3 is used for passing back captured substrings. Now it uses
1013 only the first 2/3, for compatibility. This is, of course, another edge
1014 case.
1015
1016 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
1017 main code base, and can be used by building with --enable-jit. When this is
1018 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1019 runtime --no-jit option is given.
1020
1021 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
1022 ovector, the return from the function was zero, implying that there were
1023 other matches that did not fit. The correct "exactly full" value is now
1024 returned.
1025
1026 6. If a subpattern that was called recursively or as a subroutine contained
1027 (*PRUNE) or any other control that caused it to give a non-standard return,
1028 invalid errors such as "Error -26 (nested recursion at the same subject
1029 position)" or even infinite loops could occur.
1030
1031 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1032 computing the minimum length on reaching *ACCEPT, and so ended up with the
1033 wrong value of 1 rather than 0. Further investigation indicates that
1034 computing a minimum subject length in the presence of *ACCEPT is difficult
1035 (think back references, subroutine calls), and so I have changed the code
1036 so that no minimum is registered for a pattern that contains *ACCEPT.
1037
1038 8. If (*THEN) was present in the first (true) branch of a conditional group,
1039 it was not handled as intended. [But see 16 below.]
1040
1041 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1042 Sheri Pierce.
1043
1044 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1045 the first byte in a match must be "a".
1046
1047 11. Change 17 for 8.13 increased the recursion depth for patterns like
1048 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1049 pattern contains any instances of (*THEN). If it does not, the old
1050 optimizations are restored. It would be nice to do this on a per-group
1051 basis, but at the moment that is not feasible.
1052
1053 12. In some environments, the output of pcretest -C is CRLF terminated. This
1054 broke RunTest's code that checks for the link size. A single white space
1055 character after the value is now allowed for.
1056
1057 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1058 For "fr", it uses the Windows-specific input and output files.
1059
1060 14. If (*THEN) appeared in a group that was called recursively or as a
1061 subroutine, it did not work as intended. [But see next item.]
1062
1063 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1064 pattern fragments (but not containing any | characters). If A and B are
1065 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1066 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1067 D. In other words, Perl considers parentheses that do not contain any |
1068 characters to be part of a surrounding alternative, whereas PCRE was
1069 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1070 differently. PCRE now behaves in the same way as Perl, except in the case
1071 of subroutine/recursion calls such as (?1) which have in any case always
1072 been different (but PCRE had them first :-).
1073
1074 16. Related to 15 above: Perl does not treat the | in a conditional group as
1075 creating alternatives. Such a group is treated in the same way as an
1076 ordinary group without any | characters when processing (*THEN). PCRE has
1077 been changed to match Perl's behaviour.
1078
1079 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1080 RunGrepTest script failed.
1081
1082 18. Change 22 for version 13 caused atomic groups to use more stack. This is
1083 inevitable for groups that contain captures, but it can lead to a lot of
1084 stack use in large patterns. The old behaviour has been restored for atomic
1085 groups that do not contain any capturing parentheses.
1086
1087 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1088 suppress the check for a minimum subject length at run time. (If it was
1089 given to pcre_exec() or pcre_dfa_exec() it did work.)
1090
1091 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1092 fail to work when decoding hex characters in data strings in EBCDIC
1093 environments.
1094
1095 21. It appears that in at least one Mac OS environment, the isxdigit() function
1096 is implemented as a macro that evaluates to its argument more than once,
1097 contravening the C 90 Standard (I haven't checked a later standard). There
1098 was an instance in pcretest which caused it to go wrong when processing
1099 \x{...} escapes in subject strings. The has been rewritten to avoid using
1100 things like p++ in the argument of isxdigit().
1101
1102
1103 Version 8.13 16-Aug-2011
1104 ------------------------
1105
1106 1. The Unicode data tables have been updated to Unicode 6.0.0.
1107
1108 2. Two minor typos in pcre_internal.h have been fixed.
1109
1110 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1111 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1112 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1113
1114 4. There were a number of related bugs in the code for matching backrefences
1115 caselessly in UTF-8 mode when codes for the characters concerned were
1116 different numbers of bytes. For example, U+023A and U+2C65 are an upper
1117 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1118 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1119 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1120 2-byte code at the end of the subject (it thought there wasn't enough data
1121 left).
1122
1123 5. Comprehensive information about what went wrong is now returned by
1124 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1125 as the output vector has at least 2 elements. The offset of the start of
1126 the failing character and a reason code are placed in the vector.
1127
1128 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
1129 now returned is for the first byte of the failing character, instead of the
1130 last byte inspected. This is an incompatible change, but I hope it is small
1131 enough not to be a problem. It makes the returned offset consistent with
1132 pcre_exec() and pcre_dfa_exec().
1133
1134 7. pcretest now gives a text phrase as well as the error number when
1135 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1136 failure, the offset and reason code are output.
1137
1138 8. When \R was used with a maximizing quantifier it failed to skip backwards
1139 over a \r\n pair if the subsequent match failed. Instead, it just skipped
1140 back over a single character (\n). This seems wrong (because it treated the
1141 two characters as a single entity when going forwards), conflicts with the
1142 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1143 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1144 has been changed.
1145
1146 9. Some internal refactoring has changed the processing so that the handling
1147 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1148 time (the PCRE_DOTALL option was changed this way some time ago: version
1149 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1150 which was always a bit of a fudge. It also means that there is one less
1151 argument for the match() function, which reduces its stack requirements
1152 slightly. This change also fixes an incompatibility with Perl: the pattern
1153 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1154
1155 10. More internal refactoring has drastically reduced the number of recursive
1156 calls to match() for possessively repeated groups such as (abc)++ when
1157 using pcre_exec().
1158
1159 11. While implementing 10, a number of bugs in the handling of groups were
1160 discovered and fixed:
1161
1162 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1163 (a|)*(?1) gave a compile-time internal error.
1164 ((a|)+)+ did not notice that the outer group could match an empty string.
1165 (^a|^)+ was not marked as anchored.
1166 (.*a|.*)+ was not marked as matching at start or after a newline.
1167
1168 12. Yet more internal refactoring has removed another argument from the match()
1169 function. Special calls to this function are now indicated by setting a
1170 value in a variable in the "match data" data block.
1171
1172 13. Be more explicit in pcre_study() instead of relying on "default" for
1173 opcodes that mean there is no starting character; this means that when new
1174 ones are added and accidentally left out of pcre_study(), testing should
1175 pick them up.
1176
1177 14. The -s option of pcretest has been documented for ages as being an old
1178 synonym of -m (show memory usage). I have changed it to mean "force study
1179 for every regex", that is, assume /S for every regex. This is similar to -i
1180 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1181 using it. It makes it easier to run collections of tests with and without
1182 study enabled, and thereby test pcre_study() more easily. All the standard
1183 tests are now run with and without -s (but some patterns can be marked as
1184 "never study" - see 20 below).
1185
1186 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1187 restoration of the capturing data to the outer values was not happening
1188 correctly.
1189
1190 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1191 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1192 pattern had matched an empty string, and so incorrectly returned a no
1193 match.
1194
1195 17. There was optimizing code for the last branch of non-capturing parentheses,
1196 and also for the obeyed branch of a conditional subexpression, which used
1197 tail recursion to cut down on stack usage. Unfortunately, now that there is
1198 the possibility of (*THEN) occurring in these branches, tail recursion is
1199 no longer possible because the return has to be checked for (*THEN). These
1200 two optimizations have therefore been removed. [But see 8.20/11 above.]
1201
1202 18. If a pattern containing \R was studied, it was assumed that \R always
1203 matched two bytes, thus causing the minimum subject length to be
1204 incorrectly computed because \R can also match just one byte.
1205
1206 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1207 was incorrectly computed.
1208
1209 20. If /S is present twice on a test pattern in pcretest input, it now
1210 *disables* studying, thereby overriding the use of -s on the command line
1211 (see 14 above). This is necessary for one or two tests to keep the output
1212 identical in both cases.
1213
1214 21. When (*ACCEPT) was used in an assertion that matched an empty string and
1215 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1216
1217 22. When an atomic group that contained a capturing parenthesis was
1218 successfully matched, but the branch in which it appeared failed, the
1219 capturing was not being forgotten if a higher numbered group was later
1220 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1221 group 1 to "a", when in fact it should be unset. This applied to multi-
1222 branched capturing and non-capturing groups, repeated or not, and also to
1223 positive assertions (capturing in negative assertions does not happen
1224 in PCRE) and also to nested atomic groups.
1225
1226 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1227 subject after a captured substring, to make it easier to tell which of a
1228 number of identical substrings has been captured.
1229
1230 24. The way atomic groups are processed by pcre_exec() has been changed so that
1231 if they are repeated, backtracking one repetition now resets captured
1232 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1233 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1234 "aaa". Previously, it would have been "a". As part of this code
1235 refactoring, the way recursive calls are handled has also been changed.
1236
1237 25. If an assertion condition captured any substrings, they were not passed
1238 back unless some other capturing happened later. For example, if
1239 (?(?=(a))a) was matched against "a", no capturing was returned.
1240
1241 26. When studying a pattern that contained subroutine calls or assertions,
1242 the code for finding the minimum length of a possible match was handling
1243 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1244 group 1 called group 2 while simultaneously a separate group 2 called group
1245 1). A stack overflow occurred in this case. I have fixed this by limiting
1246 the recursion depth to 10.
1247
1248 27. Updated RunTest.bat in the distribution to the version supplied by Tom
1249 Fortmann. This supports explicit test numbers on the command line, and has
1250 argument validation and error reporting.
1251
1252 28. An instance of \X with an unlimited repeat could fail if at any point the
1253 first character it looked at was a mark character.
1254
1255 29. Some minor code refactoring concerning Unicode properties and scripts
1256 should reduce the stack requirement of match() slightly.
1257
1258 30. Added the '=' option to pcretest to check the setting of unused capturing
1259 slots at the end of the pattern, which are documented as being -1, but are
1260 not included in the return count.
1261
1262 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1263 compiled something random. Now it gives a compile-time error (as does
1264 Perl).
1265
1266 32. A *MARK encountered during the processing of a positive assertion is now
1267 recorded and passed back (compatible with Perl).
1268
1269 33. If --only-matching or --colour was set on a pcregrep call whose pattern
1270 had alternative anchored branches, the search for a second match in a line
1271 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1272 matched the line "0102" twice. The same bug affected patterns that started
1273 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1274 twice.
1275
1276 34. Previously, PCRE did not allow quantification of assertions. However, Perl
1277 does, and because of capturing effects, quantifying parenthesized
1278 assertions may at times be useful. Quantifiers are now allowed for
1279 parenthesized assertions.
1280
1281 35. A minor code tidy in pcre_compile() when checking options for \R usage.
1282
1283 36. \g was being checked for fancy things in a character class, when it should
1284 just be a literal "g".
1285
1286 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1287 appearance of a nested POSIX class supersedes an apparent external class.
1288 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1289 unescaped square brackets may also appear as part of class names. For
1290 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1291 more like Perl. (But see 8.20/1 above.)
1292
1293 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1294 was because it thought it was \N{name}, which is not supported).
1295
1296 39. Add minix to OS list not supporting the -S option in pcretest.
1297
1298 40. PCRE tries to detect cases of infinite recursion at compile time, but it
1299 cannot analyze patterns in sufficient detail to catch mutual recursions
1300 such as ((?1))((?2)). There is now a runtime test that gives an error if a
1301 subgroup is called recursively as a subpattern for a second time at the
1302 same position in the subject string. In previous releases this might have
1303 been caught by the recursion limit, or it might have run out of stack.
1304
1305 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1306 happen only once. PCRE was, however incorrectly giving a compile time error
1307 "recursive call could loop indefinitely" because it cannot analyze the
1308 pattern in sufficient detail. The compile time test no longer happens when
1309 PCRE is compiling a conditional subpattern, but actual runaway loops are
1310 now caught at runtime (see 40 above).
1311
1312 42. It seems that Perl allows any characters other than a closing parenthesis
1313 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1314 has been changed to be the same.
1315
1316 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1317 as not to get warnings when autogen.sh is called. Also changed
1318 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1319
1320 44. To help people who use pcregrep to scan files containing exceedingly long
1321 lines, the following changes have been made:
1322
1323 (a) The default value of the buffer size parameter has been increased from
1324 8K to 20K. (The actual buffer used is three times this size.)
1325
1326 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1327 PCRE is built.
1328
1329 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1330 to be set at run time.
1331
1332 (d) Numerical values in pcregrep options can be followed by K or M, for
1333 example --buffer-size=50K.
1334
1335 (e) If a line being scanned overflows pcregrep's buffer, an error is now
1336 given and the return code is set to 2.
1337
1338 45. Add a pointer to the latest mark to the callout data block.
1339
1340 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1341 partial match of an empty string instead of no match. This was specific to
1342 the use of ".".
1343
1344 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1345 complete match instead of a partial match. This bug was dependent on both
1346 the PCRE_UTF8 and PCRE_DOTALL options being set.
1347
1348 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1349 starting byte set, because \b was not being ignored.
1350
1351
1352 Version 8.12 15-Jan-2011
1353 ------------------------
1354
1355 1. Fixed some typos in the markup of the man pages, and wrote a script that
1356 checks for such things as part of the documentation building process.
1357
1358 2. On a big-endian 64-bit system, pcregrep did not correctly process the
1359 --match-limit and --recursion-limit options (added for 8.11). In
1360 particular, this made one of the standard tests fail. (The integer value
1361 went into the wrong half of a long int.)
1362
1363 3. If the --colour option was given to pcregrep with -v (invert match), it
1364 did strange things, either producing crazy output, or crashing. It should,
1365 of course, ignore a request for colour when reporting lines that do not
1366 match.
1367
1368 4. Another pcregrep bug caused similar problems if --colour was specified with
1369 -M (multiline) and the pattern match finished with a line ending.
1370
1371 5. In pcregrep, when a pattern that ended with a literal newline sequence was
1372 matched in multiline mode, the following line was shown as part of the
1373 match. This seems wrong, so I have changed it.
1374
1375 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
1376 the check for further matches in the same line (so they could be coloured)
1377 to overrun the end of the current line. If another match was found, it was
1378 incorrectly shown (and then shown again when found in the next line).
1379
1380 7. If pcregrep was compiled under Windows, there was a reference to the
1381 function pcregrep_exit() before it was defined. I am assuming this was
1382 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1383 reported by a user. I've moved the definition above the reference.
1384
1385
1386 Version 8.11 10-Dec-2010
1387 ------------------------
1388
1389 1. (*THEN) was not working properly if there were untried alternatives prior
1390 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1391 backtracked to try for "b" instead of moving to the next alternative branch
1392 at the same level (in this case, to look for "c"). The Perl documentation
1393 is clear that when (*THEN) is backtracked onto, it goes to the "next
1394 alternative in the innermost enclosing group".
1395
1396 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1397 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1398 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1399 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1400 (*THEN).
1401
1402 3. If \s appeared in a character class, it removed the VT character from
1403 the class, even if it had been included by some previous item, for example
1404 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1405 of \s, but is part of the POSIX "space" class.)
1406
1407 4. A partial match never returns an empty string (because you can always
1408 match an empty string at the end of the subject); however the checking for
1409 an empty string was starting at the "start of match" point. This has been
1410 changed to the "earliest inspected character" point, because the returned
1411 data for a partial match starts at this character. This means that, for
1412 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1413 (previously it gave "no match").
1414
1415 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1416 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1417 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1418 has an implication that the given string is incomplete (because a partial
1419 match is preferred over a full match). For this reason, these items now
1420 give a partial match in this situation. [Aside: previously, the one case
1421 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1422 match rather than a full match, which was wrong by the old rules, but is
1423 now correct.]
1424
1425 6. There was a bug in the handling of #-introduced comments, recognized when
1426 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1427 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1428 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1429 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1430 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1431 places in pcre_compile().
1432
1433 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1434 comments when looking ahead for named forward references to subpatterns,
1435 the only newline sequence it recognized was NL. It now handles newlines
1436 according to the set newline convention.
1437
1438 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1439 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1440 cater for a lack of strerror(). These oversights have been fixed.
1441
1442 9. Added --match-limit and --recursion-limit to pcregrep.
1443
1444 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1445
1446 11. When the -o option was used, pcregrep was setting a return code of 1, even
1447 when matches were found, and --line-buffered was not being honoured.
1448
1449 12. Added an optional parentheses number to the -o and --only-matching options
1450 of pcregrep.
1451
1452 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1453 can match an empty string. The code to do it in pcretest and pcredemo
1454 needed fixing:
1455
1456 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1457 only one byte after an empty string match just before CRLF (this case
1458 just got forgotten; "any" and "anycrlf" were OK).
1459
1460 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1461 mode when an empty string match preceded an ASCII character followed by
1462 a non-ASCII character. (The code for advancing by one character rather
1463 than one byte was nonsense.)
1464
1465 (c) The pcredemo.c sample program did not have any code at all to handle
1466 the cases when CRLF is a valid newline sequence.
1467
1468 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1469 as a starting offset was within the subject string. There is now a new
1470 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1471 negative or greater than the length of the string. In order to test this,
1472 pcretest is extended to allow the setting of negative starting offsets.
1473
1474 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1475 starting offset points to the beginning of a UTF-8 character was
1476 unnecessarily clumsy. I tidied it up.
1477
1478 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1479 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1480
1481 17. Nobody had reported that the --include_dir option, which was added in
1482 release 7.7 should have been called --include-dir (hyphen, not underscore)
1483 for compatibility with GNU grep. I have changed it to --include-dir, but
1484 left --include_dir as an undocumented synonym, and the same for
1485 --exclude-dir, though that is not available in GNU grep, at least as of
1486 release 2.5.4.
1487
1488 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1489 characters from a string of bytes) have been redefined so as not to use
1490 loops, in order to improve performance in some environments. At the same
1491 time, I abstracted some of the common code into auxiliary macros to save
1492 repetition (this should not affect the compiled code).
1493
1494 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1495 compile-time error is now given if \c is not followed by an ASCII
1496 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1497 different, and any byte value is allowed.)
1498
1499 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1500 START_OPTIMIZE option, which is now allowed at compile time - but just
1501 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1502 to pcregrep and other applications that have no direct access to PCRE
1503 options. The new /Y option in pcretest sets this option when calling
1504 pcre_compile().
1505
1506 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1507 back references. Groups containing recursive back references were forced to
1508 be atomic by that change, but in the case of named groups, the amount of
1509 memory required was incorrectly computed, leading to "Failed: internal
1510 error: code overflow". This has been fixed.
1511
1512 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1513 pcretest.c, to avoid build problems in some Borland environments.
1514
1515
1516 Version 8.10 25-Jun-2010
1517 ------------------------
1518
1519 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1520 THEN.
1521
1522 2. (*ACCEPT) was not working when inside an atomic group.
1523
1524 3. Inside a character class, \B is treated as a literal by default, but
1525 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1526 causes the error). The code is unchanged, but I tidied the documentation.
1527
1528 4. Inside a character class, PCRE always treated \R and \X as literals,
1529 whereas Perl faults them if its -w option is set. I have changed PCRE so
1530 that it faults them when PCRE_EXTRA is set.
1531
1532 5. Added support for \N, which always matches any character other than
1533 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1534
1535 6. When compiling pcregrep with newer versions of gcc which may have
1536 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1537 declared with attribute warn_unused_result" were given. Just casting the
1538 result to (void) does not stop the warnings; a more elaborate fudge is
1539 needed. I've used a macro to implement this.
1540
1541 7. Minor change to pcretest.c to avoid a compiler warning.
