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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 115 - (show annotations) (download)
Fri Mar 9 12:23:37 2007 UTC (7 years, 1 month ago) by ph10
File size: 123620 byte(s)
Fix two obscure, but nasty bugs concerned with caseless matching with Unicode 
property support.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.1 05-Mar-07
5 ---------------------
6
7 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
8 that is more "standard", making use of automake and other autotools. There
9 is some re-arrangement of the files and adjustment of comments consequent
10 on this.
11
12 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
13 for recursive directory scanning broke on some systems because the files
14 are not scanned in any specific order and on different systems the order
15 was different. A call to "sort" has been inserted into RunGrepTest for the
16 approprate test as a short-term fix. In the longer term there may be an
17 alternative.
18
19 3. I had an email from Eric Raymond about problems translating some of PCRE's
20 man pages to HTML (despite the fact that I distribute HTML pages, some
21 people do their own conversions for various reasons). The problems
22 concerned the use of low-level troff macros .br and .in. I have therefore
23 removed all such uses from the man pages (some were redundant, some could
24 be replaced by .nf/.fi pairs). The 132html script that I use to generate
25 HTML has been updated to handle .nf/.fi and to complain if it encounters
26 .br or .in.
27
28 4. Updated comments in configure.ac that get placed in config.h.in and also
29 arranged for config.h to be included in the distribution, with the name
30 config.h.generic, for the benefit of those who have to compile without
31 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
32
33 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
34 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
35 makevp.bat for the current PCRE, using the additional files !compile.txt,
36 !linklib.txt, and pcregexp.pas.
37
38 6. A Windows user reported a minor discrepancy with test 2, which turned out
39 to be caused by a trailing space on an input line that had got lost in his
40 copy. The trailing space was an accident, so I've just removed it.
41
42 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
43 that is needed.
44
45 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
46 as "const" (a) because they are and (b) because it helps the PHP
47 maintainers who have recently made a script to detect big data structures
48 in the php code that should be moved to the .rodata section. I remembered
49 to update Builducptable as well, so it won't revert if ucptable.h is ever
50 re-created.
51
52 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
53 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
54 order to be able to cut out the UTF-8 tables in the latter when UTF-8
55 support is not required. This saves 1.5-2K of code, which is important in
56 some applications.
57
58 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
59 so as not to refer to the tables, even though these functions will never be
60 called when UTF-8 support is disabled. Otherwise there are problems with a
61 shared library.
62
63 10. The emulated memmove() function in pcre_internal.h (provided for those
64 environments that have neither memmove() nor bcopy()) was defining its
65 arguments as char * instead of void *.
66
67 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
68 and Detrail to create the HTML documentation, the .txt form of the man
69 pages, and it removes trailing spaces from listed files. It also creates
70 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
71 case, it wraps all the #defines with #ifndefs. This script should be run
72 before "make dist".
73
74 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
75 with Unicode property support.
76
77 (a) For a maximizing quantifier, if the two different cases of the
78 character were of different lengths in their UTF-8 codings (there are
79 some cases like this - I found 11), and the matching function had to
80 back up over a mixture of the two cases, it incorrectly assumed they
81 were both the same length.
82
83 (b) When PCRE was configured to use the heap rather than the stack for
84 recursion during matching, it was not correctly preserving the data for
85 the other case of a UTF-8 character when checking ahead for a match
86 while processing a minimizing repeat. If the check also involved
87 matching a wide character, but failed, corruption could cause an
88 erroneous result when trying to check for a repeat of the original
89 character.
90
91
92 Version 7.0 19-Dec-06
93 ---------------------
94
95 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
96 moving to gcc 4.1.1.
97
98 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
99 sys/time.h, which is documented as needed for this function. It doesn't
100 seem to matter on Linux, but it showed up on some releases of OS X.
101
102 3. It seems that there are systems where bytes whose values are greater than
103 127 match isprint() in the "C" locale. The "C" locale should be the
104 default when a C program starts up. In most systems, only ASCII printing
105 characters match isprint(). This difference caused the output from pcretest
106 to vary, making some of the tests fail. I have changed pcretest so that:
107
108 (a) When it is outputting text in the compiled version of a pattern, bytes
109 other than 32-126 are always shown as hex escapes.