1542
1543 8. Added four artifical Unicode properties to help with an option to make
1544 \s etc use properties (see next item). The new properties are: Xan
1545 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1546
1547 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1548 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1549 this option. Modified pcretest to add /W to test this facility. Added
1550 REG_UCP to make it available via the POSIX interface.
1551
1552 10. Added --line-buffered to pcregrep.
1553
1554 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1555 studied, and the match started with a letter with a code point greater than
1556 127 whose first byte was different to the first byte of the other case of
1557 the letter, the other case of this starting letter was not recognized
1558 (#976).
1559
1560 12. If a pattern that was studied started with a repeated Unicode property
1561 test, for example, \p{Nd}+, there was the theoretical possibility of
1562 setting up an incorrect bitmap of starting bytes, but fortunately it could
1563 not have actually happened in practice until change 8 above was made (it
1564 added property types that matched character-matching opcodes).
1565
1566 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1567 possible starting bytes for non-anchored patterns.
1568
1569 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1570 \R, and also a number of cases that involve Unicode properties, both
1571 explicit and implicit when PCRE_UCP is set.
1572
1573 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1574 input, it could crash or give wrong results if characters with values
1575 greater than 0xc0 were present in the subject string. (Detail: it assumed
1576 UTF-8 input when processing these items.)
1577
1578 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1579 size_t is 64-bit (#991).
1580
1581 17. Added a check for running out of memory when PCRE is compiled with
1582 --disable-stack-for-recursion (#990).
1583
1584 18. If the last data line in a file for pcretest does not have a newline on
1585 the end, a newline was missing in the output.
1586
1587 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1588 less than 128) in its various bitmaps. However, there is a facility for
1589 generating tables according to the current locale when PCRE is compiled. It
1590 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1591 characters, are recognized by isspace() and therefore were getting set in
1592 these tables, and indeed these tables seem to approximate to ISO 8859. This
1593 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1594 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1595 which of course cannot start UTF-8 characters. I have changed the code so
1596 that only real ASCII characters (less than 128) and the correct starting
1597 bytes for UTF-8 encodings are set for characters greater than 127 when in
1598 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1599 altogether.)
1600
1601 20. Added the /T option to pcretest so as to be able to run tests with non-
1602 standard character tables, thus making it possible to include the tests
1603 used for 19 above in the standard set of tests.
1604
1605 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1606 reference to a subpattern the other side of a comment that contains an
1607 opening parenthesis caused either an internal compiling error, or a
1608 reference to the wrong subpattern.
1609
1610
1611 Version 8.02 19-Mar-2010
1612 ------------------------
1613
1614 1. The Unicode data tables have been updated to Unicode 5.2.0.
1615
1616 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1617 configured.
1618
1619 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1620 original author of that file, following a query about its status.
1621
1622 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1623 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1624
1625 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1626 quantifier applied to a forward-referencing subroutine call, could compile
1627 incorrect code or give the error "internal error: previously-checked
1628 referenced subpattern not found".
1629
1630 6. Both MS Visual Studio and Symbian OS have problems with initializing
1631 variables to point to external functions. For these systems, therefore,
1632 pcre_malloc etc. are now initialized to local functions that call the
1633 relevant global functions.
1634
1635 7. There were two entries missing in the vectors called coptable and poptable
1636 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1637 I've fixed the data, and added a kludgy way of testing at compile time that
1638 the lengths are correct (equal to the number of opcodes).
1639
1640 8. Following on from 7, I added a similar kludge to check the length of the
1641 eint vector in pcreposix.c.
1642
1643 9. Error texts for pcre_compile() are held as one long string to avoid too
1644 much relocation at load time. To find a text, the string is searched,
1645 counting zeros. There was no check for running off the end of the string,
1646 which could happen if a new error number was added without updating the
1647 string.
1648
1649 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1650
1651 11. \K was not working if it appeared in an atomic group or in a group that
1652 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1653 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1654 the assertion is positive, but not if it is negative.
1655
1656 12. Change 11 fortuitously reduced the size of the stack frame used in the
1657 "match()" function of pcre_exec.c by one pointer. Forthcoming
1658 implementation of support for (*MARK) will need an extra pointer on the
1659 stack; I have reserved it now, so that the stack frame size does not
1660 decrease.
1661
1662 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1663 item in branch that calls a recursion is a subroutine call - as in the
1664 second branch in the above example - was incorrectly given the compile-
1665 time error "recursive call could loop indefinitely" because pcre_compile()
1666 was not correctly checking the subroutine for matching a non-empty string.
1667
1668 14. The checks for overrunning compiling workspace could trigger after an
1669 overrun had occurred. This is a "should never occur" error, but it can be
1670 triggered by pathological patterns such as hundreds of nested parentheses.
1671 The checks now trigger 100 bytes before the end of the workspace.
1672
1673 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1674
1675
1676 Version 8.01 19-Jan-2010
1677 ------------------------
1678
1679 1. If a pattern contained a conditional subpattern with only one branch (in
1680 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1681 computed the wrong minimum data length (which is of course zero for such
1682 subpatterns). This could cause incorrect "no match" results.
1683
1684 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1685 the pattern is reset in the first branch, pcre_compile() failed with
1686 "internal error: code overflow at offset...". This happened only when
1687 the reset was to the original external option setting. (An optimization
1688 abstracts leading options settings into an external setting, which was the
1689 cause of this.)
1690
1691 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1692 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1693 assertion pattern did not match (meaning that the assertion was true), it
1694 was incorrectly treated as false if the SKIP had been reached during the
1695 matching. This also applied to assertions used as conditions.
1696
1697 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1698 assertion subpattern, including such a pattern used as a condition,
1699 unpredictable results occurred, instead of the error return
1700 PCRE_ERROR_DFA_UITEM.
1701
1702 5. The C++ GlobalReplace function was not working like Perl for the special
1703 situation when an empty string is matched. It now does the fancy magic
1704 stuff that is necessary.
1705
1706 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1707 removed. (These were left over from very, very early versions of PCRE.)
1708
1709 7. Some cosmetic changes to the code to make life easier when compiling it
1710 as part of something else:
1711
1712 (a) Change DEBUG to PCRE_DEBUG.
1713
1714 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1715 called "current" as "current_branch", to prevent a collision with the
1716 Linux macro when compiled as a kernel module.
1717
1718 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1719 prevent a collision with the Linux macro when compiled as a kernel
1720 module.
1721
1722 8. In pcre_compile() there are some checks for integer overflows that used to
1723 cast potentially large values to (double). This has been changed to that
1724 when building, a check for int64_t is made, and if it is found, it is used
1725 instead, thus avoiding the use of floating point arithmetic. (There is no
1726 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1727 double.
1728
1729 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1730 2005 (difference between two addresses compared to an unsigned value).
1731
1732 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1733 custom one, because of the following reported problem in Windows:
1734
1735 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1736 under Win32.
1737 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1738 therefore missing the function definition.
1739 - The compiler thus generates a "C" signature for the test function.
1740 - The linker fails to find the "C" function.
1741 - PCRE fails to configure if asked to do so against libbz2.
1742
1743 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1744 messages were output:
1745
1746 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1747 rerunning libtoolize, to keep the correct libtool macros in-tree.
1748 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1749
1750 I have done both of these things.
1751
1752 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1753 most of the time, it *can* run out if it is given a pattern that contains a
1754 runaway infinite recursion. I updated the discussion in the pcrestack man
1755 page.
1756
1757 13. Now that we have gone to the x.xx style of version numbers, the minor
1758 version may start with zero. Using 08 or 09 is a bad idea because users
1759 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1760 interpreted as invalid octal numbers. I've updated the previous comment in
1761 configure.ac, and also added a check that gives an error if 08 or 09 are
1762 used.
1763
1764 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1765 causing partial matching to fail when the end of the subject matched \W
1766 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1767
1768 15. There were some discrepancies between the declarations in pcre_internal.h
1769 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1770 their definitions. The declarations used "const uschar *" and the
1771 definitions used USPTR. Even though USPTR is normally defined as "const
1772 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1773 reported that: "This difference in casting confuses some C++ compilers, for
1774 example, SunCC recognizes above declarations as different functions and
1775 generates broken code for hbpcre." I have changed the declarations to use
1776 USPTR.
1777
1778 16. GNU libtool is named differently on some systems. The autogen.sh script now
1779 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1780 (FreeBSD).
1781
1782 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1783 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1784 comment: "Figure out how to create a longlong from a string: strtoll and
1785 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1786 instance, but it only takes 2 args instead of 3!"
1787
1788 18. A subtle bug concerned with back references has been fixed by a change of
1789 specification, with a corresponding code fix. A pattern such as
1790 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1791 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1792 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1793 same bug. Such groups have to be quantified to be useful, or contained
1794 inside another quantified group. (If there's no repetition, the reference
1795 can never match.) The problem arises because, having left the group and
1796 moved on to the rest of the pattern, a later failure that backtracks into
1797 the group uses the captured value from the final iteration of the group
1798 rather than the correct earlier one. I have fixed this in PCRE by forcing
1799 any group that contains a reference to itself to be an atomic group; that
1800 is, there cannot be any backtracking into it once it has completed. This is
1801 similar to recursive and subroutine calls.
1802
1803
1804 Version 8.00 19-Oct-09
1805 ----------------------
1806
1807 1. The table for translating pcre_compile() error codes into POSIX error codes
1808 was out-of-date, and there was no check on the pcre_compile() error code
1809 being within the table. This could lead to an OK return being given in
1810 error.
1811
1812 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1813 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1814 in a Windows environment.
1815
1816 3. The pcregrep --count option prints the count for each file even when it is
1817 zero, as does GNU grep. However, pcregrep was also printing all files when
1818 --files-with-matches was added. Now, when both options are given, it prints
1819 counts only for those files that have at least one match. (GNU grep just
1820 prints the file name in this circumstance, but including the count seems
1821 more useful - otherwise, why use --count?) Also ensured that the
1822 combination -clh just lists non-zero counts, with no names.
1823
1824 4. The long form of the pcregrep -F option was incorrectly implemented as
1825 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1826 but it seems right to fix it, and I didn't think it was worth preserving
1827 the old behaviour.
1828
1829 5. The command line items --regex=pattern and --regexp=pattern were not
1830 recognized by pcregrep, which required --regex pattern or --regexp pattern
1831 (with a space rather than an '='). The man page documented the '=' forms,
1832 which are compatible with GNU grep; these now work.
1833
1834 6. No libpcreposix.pc file was created for pkg-config; there was just
1835 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1836
1837 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1838 when UCP support is not needed, by modifying the Python script that
1839 generates it from Unicode data files. This should not matter if the module
1840 is correctly used as a library, but I received one complaint about 50K of
1841 unwanted data. My guess is that the person linked everything into his
1842 program rather than using a library. Anyway, it does no harm.
1843
1844 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1845 was a minimum greater than 1 for a wide character in a possessive
1846 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1847 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1848 character. Chaos in the form of incorrect output or a compiling loop could
1849 result.
1850
1851 9. The restrictions on what a pattern can contain when partial matching is
1852 requested for pcre_exec() have been removed. All patterns can now be
1853 partially matched by this function. In addition, if there are at least two
1854 slots in the offset vector, the offset of the earliest inspected character
1855 for the match and the offset of the end of the subject are set in them when
1856 PCRE_ERROR_PARTIAL is returned.
1857
1858 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1859 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1860 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1861 and may be more useful for multi-segment matching.
1862
1863 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1864 used to be given if ever the end of the subject was reached; now it is
1865 given only if matching could not proceed because another character was
1866 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1867 string "Z", which now yields "no match" instead of "partial match". In the
1868 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1869 final character ended with (*FAIL).
1870
1871 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1872 if the pattern had a "must contain" character that was already found in the
1873 earlier partial match, unless partial matching was again requested. For
1874 example, with the pattern /dog.(body)?/, the "must contain" character is
1875 "g". If the first part-match was for the string "dog", restarting with
1876 "sbody" failed. This bug has been fixed.
1877
1878 13. The string returned by pcre_dfa_exec() after a partial match has been
1879 changed so that it starts at the first inspected character rather than the
1880 first character of the match. This makes a difference only if the pattern
1881 starts with a lookbehind assertion or \b or \B (\K is not supported by
1882 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1883 matching functions compatible, and I think it's the right thing to do.
1884
1885 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1886 so that the demonstration program is easily available in environments where
1887 PCRE has not been installed from source.
1888
1889 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1890 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1891 library.
1892
1893 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1894 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1895 is not the first non-POSIX option to be added. Clearly some people find
1896 these options useful.
1897
1898 17. If a caller to the POSIX matching function regexec() passes a non-zero
1899 value for nmatch with a NULL value for pmatch, the value of
1900 nmatch is forced to zero.
1901
1902 18. RunGrepTest did not have a test for the availability of the -u option of
1903 the diff command, as RunTest does. It now checks in the same way as
1904 RunTest, and also checks for the -b option.
1905
1906 19. If an odd number of negated classes containing just a single character
1907 interposed, within parentheses, between a forward reference to a named
1908 subpattern and the definition of the subpattern, compilation crashed with
1909 an internal error, complaining that it could not find the referenced
1910 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1911 [The bug was that it was starting one character too far in when skipping
1912 over the character class, thus treating the ] as data rather than
1913 terminating the class. This meant it could skip too much.]
1914
1915 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1916 /g option in pcretest when the pattern contains \K, which makes it possible
1917 to have an empty string match not at the start, even when the pattern is
1918 anchored. Updated pcretest and pcredemo to use this option.
1919
1920 21. If the maximum number of capturing subpatterns in a recursion was greater
1921 than the maximum at the outer level, the higher number was returned, but
1922 with unset values at the outer level. The correct (outer level) value is
1923 now given.
1924
1925 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1926 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1927 make it do so. The string so far is captured, making this feature
1928 compatible with Perl.
1929
1930 23. The tests have been re-organized, adding tests 11 and 12, to make it
1931 possible to check the Perl 5.10 features against Perl 5.10.
1932
1933 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1934 pattern matches a fixed length string. PCRE did not allow this; now it
1935 does. Neither allows recursion.
1936
1937 25. I finally figured out how to implement a request to provide the minimum
1938 length of subject string that was needed in order to match a given pattern.
1939 (It was back references and recursion that I had previously got hung up
1940 on.) This code has now been added to pcre_study(); it finds a lower bound
1941 to the length of subject needed. It is not necessarily the greatest lower
1942 bound, but using it to avoid searching strings that are too short does give
1943 some useful speed-ups. The value is available to calling programs via
1944 pcre_fullinfo().
1945
1946 26. While implementing 25, I discovered to my embarrassment that pcretest had
1947 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1948 study optimizations had never been tested with that matching function.
1949 Oops. What is worse, even when it was passed study data, there was a bug in
1950 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1951 were also very few tests of studied patterns with pcre_dfa_exec().
1952
1953 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1954 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1955 on the other side of the coin, they are no longer allowed to have different
1956 names, because these cannot be distinguished in PCRE, and this has caused
1957 confusion. (This is a difference from Perl.)
1958
1959 28. When duplicate subpattern names are present (necessarily with different
1960 numbers, as required by 27 above), and a test is made by name in a
1961 conditional pattern, either for a subpattern having been matched, or for
1962 recursion in such a pattern, all the associated numbered subpatterns are
1963 tested, and the overall condition is true if the condition is true for any
1964 one of them. This is the way Perl works, and is also more like the way
1965 testing by number works.
1966
1967
1968 Version 7.9 11-Apr-09
1969 ---------------------
1970
1971 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1972 (pcretest), all targets were linked against these libraries. This included
1973 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1974 libraries. This caused unwanted dependencies to be created. This problem
1975 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1976 pcretest is linked with readline.
1977
1978 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1979 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1980 moved outside it again, because FALSE and TRUE are already defined in AIX,
1981 but BOOL is not.
1982
1983 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1984 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1985
1986 4. The pcregrep documentation said spaces were inserted as well as colons (or
1987 hyphens) following file names and line numbers when outputting matching
1988 lines. This is not true; no spaces are inserted. I have also clarified the
1989 wording for the --colour (or --color) option.
1990
1991 5. In pcregrep, when --colour was used with -o, the list of matching strings
1992 was not coloured; this is different to GNU grep, so I have changed it to be
1993 the same.
1994
1995 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1996 each matching line was coloured. Now it goes on to look for further matches
1997 of any of the test patterns, which is the same behaviour as GNU grep.
1998
1999 7. A pattern that could match an empty string could cause pcregrep to loop; it
2000 doesn't make sense to accept an empty string match in pcregrep, so I have
2001 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2002 seems to be how GNU grep behaves. [But see later change 40 for release
2003 8.33.]
2004
2005 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2006 start or after a newline", because the conditional assertion was not being
2007 correctly handled. The rule now is that both the assertion and what follows
2008 in the first alternative must satisfy the test.
2009
2010 9. If auto-callout was enabled in a pattern with a conditional group whose
2011 condition was an assertion, PCRE could crash during matching, both with
2012 pcre_exec() and pcre_dfa_exec().
2013
2014 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2015 used for matching.
2016
2017 11. Unicode property support in character classes was not working for
2018 characters (bytes) greater than 127 when not in UTF-8 mode.
2019
2020 12. Added the -M command line option to pcretest.
2021
2022 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2023
2024 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2025
2026 16. Added comments and documentation about mis-use of no_arg in the C++
2027 wrapper.
2028
2029 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2030 from Martin Jerabek that uses macro names for all relevant character and
2031 string constants.
2032
2033 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2034 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2035 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2036 these, but not everybody uses configure.
2037
2038 19. A conditional group that had only one branch was not being correctly
2039 recognized as an item that could match an empty string. This meant that an
2040 enclosing group might also not be so recognized, causing infinite looping
2041 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2042 with the subject "ab", where knowledge that the repeated group can match
2043 nothing is needed in order to break the loop.
2044
2045 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2046 exec(), but without supplying a callout function, matching went wrong.
2047
2048 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2049 leak if the size of the offset vector was greater than 30. When the vector
2050 is smaller, the saved offsets during recursion go onto a local stack
2051 vector, but for larger vectors malloc() is used. It was failing to free
2052 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2053 error, in fact).
2054
2055 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2056 heapframe that is used only when UTF-8 support is enabled. This caused no
2057 problem, but was untidy.
2058
2059 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2060 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2061 included within another project.
2062
2063 24. Steven Van Ingelgem's patches to add more options to the CMake support,
2064 slightly modified by me:
2065
2066 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2067 not building pcregrep.
2068
2069 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2070 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2071
2072 25. Forward references, both numeric and by name, in patterns that made use of
2073 duplicate group numbers, could behave incorrectly or give incorrect errors,
2074 because when scanning forward to find the reference group, PCRE was not
2075 taking into account the duplicate group numbers. A pattern such as
2076 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2077
2078 26. Changed a few more instances of "const unsigned char *" to USPTR, making
2079 the feature of a custom pointer more persuasive (as requested by a user).
2080
2081 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2082 pcretest.c, inside #ifndefs, because it seems they are sometimes already
2083 pre-defined.
2084
2085 28. Added support for (*UTF8) at the start of a pattern.
2086
2087 29. Arrange for flags added by the "release type" setting in CMake to be shown
2088 in the configuration summary.
2089
2090
2091 Version 7.8 05-Sep-08
2092 ---------------------
2093
2094 1. Replaced UCP searching code with optimized version as implemented for Ad
2095 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2096 stage table and inline lookup instead of a function, giving speed ups of 2
2097 to 5 times on some simple patterns that I tested. Permission was given to
2098 distribute the MultiStage2.py script that generates the tables (it's not in
2099 the tarball, but is in the Subversion repository).
2100
2101 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2102 scripts.
2103
2104 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2105 a group with a zero qualifier. The result of the study could be incorrect,
2106 or the function might crash, depending on the pattern.
2107
2108 4. Caseless matching was not working for non-ASCII characters in back
2109 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2110 It now works when Unicode Property Support is available.
2111
2112 5. In pcretest, an escape such as \x{de} in the data was always generating
2113 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2114 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2115 truncation.
2116
2117 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2118
2119 7. Added two (int) casts to pcregrep when printing the difference of two
2120 pointers, in case they are 64-bit values.
2121
2122 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
2123 test 2 if it fails.
2124
2125 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2126 and a #define of that name to empty if it is not externally set. This is to
2127 allow users of MSVC to set it if necessary.
2128
2129 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2130 the convenience functions in the pcre_get.c source file.
2131
2132 11. An option change at the start of a pattern that had top-level alternatives
2133 could cause overwriting and/or a crash. This command provoked a crash in
2134 some environments:
2135
2136 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2137
2138 This potential security problem was recorded as CVE-2008-2371.
2139
2140 12. For a pattern where the match had to start at the beginning or immediately
2141 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2142 pcre_dfa_exec() could read past the end of the passed subject if there was
2143 no match. To help with detecting such bugs (e.g. with valgrind), I modified
2144 pcretest so that it places the subject at the end of its malloc-ed buffer.
2145
2146 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2147 exec() might read past the end of the data buffer in UTF-8 mode.
2148
2149 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2150 the data contained the byte 0x85 as part of a UTF-8 character within its
2151 first line. This applied both to normal and DFA matching.
2152
2153 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2154 /^[^d]*?$/8 failed to match "abc".
2155
2156 16. Added a missing copyright notice to pcrecpp_internal.h.
2157
2158 17. Make it more clear in the documentation that values returned from
2159 pcre_exec() in ovector are byte offsets, not character counts.
2160
2161 18. Tidied a few places to stop certain compilers from issuing warnings.
2162
2163 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2164 supplied by Stefan Weber. I made a further small update for 7.8 because
2165 there is a change of source arrangements: the pcre_searchfuncs.c module is
2166 replaced by pcre_ucd.c.
2167
2168
2169 Version 7.7 07-May-08
2170 ---------------------
2171
2172 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
2173 a string to a long long, pretend we don't even have a long long." This is
2174 done by checking for the strtoq, strtoll, and _strtoi64 functions.
2175
2176 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2177 pre-7.6 versions, which defined a global no_arg variable instead of putting
2178 it in the RE class. (See also #8 below.)
2179
2180 3. Remove a line of dead code, identified by coverity and reported by Nuno
2181 Lopes.
2182
2183 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
2184
2185 (1) The include/exclude patterns were being applied to the whole pathnames
2186 of files, instead of just to the final components.
2187
2188 (2) If there was more than one level of directory, the subdirectories were
2189 skipped unless they satisfied the include/exclude conditions. This is
2190 inconsistent with GNU grep (and could even be seen as contrary to the
2191 pcregrep specification - which I improved to make it absolutely clear).
2192 The action now is always to scan all levels of directory, and just
2193 apply the include/exclude patterns to regular files.
2194
2195 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2196 --exclude_dir in the tests to avoid scanning .svn directories.
2197
2198 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
2199 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2200 doesn't support NULs in patterns.
2201
2202 7. Added some missing "const"s to declarations of static tables in
2203 pcre_compile.c and pcre_dfa_exec.c.
2204
2205 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2206 caused by fix #2 above. (Subsequently also a second patch to fix the
2207 first patch. And a third patch - this was a messy problem.)
2208
2209 9. Applied Craig's patch to remove the use of push_back().
2210
2211 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2212 matching function regexec().
2213
2214 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2215 which, however, unlike Perl's \g{...}, are subroutine calls, not back
2216 references. PCRE supports relative numbers with this syntax (I don't think
2217 Oniguruma does).
2218
2219 12. Previously, a group with a zero repeat such as (...){0} was completely
2220 omitted from the compiled regex. However, this means that if the group
2221 was called as a subroutine from elsewhere in the pattern, things went wrong
2222 (an internal error was given). Such groups are now left in the compiled
2223 pattern, with a new opcode that causes them to be skipped at execution
2224 time.
2225
2226 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2227 to the way PCRE behaves:
2228
2229 (a) A lone ] character is dis-allowed (Perl treats it as data).
2230
2231 (b) A back reference to an unmatched subpattern matches an empty string
2232 (Perl fails the current match path).
2233
2234 (c) A data ] in a character class must be notated as \] because if the
2235 first data character in a class is ], it defines an empty class. (In
2236 Perl it is not possible to have an empty class.) The empty class []
2237 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2238 The negative empty class [^] matches any one character, independently
2239 of the DOTALL setting.
2240
2241 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2242 non-existent subpattern following a character class starting with ']' and
2243 containing () gave an internal compiling error instead of "reference to
2244 non-existent subpattern". Fortunately, when the pattern did exist, the
2245 compiled code was correct. (When scanning forwards to check for the
2246 existence of the subpattern, it was treating the data ']' as terminating
2247 the class, so got the count wrong. When actually compiling, the reference
2248 was subsequently set up correctly.)
2249
2250 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2251 it was being rejected as not supported by pcre_dfa_exec(), even though
2252 other assertions are supported. I have made pcre_dfa_exec() support
2253 (*FAIL).
2254
2255 16. The implementation of 13c above involved the invention of a new opcode,
2256 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2257 cannot be changed at match time, I realized I could make a small
2258 improvement to matching performance by compiling OP_ALLANY instead of
2259 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2260 on the OP_ANY path.
2261
2262 17. Compiling pcretest on Windows with readline support failed without the
2263 following two fixes: (1) Make the unistd.h include conditional on
2264 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2265
2266 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2267 ncurses library to be included for pcretest when ReadLine support is
2268 requested, but also to allow for it to be overridden. This patch came from
2269 Daniel Bergström.
2270
2271 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2272 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2273 any errors with the current Unicode tables. Thanks to Peter Kankowski for
2274 spotting this.
2275
2276
2277 Version 7.6 28-Jan-08
2278 ---------------------
2279
2280 1. A character class containing a very large number of characters with
2281 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2282 overflow.
2283
2284 2. Patch to cut out the "long long" test in pcrecpp_unittest when
2285 HAVE_LONG_LONG is not defined.
2286
2287 3. Applied Christian Ehrlicher's patch to update the CMake build files to
2288 bring them up to date and include new features. This patch includes:
2289
2290 - Fixed PH's badly added libz and libbz2 support.
2291 - Fixed a problem with static linking.
2292 - Added pcredemo. [But later removed - see 7 below.]
2293 - Fixed dftables problem and added an option.
2294 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2295 HAVE_LONG_LONG.
2296 - Added readline support for pcretest.
2297 - Added an listing of the option settings after cmake has run.
2298
2299 4. A user submitted a patch to Makefile that makes it easy to create
2300 "pcre.dll" under mingw when using Configure/Make. I added stuff to
2301 Makefile.am that cause it to include this special target, without
2302 affecting anything else. Note that the same mingw target plus all
2303 the other distribution libraries and programs are now supported
2304 when configuring with CMake (see 6 below) instead of with
2305 Configure/Make.
2306
2307 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2308 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2309 exported in the Windows port". It has not yet been confirmed that the patch
2310 solves the problem, but it does no harm.
2311
2312 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2313 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2314 with CMake, and also correct the comment about stack recursion.
2315
2316 7. Remove the automatic building of pcredemo from the ./configure system and
2317 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2318 of a program that users should build themselves after PCRE is installed, so
2319 building it automatically is not really right. What is more, it gave
2320 trouble in some build environments.
2321
2322 8. Further tidies to CMakeLists.txt from Sheri and Christian.
2323
2324
2325 Version 7.5 10-Jan-08
2326 ---------------------
2327
2328 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
2329 values in parens when parsing an RE using the C++ wrapper."
2330
2331 2. Negative specials like \S did not work in character classes in UTF-8 mode.
2332 Characters greater than 255 were excluded from the class instead of being
2333 included.
2334
2335 3. The same bug as (2) above applied to negated POSIX classes such as
2336 [:^space:].
2337
2338 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2339 defined or documented. It seems to have been a typo for PCRE_STATIC, so
2340 I have changed it.
2341
2342 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
2343 first named subpattern) and a construct such as (?&a) would reference the
2344 first named subpattern whose name started with "a" (in other words, the
2345 length check was missing). Both these problems are fixed. "Subpattern name
2346 expected" is now given for (?&) (a zero-length name), and this patch also
2347 makes it give the same error for \k'' (previously it complained that that
2348 was a reference to a non-existent subpattern).
2349
2350 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
2351 this is right because (?- can be followed by option settings as well as by
2352 digits. I have, however, made the messages clearer.
2353
2354 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2355 than the number used in the conditional) now cause a compile-time error.
2356 This is actually not compatible with Perl, which accepts such patterns, but
2357 treats the conditional as always being FALSE (as PCRE used to), but it
2358 seems to me that giving a diagnostic is better.
2359
2360 8. Change "alphameric" to the more common word "alphanumeric" in comments
2361 and messages.
2362
2363 9. Fix two occurrences of "backslash" in comments that should have been
2364 "backspace".
2365
2366 10. Remove two redundant lines of code that can never be obeyed (their function
2367 was moved elsewhere).
2368
2369 11. The program that makes PCRE's Unicode character property table had a bug
2370 which caused it to generate incorrect table entries for sequences of
2371 characters that have the same character type, but are in different scripts.
2372 It amalgamated them into a single range, with the script of the first of
2373 them. In other words, some characters were in the wrong script. There were
2374 thirteen such cases, affecting characters in the following ranges:
2375
2376 U+002b0 - U+002c1
2377 U+0060c - U+0060d
2378 U+0061e - U+00612
2379 U+0064b - U+0065e
2380 U+0074d - U+0076d
2381 U+01800 - U+01805
2382 U+01d00 - U+01d77
2383 U+01d9b - U+01dbf
2384 U+0200b - U+0200f
2385 U+030fc - U+030fe
2386 U+03260 - U+0327f
2387 U+0fb46 - U+0fbb1
2388 U+10450 - U+1049d
2389
2390 12. The -o option (show only the matching part of a line) for pcregrep was not
2391 compatible with GNU grep in that, if there was more than one match in a
2392 line, it showed only the first of them. It now behaves in the same way as
2393 GNU grep.
2394
2395 13. If the -o and -v options were combined for pcregrep, it printed a blank
2396 line for every non-matching line. GNU grep prints nothing, and pcregrep now
2397 does the same. The return code can be used to tell if there were any
2398 non-matching lines.
2399
2400 14. Added --file-offsets and --line-offsets to pcregrep.
2401
2402 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2403 infinitely looping recursion. The bug was that positive lookaheads were not
2404 being skipped when checking for a possible empty match (negative lookaheads
2405 and both kinds of lookbehind were skipped).
2406
2407 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2408 inclusion of <windows.h> to before rather than after the definition of
2409 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2410
2411 17. Specifying a possessive quantifier with a specific limit for a Unicode
2412 character property caused pcre_compile() to compile bad code, which led at
2413 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2414 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2415 caused the error; without that there was no problem.
2416
2417 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2418
2419 19. Added --enable-pcretest-libreadline.
2420
2421 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2422 RE::GlobalReplace(). As a result, the number of replacements returned was
2423 double what it should be. I removed one of the increments, but Craig sent a
2424 later patch that removed the other one (the right fix) and added unit tests
2425 that check the return values (which was not done before).
2426
2427 21. Several CMake things:
2428
2429 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2430 the names libpcre and libpcreposix, not just pcre and pcreposix.
2431
2432 (2) The above change means that pcretest and pcregrep are now correctly
2433 linked with the newly-built libraries, not previously installed ones.
2434
2435 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2436
2437 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2438 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2439 UTF-8 newline character). The key issue is that the pattern starts .*;
2440 this means that the match must be either at the beginning, or after a
2441 newline. The bug was in the code for advancing after a failed match and
2442 checking that the new position followed a newline. It was not taking
2443 account of UTF-8 characters correctly.
2444
2445 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2446 character classes. PCRE was not treating the sequence [:...:] as a
2447 character class unless the ... were all letters. Perl, however, seems to
2448 allow any characters between [: and :], though of course it rejects as
2449 unknown any "names" that contain non-letters, because all the known class
2450 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2451 for example, whereas PCRE did not - it did not recognize a POSIX character
2452 class. This seemed a bit dangerous, so the code has been changed to be
2453 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2454 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2455 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2456 Perl does, and where it didn't before.
2457
2458 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2459 Windows environments %n is disabled by default.
2460
2461
2462 Version 7.4 21-Sep-07
2463 ---------------------
2464
2465 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2466 means that a class such as [\s] counted as "explicit reference to CR or
2467 LF". That isn't really right - the whole point of the change was to try to
2468 help when there was an actual mention of one of the two characters. So now
2469 the change happens only if \r or \n (or a literal CR or LF) character is
2470 encountered.
2471
2472 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2473 of both had grown to the point where there were only 3 bits left.
2474 Fortunately, there was spare space in the data structure, and so I have
2475 moved the internal flags into a new 16-bit field to free up more option
2476 bits.
2477
2478 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2479 but did not set the internal JCHANGED flag - either of these is enough to
2480 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2481 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2482 start sets both bits.
2483
2484 4. Added options (at build time, compile time, exec time) to change \R from
2485 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2486
2487 5. doc/pcresyntax.html was missing from the distribution.
2488
2489 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2490 compatibility, even though it is no longer used.
2491
2492 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2493 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2494 windows.h file is present (where different names are used). [This was
2495 reversed later after testing - see 16 below.]
2496
2497 8. Changed all #include <config.h> to #include "config.h". There were also
2498 some further <pcre.h> cases that I changed to "pcre.h".
2499
2500 9. When pcregrep was used with the --colour option, it missed the line ending
2501 sequence off the lines that it output.
2502
2503 10. It was pointed out to me that arrays of string pointers cause lots of
2504 relocations when a shared library is dynamically loaded. A technique of
2505 using a single long string with a table of offsets can drastically reduce
2506 these. I have refactored PCRE in four places to do this. The result is
2507 dramatic:
2508
2509 Originally: 290
2510 After changing UCP table: 187
2511 After changing error message table: 43
2512 After changing table of "verbs" 36
2513 After changing table of Posix names 22
2514
2515 Thanks to the folks working on Gregex for glib for this insight.
2516
2517 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2518 unicode-properties was also set.
2519
2520 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2521
2522 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2523 checked only for CRLF.
2524
2525 14. Added casts to pcretest.c to avoid compiler warnings.
2526
2527 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2528
2529 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2530 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2531 entirely. This removes changes made in 7 above.
2532
2533 17. The CMake files have been updated, and there is now more information about
2534 building with CMake in the NON-UNIX-USE document.
2535
2536
2537 Version 7.3 28-Aug-07
2538 ---------------------
2539
2540 1. In the rejigging of the build system that eventually resulted in 7.1, the
2541 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2542 brackets there is not right, since it causes compilers to look for an
2543 installed pcre.h, not the version that is in the source that is being
2544 compiled (which of course may be different). I have changed it back to:
2545
2546 #include "pcre.h"
2547
2548 I have a vague recollection that the change was concerned with compiling in
2549 different directories, but in the new build system, that is taken care of
2550 by the VPATH setting the Makefile.
2551
2552 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2553 when the subject happened to end in the byte 0x85 (e.g. if the last
2554 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2555 characters but of course it shouldn't be taken as a newline when it is part
2556 of another character. The bug was that, for an unlimited repeat of . in
2557 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2558 characters when looking for a newline.
2559
2560 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2561
2562 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2563 in debug output.
2564
2565 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2566 long printing in the pcrecpp unittest when running under MinGW.
2567
2568 6. ESC_K was left out of the EBCDIC table.
2569
2570 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2571 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2572 limit also applies to "virtual nesting" when a pattern is recursive, and in
2573 this case 1000 isn't so big. I have been able to remove this limit at the
2574 expense of backing off one optimization in certain circumstances. Normally,
2575 when pcre_exec() would call its internal match() function recursively and
2576 immediately return the result unconditionally, it uses a "tail recursion"
2577 feature to save stack. However, when a subpattern that can match an empty
2578 string has an unlimited repetition quantifier, it no longer makes this
2579 optimization. That gives it a stack frame in which to save the data for
2580 checking that an empty string has been matched. Previously this was taken
2581 from the 1000-entry workspace that had been reserved. So now there is no
2582 explicit limit, but more stack is used.
2583
2584 8. Applied Daniel's patches to solve problems with the import/export magic
2585 syntax that is required for Windows, and which was going wrong for the
2586 pcreposix and pcrecpp parts of the library. These were overlooked when this
2587 problem was solved for the main library.
2588
2589 9. There were some crude static tests to avoid integer overflow when computing
2590 the size of patterns that contain repeated groups with explicit upper
2591 limits. As the maximum quantifier is 65535, the maximum group length was
2592 set at 30,000 so that the product of these two numbers did not overflow a
2593 32-bit integer. However, it turns out that people want to use groups that
2594 are longer than 30,000 bytes (though not repeat them that many times).
2595 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2596 made it possible to implement the integer overflow checks in a much more
2597 dynamic way, which I have now done. The artificial limitation on group
2598 length has been removed - we now have only the limit on the total length of
2599 the compiled pattern, which depends on the LINK_SIZE setting.
2600
2601 10. Fixed a bug in the documentation for get/copy named substring when
2602 duplicate names are permitted. If none of the named substrings are set, the
2603 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2604 empty string.
2605
2606 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2607 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2608 because the ] is interpreted as the first data character and the
2609 terminating ] is not found. PCRE has been made compatible with Perl in this
2610 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2611 cause memory overwriting.
2612
2613 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2614 string has been matched (to stop an infinite loop). It was not recognizing
2615 a conditional subpattern that could match an empty string if that
2616 subpattern was within another subpattern. For example, it looped when
2617 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2618 condition was not nested. This bug has been fixed.
2619
2620 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2621 past the start of the subject in the presence of bytes with the top bit
2622 set, for example "\x8aBCD".
2623
2624 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2625 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2626
2627 14. Optimized (?!) to (*FAIL).
2628
2629 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2630 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2631 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2632 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2633 does: it's just the validity check that is more restrictive.
2634
2635 16. Inserted checks for integer overflows during escape sequence (backslash)
2636 processing, and also fixed erroneous offset values for syntax errors during
2637 backslash processing.
2638
2639 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2640 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2641
2642 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2643 caused an overrun.
2644
2645 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2646 something other than just ASCII characters) inside a group that had an
2647 unlimited repeat caused a loop at compile time (while checking to see
2648 whether the group could match an empty string).
2649
2650 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2651 [\P{Any}] did so. (Error in the code for printing property names.)
2652
2653 21. An orphan \E inside a character class could cause a crash.
2654
2655 22. A repeated capturing bracket such as (A)? could cause a wild memory
2656 reference during compilation.
2657
2658 23. There are several functions in pcre_compile() that scan along a compiled
2659 expression for various reasons (e.g. to see if it's fixed length for look
2660 behind). There were bugs in these functions when a repeated \p or \P was
2661 present in the pattern. These operators have additional parameters compared
2662 with \d, etc, and these were not being taken into account when moving along
2663 the compiled data. Specifically:
2664
2665 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2666 length.
2667
2668 (b) An item such as \pL+ within a repeated group could cause crashes or
2669 loops.