110
111 (b) When it is outputting text that is a matched part of a subject string,
112 it does the same, unless a different locale has been set for the match
113 (using the /L modifier). In this case, it uses isprint() to decide.
114
115 4. Fixed a major bug that caused incorrect computation of the amount of memory
116 required for a compiled pattern when options that changed within the
117 pattern affected the logic of the preliminary scan that determines the
118 length. The relevant options are -x, and -i in UTF-8 mode. The result was
119 that the computed length was too small. The symptoms of this bug were
120 either the PCRE error "internal error: code overflow" from pcre_compile(),
121 or a glibc crash with a message such as "pcretest: free(): invalid next
122 size (fast)". Examples of patterns that provoked this bug (shown in
123 pcretest format) are:
124
125 /(?-x: )/x
126 /(?x)(?-x: \s*#\s*)/
127 /((?i)[\x{c0}])/8
128 /(?i:[\x{c0}])/8
129
130 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
131 is now done differently.
132
133 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
134 wrapper classes; (b) implement a new function in the C++ scanner that is
135 more efficient than the old way of doing things because it avoids levels of
136 recursion in the regex matching; (c) add a paragraph to the documentation
137 for the FullMatch() function.
138
139 6. The escape sequence \n was being treated as whatever was defined as
140 "newline". Not only was this contrary to the documentation, which states
141 that \n is character 10 (hex 0A), but it also went horribly wrong when
142 "newline" was defined as CRLF. This has been fixed.
143
144 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
145 was being set to -1 for the "end of line" case (supposedly a value that no
146 character can have). Though this value is never used (the check for end of
147 line is "zero bytes in current character"), it caused compiler complaints.
148 I've changed it to 0xffffffff.
149
150 8. In pcre_version.c, the version string was being built by a sequence of
151 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
152 string (as it is for production releases) called a macro with an empty
153 argument. The C standard says the result of this is undefined. The gcc
154 compiler treats it as an empty string (which was what was wanted) but it is
155 reported that Visual C gives an error. The source has been hacked around to
156 avoid this problem.
157
158 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
159 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
160 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
161 of them did).
162
163 10. Originally, pcretest opened its input and output without "b"; then I was
164 told that "b" was needed in some environments, so it was added for release
165 5.0 to both the input and output. (It makes no difference on Unix-like
166 systems.) Later I was told that it is wrong for the input on Windows. I've
167 now abstracted the modes into two macros, to make it easier to fiddle with
168 them, and removed "b" from the input mode under Windows.
169
170 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
171
172 12. Added -help and --help to pcretest as an official way of being reminded
173 of the options.
174
175 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
176 and pcrecpp.cc because they annoy compilers at high warning levels.
177
178 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
179
180 15. Fixed an occurrence of == in configure.ac that should have been = (shell
181 scripts are not C programs :-) and which was not noticed because it works
182 on Linux.
183
184 16. pcretest is supposed to handle any length of pattern and data line (as one
185 line or as a continued sequence of lines) by extending its input buffer if
186 necessary. This feature was broken for very long pattern lines, leading to
187 a string of junk being passed to pcre_compile() if the pattern was longer
188 than about 50K.
189
190 17. I have done a major re-factoring of the way pcre_compile() computes the
191 amount of memory needed for a compiled pattern. Previously, there was code
192 that made a preliminary scan of the pattern in order to do this. That was
193 OK when PCRE was new, but as the facilities have expanded, it has become
194 harder and harder to keep it in step with the real compile phase, and there
195 have been a number of bugs (see for example, 4 above). I have now found a
196 cunning way of running the real compile function in a "fake" mode that
197 enables it to compute how much memory it would need, while actually only
198 ever using a few hundred bytes of working memory and without too many
199 tests of the mode. This should make future maintenance and development
200 easier. A side effect of this work is that the limit of 200 on the nesting
201 depth of parentheses has been removed (though this was never a serious
202 limitation, I suspect). However, there is a downside: pcre_compile() now
203 runs more slowly than before (30% or more, depending on the pattern). I
204 hope this isn't a big issue. There is no effect on runtime performance.