2670
2671 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2672 "reference to non-existent subpattern" error.
2673
2674 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2675
2676 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2677 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2678
2679 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2680
2681 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2682 character were causing crashes (broken optimization).
2683
2684 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2685 \p or \P) caused a compile-time loop.
2686
2687 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2688 break. For example, the unstudied pattern [\r\n]A does not match the string
2689 "\r\nA" because change 7.0/46 below moves the current point on by two
2690 characters after failing to match at the start. However, the pattern \nA
2691 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2692 the same is true. There doesn't seem any very clean way out of this, but
2693 what I have chosen to do makes the common cases work: PCRE now takes note
2694 of whether there can be an explicit match for \r or \n anywhere in the
2695 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2696 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2697 pattern has explicit CR or LF references.
2698
2699 29. Added (*CR) etc for changing newline setting at start of pattern.
2700
2701
2702 Version 7.2 19-Jun-07
2703 ---------------------
2704
2705 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2706 which is apparently normally available under Windows.
2707
2708 2. Re-jig the pcregrep tests with different newline settings in an attempt
2709 to make them independent of the local environment's newline setting.
2710
2711 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2712
2713 4. Some of the "internals" tests were previously cut out when the link size
2714 was not 2, because the output contained actual offsets. The recent new
2715 "Z" feature of pcretest means that these can be cut out, making the tests
2716 usable with all link sizes.
2717
2718 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2719 stack recursion. This gives a massive performance boost under BSD, but just
2720 a small improvement under Linux. However, it saves one field in the frame
2721 in all cases.
2722
2723 6. Added more features from the forthcoming Perl 5.10:
2724
2725 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2726 recursion call. It refers to the nth most recently opened parentheses.
2727
2728 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2729 to be opened parentheses.
2730
2731 (c) Conditions that refer to capturing parentheses can be specified
2732 relatively, for example, (?(-2)... or (?(+3)...
2733
2734 (d) \K resets the start of the current match so that everything before
2735 is not part of it.
2736
2737 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2738
2739 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2740 reference syntax.
2741
2742 (g) (?| introduces a group in which the numbering of parentheses in each
2743 alternative starts with the same number.
2744
2745 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2746
2747 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2748 PCRE_INFO_JCHANGED.
2749
2750 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2751 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2752 for detecting groups that can match an empty string.
2753
2754 9. A pattern with a very large number of alternatives (more than several
2755 hundred) was running out of internal workspace during the pre-compile
2756 phase, where pcre_compile() figures out how much memory will be needed. A
2757 bit of new cunning has reduced the workspace needed for groups with
2758 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2759 workspace instead of running out of the 4096 that are available.
2760
2761 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2762
2763 11. Applied patch from Google to remove an optimization that didn't quite work.
2764 The report of the bug said:
2765
2766 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2767 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2768 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2769
2770 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2771 it matched the wrong number of bytes.
2772
2773
2774 Version 7.1 24-Apr-07
2775 ---------------------
2776
2777 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2778 that is more "standard", making use of automake and other Autotools. There
2779 is some re-arrangement of the files and adjustment of comments consequent
2780 on this.
2781
2782 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2783 for recursive directory scanning broke on some systems because the files
2784 are not scanned in any specific order and on different systems the order
2785 was different. A call to "sort" has been inserted into RunGrepTest for the
2786 approprate test as a short-term fix. In the longer term there may be an
2787 alternative.
2788
2789 3. I had an email from Eric Raymond about problems translating some of PCRE's
2790 man pages to HTML (despite the fact that I distribute HTML pages, some
2791 people do their own conversions for various reasons). The problems
2792 concerned the use of low-level troff macros .br and .in. I have therefore
2793 removed all such uses from the man pages (some were redundant, some could
2794 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2795 HTML has been updated to handle .nf/.fi and to complain if it encounters
2796 .br or .in.
2797
2798 4. Updated comments in configure.ac that get placed in config.h.in and also
2799 arranged for config.h to be included in the distribution, with the name
2800 config.h.generic, for the benefit of those who have to compile without
2801 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2802
2803 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2804 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2805 makevp.bat for the current PCRE, using the additional files
2806 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2807
2808 6. A Windows user reported a minor discrepancy with test 2, which turned out
2809 to be caused by a trailing space on an input line that had got lost in his
2810 copy. The trailing space was an accident, so I've just removed it.
2811
2812 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2813 that is needed.
2814
2815 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2816 as "const" (a) because they are and (b) because it helps the PHP
2817 maintainers who have recently made a script to detect big data structures
2818 in the php code that should be moved to the .rodata section. I remembered
2819 to update Builducptable as well, so it won't revert if ucptable.h is ever
2820 re-created.
2821
2822 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2823 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2824 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2825 support is not required. This saves 1.5-2K of code, which is important in
2826 some applications.
2827
2828 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2829 so as not to refer to the tables, even though these functions will never be
2830 called when UTF-8 support is disabled. Otherwise there are problems with a
2831 shared library.
2832
2833 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2834
2835 (a) It was defining its arguments as char * instead of void *.
2836
2837 (b) It was assuming that all moves were upwards in memory; this was true
2838 a long time ago when I wrote it, but is no longer the case.
2839
2840 The emulated memove() is provided for those environments that have neither
2841 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2842 is clearly not the case, as these two bugs were recently reported.
2843
2844 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2845 and Detrail to create the HTML documentation, the .txt form of the man
2846 pages, and it removes trailing spaces from listed files. It also creates
2847 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2848 case, it wraps all the #defines with #ifndefs. This script should be run
2849 before "make dist".
2850
2851 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2852 with Unicode property support.
2853
2854 (a) For a maximizing quantifier, if the two different cases of the
2855 character were of different lengths in their UTF-8 codings (there are
2856 some cases like this - I found 11), and the matching function had to
2857 back up over a mixture of the two cases, it incorrectly assumed they
2858 were both the same length.
2859
2860 (b) When PCRE was configured to use the heap rather than the stack for
2861 recursion during matching, it was not correctly preserving the data for
2862 the other case of a UTF-8 character when checking ahead for a match
2863 while processing a minimizing repeat. If the check also involved
2864 matching a wide character, but failed, corruption could cause an
2865 erroneous result when trying to check for a repeat of the original
2866 character.
2867
2868 13. Some tidying changes to the testing mechanism:
2869
2870 (a) The RunTest script now detects the internal link size and whether there
2871 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2872 values substituted by "configure". (The RunGrepTest script already did
2873 this for UTF-8.) The configure.ac script no longer substitutes the
2874 relevant variables.
2875
2876 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2877 with length and offset values. This means that the output is different
2878 for different internal link sizes. Test 2 is skipped for link sizes
2879 other than 2 because of this, bypassing the problem. Unfortunately,
2880 there was also a test in test 3 (the locale tests) that used /B and
2881 failed for link sizes other than 2. Rather than cut the whole test out,
2882 I have added a new /Z option to pcretest that replaces the length and
2883 offset values with spaces. This is now used to make test 3 independent
2884 of link size. (Test 2 will be tidied up later.)
2885
2886 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2887 segmentation fault instead of returning the appropriate error message.
2888
2889 15. In multiline mode when the newline sequence was set to "any", the pattern
2890 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2891 This doesn't seem right; it now treats the CRLF combination as the line
2892 ending, and so does not match in that case. It's only a pattern such as ^$
2893 that would hit this one: something like ^ABC$ would have failed after \r
2894 and then tried again after \r\n.
2895
2896 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2897 in an attempt to make files that differ only in their line terminators
2898 compare equal. This works on Linux.
2899
2900 17. Under certain error circumstances pcregrep might try to free random memory
2901 as it exited. This is now fixed, thanks to valgrind.
2902
2903 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2904 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2905 was because its rules for how to advance for /g after matching an empty
2906 string at the end of a line did not allow for this case. They now check for
2907 it specially.
2908
2909 20. pcretest is supposed to handle patterns and data of any length, by
2910 extending its buffers when necessary. It was getting this wrong when the
2911 buffer for a data line had to be extended.
2912
2913 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2914 CRLF as a newline sequence.
2915
2916 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2917 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2918 I have nevertheless tidied it up.
2919
2920 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2921
2922 24. Added a man page for pcre-config.
2923
2924
2925 Version 7.0 19-Dec-06
2926 ---------------------
2927
2928 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2929 moving to gcc 4.1.1.
2930
2931 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2932 sys/time.h, which is documented as needed for this function. It doesn't
2933 seem to matter on Linux, but it showed up on some releases of OS X.
2934
2935 3. It seems that there are systems where bytes whose values are greater than
2936 127 match isprint() in the "C" locale. The "C" locale should be the
2937 default when a C program starts up. In most systems, only ASCII printing
2938 characters match isprint(). This difference caused the output from pcretest
2939 to vary, making some of the tests fail. I have changed pcretest so that:
2940
2941 (a) When it is outputting text in the compiled version of a pattern, bytes
2942 other than 32-126 are always shown as hex escapes.
2943
2944 (b) When it is outputting text that is a matched part of a subject string,
2945 it does the same, unless a different locale has been set for the match
2946 (using the /L modifier). In this case, it uses isprint() to decide.
2947
2948 4. Fixed a major bug that caused incorrect computation of the amount of memory
2949 required for a compiled pattern when options that changed within the
2950 pattern affected the logic of the preliminary scan that determines the
2951 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2952 that the computed length was too small. The symptoms of this bug were
2953 either the PCRE error "internal error: code overflow" from pcre_compile(),
2954 or a glibc crash with a message such as "pcretest: free(): invalid next
2955 size (fast)". Examples of patterns that provoked this bug (shown in
2956 pcretest format) are:
2957
2958 /(?-x: )/x
2959 /(?x)(?-x: \s*#\s*)/
2960 /((?i)[\x{c0}])/8
2961 /(?i:[\x{c0}])/8
2962
2963 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2964 is now done differently.
2965
2966 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2967 wrapper classes; (b) implement a new function in the C++ scanner that is
2968 more efficient than the old way of doing things because it avoids levels of
2969 recursion in the regex matching; (c) add a paragraph to the documentation
2970 for the FullMatch() function.
2971
2972 6. The escape sequence \n was being treated as whatever was defined as
2973 "newline". Not only was this contrary to the documentation, which states
2974 that \n is character 10 (hex 0A), but it also went horribly wrong when
2975 "newline" was defined as CRLF. This has been fixed.
2976
2977 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2978 was being set to -1 for the "end of line" case (supposedly a value that no
2979 character can have). Though this value is never used (the check for end of
2980 line is "zero bytes in current character"), it caused compiler complaints.
2981 I've changed it to 0xffffffff.
2982
2983 8. In pcre_version.c, the version string was being built by a sequence of
2984 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2985 string (as it is for production releases) called a macro with an empty
2986 argument. The C standard says the result of this is undefined. The gcc
2987 compiler treats it as an empty string (which was what was wanted) but it is
2988 reported that Visual C gives an error. The source has been hacked around to
2989 avoid this problem.
2990
2991 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2992 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2993 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2994 of them did).
2995
2996 10. Originally, pcretest opened its input and output without "b"; then I was
2997 told that "b" was needed in some environments, so it was added for release
2998 5.0 to both the input and output. (It makes no difference on Unix-like
2999 systems.) Later I was told that it is wrong for the input on Windows. I've
3000 now abstracted the modes into two macros, to make it easier to fiddle with
3001 them, and removed "b" from the input mode under Windows.
3002
3003 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
3004
3005 12. Added -help and --help to pcretest as an official way of being reminded
3006 of the options.
3007
3008 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
3009 and pcrecpp.cc because they annoy compilers at high warning levels.
3010
3011 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
3012
3013 15. Fixed an occurrence of == in configure.ac that should have been = (shell
3014 scripts are not C programs :-) and which was not noticed because it works
3015 on Linux.
3016
3017 16. pcretest is supposed to handle any length of pattern and data line (as one
3018 line or as a continued sequence of lines) by extending its input buffer if
3019 necessary. This feature was broken for very long pattern lines, leading to
3020 a string of junk being passed to pcre_compile() if the pattern was longer
3021 than about 50K.
3022
3023 17. I have done a major re-factoring of the way pcre_compile() computes the
3024 amount of memory needed for a compiled pattern. Previously, there was code
3025 that made a preliminary scan of the pattern in order to do this. That was
3026 OK when PCRE was new, but as the facilities have expanded, it has become
3027 harder and harder to keep it in step with the real compile phase, and there
3028 have been a number of bugs (see for example, 4 above). I have now found a
3029 cunning way of running the real compile function in a "fake" mode that
3030 enables it to compute how much memory it would need, while actually only
3031 ever using a few hundred bytes of working memory and without too many
3032 tests of the mode. This should make future maintenance and development
3033 easier. A side effect of this work is that the limit of 200 on the nesting
3034 depth of parentheses has been removed (though this was never a serious
3035 limitation, I suspect). However, there is a downside: pcre_compile() now
3036 runs more slowly than before (30% or more, depending on the pattern). I
3037 hope this isn't a big issue. There is no effect on runtime performance.
3038
3039 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
3040 newline (only possible for the last line of a file) and it was a
3041 pattern that set a locale (followed by /Lsomething), pcretest crashed.
3042
3043 19. Added additional timing features to pcretest. (1) The -tm option now times
3044 matching only, not compiling. (2) Both -t and -tm can be followed, as a
3045 separate command line item, by a number that specifies the number of
3046 repeats to use when timing. The default is 50000; this gives better
3047 precision, but takes uncomfortably long for very large patterns.
3048
3049 20. Extended pcre_study() to be more clever in cases where a branch of a
3050 subpattern has no definite first character. For example, (a*|b*)[cd] would
3051 previously give no result from pcre_study(). Now it recognizes that the
3052 first character must be a, b, c, or d.
3053
3054 21. There was an incorrect error "recursive call could loop indefinitely" if
3055 a subpattern (or the entire pattern) that was being tested for matching an
3056 empty string contained only one non-empty item after a nested subpattern.
3057 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
3058 incorrectly, because the \d was being skipped in the check.
3059
3060 22. The pcretest program now has a new pattern option /B and a command line
3061 option -b, which is equivalent to adding /B to every pattern. This causes
3062 it to show the compiled bytecode, without the additional information that
3063 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
3064 is the same as /B/I).
3065
3066 23. A new optimization is now able automatically to treat some sequences such
3067 as a*b as a*+b. More specifically, if something simple (such as a character
3068 or a simple class like \d) has an unlimited quantifier, and is followed by
3069 something that cannot possibly match the quantified thing, the quantifier
3070 is automatically "possessified".
3071
3072 24. A recursive reference to a subpattern whose number was greater than 39
3073 went wrong under certain circumstances in UTF-8 mode. This bug could also
3074 have affected the operation of pcre_study().
3075
3076 25. Realized that a little bit of performance could be had by replacing
3077 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
3078
3079 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
3080
3081 27. Possessive quantifiers such as a++ were previously implemented by turning
3082 them into atomic groups such as ($>a+). Now they have their own opcodes,
3083 which improves performance. This includes the automatically created ones
3084 from 23 above.
3085
3086 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
3087 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
3088 the first matched character to be a colon. This applied both to named and
3089 numbered groups.
3090
3091 29. The ucpinternal.h header file was missing its idempotency #ifdef.
3092
3093 30. I was sent a "project" file called libpcre.a.dev which I understand makes
3094 building PCRE on Windows easier, so I have included it in the distribution.
3095
3096 31. There is now a check in pcretest against a ridiculously large number being
3097 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
3098 loop, the loop is abandoned.
3099
3100 32. Forward references to subpatterns in conditions such as (?(2)...) where
3101 subpattern 2 is defined later cause pcre_compile() to search forwards in
3102 the pattern for the relevant set of parentheses. This search went wrong
3103 when there were unescaped parentheses in a character class, parentheses
3104 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
3105
3106 33. "Subroutine" calls and backreferences were previously restricted to
3107 referencing subpatterns earlier in the regex. This restriction has now
3108 been removed.
3109
3110 34. Added a number of extra features that are going to be in Perl 5.10. On the
3111 whole, these are just syntactic alternatives for features that PCRE had
3112 previously implemented using the Python syntax or my own invention. The
3113 other formats are all retained for compatibility.
3114
3115 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
3116 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
3117 also .NET compatible.
3118
3119 (b) A recursion or subroutine call to a named group can now be defined as
3120 (?&name) as well as (?P>name).
3121
3122 (c) A backreference to a named group can now be defined as \k<name> or
3123 \k'name' as well as (?P=name). The new forms, as well as being in Perl
3124 5.10, are also .NET compatible.
3125
3126 (d) A conditional reference to a named group can now use the syntax
3127 (?(<name>) or (?('name') as well as (?(name).
3128
3129 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
3130 groups (named and numbered) that are never evaluated inline, but can be
3131 called as "subroutines" from elsewhere. In effect, the DEFINE condition
3132 is always false. There may be only one alternative in such a group.
3133
3134 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
3135 as the simple (?(R). The condition is true only if the most recent
3136 recursion is that of the given number or name. It does not search out
3137 through the entire recursion stack.
3138
3139 (g) The escape \gN or \g{N} has been added, where N is a positive or
3140 negative number, specifying an absolute or relative reference.
3141
3142 35. Tidied to get rid of some further signed/unsigned compiler warnings and
3143 some "unreachable code" warnings.
3144
3145 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
3146 things, this adds five new scripts.
3147
3148 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
3149 There were also incompatibilities regarding the handling of \Q..\E inside
3150 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
3151 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
3152
3153 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
3154 matches an empty string, and forcibly breaks the loop. There were bugs in
3155 this code in non-simple cases. For a pattern such as ^(a()*)* matched
3156 against aaaa the result was just "a" rather than "aaaa", for example. Two
3157 separate and independent bugs (that affected different cases) have been
3158 fixed.
3159
3160 39. Refactored the code to abolish the use of different opcodes for small
3161 capturing bracket numbers. This is a tidy that I avoided doing when I
3162 removed the limit on the number of capturing brackets for 3.5 back in 2001.
3163 The new approach is not only tidier, it makes it possible to reduce the
3164 memory needed to fix the previous bug (38).
3165
3166 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3167 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3168 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3169 mode.
3170
3171 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3172 report.
3173
3174 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3175 copy construction and assignment in the C++ wrapper.
3176
3177 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3178 couple of bugs that could have given wrong results in the "--newline=crlf"
3179 case.
3180
3181 44. Added a number of casts and did some reorganization of signed/unsigned int
3182 variables following suggestions from Dair Grant. Also renamed the variable
3183 "this" as "item" because it is a C++ keyword.
3184
3185 45. Arranged for dftables to add
3186
3187 #include "pcre_internal.h"
3188
3189 to pcre_chartables.c because without it, gcc 4.x may remove the array
3190 definition from the final binary if PCRE is built into a static library and
3191 dead code stripping is activated.
3192
3193 46. For an unanchored pattern, if a match attempt fails at the start of a
3194 newline sequence, and the newline setting is CRLF or ANY, and the next two
3195 characters are CRLF, advance by two characters instead of one.
3196
3197
3198 Version 6.7 04-Jul-06
3199 ---------------------
3200
3201 1. In order to handle tests when input lines are enormously long, pcretest has
3202 been re-factored so that it automatically extends its buffers when
3203 necessary. The code is crude, but this _is_ just a test program. The
3204 default size has been increased from 32K to 50K.
3205
3206 2. The code in pcre_study() was using the value of the re argument before
3207 testing it for NULL. (Of course, in any sensible call of the function, it
3208 won't be NULL.)
3209
3210 3. The memmove() emulation function in pcre_internal.h, which is used on
3211 systems that lack both memmove() and bcopy() - that is, hardly ever -
3212 was missing a "static" storage class specifier.