205
206 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
207 newline (only possible for the last line of a file) and it was a
208 pattern that set a locale (followed by /Lsomething), pcretest crashed.
209
210 19. Added additional timing features to pcretest. (1) The -tm option now times
211 matching only, not compiling. (2) Both -t and -tm can be followed, as a
212 separate command line item, by a number that specifies the number of
213 repeats to use when timing. The default is 50000; this gives better
214 precision, but takes uncomfortably long for very large patterns.
215
216 20. Extended pcre_study() to be more clever in cases where a branch of a
217 subpattern has no definite first character. For example, (a*|b*)[cd] would
218 previously give no result from pcre_study(). Now it recognizes that the
219 first character must be a, b, c, or d.
220
221 21. There was an incorrect error "recursive call could loop indefinitely" if
222 a subpattern (or the entire pattern) that was being tested for matching an
223 empty string contained only one non-empty item after a nested subpattern.
224 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
225 incorrectly, because the \d was being skipped in the check.
226
227 22. The pcretest program now has a new pattern option /B and a command line
228 option -b, which is equivalent to adding /B to every pattern. This causes
229 it to show the compiled bytecode, without the additional information that
230 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
231 is the same as /B/I).
232
233 23. A new optimization is now able automatically to treat some sequences such
234 as a*b as a*+b. More specifically, if something simple (such as a character
235 or a simple class like \d) has an unlimited quantifier, and is followed by
236 something that cannot possibly match the quantified thing, the quantifier
237 is automatically "possessified".
238
239 24. A recursive reference to a subpattern whose number was greater than 39
240 went wrong under certain circumstances in UTF-8 mode. This bug could also
241 have affected the operation of pcre_study().
242
243 25. Realized that a little bit of performance could be had by replacing
244 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
245
246 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
247
248 27. Possessive quantifiers such as a++ were previously implemented by turning
249 them into atomic groups such as ($>a+). Now they have their own opcodes,
250 which improves performance. This includes the automatically created ones
251 from 23 above.
252
253 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
254 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
255 the first matched character to be a colon. This applied both to named and
256 numbered groups.
257
258 29. The ucpinternal.h header file was missing its idempotency #ifdef.
259
260 30. I was sent a "project" file called libpcre.a.dev which I understand makes
261 building PCRE on Windows easier, so I have included it in the distribution.
262
263 31. There is now a check in pcretest against a ridiculously large number being
264 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
265 loop, the loop is abandoned.
266
267 32. Forward references to subpatterns in conditions such as (?(2)...) where
268 subpattern 2 is defined later cause pcre_compile() to search forwards in
269 the pattern for the relevant set of parentheses. This search went wrong
270 when there were unescaped parentheses in a character class, parentheses
271 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
272
273 33. "Subroutine" calls and backreferences were previously restricted to
274 referencing subpatterns earlier in the regex. This restriction has now
275 been removed.
276
277 34. Added a number of extra features that are going to be in Perl 5.10. On the
278 whole, these are just syntactic alternatives for features that PCRE had
279 previously implemented using the Python syntax or my own invention. The
280 other formats are all retained for compatibility.
281
282 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
283 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
284 also .NET compatible.
285
286 (b) A recursion or subroutine call to a named group can now be defined as
287 (?&name) as well as (?P>name).
288
289 (c) A backreference to a named group can now be defined as \k<name> or
290 \k'name' as well as (?P=name). The new forms, as well as being in Perl
291 5.10, are also .NET compatible.
292
293 (d) A conditional reference to a named group can now use the syntax
294 (?(<name>) or (?('name') as well as (?(name).
295
296 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
297 groups (named and numbered) that are never evaluated inline, but can be
298 called as "subroutines" from elsewhere. In effect, the DEFINE condition
299 is always false. There may be only one alternative in such a group.