3213
3214 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3215 containing an extended class (one that cannot be represented by a bitmap
3216 because it contains high-valued characters or Unicode property items, e.g.
3217 [\pZ]). Almost always one would set UTF-8 mode when processing such a
3218 pattern, but PCRE should not loop if you do not (it no longer does).
3219 [Detail: two cases were found: (a) a repeated subpattern containing an
3220 extended class; (b) a recursive reference to a subpattern that followed a
3221 previous extended class. It wasn't skipping over the extended class
3222 correctly when UTF-8 mode was not set.]
3223
3224 5. A negated single-character class was not being recognized as fixed-length
3225 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3226 compile error "lookbehind assertion is not fixed length".
3227
3228 6. The RunPerlTest auxiliary script was showing an unexpected difference
3229 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3230 write a Perl script that can interpret lines of an input file either as
3231 byte characters or as UTF-8, which is what "perltest" was being required to
3232 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3233 can't do is switch easily at run time between having the "use utf8;" pragma
3234 or not. In the end, I fudged it by using the RunPerlTest script to insert
3235 "use utf8;" explicitly for the UTF-8 tests.
3236
3237 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3238 the end of the subject string, contrary to the documentation and to what
3239 Perl does. This was true of both matching functions. Now it matches only at
3240 the start of the subject and immediately after *internal* newlines.
3241
3242 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3243 a pointer to an int instead of a pointer to an unsigned long int. This
3244 caused problems on 64-bit systems.
3245
3246 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3247 instance of the 'standard' template library not being so standard".
3248
3249 10. There was no check on the number of named subpatterns nor the maximum
3250 length of a subpattern name. The product of these values is used to compute
3251 the size of the memory block for a compiled pattern. By supplying a very
3252 long subpattern name and a large number of named subpatterns, the size
3253 computation could be caused to overflow. This is now prevented by limiting
3254 the length of names to 32 characters, and the number of named subpatterns
3255 to 10,000.
3256
3257 11. Subpatterns that are repeated with specific counts have to be replicated in
3258 the compiled pattern. The size of memory for this was computed from the
3259 length of the subpattern and the repeat count. The latter is limited to
3260 65535, but there was no limit on the former, meaning that integer overflow
3261 could in principle occur. The compiled length of a repeated subpattern is
3262 now limited to 30,000 bytes in order to prevent this.
3263
3264 12. Added the optional facility to have named substrings with the same name.
3265
3266 13. Added the ability to use a named substring as a condition, using the
3267 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3268 are numbers (not recommended). Forward references are permitted.
3269
3270 14. Added forward references in named backreferences (if you see what I mean).
3271
3272 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3273 pattern could run off the end of the subject. For example, the pattern
3274 "(?s)(.{1,5})"8 did this with the subject "ab".
3275
3276 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3277 PCRE_CASELESS was set when matching characters that were quantified with ?
3278 or *.
3279
3280 17. A character class other than a single negated character that had a minimum
3281 but no maximum quantifier - for example [ab]{6,} - was not handled
3282 correctly by pce_dfa_exec(). It would match only one character.
3283
3284 18. A valid (though odd) pattern that looked like a POSIX character
3285 class but used an invalid character after [ (for example [[,abc,]]) caused
3286 pcre_compile() to give the error "Failed: internal error: code overflow" or
3287 in some cases to crash with a glibc free() error. This could even happen if
3288 the pattern terminated after [[ but there just happened to be a sequence of
3289 letters, a binary zero, and a closing ] in the memory that followed.
3290
3291 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3292 over the years. Originally (before any Unicode support), just the bottom 8
3293 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3294 output from "man perlunicode" includes this:
3295
3296 The regular expression compiler produces polymorphic opcodes. That
3297 is, the pattern adapts to the data and automatically switches to
3298 the Unicode character scheme when presented with Unicode data--or
3299 instead uses a traditional byte scheme when presented with byte
3300 data.
3301
3302 Sadly, a wide octal escape does not cause a switch, and in a string with
3303 no other multibyte characters, these octal escapes are treated as before.
3304 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
3305 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3306 Unicode string.
3307
3308 I have not perpetrated such confusion in PCRE. Up till now, it took just
3309 the bottom 8 bits, as in old Perl. I have now made octal escapes with
3310 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3311 translate to the appropriate multibyte character.
3312
3313 29. Applied some refactoring to reduce the number of warnings from Microsoft
3314 and Borland compilers. This has included removing the fudge introduced
3315 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3316 a warning about an unused variable.
3317
3318 21. PCRE has not included VT (character 0x0b) in the set of whitespace
3319 characters since release 4.0, because Perl (from release 5.004) does not.
3320 [Or at least, is documented not to: some releases seem to be in conflict
3321 with the documentation.] However, when a pattern was studied with
3322 pcre_study() and all its branches started with \s, PCRE still included VT
3323 as a possible starting character. Of course, this did no harm; it just
3324 caused an unnecessary match attempt.
3325
3326 22. Removed a now-redundant internal flag bit that recorded the fact that case
3327 dependency changed within the pattern. This was once needed for "required
3328 byte" processing, but is no longer used. This recovers a now-scarce options
3329 bit. Also moved the least significant internal flag bit to the most-
3330 significant bit of the word, which was not previously used (hangover from
3331 the days when it was an int rather than a uint) to free up another bit for
3332 the future.
3333
3334 23. Added support for CRLF line endings as well as CR and LF. As well as the
3335 default being selectable at build time, it can now be changed at runtime
3336 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3337 specify that it is scanning data with non-default line endings.
3338
3339 24. Changed the definition of CXXLINK to make it agree with the definition of
3340 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3341
3342 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3343 recursions. This makes a big different to stack usage for some patterns.
3344
3345 26. If a subpattern containing a named recursion or subroutine reference such
3346 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3347 the space required for the compiled pattern went wrong and gave too small a
3348 value. Depending on the environment, this could lead to "Failed: internal
3349 error: code overflow at offset 49" or "glibc detected double free or
3350 corruption" errors.
3351
3352 27. Applied patches from Google (a) to support the new newline modes and (b) to
3353 advance over multibyte UTF-8 characters in GlobalReplace.
3354
3355 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3356 difference for some implementation of PCRE in some Windows version.
3357
3358 29. Added some extra testing facilities to pcretest:
3359
3360 \q<number> in a data line sets the "match limit" value
3361 \Q<number> in a data line sets the "match recursion limt" value
3362 -S <number> sets the stack size, where <number> is in megabytes
3363
3364 The -S option isn't available for Windows.
3365
3366
3367 Version 6.6 06-Feb-06
3368 ---------------------
3369
3370 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3371 in pcreposix.h. I have copied the definition from pcre.h.
3372
3373 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3374 because pcre.h is no longer a built file.
3375
3376 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3377 not normally included in the compiled code.
3378
3379
3380 Version 6.5 01-Feb-06
3381 ---------------------
3382
3383 1. When using the partial match feature with pcre_dfa_exec(), it was not
3384 anchoring the second and subsequent partial matches at the new starting
3385 point. This could lead to incorrect results. For example, with the pattern
3386 /1234/, partially matching against "123" and then "a4" gave a match.
3387
3388 2. Changes to pcregrep:
3389
3390 (a) All non-match returns from pcre_exec() were being treated as failures
3391 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3392 error message is output. Some extra information is given for the
3393 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3394 probably the only errors that are likely to be caused by users (by
3395 specifying a regex that has nested indefinite repeats, for instance).
3396 If there are more than 20 of these errors, pcregrep is abandoned.
3397
3398 (b) A binary zero was treated as data while matching, but terminated the
3399 output line if it was written out. This has been fixed: binary zeroes
3400 are now no different to any other data bytes.
3401
3402 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3403 used to set a locale for matching. The --locale=xxxx long option has
3404 been added (no short equivalent) to specify a locale explicitly on the
3405 pcregrep command, overriding the environment variables.
3406
3407 (d) When -B was used with -n, some line numbers in the output were one less
3408 than they should have been.
3409
3410 (e) Added the -o (--only-matching) option.
3411
3412 (f) If -A or -C was used with -c (count only), some lines of context were
3413 accidentally printed for the final match.
3414
3415 (g) Added the -H (--with-filename) option.
3416
3417 (h) The combination of options -rh failed to suppress file names for files
3418 that were found from directory arguments.
3419
3420 (i) Added the -D (--devices) and -d (--directories) options.
3421
3422 (j) Added the -F (--fixed-strings) option.
3423
3424 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3425
3426 (l) Added the --colo(u)r option.
3427
3428 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3429 is not present by default.
3430
3431 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3432 items such as (?R) or (?1), when the recursion could match a number of
3433 alternatives. If it matched one of the alternatives, but subsequently,
3434 outside the recursion, there was a failure, the code tried to back up into
3435 the recursion. However, because of the way PCRE is implemented, this is not
3436 possible, and the result was an incorrect result from the match.
3437
3438 In order to prevent this happening, the specification of recursion has
3439 been changed so that all such subpatterns are automatically treated as
3440 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3441
3442 4. I had overlooked the fact that, in some locales, there are characters for
3443 which isalpha() is true but neither isupper() nor islower() are true. In
3444 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3445 and ordfeminine) are like this. This affected the treatment of \w and \W
3446 when they appeared in character classes, but not when they appeared outside
3447 a character class. The bit map for "word" characters is now created
3448 separately from the results of isalnum() instead of just taking it from the
3449 upper, lower, and digit maps. (Plus the underscore character, of course.)
3450
3451 5. The above bug also affected the handling of POSIX character classes such as
3452 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3453 permanent tables. Instead, the bit maps for such a class were previously
3454 created as the appropriate unions of the upper, lower, and digit bitmaps.
3455 Now they are created by subtraction from the [[:word:]] class, which has
3456 its own bitmap.
3457
3458 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3459 It is created by subtracting the vertical space characters (\x09, \x0a,
3460 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3461 subtraction was done in the overall bitmap for a character class, meaning
3462 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3463 be recognized. This bug has been fixed.
3464
3465 7. Patches from the folks at Google:
3466
3467 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3468 real life, but is still worth protecting against".
3469
3470 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3471 regular expressions".
3472
3473 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3474 have it.
3475
3476 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3477 "configure" and the latter not, in order to fix a problem somebody had
3478 with compiling the Arg class on HP-UX.
3479
3480 (e) Improve the error-handling of the C++ wrapper a little bit.
3481
3482 (f) New tests for checking recursion limiting.
3483
3484 8. The pcre_memmove() function, which is used only if the environment does not
3485 have a standard memmove() function (and is therefore rarely compiled),
3486 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3487 returning a result (though PCRE never actually uses the result).
3488
3489 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3490 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3491 returned instead of calling malloc() with an overflowing number that would
3492 most likely cause subsequent chaos.
3493
3494 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3495
3496 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3497 with this option is matched, the nmatch and pmatch options of regexec() are
3498 ignored.
3499
3500 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3501 provided in case anyone wants to the the POSIX interface with UTF-8
3502 strings.
3503
3504 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3505 C++ linking (needed for some HP-UX environments).
3506
3507 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3508 (unused parameter) and in the pcre_printint() function (omitted "default"
3509 switch label when the default is to do nothing).
3510
3511 15. Added some code to make it possible, when PCRE is compiled as a C++
3512 library, to replace subject pointers for pcre_exec() with a smart pointer
3513 class, thus making it possible to process discontinuous strings.
3514
3515 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3516 much the same function. They were added by different people who were trying
3517 to make PCRE easy to compile on non-Unix systems. It has been suggested
3518 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3519 for compiling on Windows systems. I have therefore replaced it with
3520 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3521 defaults to "extern" for C or "extern C" for C++, which works fine on
3522 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3523 SCOPE with something explicit in config.h. In addition:
3524
3525 (a) pcreposix.h still had just "extern" instead of either of these macros;
3526 I have replaced it with PCRE_DATA_SCOPE.
3527
3528 (b) Functions such as _pcre_xclass(), which are internal to the library,
3529 but external in the C sense, all had PCRE_EXPORT in their definitions.
3530 This is apparently wrong for the Windows case, so I have removed it.
3531 (It makes no difference on Unix-like systems.)
3532
3533 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3534 of recursive calls to match(). This is different to MATCH_LIMIT because
3535 that limits the total number of calls to match(), not all of which increase
3536 the depth of recursion. Limiting the recursion depth limits the amount of
3537 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3538 when PCRE is compiled, and changed at run time. A patch from Google adds
3539 this functionality to the C++ interface.
3540
3541 18. Changes to the handling of Unicode character properties:
3542
3543 (a) Updated the table to Unicode 4.1.0.
3544
3545 (b) Recognize characters that are not in the table as "Cn" (undefined).
3546
3547 (c) I revised the way the table is implemented to a much improved format
3548 which includes recognition of ranges. It now supports the ranges that
3549 are defined in UnicodeData.txt, and it also amalgamates other
3550 characters into ranges. This has reduced the number of entries in the
3551 table from around 16,000 to around 3,000, thus reducing its size
3552 considerably. I realized I did not need to use a tree structure after
3553 all - a binary chop search is just as efficient. Having reduced the
3554 number of entries, I extended their size from 6 bytes to 8 bytes to
3555 allow for more data.
3556
3557 (d) Added support for Unicode script names via properties such as \p{Han}.
3558
3559 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3560 matching that character.
3561
3562 20. When matching a repeated Unicode property with a minimum greater than zero,
3563 (for example \pL{2,}), PCRE could look past the end of the subject if it
3564 reached it while seeking the minimum number of characters. This could
3565 happen only if some of the characters were more than one byte long, because
3566 there is a check for at least the minimum number of bytes.
3567
3568 21. Refactored the implementation of \p and \P so as to be more general, to
3569 allow for more different types of property in future. This has changed the
3570 compiled form incompatibly. Anybody with saved compiled patterns that use
3571 \p or \P will have to recompile them.
3572
3573 22. Added "Any" and "L&" to the supported property types.
3574
3575 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3576 but give a compile time error if the value is greater than 0xff.
3577
3578 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3579 accidentally not being installed or uninstalled.
3580
3581 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3582 made were to insert the current release number. This seemed silly, because
3583 it made things harder for people building PCRE on systems that don't run
3584 "configure". I have turned pcre.h into a distributed file, no longer built
3585 by "configure", with the version identification directly included. There is
3586 no longer a pcre.h.in file.
3587
3588 However, this change necessitated a change to the pcre-config script as
3589 well. It is built from pcre-config.in, and one of the substitutions was the
3590 release number. I have updated configure.ac so that ./configure now finds
3591 the release number by grepping pcre.h.
3592
3593 26. Added the ability to run the tests under valgrind.
3594
3595
3596 Version 6.4 05-Sep-05
3597 ---------------------
3598
3599 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3600 "--" to be printed when multiple files were scanned, even when none of the
3601 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3602 consider it to be a bug, and have restored the previous behaviour.
3603
3604 2. A couple of code tidies to get rid of compiler warnings.
3605
3606 3. The pcretest program used to cheat by referring to symbols in the library
3607 whose names begin with _pcre_. These are internal symbols that are not
3608 really supposed to be visible externally, and in some environments it is
3609 possible to suppress them. The cheating is now confined to including
3610 certain files from the library's source, which is a bit cleaner.
3611
3612 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3613 file's purpose clearer.
3614
3615 5. Reorganized pcre_ucp_findchar().
3616
3617
3618 Version 6.3 15-Aug-05
3619 ---------------------
3620
3621 1. The file libpcre.pc.in did not have general read permission in the tarball.
3622
3623 2. There were some problems when building without C++ support:
3624
3625 (a) If C++ support was not built, "make install" and "make test" still
3626 tried to test it.
3627
3628 (b) There were problems when the value of CXX was explicitly set. Some
3629 changes have been made to try to fix these, and ...
3630
3631 (c) --disable-cpp can now be used to explicitly disable C++ support.
3632
3633 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3634 backslash in a target when C++ was disabled. This confuses some
3635 versions of "make", apparently. Using an intermediate variable solves
3636 this. (Same for CPP_LOBJ.)
3637
3638 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3639 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3640 necessary on certain architectures.
3641
3642 4. Added a setting of -export-symbols-regex to the link command to remove
3643 those symbols that are exported in the C sense, but actually are local
3644 within the library, and not documented. Their names all begin with
3645 "_pcre_". This is not a perfect job, because (a) we have to except some
3646 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3647 available (and never for static libraries). I have made a note to try to
3648 find a way round (a) in the future.
3649
3650
3651 Version 6.2 01-Aug-05
3652 ---------------------
3653
3654 1. There was no test for integer overflow of quantifier values. A construction
3655 such as {1111111111111111} would give undefined results. What is worse, if
3656 a minimum quantifier for a parenthesized subpattern overflowed and became
3657 negative, the calculation of the memory size went wrong. This could have
3658 led to memory overwriting.
3659
3660 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3661
3662 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3663 operating environments where this matters.
3664
3665 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3666 PCRE options from within the C++ wrapper.
3667
3668 5. Named capturing subpatterns were not being correctly counted when a pattern
3669 was compiled. This caused two problems: (a) If there were more than 100
3670 such subpatterns, the calculation of the memory needed for the whole
3671 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3672 back references of the form \12, where the number was greater than 9, were
3673 not recognized as back references, even though there were sufficient
3674 previous subpatterns.
3675
3676 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3677 versions of gcc, e.g. 2.95.4.
3678
3679
3680 Version 6.1 21-Jun-05
3681 ---------------------
3682
3683 1. There was one reference to the variable "posix" in pcretest.c that was not
3684 surrounded by "#if !defined NOPOSIX".
3685
3686 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3687 the cross-check on the old pcre_info() function, for the benefit of the
3688 cut-down version of PCRE that is currently imported into Exim.
3689
3690 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3691 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3692 patterns that start (?i) but I don't think that matters. The use of (?i) is
3693 just an example; this all applies to the other options as well.
3694
3695 4. Since libtool seems to echo the compile commands it is issuing, the output
3696 from "make" can be reduced a bit by putting "@" in front of each libtool
3697 compile command.
3698
3699 5. Patch from the folks at Google for configure.in to be a bit more thorough
3700 in checking for a suitable C++ installation before trying to compile the
3701 C++ stuff. This should fix a reported problem when a compiler was present,
3702 but no suitable headers.
3703
3704 6. The man pages all had just "PCRE" as their title. I have changed them to
3705 be the relevant file name. I have also arranged that these names are
3706 retained in the file doc/pcre.txt, which is a concatenation in text format
3707 of all the man pages except the little individual ones for each function.
3708
3709 7. The NON-UNIX-USE file had not been updated for the different set of source
3710 files that come with release 6. I also added a few comments about the C++
3711 wrapper.
3712
3713
3714 Version 6.0 07-Jun-05
3715 ---------------------
3716
3717 1. Some minor internal re-organization to help with my DFA experiments.
3718
3719 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3720 didn't matter for the library itself when fully configured, but did matter
3721 when compiling without UCP support, or within Exim, where the ucp files are
3722 not imported.
3723
3724 3. Refactoring of the library code to split up the various functions into
3725 different source modules. The addition of the new DFA matching code (see
3726 below) to a single monolithic source would have made it really too
3727 unwieldy, quite apart from causing all the code to be include in a
3728 statically linked application, when only some functions are used. This is
3729 relevant even without the DFA addition now that patterns can be compiled in
3730 one application and matched in another.
3731
3732 The downside of splitting up is that there have to be some external
3733 functions and data tables that are used internally in different modules of
3734 the library but which are not part of the API. These have all had their
3735 names changed to start with "_pcre_" so that they are unlikely to clash
3736 with other external names.