300
301 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
302 as the simple (?(R). The condition is true only if the most recent
303 recursion is that of the given number or name. It does not search out
304 through the entire recursion stack.
305
306 (g) The escape \gN or \g{N} has been added, where N is a positive or
307 negative number, specifying an absolute or relative reference.
308
309 35. Tidied to get rid of some further signed/unsigned compiler warnings and
310 some "unreachable code" warnings.
311
312 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
313 things, this adds five new scripts.
314
315 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
316 There were also incompatibilities regarding the handling of \Q..\E inside
317 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
318 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
319
320 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
321 matches an empty string, and forcibly breaks the loop. There were bugs in
322 this code in non-simple cases. For a pattern such as ^(a()*)* matched
323 against aaaa the result was just "a" rather than "aaaa", for example. Two
324 separate and independent bugs (that affected different cases) have been
325 fixed.
326
327 39. Refactored the code to abolish the use of different opcodes for small
328 capturing bracket numbers. This is a tidy that I avoided doing when I
329 removed the limit on the number of capturing brackets for 3.5 back in 2001.
330 The new approach is not only tidier, it makes it possible to reduce the
331 memory needed to fix the previous bug (38).
332
333 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
334 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
335 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
336 mode.
337
338 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
339 report.
340
341 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
342 copy construction and assignment in the C++ wrapper.
343
344 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
345 couple of bugs that could have given wrong results in the "--newline=crlf"
346 case.
347
348 44. Added a number of casts and did some reorganization of signed/unsigned int
349 variables following suggestions from Dair Grant. Also renamed the variable
350 "this" as "item" because it is a C++ keyword.
351
352 45. Arranged for dftables to add
353
354 #include "pcre_internal.h"
355
356 to pcre_chartables.c because without it, gcc 4.x may remove the array
357 definition from the final binary if PCRE is built into a static library and
358 dead code stripping is activated.
359
360 46. For an unanchored pattern, if a match attempt fails at the start of a
361 newline sequence, and the newline setting is CRLF or ANY, and the next two
362 characters are CRLF, advance by two characters instead of one.
363
364
365 Version 6.7 04-Jul-06
366 ---------------------
367
368 1. In order to handle tests when input lines are enormously long, pcretest has
369 been re-factored so that it automatically extends its buffers when
370 necessary. The code is crude, but this _is_ just a test program. The
371 default size has been increased from 32K to 50K.
372
373 2. The code in pcre_study() was using the value of the re argument before
374 testing it for NULL. (Of course, in any sensible call of the function, it
375 won't be NULL.)
376
377 3. The memmove() emulation function in pcre_internal.h, which is used on
378 systems that lack both memmove() and bcopy() - that is, hardly ever -
379 was missing a "static" storage class specifier.
380
381 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
382 containing an extended class (one that cannot be represented by a bitmap
383 because it contains high-valued characters or Unicode property items, e.g.
384 [\pZ]). Almost always one would set UTF-8 mode when processing such a
385 pattern, but PCRE should not loop if you do not (it no longer does).
386 [Detail: two cases were found: (a) a repeated subpattern containing an
387 extended class; (b) a recursive reference to a subpattern that followed a
388 previous extended class. It wasn't skipping over the extended class
389 correctly when UTF-8 mode was not set.]
390
391 5. A negated single-character class was not being recognized as fixed-length
392 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
393 compile error "lookbehind assertion is not fixed length".
394
395 6. The RunPerlTest auxiliary script was showing an unexpected difference
396 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
397 write a Perl script that can interpret lines of an input file either as
398 byte characters or as UTF-8, which is what "perltest" was being required to
399 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
400 can't do is switch easily at run time between having the "use utf8;" pragma
401 or not. In the end, I fudged it by using the RunPerlTest script to insert
402 "use utf8;" explicitly for the UTF-8 tests.
403
404 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
405 the end of the subject string, contrary to the documentation and to what
406 Perl does. This was true of both matching functions. Now it matches only at
407 the start of the subject and immediately after *internal* newlines.