3737
3738 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3739 a different (DFA) algorithm. Although it is slower than the original
3740 function, it does have some advantages for certain types of matching
3741 problem.
3742
3743 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3744 including restarting after a partial match.
3745
3746 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3747 defined when compiling for Windows was sent to me. I have put it into the
3748 code, though I have no means of testing or verifying it.
3749
3750 7. Added the pcre_refcount() auxiliary function.
3751
3752 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3753 match before or at the first newline in the subject string. In pcretest,
3754 the /f option on a pattern can be used to set this.
3755
3756 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3757 would behave wrongly. This has been present in PCRE since release 4.0.
3758
3759 10. A number of changes to the pcregrep command:
3760
3761 (a) Refactored how -x works; insert ^(...)$ instead of setting
3762 PCRE_ANCHORED and checking the length, in preparation for adding
3763 something similar for -w.
3764
3765 (b) Added the -w (match as a word) option.
3766
3767 (c) Refactored the way lines are read and buffered so as to have more
3768 than one at a time available.
3769
3770 (d) Implemented a pcregrep test script.
3771
3772 (e) Added the -M (multiline match) option. This allows patterns to match
3773 over several lines of the subject. The buffering ensures that at least
3774 8K, or the rest of the document (whichever is the shorter) is available
3775 for matching (and similarly the previous 8K for lookbehind assertions).
3776
3777 (f) Changed the --help output so that it now says
3778
3779 -w, --word-regex(p)
3780
3781 instead of two lines, one with "regex" and the other with "regexp"
3782 because that confused at least one person since the short forms are the
3783 same. (This required a bit of code, as the output is generated
3784 automatically from a table. It wasn't just a text change.)
3785
3786 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3787 option but starts with a hyphen. Could be a pattern or a path name
3788 starting with a hyphen, for instance.
3789
3790 (h) "-" can be given as a file name to represent stdin.
3791
3792 (i) When file names are being printed, "(standard input)" is used for
3793 the standard input, for compatibility with GNU grep. Previously
3794 "<stdin>" was used.
3795
3796 (j) The option --label=xxx can be used to supply a name to be used for
3797 stdin when file names are being printed. There is no short form.
3798
3799 (k) Re-factored the options decoding logic because we are going to add
3800 two more options that take data. Such options can now be given in four
3801 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3802
3803 (l) Added the -A, -B, and -C options for requesting that lines of context
3804 around matches be printed.
3805
3806 (m) Added the -L option to print the names of files that do not contain
3807 any matching lines, that is, the complement of -l.
3808
3809 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3810 continue to scan other files.
3811
3812 (o) The -s option was incorrectly implemented. For compatibility with other
3813 greps, it now suppresses the error message for a non-existent or non-
3814 accessible file (but not the return code). There is a new option called
3815 -q that suppresses the output of matching lines, which was what -s was
3816 previously doing.
3817
3818 (p) Added --include and --exclude options to specify files for inclusion
3819 and exclusion when recursing.
3820
3821 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3822 Hopefully, it now does.
3823
3824 12. Missing cast in pcre_study().
3825
3826 13. Added an "uninstall" target to the makefile.
3827
3828 14. Replaced "extern" in the function prototypes in Makefile.in with
3829 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3830 world, but is set differently for Windows.
3831
3832 15. Added a second compiling function called pcre_compile2(). The only
3833 difference is that it has an extra argument, which is a pointer to an
3834 integer error code. When there is a compile-time failure, this is set
3835 non-zero, in addition to the error test pointer being set to point to an
3836 error message. The new argument may be NULL if no error number is required
3837 (but then you may as well call pcre_compile(), which is now just a
3838 wrapper). This facility is provided because some applications need a
3839 numeric error indication, but it has also enabled me to tidy up the way
3840 compile-time errors are handled in the POSIX wrapper.
3841
3842 16. Added VPATH=.libs to the makefile; this should help when building with one
3843 prefix path and installing with another. (Or so I'm told by someone who
3844 knows more about this stuff than I do.)
3845
3846 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3847 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3848 match everything, including newlines. This is not POSIX-compatible, but
3849 somebody wanted the feature. From pcretest it can be activated by using
3850 both the P and the s flags.
3851
3852 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3853
3854 19. libpcre.pc was being incorrectly installed as executable.
3855
3856 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3857 it now also looks for '\r' so that it will work unmodified on Windows.
3858
3859 21. Added Google's contributed C++ wrapper to the distribution.
3860
3861 22. Added some untidy missing memory free() calls in pcretest, to keep
3862 Electric Fence happy when testing.
3863
3864
3865
3866 Version 5.0 13-Sep-04
3867 ---------------------
3868
3869 1. Internal change: literal characters are no longer packed up into items
3870 containing multiple characters in a single byte-string. Each character
3871 is now matched using a separate opcode. However, there may be more than one
3872 byte in the character in UTF-8 mode.
3873
3874 2. The pcre_callout_block structure has two new fields: pattern_position and
3875 next_item_length. These contain the offset in the pattern to the next match
3876 item, and its length, respectively.
3877
3878 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3879 insertion of callouts before each pattern item. Added the /C option to
3880 pcretest to make use of this.
3881
3882 4. On the advice of a Windows user, the lines
3883
3884 #if defined(_WIN32) || defined(WIN32)
3885 _setmode( _fileno( stdout ), 0x8000 );
3886 #endif /* defined(_WIN32) || defined(WIN32) */
3887
3888 have been added to the source of pcretest. This apparently does useful
3889 magic in relation to line terminators.
3890
3891 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3892 for the benefit of those environments where the "b" makes a difference.
3893
3894 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3895 to know about it. I have put a hack into configure.in that adds in code
3896 to set GCC=yes if CC=icc. This seems to end up at a point in the
3897 generated configure script that is early enough to affect the setting of
3898 compiler options, which is what is needed, but I have no means of testing
3899 whether it really works. (The user who reported this had patched the
3900 generated configure script, which of course I cannot do.)
3901
3902 LATER: After change 22 below (new libtool files), the configure script
3903 seems to know about icc (and also ecc). Therefore, I have commented out
3904 this hack in configure.in.
3905
3906 7. Added support for pkg-config (2 patches were sent in).
3907
3908 8. Negated POSIX character classes that used a combination of internal tables
3909 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3910 [[:^ascii]]. Typically, they would match almost any characters. The other
3911 POSIX classes were not broken in this way.
3912
3913 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3914 to find the match, as PCRE was deluded into thinking that the match had to
3915 start at the start point or following a newline. The same bug applied to
3916 patterns with negative forward assertions or any backward assertions
3917 preceding ".*" at the start, unless the pattern required a fixed first
3918 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3919
3920 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3921 starting at the last subject character, bytes beyond the end of the subject
3922 string were read.
3923
3924 11. Renamed the variable "class" as "classbits" to make life easier for C++
3925 users. (Previously there was a macro definition, but it apparently wasn't
3926 enough.)
3927
3928 12. Added the new field "tables" to the extra data so that tables can be passed
3929 in at exec time, or the internal tables can be re-selected. This allows
3930 a compiled regex to be saved and re-used at a later time by a different
3931 program that might have everything at different addresses.
3932
3933 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3934 -R library as well as a -L library.
3935
3936 14. The debugging options of pcretest (-d on the command line or D on a
3937 pattern) showed incorrect output for anything following an extended class
3938 that contained multibyte characters and which was followed by a quantifier.
3939
3940 15. Added optional support for general category Unicode character properties
3941 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3942 support. It adds about 90K to the size of the library. The meanings of the
3943 inbuilt class escapes such as \d and \s have NOT been changed.
3944
3945 16. Updated pcredemo.c to include calls to free() to release the memory for the
3946 compiled pattern.
3947
3948 17. The generated file chartables.c was being created in the source directory
3949 instead of in the building directory. This caused the build to fail if the
3950 source directory was different from the building directory, and was
3951 read-only.
3952
3953 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3954 file. No doubt somebody will tell me if they don't make sense... Also added
3955 Dan Mooney's comments about building on OpenVMS.
3956
3957 19. Added support for partial matching via the PCRE_PARTIAL option for
3958 pcre_exec() and the \P data escape in pcretest.
3959
3960 20. Extended pcretest with 3 new pattern features:
3961
3962 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3963 write the compiled pattern to the file whose name is "rest-of-line".
3964 This is a straight binary dump of the data, with the saved pointer to
3965 the character tables forced to be NULL. The study data, if any, is
3966 written too. After writing, pcretest reads a new pattern.
3967
3968 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3969 compiled pattern from the given file. There must not be any
3970 occurrences of "<" in the file name (pretty unlikely); if there are,
3971 pcretest will instead treat the initial "<" as a pattern delimiter.
3972 After reading in the pattern, pcretest goes on to read data lines as
3973 usual.
3974
3975 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3976 and 16-bit fields in a compiled pattern, to simulate a pattern that
3977 was compiled on a host of opposite endianness.
3978
3979 21. The pcre-exec() function can now cope with patterns that were compiled on
3980 hosts of opposite endianness, with this restriction:
3981
3982 As for any compiled expression that is saved and used later, the tables
3983 pointer field cannot be preserved; the extra_data field in the arguments
3984 to pcre_exec() should be used to pass in a tables address if a value
3985 other than the default internal tables were used at compile time.
3986
3987 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3988 now diagnosed as an error. Previously, most of the time, a negative number
3989 would have been treated as zero, but if in addition "ovector" was passed as
3990 NULL, a crash could occur.
3991
3992 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3993 new versions from the libtool 1.5 distribution (the last one is a copy of
3994 a file called libtool.m4). This seems to have fixed the need to patch
3995 "configure" to support Darwin 1.3 (which I used to do). However, I still
3996 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3997 workstation).
3998
3999 24. Changed the PCRE licence to be the more standard "BSD" licence.
4000
4001
4002 Version 4.5 01-Dec-03
4003 ---------------------
4004
4005 1. There has been some re-arrangement of the code for the match() function so
4006 that it can be compiled in a version that does not call itself recursively.
4007 Instead, it keeps those local variables that need separate instances for
4008 each "recursion" in a frame on the heap, and gets/frees frames whenever it
4009 needs to "recurse". Keeping track of where control must go is done by means
4010 of setjmp/longjmp. The whole thing is implemented by a set of macros that
4011 hide most of the details from the main code, and operates only if
4012 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
4013 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
4014 operating.
4015
4016 To make it easier for callers to provide specially tailored get/free
4017 functions for this usage, two new functions, pcre_stack_malloc, and
4018 pcre_stack_free, are used. They are always called in strict stacking order,
4019 and the size of block requested is always the same.
4020
4021 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
4022 PCRE has been compiled to use the stack or the heap for recursion. The
4023 -C option of pcretest uses this to show which version is compiled.
4024
4025 A new data escape \S, is added to pcretest; it causes the amounts of store
4026 obtained and freed by both kinds of malloc/free at match time to be added
4027 to the output.
4028
4029 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
4030 what's available on my current Linux desktop machine.
4031
4032 3. When matching a UTF-8 string, the test for a valid string at the start has
4033 been extended. If start_offset is not zero, PCRE now checks that it points
4034 to a byte that is the start of a UTF-8 character. If not, it returns
4035 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
4036 this is necessary because there may be backward assertions in the pattern.
4037 When matching the same subject several times, it may save resources to use
4038 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
4039
4040 4. The code for checking the validity of UTF-8 strings has been tightened so
4041 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
4042 containing "overlong sequences".
4043
4044 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
4045 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
4046 should have been "&", but it just so happened that all the cases this let
4047 through by mistake were picked up later in the function.
4048
4049 6. I had used a variable called "isblank" - this is a C99 function, causing
4050 some compilers to warn. To avoid this, I renamed it (as "blankclass").
4051
4052 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
4053 prompting; (b) run "./pcretest /dev/null" at the start of the test script
4054 so the version is shown; (c) stop "make test" echoing "./RunTest".
4055
4056 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
4057
4058 9. The prototype for memmove() for systems that don't have it was using
4059 size_t, but the inclusion of the header that defines size_t was later. I've
4060 moved the #includes for the C headers earlier to avoid this.
4061
4062 10. Added some adjustments to the code to make it easier to compiler on certain
4063 special systems:
4064
4065 (a) Some "const" qualifiers were missing.
4066 (b) Added the macro EXPORT before all exported functions; by default this
4067 is defined to be empty.
4068 (c) Changed the dftables auxiliary program (that builds chartables.c) so
4069 that it reads its output file name as an argument instead of writing
4070 to the standard output and assuming this can be redirected.
4071
4072 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
4073 class containing characters with values greater than 255, PCRE compilation
4074 went into a loop.
4075
4076 12. A recursive reference to a subpattern that was within another subpattern
4077 that had a minimum quantifier of zero caused PCRE to crash. For example,
4078 (x(y(?2))z)? provoked this bug with a subject that got as far as the
4079 recursion. If the recursively-called subpattern itself had a zero repeat,
4080 that was OK.
4081
4082 13. In pcretest, the buffer for reading a data line was set at 30K, but the
4083 buffer into which it was copied (for escape processing) was still set at
4084 1024, so long lines caused crashes.
4085
4086 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
4087 "internal error: code overflow...". This applied to any character class
4088 that was followed by a possessive quantifier.
4089
4090 15. Modified the Makefile to add libpcre.la as a prerequisite for
4091 libpcreposix.la because I was told this is needed for a parallel build to
4092 work.
4093
4094 16. If a pattern that contained .* following optional items at the start was
4095 studied, the wrong optimizing data was generated, leading to matching
4096 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
4097 matching string must start with a or b or c. The correct conclusion for
4098 this pattern is that a match can start with any character.
4099
4100
4101 Version 4.4 13-Aug-03
4102 ---------------------
4103
4104 1. In UTF-8 mode, a character class containing characters with values between
4105 127 and 255 was not handled correctly if the compiled pattern was studied.
4106 In fixing this, I have also improved the studying algorithm for such
4107 classes (slightly).
4108
4109 2. Three internal functions had redundant arguments passed to them. Removal
4110 might give a very teeny performance improvement.
4111
4112 3. Documentation bug: the value of the capture_top field in a callout is *one
4113 more than* the number of the hightest numbered captured substring.
4114
4115 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
4116 in incorrectly linking with a previously installed version. They now link
4117 explicitly with libpcre.la.
4118
4119 5. configure.in no longer needs to recognize Cygwin specially.
4120
4121 6. A problem in pcre.in for Windows platforms is fixed.
4122
4123 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
4124 pcretest, it used to include the size of the study block as part of its
4125 output. Unfortunately, the structure contains a field that has a different
4126 size on different hardware architectures. This meant that the tests that
4127 showed this size failed. As the block is currently always of a fixed size,
4128 this information isn't actually particularly useful in pcretest output, so
4129 I have just removed it.
4130
4131 8. Three pre-processor statements accidentally did not start in column 1.
4132 Sadly, there are *still* compilers around that complain, even though
4133 standard C has not required this for well over a decade. Sigh.
4134
4135 9. In pcretest, the code for checking callouts passed small integers in the
4136 callout_data field, which is a void * field. However, some picky compilers
4137 complained about the casts involved for this on 64-bit systems. Now
4138 pcretest passes the address of the small integer instead, which should get
4139 rid of the warnings.
4140
4141 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
4142 both compile and run time, and gives an error if an invalid UTF-8 sequence
4143 is found. There is a option for disabling this check in cases where the
4144 string is known to be correct and/or the maximum performance is wanted.
4145
4146 11. In response to a bug report, I changed one line in Makefile.in from
4147
4148 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
4149 to
4150 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
4151
4152 to look similar to other lines, but I have no way of telling whether this
4153 is the right thing to do, as I do not use Windows. No doubt I'll get told
4154 if it's wrong...
4155
4156
4157 Version 4.3 21-May-03
4158 ---------------------
4159
4160 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
4161 Makefile.
4162
4163 2. Some refactoring to improve the quality of the code:
4164
4165 (i) The utf8_table... variables are now declared "const".
4166
4167 (ii) The code for \cx, which used the "case flipping" table to upper case
4168 lower case letters, now just substracts 32. This is ASCII-specific,
4169 but the whole concept of \cx is ASCII-specific, so it seems
4170 reasonable.
4171
4172 (iii) PCRE was using its character types table to recognize decimal and
4173 hexadecimal digits in the pattern. This is silly, because it handles
4174 only 0-9, a-f, and A-F, but the character types table is locale-
4175 specific, which means strange things might happen. A private
4176 table is now used for this - though it costs 256 bytes, a table is
4177 much faster than multiple explicit tests. Of course, the standard
4178 character types table is still used for matching digits in subject
4179 strings against \d.
4180
4181 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
4182 ending in _t are). So I've renamed it as ESC_tee.
4183
4184 3. The first argument for regexec() in the POSIX wrapper should have been
4185 defined as "const".
4186
4187 4. Changed pcretest to use malloc() for its buffers so that they can be
4188 Electric Fenced for debugging.
4189
4190 5. There were several places in the code where, in UTF-8 mode, PCRE would try
4191 to read one or more bytes before the start of the subject string. Often this
4192 had no effect on PCRE's behaviour, but in some circumstances it could
4193 provoke a segmentation fault.
4194
4195 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
4196 to try to read one or more bytes before the start of the subject string.
4197
4198 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
4199 UTF-8 support could misbehave in various ways if the subject string
4200 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
4201 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
4202 back over UTF-8 characters.)
4203
4204
4205 Version 4.2 14-Apr-03
4206 ---------------------
4207
4208 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
4209
4210 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
4211 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
4212 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
4213 [WIN_PREFIX]: new variable, "cyg" for Cygwin
4214 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
4215 and BUILD_EXEEXT
4216 Note: automatic setting of the BUILD variables is not yet working
4217 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
4218 compile-time but not at link-time
4219 [LINK]: use for linking executables only
4220 make different versions for Windows and non-Windows
4221 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
4222 libraries
4223 [LINK_FOR_BUILD]: new variable
4224 [OBJEXT]: use throughout
4225 [EXEEXT]: use throughout
4226 <winshared>: new target
4227 <wininstall>: new target
4228 <dftables.o>: use native compiler
4229 <dftables>: use native linker
4230 <install>: handle Windows platform correctly
4231 <clean>: ditto
4232 <check>: ditto
4233 copy DLL to top builddir before testing
4234
4235 As part of these changes, -no-undefined was removed again. This was reported
4236 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
4237 in any case.
4238
4239 3. Some tidies to get rid of compiler warnings:
4240
4241 . In the match_data structure, match_limit was an unsigned long int, whereas
4242 match_call_count was an int. I've made them both unsigned long ints.
4243
4244 . In pcretest the fact that a const uschar * doesn't automatically cast to
4245 a void * provoked a warning.
4246
4247 . Turning on some more compiler warnings threw up some "shadow" variables
4248 and a few more missing casts.
4249
4250 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
4251 option, a class that contained a single character with a value between 128
4252 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
4253
4254 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
4255 option, a class that contained several characters, but with at least one
4256 whose value was between 128 and 255 caused PCRE to crash.
4257
4258
4259 Version 4.1 12-Mar-03
4260 ---------------------
4261
4262 1. Compiling with gcc -pedantic found a couple of places where casts were
4263 needed, and a string in dftables.c that was longer than standard compilers are
4264 required to support.
4265
4266 2. Compiling with Sun's compiler found a few more places where the code could
4267 be tidied up in order to avoid warnings.
4268
4269 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
4270 first of these names is deprecated in the latest Autoconf in favour of the name
4271 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
4272 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
4273 analogy I have changed it to CFLAGS_FOR_BUILD.