408
409 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
410 a pointer to an int instead of a pointer to an unsigned long int. This
411 caused problems on 64-bit systems.
412
413 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
414 instance of the 'standard' template library not being so standard".
415
416 10. There was no check on the number of named subpatterns nor the maximum
417 length of a subpattern name. The product of these values is used to compute
418 the size of the memory block for a compiled pattern. By supplying a very
419 long subpattern name and a large number of named subpatterns, the size
420 computation could be caused to overflow. This is now prevented by limiting
421 the length of names to 32 characters, and the number of named subpatterns
422 to 10,000.
423
424 11. Subpatterns that are repeated with specific counts have to be replicated in
425 the compiled pattern. The size of memory for this was computed from the
426 length of the subpattern and the repeat count. The latter is limited to
427 65535, but there was no limit on the former, meaning that integer overflow
428 could in principle occur. The compiled length of a repeated subpattern is
429 now limited to 30,000 bytes in order to prevent this.
430
431 12. Added the optional facility to have named substrings with the same name.
432
433 13. Added the ability to use a named substring as a condition, using the
434 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
435 are numbers (not recommended). Forward references are permitted.
436
437 14. Added forward references in named backreferences (if you see what I mean).
438
439 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
440 pattern could run off the end of the subject. For example, the pattern
441 "(?s)(.{1,5})"8 did this with the subject "ab".
442
443 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
444 PCRE_CASELESS was set when matching characters that were quantified with ?
445 or *.
446
447 17. A character class other than a single negated character that had a minimum
448 but no maximum quantifier - for example [ab]{6,} - was not handled
449 correctly by pce_dfa_exec(). It would match only one character.
450
451 18. A valid (though odd) pattern that looked like a POSIX character
452 class but used an invalid character after [ (for example [[,abc,]]) caused
453 pcre_compile() to give the error "Failed: internal error: code overflow" or
454 in some cases to crash with a glibc free() error. This could even happen if
455 the pattern terminated after [[ but there just happened to be a sequence of
456 letters, a binary zero, and a closing ] in the memory that followed.
457
458 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
459 over the years. Originally (before any Unicode support), just the bottom 8
460 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
461 output from "man perlunicode" includes this:
462
463 The regular expression compiler produces polymorphic opcodes. That
464 is, the pattern adapts to the data and automatically switches to
465 the Unicode character scheme when presented with Unicode data--or
466 instead uses a traditional byte scheme when presented with byte
467 data.
468
469 Sadly, a wide octal escape does not cause a switch, and in a string with
470 no other multibyte characters, these octal escapes are treated as before.
471 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
472 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
473 Unicode string.
474
475 I have not perpetrated such confusion in PCRE. Up till now, it took just
476 the bottom 8 bits, as in old Perl. I have now made octal escapes with
477 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
478 translate to the appropriate multibyte character.
479
480 29. Applied some refactoring to reduce the number of warnings from Microsoft
481 and Borland compilers. This has included removing the fudge introduced
482 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
483 a warning about an unused variable.
484
485 21. PCRE has not included VT (character 0x0b) in the set of whitespace
486 characters since release 4.0, because Perl (from release 5.004) does not.
487 [Or at least, is documented not to: some releases seem to be in conflict
488 with the documentation.] However, when a pattern was studied with
489 pcre_study() and all its branches started with \s, PCRE still included VT
490 as a possible starting character. Of course, this did no harm; it just
491 caused an unnecessary match attempt.
492
493 22. Removed a now-redundant internal flag bit that recorded the fact that case
494 dependency changed within the pattern. This was once needed for "required
495 byte" processing, but is no longer used. This recovers a now-scarce options
496 bit. Also moved the least significant internal flag bit to the most-
497 significant bit of the word, which was not previously used (hangover from
498 the days when it was an int rather than a uint) to free up another bit for
499 the future.
500
501 23. Added support for CRLF line endings as well as CR and LF. As well as the
502 default being selectable at build time, it can now be changed at runtime
503 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
504 specify that it is scanning data with non-default line endings.
505
506 24. Changed the definition of CXXLINK to make it agree with the definition of
507 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
508
509 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
510 recursions. This makes a big different to stack usage for some patterns.
511
512 26. If a subpattern containing a named recursion or subroutine reference such
513 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
514 the space required for the compiled pattern went wrong and gave too small a
515 value. Depending on the environment, this could lead to "Failed: internal
516 error: code overflow at offset 49" or "glibc detected double free or
517 corruption" errors.
518
519 27. Applied patches from Google (a) to support the new newline modes and (b) to
520 advance over multibyte UTF-8 characters in GlobalReplace.
521
522 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
523 difference for some implementation of PCRE in some Windows version.
524
525 29. Added some extra testing facilities to pcretest:
526
527 \q<number> in a data line sets the "match limit" value
528 \Q<number> in a data line sets the "match recursion limt" value
529 -S <number> sets the stack size, where <number> is in megabytes
530
531 The -S option isn't available for Windows.
532
533
534 Version 6.6 06-Feb-06
535 ---------------------
536
537 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
538 in pcreposix.h. I have copied the definition from pcre.h.
539
540 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
541 because pcre.h is no longer a built file.
542
543 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
544 not normally included in the compiled code.
545
546
547 Version 6.5 01-Feb-06
548 ---------------------
549
550 1. When using the partial match feature with pcre_dfa_exec(), it was not
551 anchoring the second and subsequent partial matches at the new starting
552 point. This could lead to incorrect results. For example, with the pattern
553 /1234/, partially matching against "123" and then "a4" gave a match.
554
555 2. Changes to pcregrep:
556
557 (a) All non-match returns from pcre_exec() were being treated as failures
558 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
559 error message is output. Some extra information is given for the
560 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
561 probably the only errors that are likely to be caused by users (by
562 specifying a regex that has nested indefinite repeats, for instance).
563 If there are more than 20 of these errors, pcregrep is abandoned.
564
565 (b) A binary zero was treated as data while matching, but terminated the
566 output line if it was written out. This has been fixed: binary zeroes
567 are now no different to any other data bytes.
568
569 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
570 used to set a locale for matching. The --locale=xxxx long option has
571 been added (no short equivalent) to specify a locale explicitly on the
572 pcregrep command, overriding the environment variables.
573
574 (d) When -B was used with -n, some line numbers in the output were one less
575 than they should have been.
576
577 (e) Added the -o (--only-matching) option.
578
579 (f) If -A or -C was used with -c (count only), some lines of context were
580 accidentally printed for the final match.
581
582 (g) Added the -H (--with-filename) option.
583
584 (h) The combination of options -rh failed to suppress file names for files
585 that were found from directory arguments.
586
587 (i) Added the -D (--devices) and -d (--directories) options.
588
589 (j) Added the -F (--fixed-strings) option.
590
591 (k) Allow "-" to be used as a file name for -f as well as for a data file.
592
593 (l) Added the --colo(u)r option.
594
595 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
596 is not present by default.
597
598 3. A nasty bug was discovered in the handling of recursive patterns, that is,
599 items such as (?R) or (?1), when the recursion could match a number of
600 alternatives. If it matched one of the alternatives, but subsequently,
601 outside the recursion, there was a failure, the code tried to back up into
602 the recursion. However, because of the way PCRE is implemented, this is not
603 possible, and the result was an incorrect result from the match.
604
605 In order to prevent this happening, the specification of recursion has
606 been changed so that all such subpatterns are automatically treated as
607 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
608
609 4. I had overlooked the fact that, in some locales, there are characters for
610 which isalpha() is true but neither isupper() nor islower() are true. In
611 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
612 and ordfeminine) are like this. This affected the treatment of \w and \W
613 when they appeared in character classes, but not when they appeared outside
614 a character class. The bit map for "word" characters is now created
615 separately from the results of isalnum() instead of just taking it from the
616 upper, lower, and digit maps. (Plus the underscore character, of course.)