4274
4275 4. Added -no-undefined to the linking command in the Makefile, because this is
4276 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
4277 linking step for the pcreposix library.
4278
4279 5. PCRE was failing to diagnose the case of two named groups with the same
4280 name.
4281
4282 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
4283 literal character that is needed in the subject for a match, and scans along to
4284 ensure that it is present before embarking on the full matching process. This
4285 saves time in cases of nested unlimited repeats that are never going to match.
4286 Problem: the scan can take a lot of time if the subject is very long (e.g.
4287 megabytes), thus penalizing straightforward matches. It is now done only if the
4288 amount of subject to be scanned is less than 1000 bytes.
4289
4290 7. A lesser problem with the same optimization is that it was recording the
4291 first character of an anchored pattern as "needed", thus provoking a search
4292 right along the subject, even when the first match of the pattern was going to
4293 fail. The "needed" character is now not set for anchored patterns, unless it
4294 follows something in the pattern that is of non-fixed length. Thus, it still
4295 fulfils its original purpose of finding quick non-matches in cases of nested
4296 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
4297
4298
4299 Version 4.0 17-Feb-03
4300 ---------------------
4301
4302 1. If a comment in an extended regex that started immediately after a meta-item
4303 extended to the end of string, PCRE compiled incorrect data. This could lead to
4304 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
4305
4306 2. Moved to autoconf 2.53 and libtool 1.4.2.
4307
4308 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
4309 the special perltest8 script is no longer needed - all the tests can be run
4310 from a single perltest script.
4311
4312 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
4313 by \s. It has now been removed in PCRE. This means it isn't recognized as
4314 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
4315 class [:space:] *does* include VT, thereby creating a mess.
4316
4317 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
4318 space and tab.
4319
4320 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
4321 its new features into the main test script, reducing the number of scripts.
4322
4323 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
4324 were backward compatible, and made the (?i) apply to the whole pattern, as if
4325 /i were given. Now it behaves more logically, and applies the option setting
4326 only to what follows. PCRE has been changed to follow suit. However, if it
4327 finds options settings right at the start of the pattern, it extracts them into
4328 the global options, as before. Thus, they show up in the info data.
4329
4330 8. Added support for the \Q...\E escape sequence. Characters in between are
4331 treated as literals. This is slightly different from Perl in that $ and @ are
4332 also handled as literals inside the quotes. In Perl, they will cause variable
4333 interpolation. Note the following examples:
4334
4335 Pattern PCRE matches Perl matches
4336
4337 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
4338 \Qabc\$xyz\E abc\$xyz abc\$xyz
4339 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
4340
4341 For compatibility with Perl, \Q...\E sequences are recognized inside character
4342 classes as well as outside them.
4343
4344 9. Re-organized 3 code statements in pcretest to avoid "overflow in
4345 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
4346 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
4347 signed/unsigned warnings.
4348
4349 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
4350 option for pcretest, so I've replaced it by a simple function that does just
4351 that job.
4352
4353 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
4354 "pcregrep -".
4355
4356 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
4357 Java package. This provides some syntactic sugar for simple cases of what my
4358 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
4359 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
4360 item, you can use this simplified notation. Note that only makes sense with
4361 greedy quantifiers. Consequently, the use of the possessive quantifier forces
4362 greediness, whatever the setting of the PCRE_UNGREEDY option.
4363
4364 13. A change of greediness default within a pattern was not taking effect at
4365 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
4366 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
4367 was abstracted outside.
4368
4369 14. PCRE now supports the \G assertion. It is true when the current matching
4370 position is at the start point of the match. This differs from \A when the
4371 starting offset is non-zero. Used with the /g option of pcretest (or similar
4372 code), it works in the same way as it does for Perl's /g option. If all
4373 alternatives of a regex begin with \G, the expression is anchored to the start
4374 match position, and the "anchored" flag is set in the compiled expression.
4375
4376 15. Some bugs concerning the handling of certain option changes within patterns
4377 have been fixed. These applied to options other than (?ims). For example,
4378 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
4379 been the other way round. Some of this was related to change 7 above.
4380
4381 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
4382 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
4383 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
4384 POSIX classes only within a class (e.g. /[[:alpha:]]/).
4385
4386 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
4387 mode. Unlike ".", it always matches newline, whatever the setting of
4388 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
4389 assertions. Perl allows it, but it doesn't (in general) work because it can't
4390 calculate the length of the lookbehind. At least, that's the case for Perl
4391 5.8.0 - I've been told they are going to document that it doesn't work in
4392 future.
4393
4394 18. Added an error diagnosis for escapes that PCRE does not support: these are
4395 \L, \l, \N, \P, \p, \U, \u, and \X.
4396
4397 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4398 reading past the end of the pattern in cases such as /[abcd/.
4399
4400 20. PCRE was getting more memory than necessary for patterns with classes that
4401 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4402
4403 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4404 compiling PCRE for use with Virtual Pascal.
4405
4406 22. Small fix to the Makefile to make it work properly if the build is done
4407 outside the source tree.
4408
4409 23. Added a new extension: a condition to go with recursion. If a conditional
4410 subpattern starts with (?(R) the "true" branch is used if recursion has
4411 happened, whereas the "false" branch is used only at the top level.
4412
4413 24. When there was a very long string of literal characters (over 255 bytes
4414 without UTF support, over 250 bytes with UTF support), the computation of how
4415 much memory was required could be incorrect, leading to segfaults or other
4416 strange effects.
4417
4418 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4419 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4420 there was a subsequent back reference to those brackets. This meant that, for
4421 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4422 possible to check for precisely this case. All we can do is abandon the
4423 optimization if .* occurs inside capturing brackets when there are any back
4424 references whatsoever. (See below for a better fix that came later.)
4425
4426 26. The handling of the optimization for finding the first character of a
4427 non-anchored pattern, and for finding a character that is required later in the
4428 match were failing in some cases. This didn't break the matching; it just
4429 failed to optimize when it could. The way this is done has been re-implemented.
4430
4431 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4432
4433 28. Added a new feature that provides some of the functionality that Perl
4434 provides with (?{...}). The facility is termed a "callout". The way it is done
4435 in PCRE is for the caller to provide an optional function, by setting
4436 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4437 global variable. By default it is unset, which disables all calling out. To get
4438 the function called, the regex must include (?C) at appropriate points. This
4439 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4440 This provides a means of identifying different callout points. When PCRE
4441 reaches such a point in the regex, if pcre_callout has been set, the external
4442 function is called. It is provided with data in a structure called
4443 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4444 matching continues; if it returns a non-zero value, the match at the current
4445 point fails. However, backtracking will occur if possible. [This was changed
4446 later and other features added - see item 49 below.]
4447
4448 29. pcretest is upgraded to test the callout functionality. It provides a
4449 callout function that displays information. By default, it shows the start of
4450 the match and the current position in the text. There are some new data escapes
4451 to vary what happens:
4452
4453 \C+ in addition, show current contents of captured substrings
4454 \C- do not supply a callout function
4455 \C!n return 1 when callout number n is reached
4456 \C!n!m return 1 when callout number n is reached for the mth time
4457
4458 30. If pcregrep was called with the -l option and just a single file name, it
4459 output "<stdin>" if a match was found, instead of the file name.
4460
4461 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4462 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4463 pcre_exec(). This saves a malloc/free per call. The default value of
4464 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4465 when configuring.
4466
4467 32. The default maximum size of a compiled pattern is 64K. There have been a
4468 few cases of people hitting this limit. The code now uses macros to handle the
4469 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4470 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4471 configuring. Tests 2 and 5 work only with 2-byte links because they output
4472 debugging information about compiled patterns.
4473
4474 33. Internal code re-arrangements:
4475
4476 (a) Moved the debugging function for printing out a compiled regex into
4477 its own source file (printint.c) and used #include to pull it into
4478 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4479 separate copies.
4480
4481 (b) Defined the list of op-code names for debugging as a macro in
4482 internal.h so that it is next to the definition of the opcodes.
4483
4484 (c) Defined a table of op-code lengths for simpler skipping along compiled
4485 code. This is again a macro in internal.h so that it is next to the
4486 definition of the opcodes.
4487
4488 34. Added support for recursive calls to individual subpatterns, along the
4489 lines of Robin Houston's patch (but implemented somewhat differently).
4490
4491 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4492 allow it to read and process whole directories in Win32. This code was
4493 contributed by Lionel Fourquaux; it has not been tested by me.
4494
4495 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4496 used to name a group. Names consist of alphanumerics and underscores, and must
4497 be unique. Back references use the syntax (?P=name) and recursive calls use
4498 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4499 numbers. The function pcre_fullinfo() can be used after compilation to extract
4500 a name/number map. There are three relevant calls:
4501
4502 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4503 PCRE_INFO_NAMECOUNT yields the number of entries
4504 PCRE_INFO_NAMETABLE yields a pointer to the map.
4505
4506 The map is a vector of fixed-size entries. The size of each entry depends on
4507 the length of the longest name used. The first two bytes of each entry are the
4508 group number, most significant byte first. There follows the corresponding
4509 name, zero terminated. The names are in alphabetical order.
4510
4511 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4512 case instead of 255. Making it the same both with and without UTF-8 support
4513 means that the same test output works with both.
4514
4515 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4516 calling malloc() with a zero argument.
4517
4518 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4519 optimization. I've improved things by keeping a bitmap of backreferences with
4520 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4521 fact referenced, the optimization can be applied. It is unlikely that a
4522 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4523 the match to follow \n) will appear inside brackets with a number greater than
4524 31, but if it does, any back reference > 31 suppresses the optimization.
4525
4526 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4527 of disabling numbered capturing parentheses. Any opening parenthesis that is
4528 not followed by ? behaves as if it were followed by ?: but named parentheses
4529 can still be used for capturing (and they will acquire numbers in the usual
4530 way).
4531
4532 41. Redesigned the return codes from the match() function into yes/no/error so
4533 that errors can be passed back from deep inside the nested calls. A malloc
4534 failure while inside a recursive subpattern call now causes the
4535 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4536
4537 42. It is now possible to set a limit on the number of times the match()
4538 function is called in a call to pcre_exec(). This facility makes it possible to
4539 limit the amount of recursion and backtracking, though not in a directly
4540 obvious way, because the match() function is used in a number of different
4541 circumstances. The count starts from zero for each position in the subject
4542 string (for non-anchored patterns). The default limit is, for compatibility, a
4543 large number, namely 10 000 000. You can change this in two ways:
4544
4545 (a) When configuring PCRE before making, you can use --with-match-limit=n
4546 to set a default value for the compiled library.
4547
4548 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4549 a different value is set. See 45 below.
4550
4551 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4552
4553 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4554 of things that can be changed at compile time. The first argument specifies
4555 what is wanted and the second points to where the information is to be placed.
4556 The current list of available information is:
4557
4558 PCRE_CONFIG_UTF8
4559
4560 The output is an integer that is set to one if UTF-8 support is available;
4561 otherwise it is set to zero.
4562
4563 PCRE_CONFIG_NEWLINE
4564
4565 The output is an integer that it set to the value of the code that is used for
4566 newline. It is either LF (10) or CR (13).
4567
4568 PCRE_CONFIG_LINK_SIZE
4569
4570 The output is an integer that contains the number of bytes used for internal
4571 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4572
4573 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
4574
4575 The output is an integer that contains the threshold above which the POSIX
4576 interface uses malloc() for output vectors. See item 31 above.
4577
4578 PCRE_CONFIG_MATCH_LIMIT
4579
4580 The output is an unsigned integer that contains the default limit of the number
4581 of match() calls in a pcre_exec() execution. See 42 above.
4582
4583 44. pcretest has been upgraded by the addition of the -C option. This causes it
4584 to extract all the available output from the new pcre_config() function, and to
4585 output it. The program then exits immediately.
4586
4587 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4588 order to support additional features. One way would have been to define
4589 pcre_exec2() (for example) with extra arguments, but this would not have been
4590 extensible, and would also have required all calls to the original function to
4591 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4592 is used for passing in "extra" data from pcre_study().
4593
4594 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4595 contains the following fields:
4596
4597 flags a bitmap indicating which of the following fields are set
4598 study_data opaque data from pcre_study()
4599 match_limit a way of specifying a limit on match() calls for a specific
4600 call to pcre_exec()
4601 callout_data data for callouts (see 49 below)
4602
4603 The flag bits are also defined in pcre.h, and are
4604
4605 PCRE_EXTRA_STUDY_DATA
4606 PCRE_EXTRA_MATCH_LIMIT
4607 PCRE_EXTRA_CALLOUT_DATA
4608
4609 The pcre_study() function now returns one of these new pcre_extra blocks, with
4610 the actual study data pointed to by the study_data field, and the
4611 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4612 before. That is, this change is entirely upwards-compatible and requires no
4613 change to existing code.
4614
4615 If you want to pass in additional data to pcre_exec(), you can either place it
4616 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4617 block.
4618
4619 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4620 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4621 times with different match limits, until it finds the minimum value needed for
4622 pcre_exec() to complete. The value is then output. This can be instructive; for
4623 most simple matches the number is quite small, but for pathological cases it
4624 gets very large very quickly.
4625
4626 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4627 returns the size of the data block pointed to by the study_data field in a
4628 pcre_extra block, that is, the value that was passed as the argument to
4629 pcre_malloc() when PCRE was getting memory in which to place the information
4630 created by pcre_study(). The fourth argument should point to a size_t variable.
4631 pcretest has been extended so that this information is shown after a successful
4632 pcre_study() call when information about the compiled regex is being displayed.
4633
4634 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4635 because what follows is always an absolute path. (Later: it turns out that this
4636 is more than cosmetic for MinGW, because it doesn't like empty path
4637 components.)
4638
4639 49. Some changes have been made to the callout feature (see 28 above):
4640
4641 (i) A callout function now has three choices for what it returns:
4642
4643 0 => success, carry on matching
4644 > 0 => failure at this point, but backtrack if possible
4645 < 0 => serious error, return this value from pcre_exec()
4646
4647 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4648 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4649 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4650 use by callout functions. It will never be used by PCRE itself.
4651
4652 (ii) The pcre_extra structure (see 45 above) has a void * field called
4653 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4654 pcre_callout_block structure has a field of the same name. The contents of
4655 the field passed in the pcre_extra structure are passed to the callout
4656 function in the corresponding field in the callout block. This makes it
4657 easier to use the same callout-containing regex from multiple threads. For
4658 testing, the pcretest program has a new data escape
4659
4660 \C*n pass the number n (may be negative) as callout_data
4661
4662 If the callout function in pcretest receives a non-zero value as
4663 callout_data, it returns that value.
4664
4665 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4666 there were some redundant $(CFLAGS) in commands that are now specified as
4667 $(LINK), which already includes $(CFLAGS).
4668
4669 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4670 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4671 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4672 one-byte characters throughout. Note that case-insensitive matching applies
4673 only to characters whose values are less than 256. PCRE doesn't support the
4674 notion of cases for higher-valued characters.
4675
4676 (i) A character class whose characters are all within 0-255 is handled as
4677 a bit map, and the map is inverted for negative classes. Previously, a
4678 character > 255 always failed to match such a class; however it should
4679 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4680
4681 (ii) A negated character class with a single character < 255 is coded as
4682 "not this character" (OP_NOT). This wasn't working properly when the test
4683 character was multibyte, either singly or repeated.
4684
4685 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4686 mode, for example: \x{100}{2,3}.
4687
4688 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4689 singly or repeated) now correctly test multibyte characters. However,
4690 PCRE doesn't recognize any characters with values greater than 255 as
4691 digits, spaces, or word characters. Such characters always match \D, \S,
4692 and \W, and never match \d, \s, or \w.
4693
4694 (v) Classes may now contain characters and character ranges with values
4695 greater than 255. For example: [ab\x{100}-\x{400}].
4696
4697 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4698 PCRE in UTF-8 mode.
4699
4700 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4701 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4702 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4703 value.)
4704
4705 53. The single man page has become too large. I have therefore split it up into
4706 a number of separate man pages. These also give rise to individual HTML pages;
4707 these are now put in a separate directory, and there is an index.html page that
4708 lists them all. Some hyperlinking between the pages has been installed.
4709
4710 54. Added convenience functions for handling named capturing parentheses.
4711
4712 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4713 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4714 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4715 are faulted.
4716
4717 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4718 calling configure. These values are used when compiling the dftables.c program
4719 which is run to generate the source of the default character tables. They
4720 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4721 you will need to set these values.
4722
4723 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4724
4725
4726 Version 3.9 02-Jan-02
4727 ---------------------
4728
4729 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4730
4731 2. If --disable-static was given, the building process failed when trying to
4732 build pcretest and pcregrep. (For some reason it was using libtool to compile
4733 them, which is not right, as they aren't part of the library.)
4734
4735
4736 Version 3.8 18-Dec-01
4737 ---------------------
4738
4739 1. The experimental UTF-8 code was completely screwed up. It was packing the
4740 bytes in the wrong order. How dumb can you get?
4741
4742
4743 Version 3.7 29-Oct-01
4744 ---------------------
4745
4746 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4747 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4748 this didn't happen under Solaris 8, where I normally test things.
4749
4750 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4751 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4752 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4753 this fix an hour or so after the initial 3.7 release.)
4754
4755
4756 Version 3.6 23-Oct-01
4757 ---------------------
4758
4759 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4760 offsets passed as NULL with zero offset count.
4761
4762 2. The config.guess and config.sub files had not been updated when I moved to
4763 the latest autoconf.
4764
4765
4766 Version 3.5 15-Aug-01
4767 ---------------------
4768
4769 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4770 had been forgotten.
4771
4772 2. By using declared but undefined structures, we can avoid using "void"
4773 definitions in pcre.h while keeping the internal definitions of the structures
4774 private.
4775
4776 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4777 user point of view, this means that both static and shared libraries are built
4778 by default, but this can be individually controlled. More of the work of
4779 handling this static/shared cases is now inside libtool instead of PCRE's make
4780 file.
4781
4782 4. The pcretest utility is now installed along with pcregrep because it is
4783 useful for users (to test regexs) and by doing this, it automatically gets
4784 relinked by libtool. The documentation has been turned into a man page, so
4785 there are now .1, .txt, and .html versions in /doc.
4786
4787 5. Upgrades to pcregrep:
4788 (i) Added long-form option names like gnu grep.
4789 (ii) Added --help to list all options with an explanatory phrase.
4790 (iii) Added -r, --recursive to recurse into sub-directories.
4791 (iv) Added -f, --file to read patterns from a file.
4792
4793 6. pcre_exec() was referring to its "code" argument before testing that
4794 argument for NULL (and giving an error if it was NULL).
4795
4796 7. Upgraded Makefile.in to allow for compiling in a different directory from
4797 the source directory.
4798
4799 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4800 options bits, the pointer it was passed was to an int instead of to an unsigned
4801 long int. This mattered only on 64-bit systems.
4802
4803 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4804 generated) instead of pcre.in, which it its source. Also made the same change
4805 in several of the .c files.
4806
4807 10. A new release of gcc defines printf() as a macro, which broke pcretest
4808 because it had an ifdef in the middle of a string argument for printf(). Fixed
4809 by using separate calls to printf().
4810
4811 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4812 script, to force use of CR or LF instead of \n in the source. On non-Unix
4813 systems, the value can be set in config.h.
4814
4815 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4816 absolute limit. Changed the text of the error message to make this clear, and
4817 likewise updated the man page.
4818
4819 13. The limit of 99 on the number of capturing subpatterns has been removed.