617
618 5. The above bug also affected the handling of POSIX character classes such as
619 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
620 permanent tables. Instead, the bit maps for such a class were previously
621 created as the appropriate unions of the upper, lower, and digit bitmaps.
622 Now they are created by subtraction from the [[:word:]] class, which has
623 its own bitmap.
624
625 6. The [[:blank:]] character class matches horizontal, but not vertical space.
626 It is created by subtracting the vertical space characters (\x09, \x0a,
627 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
628 subtraction was done in the overall bitmap for a character class, meaning
629 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
630 be recognized. This bug has been fixed.
631
632 7. Patches from the folks at Google:
633
634 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
635 real life, but is still worth protecting against".
636
637 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
638 regular expressions".
639
640 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
641 have it.
642
643 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
644 "configure" and the latter not, in order to fix a problem somebody had
645 with compiling the Arg class on HP-UX.
646
647 (e) Improve the error-handling of the C++ wrapper a little bit.
648
649 (f) New tests for checking recursion limiting.
650
651 8. The pcre_memmove() function, which is used only if the environment does not
652 have a standard memmove() function (and is therefore rarely compiled),
653 contained two bugs: (a) use of int instead of size_t, and (b) it was not
654 returning a result (though PCRE never actually uses the result).
655
656 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
657 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
658 returned instead of calling malloc() with an overflowing number that would
659 most likely cause subsequent chaos.
660
661 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
662
663 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
664 with this option is matched, the nmatch and pmatch options of regexec() are
665 ignored.
666
667 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
668 provided in case anyone wants to the the POSIX interface with UTF-8
669 strings.
670
671 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
672 C++ linking (needed for some HP-UX environments).
673
674 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
675 (unused parameter) and in the pcre_printint() function (omitted "default"
676 switch label when the default is to do nothing).
677
678 15. Added some code to make it possible, when PCRE is compiled as a C++
679 library, to replace subject pointers for pcre_exec() with a smart pointer
680 class, thus making it possible to process discontinuous strings.
681
682 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
683 much the same function. They were added by different people who were trying
684 to make PCRE easy to compile on non-Unix systems. It has been suggested
685 that PCRE_EXPORT be abolished now that there is more automatic apparatus
686 for compiling on Windows systems. I have therefore replaced it with
687 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
688 defaults to "extern" for C or "extern C" for C++, which works fine on
689 Unix-like systems. It is now possible to override the value of PCRE_DATA_
690 SCOPE with something explicit in config.h. In addition:
691
692 (a) pcreposix.h still had just "extern" instead of either of these macros;
693 I have replaced it with PCRE_DATA_SCOPE.
694
695 (b) Functions such as _pcre_xclass(), which are internal to the library,
696 but external in the C sense, all had PCRE_EXPORT in their definitions.
697 This is apparently wrong for the Windows case, so I have removed it.
698 (It makes no difference on Unix-like systems.)
699
700 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
701 of recursive calls to match(). This is different to MATCH_LIMIT because
702 that limits the total number of calls to match(), not all of which increase
703 the depth of recursion. Limiting the recursion depth limits the amount of
704 stack (or heap if NO_RECURSE is set) that is used. The default can be set
705 when PCRE is compiled, and changed at run time. A patch from Google adds
706 this functionality to the C++ interface.
707
708 18. Changes to the handling of Unicode character properties:
709
710 (a) Updated the table to Unicode 4.1.0.
711
712 (b) Recognize characters that are not in the table as "Cn" (undefined).
713
714 (c) I revised the way the table is implemented to a much improved format
715 which includes recognition of ranges. It now supports the ranges that
716 are defined in UnicodeData.txt, and it also amalgamates other
717 characters into ranges. This has reduced the number of entries in the
718 table from around 16,000 to around 3,000, thus reducing its size
719 considerably. I realized I did not need to use a tree structure after
720 all - a binary chop search is just as efficient. Having reduced the
721 number of entries, I extended their size from 6 bytes to 8 bytes to
722 allow for more data.
723
724 (d) Added support for Unicode script names via properties such as \p{Han}.