4820 The new limit is 65535, which I hope will not be a "real" limit.
4821
4822
4823 Version 3.4 22-Aug-00
4824 ---------------------
4825
4826 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4827
4828 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4829
4830
4831 Version 3.3 01-Aug-00
4832 ---------------------
4833
4834 1. If an octal character was given, but the value was greater than \377, it
4835 was not getting masked to the least significant bits, as documented. This could
4836 lead to crashes in some systems.
4837
4838 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4839 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4840
4841 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4842 These just pass their arguments on to (pcre_free)(), but they are provided
4843 because some uses of PCRE bind it to non-C systems that can call its functions,
4844 but cannot call free() or pcre_free() directly.
4845
4846 4. Add "make test" as a synonym for "make check". Corrected some comments in
4847 the Makefile.
4848
4849 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4850 Makefile.
4851
4852 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4853 command called pgrep for grepping around the active processes.
4854
4855 7. Added the beginnings of support for UTF-8 character strings.
4856
4857 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4858 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4859 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4860 out for the ar command.)
4861
4862
4863 Version 3.2 12-May-00
4864 ---------------------
4865
4866 This is purely a bug fixing release.
4867
4868 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4869 of ZA. This was just one example of several cases that could provoke this bug,
4870 which was introduced by change 9 of version 2.00. The code for breaking
4871 infinite loops after an iteration that matches an empty string was't working
4872 correctly.
4873
4874 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4875 when matched against abbab (for example). After matching an empty string, it
4876 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4877 caused it to match further down the string than it should.
4878
4879 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4880 was there because it doesn't seem to be needed, and it causes trouble on some
4881 systems, as it is not a Standard C header. It has been removed.
4882
4883 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4884 were reported on the Macintosh. The changes were from
4885
4886 while ((c = *(++ptr)) != 0 && c != '\n');
4887 to
4888 while ((c = *(++ptr)) != 0 && c != '\n') ;
4889
4890 Totally extraordinary, but if that's what it takes...
4891
4892 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4893 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4894 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4895 assumes the way PCRE uses memmove() (always moving upwards).
4896
4897 6. PCRE is being used in one environment where strchr() is not available. There
4898 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4899 faster code anyway.
4900
4901
4902 Version 3.1 09-Feb-00
4903 ---------------------
4904
4905 The only change in this release is the fixing of some bugs in Makefile.in for
4906 the "install" target:
4907
4908 (1) It was failing to install pcreposix.h.
4909
4910 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4911
4912
4913 Version 3.0 01-Feb-00
4914 ---------------------
4915
4916 1. Add support for the /+ modifier to perltest (to output $` like it does in
4917 pcretest).
4918
4919 2. Add support for the /g modifier to perltest.
4920
4921 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4922 matches null strings.
4923
4924 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4925 pattern. Perl treats empty patterns specially - it reuses the most recent
4926 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4927 effect.
4928
4929 5. The POSIX interface was broken in that it was just handing over the POSIX
4930 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4931 required a bigger vector, with some working space on the end. This means that
4932 the POSIX wrapper now has to get and free some memory, and copy the results.
4933
4934 6. Added some simple autoconf support, placing the test data and the
4935 documentation in separate directories, re-organizing some of the
4936 information files, and making it build pcre-config (a GNU standard). Also added
4937 libtool support for building PCRE as a shared library, which is now the
4938 default.
4939
4940 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4941 09 are not valid octal constants. Single digits will be used for minor values
4942 less than 10.
4943
4944 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4945 existing programs that set these in the POSIX interface can use PCRE without
4946 modification.
4947
4948 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4949 return all that pcre_info() returns, plus additional data. The pcre_info()
4950 function is retained for compatibility, but is considered to be obsolete.
4951
4952 10. Added experimental recursion feature (?R) to handle one common case that
4953 Perl 5.6 will be able to do with (?p{...}).
4954
4955 11. Added support for POSIX character classes like [:alpha:], which Perl is
4956 adopting.
4957
4958
4959 Version 2.08 31-Aug-99
4960 ----------------------
4961
4962 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4963 trying to match at the startoffset position, but instead was moving forward to
4964 the next newline as if a previous match had failed.
4965
4966 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4967 and could get into a loop if a null string was matched other than at the start
4968 of the subject.
4969
4970 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4971 be distinguished at compile time, and for completeness also added PCRE_DATE.
4972
4973 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4974 in GnuWin32 environments.
4975
4976
4977 Version 2.07 29-Jul-99
4978 ----------------------
4979
4980 1. The documentation is now supplied in plain text form and HTML as well as in
4981 the form of man page sources.
4982
4983 2. C++ compilers don't like assigning (void *) values to other pointer types.
4984 In particular this affects malloc(). Although there is no problem in Standard
4985 C, I've put in casts to keep C++ compilers happy.
4986
4987 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4988 should be (const char *).
4989
4990 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4991 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4992 However, I haven't made this a standard facility. The documentation doesn't
4993 mention it, and the Makefile doesn't support it.
4994
4995 5. The Makefile now contains an "install" target, with editable destinations at
4996 the top of the file. The pcretest program is not installed.
4997
4998 6. pgrep -V now gives the PCRE version number and date.
4999
5000 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
5001 causing the entire string to be ignored, instead of just the last character.
5002
5003 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
5004 non-matching string, it can take a very, very long time, even for strings of
5005 quite modest length, because of the nested recursion. PCRE now does better in
5006 some of these cases. It does this by remembering the last required literal
5007 character in the pattern, and pre-searching the subject to ensure it is present
5008 before running the real match. In other words, it applies a heuristic to detect
5009 some types of certain failure quickly, and in the above example, if presented
5010 with a string that has no trailing " it gives "no match" very quickly.
5011
5012 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
5013 other alternatives are tried instead.
5014
5015
5016 Version 2.06 09-Jun-99
5017 ----------------------
5018
5019 1. Change pcretest's output for amount of store used to show just the code
5020 space, because the remainder (the data block) varies in size between 32-bit and
5021 64-bit systems.
5022
5023 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
5024 start matching at. This allows lookbehinds to work when searching for multiple
5025 occurrences in a string.
5026
5027 3. Added additional options to pcretest for testing multiple occurrences:
5028
5029 /+ outputs the rest of the string that follows a match
5030 /g loops for multiple occurrences, using the new startoffset argument
5031 /G loops for multiple occurrences by passing an incremented pointer
5032
5033 4. PCRE wasn't doing the "first character" optimization for patterns starting
5034 with \b or \B, though it was doing it for other lookbehind assertions. That is,
5035 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
5036 the letter 'x'. On long subject strings, this gives a significant speed-up.
5037
5038
5039 Version 2.05 21-Apr-99
5040 ----------------------
5041
5042 1. Changed the type of magic_number from int to long int so that it works
5043 properly on 16-bit systems.
5044
5045 2. Fixed a bug which caused patterns starting with .* not to work correctly
5046 when the subject string contained newline characters. PCRE was assuming
5047 anchoring for such patterns in all cases, which is not correct because .* will
5048 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
5049 DOTALL is set at top level; otherwise it knows that patterns starting with .*
5050 must be retried after every newline in the subject.
5051
5052
5053 Version 2.04 18-Feb-99
5054 ----------------------
5055
5056 1. For parenthesized subpatterns with repeats whose minimum was zero, the
5057 computation of the store needed to hold the pattern was incorrect (too large).
5058 If such patterns were nested a few deep, this could multiply and become a real
5059 problem.
5060
5061 2. Added /M option to pcretest to show the memory requirement of a specific
5062 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
5063
5064 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
5065 compiled in such a way that the backtracking after subsequent failure was
5066 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
5067 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
5068
5069
5070 Version 2.03 02-Feb-99
5071 ----------------------
5072
5073 1. Fixed typo and small mistake in man page.
5074
5075 2. Added 4th condition (GPL supersedes if conflict) and created separate
5076 LICENCE file containing the conditions.
5077
5078 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
5079 Perl, that is the internal \ allows the delimiter to be included in the
5080 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
5081 the final delimiter, add \ to the end of the pattern (to test the error).
5082
5083 4. Added the convenience functions for extracting substrings after a successful
5084 match. Updated pcretest to make it able to test these functions.
5085
5086
5087 Version 2.02 14-Jan-99
5088 ----------------------
5089
5090 1. Initialized the working variables associated with each extraction so that
5091 their saving and restoring doesn't refer to uninitialized store.
5092
5093 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
5094 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
5095 fix the problem.
5096
5097 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
5098 calls, and wasn't printing the correct value for compiling calls. Increased the
5099 default value of LOOPREPEAT, and the number of significant figures in the
5100 times.
5101
5102 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
5103
5104 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
5105 a building problem on Windows NT with a FAT file system.
5106
5107
5108 Version 2.01 21-Oct-98
5109 ----------------------
5110
5111 1. Changed the API for pcre_compile() to allow for the provision of a pointer
5112 to character tables built by pcre_maketables() in the current locale. If NULL
5113 is passed, the default tables are used.
5114
5115
5116 Version 2.00 24-Sep-98
5117 ----------------------
5118
5119 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
5120 it any more.
5121
5122 2. Allow quantification of (?>) groups, and make it work correctly.
5123
5124 3. The first character computation wasn't working for (?>) groups.
5125
5126 4. Correct the implementation of \Z (it is permitted to match on the \n at the
5127 end of the subject) and add 5.005's \z, which really does match only at the
5128 very end of the subject.
5129
5130 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
5131
5132 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
5133 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
5134 localized options. All options to pcre_study() were also removed.
5135
5136 7. Add other new features from 5.005:
5137
5138 $(?<= positive lookbehind
5139 $(?<! negative lookbehind
5140 (?imsx-imsx) added the unsetting capability
5141 such a setting is global if at outer level; local otherwise
5142 (?imsx-imsx:) non-capturing groups with option setting
5143 (?(cond)re|re) conditional pattern matching
5144
5145 A backreference to itself in a repeated group matches the previous
5146 captured string.
5147
5148 8. General tidying up of studying (both automatic and via "study")
5149 consequential on the addition of new assertions.
5150
5151 9. As in 5.005, unlimited repeated groups that could match an empty substring
5152 are no longer faulted at compile time. Instead, the loop is forcibly broken at
5153 runtime if any iteration does actually match an empty substring.
5154
5155 10. Include the RunTest script in the distribution.
5156
5157 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
5158 discrepancies, some of which were old and were also with respect to 5.004. They
5159 have now been fixed.
5160
5161
5162 Version 1.09 28-Apr-98
5163 ----------------------
5164
5165 1. A negated single character class followed by a quantifier with a minimum
5166 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
5167 program crashes, or just wrong answers. This did not apply to negated classes
5168 containing more than one character, or to minima other than one.
5169
5170
5171 Version 1.08 27-Mar-98
5172 ----------------------
5173
5174 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
5175
5176 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
5177 latter must appear before anything that relies on it in the pattern.
5178
5179
5180 Version 1.07 16-Feb-98
5181 ----------------------
5182
5183 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
5184 repeat of a potentially empty string).
5185
5186
5187 Version 1.06 23-Jan-98
5188 ----------------------
5189
5190 1. Added Markus Oberhumer's little patches for C++.
5191
5192 2. Literal strings longer than 255 characters were broken.
5193
5194
5195 Version 1.05 23-Dec-97
5196 ----------------------
5197
5198 1. Negated character classes containing more than one character were failing if
5199 PCRE_CASELESS was set at run time.
5200
5201
5202 Version 1.04 19-Dec-97
5203 ----------------------
5204
5205 1. Corrected the man page, where some "const" qualifiers had been omitted.
5206
5207 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
5208 input syntax.
5209
5210 3. Fixed memory leak which occurred when a regex with back references was
5211 matched with an offsets vector that wasn't big enough. The temporary memory
5212 that is used in this case wasn't being freed if the match failed.
5213
5214 4. Tidied pcretest to ensure it frees memory that it gets.
5215
5216 5. Temporary memory was being obtained in the case where the passed offsets
5217 vector was exactly big enough.
5218
5219 6. Corrected definition of offsetof() from change 5 below.
5220
5221 7. I had screwed up change 6 below and broken the rules for the use of
5222 setjmp(). Now fixed.
5223
5224
5225 Version 1.03 18-Dec-97
5226 ----------------------
5227
5228 1. A erroneous regex with a missing opening parenthesis was correctly
5229 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
5230 on some systems.
5231
5232 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
5233 it was reported that one broken compiler failed on the former because "code" is
5234 also an independent variable.
5235
5236 3. The erroneous regex a[]b caused an array overrun reference.
5237
5238 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
5239 fail on data ending with that character. (It was going on too far, and checking
5240 the next character, typically a binary zero.) This was specific to the
5241 optimized code for single-character negative classes.
5242
5243 5. Added a contributed patch from the TIN world which does the following:
5244
5245 + Add an undef for memmove, in case the the system defines a macro for it.
5246
5247 + Add a definition of offsetof(), in case there isn't one. (I don't know
5248 the reason behind this - offsetof() is part of the ANSI standard - but
5249 it does no harm).
5250
5251 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
5252 most of the places where whitespace preceded '#'. I have given up and
5253 allowed the remaining 2 cases to be at the margin.
5254
5255 + Rename some variables in pcre to eliminate shadowing. This seems very
5256 pedantic, but does no harm, of course.
5257
5258 6. Moved the call to setjmp() into its own function, to get rid of warnings
5259 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
5260
5261 7. Constructs such as \d{8,} were compiling into the equivalent of
5262 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
5263 outcome, but in this particular case used more store than had been allocated,
5264 which caused the bug to be discovered because it threw up an internal error.
5265
5266 8. The debugging code in both pcre and pcretest for outputting the compiled
5267 form of a regex was going wrong in the case of back references followed by
5268 curly-bracketed repeats.
5269
5270
5271 Version 1.02 12-Dec-97
5272 ----------------------
5273
5274 1. Typos in pcre.3 and comments in the source fixed.
5275
5276 2. Applied a contributed patch to get rid of places where it used to remove
5277 'const' from variables, and fixed some signed/unsigned and uninitialized
5278 variable warnings.
5279
5280 3. Added the "runtest" target to Makefile.
5281
5282 4. Set default compiler flag to -O2 rather than just -O.
5283
5284
5285 Version 1.01 19-Nov-97
5286 ----------------------
5287
5288 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
5289 like /([ab]*)*/, that is, for classes with more than one character in them.
5290
5291 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
5292 as /((?>a*))*/ (a PCRE_EXTRA facility).
5293
5294
5295 Version 1.00 18-Nov-97
5296 ----------------------
5297
5298 1. Added compile-time macros to support systems such as SunOS4 which don't have
5299 memmove() or strerror() but have other things that can be used instead.
5300
5301 2. Arranged that "make clean" removes the executables.
5302
5303
5304 Version 0.99 27-Oct-97
5305 ----------------------
5306
5307 1. Fixed bug in code for optimizing classes with only one character. It was
5308 initializing a 32-byte map regardless, which could cause it to run off the end
5309 of the memory it had got.
5310
5311 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
5312
5313
5314 Version 0.98 22-Oct-97
5315 ----------------------
5316
5317 1. Fixed bug in code for handling temporary memory usage when there are more
5318 back references than supplied space in the ovector. This could cause segfaults.
5319
5320
5321 Version 0.97 21-Oct-97
5322 ----------------------
5323
5324 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
5325
5326 2. Optimized negated single characters not to use a bit map.
5327
5328 3. Brought error texts together as macro definitions; clarified some of them;
5329 fixed one that was wrong - it said "range out of order" when it meant "invalid
5330 escape sequence".
5331
5332 4. Changed some char * arguments to const char *.
5333
5334 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
5335
5336 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
5337 pcretest.
5338
5339
5340 Version 0.96 16-Oct-97
5341 ----------------------
5342
5343 1. Added a simple "pgrep" utility to the distribution.
5344
5345 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
5346 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
5347 where "ddd" means "one or more decimal digits".
5348
5349 3. Fixed serious bug. If a pattern had a back reference, but the call to
5350 pcre_exec() didn't supply a large enough ovector to record the related
5351 identifying subpattern, the match always failed. PCRE now remembers the number
5352 of the largest back reference, and gets some temporary memory in which to save
5353 the offsets during matching if necessary, in order to ensure that
5354 backreferences always work.
5355
5356 4. Increased the compatibility with Perl in a number of ways:
5357
5358 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
5359 to request this handling. The option can be set at compile or exec time.
5360
5361 (b) $ matches before a terminating newline by default; an option
5362 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
5363 mode). The option can be set at compile or exec time.
5364
5365 (c) The handling of \ followed by a digit other than 0 is now supposed to be
5366 the same as Perl's. If the decimal number it represents is less than 10
5367 or there aren't that many previous left capturing parentheses, an octal
5368 escape is read. Inside a character class, it's always an octal escape,
5369 even if it is a single digit.
5370
5371 (d) An escaped but undefined alphabetic character is taken as a literal,
5372 unless PCRE_EXTRA is set. Currently this just reserves the remaining
5373 escapes.
5374
5375 (e) {0} is now permitted. (The previous item is removed from the compiled
5376 pattern).
5377
5378 5. Changed all the names of code files so that the basic parts are no longer
5379 than 10 characters, and abolished the teeny "globals.c" file.
5380
5381 6. Changed the handling of character classes; they are now done with a 32-byte
5382 bit map always.
5383
5384 7. Added the -d and /D options to pcretest to make it possible to look at the
5385 internals of compilation without having to recompile pcre.
5386
5387
5388 Version 0.95 23-Sep-97
5389 ----------------------
5390
5391 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
5392 \x20 at the start of a run of normal characters. These were being treated as
5393 real characters, instead of the source characters being re-checked.
5394
5395
5396 Version 0.94 18-Sep-97
5397 ----------------------
5398
5399 1. The functions are now thread-safe, with the caveat that the global variables
5400 containing pointers to malloc() and free() or alternative functions are the
5401 same for all threads.
5402
5403 2. Get pcre_study() to generate a bitmap of initial characters for non-
5404 anchored patterns when this is possible, and use it if passed to pcre_exec().
5405
5406
5407 Version 0.93 15-Sep-97
5408 ----------------------
5409
5410 1. /(b)|(:+)/ was computing an incorrect first character.
5411
5412 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5413 but not actually doing anything yet.
5414
5415 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5416 as Perl does (e.g. [-az] or [az-]).
5417
5418 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5419 all possible positions.
5420
5421 5. Split up into different modules to avoid including unneeded functions in a
5422 compiled binary. However, compile and exec are still in one module. The "study"
5423 function is split off.
5424
5425 6. The character tables are now in a separate module whose source is generated
5426 by an auxiliary program - but can then be edited by hand if required. There are
5427 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5428 toupper() in the code.
5429
5430 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5431 make them global. Abolish the function for setting them, as the caller can now
5432 set them directly.
5433
5434
5435 Version 0.92 11-Sep-97
5436 ----------------------
5437
5438 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5439 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5440
5441 2. Caseless matching was not working in character classes if the characters in
5442 the pattern were in upper case.
5443
5444 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5445
5446 4. Make PCRE_ANCHORED public and accept as a compile option.
5447
5448 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5449 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5450 pass them.
5451
5452 6. Give an error if bad option bits passed at compile or run time.
5453
5454 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5455 pcretest to cause it to pass that flag.
5456
5457 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5458 options, and the first character, if set.
5459
5460 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5461
5462
5463 Version 0.91 10-Sep-97
5464 ----------------------
5465
5466 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5467 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5468
5469 2. PCRE was looping on encountering an indefinitely repeated back reference to
5470 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5471 Perl does - treats the match as successful.
5472
5473 ****

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

  ViewVC Help
Powered by ViewVC 1.1.5