725
726 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
727 matching that character.
728
729 20. When matching a repeated Unicode property with a minimum greater than zero,
730 (for example \pL{2,}), PCRE could look past the end of the subject if it
731 reached it while seeking the minimum number of characters. This could
732 happen only if some of the characters were more than one byte long, because
733 there is a check for at least the minimum number of bytes.
734
735 21. Refactored the implementation of \p and \P so as to be more general, to
736 allow for more different types of property in future. This has changed the
737 compiled form incompatibly. Anybody with saved compiled patterns that use
738 \p or \P will have to recompile them.
739
740 22. Added "Any" and "L&" to the supported property types.
741
742 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
743 but give a compile time error if the value is greater than 0xff.
744
745 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
746 accidentally not being installed or uninstalled.
747
748 25. The pcre.h file was built from pcre.h.in, but the only changes that were
749 made were to insert the current release number. This seemed silly, because
750 it made things harder for people building PCRE on systems that don't run
751 "configure". I have turned pcre.h into a distributed file, no longer built
752 by "configure", with the version identification directly included. There is
753 no longer a pcre.h.in file.
754
755 However, this change necessitated a change to the pcre-config script as
756 well. It is built from pcre-config.in, and one of the substitutions was the
757 release number. I have updated configure.ac so that ./configure now finds
758 the release number by grepping pcre.h.
759
760 26. Added the ability to run the tests under valgrind.
761
762
763 Version 6.4 05-Sep-05
764 ---------------------
765
766 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
767 "--" to be printed when multiple files were scanned, even when none of the
768 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
769 consider it to be a bug, and have restored the previous behaviour.
770
771 2. A couple of code tidies to get rid of compiler warnings.
772
773 3. The pcretest program used to cheat by referring to symbols in the library
774 whose names begin with _pcre_. These are internal symbols that are not
775 really supposed to be visible externally, and in some environments it is
776 possible to suppress them. The cheating is now confined to including
777 certain files from the library's source, which is a bit cleaner.
778
779 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
780 file's purpose clearer.
781
782 5. Reorganized pcre_ucp_findchar().
783
784
785 Version 6.3 15-Aug-05
786 ---------------------
787
788 1. The file libpcre.pc.in did not have general read permission in the tarball.
789
790 2. There were some problems when building without C++ support:
791
792 (a) If C++ support was not built, "make install" and "make test" still
793 tried to test it.
794
795 (b) There were problems when the value of CXX was explicitly set. Some
796 changes have been made to try to fix these, and ...
797
798 (c) --disable-cpp can now be used to explicitly disable C++ support.
799
800 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
801 backslash in a target when C++ was disabled. This confuses some
802 versions of "make", apparently. Using an intermediate variable solves
803 this. (Same for CPP_LOBJ.)
804
805 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
806 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
807 necessary on certain architectures.
808
809 4. Added a setting of -export-symbols-regex to the link command to remove
810 those symbols that are exported in the C sense, but actually are local
811 within the library, and not documented. Their names all begin with
812 "_pcre_". This is not a perfect job, because (a) we have to except some
813 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
814 available (and never for static libraries). I have made a note to try to
815 find a way round (a) in the future.
816
817
818 Version 6.2 01-Aug-05
819 ---------------------
820
821 1. There was no test for integer overflow of quantifier values. A construction
822 such as {1111111111111111} would give undefined results. What is worse, if
823 a minimum quantifier for a parenthesized subpattern overflowed and became
824 negative, the calculation of the memory size went wrong. This could have
825 led to memory overwriting.
826
827 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
828
829 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
830 operating environments where this matters.
831
832 4. Applied Giuseppe Maxia's patch to add additional features for controlling
833 PCRE options from within the C++ wrapper.
834
835 5. Named capturing subpatterns were not being correctly counted when a pattern
836 was compiled. This caused two problems: (a) If there were more than 100
837 such subpatterns, the calculation of the memory needed for the whole
838 compiled pattern went wrong, leading to an overflow error. (b) Numerical
839 back references of the form \12, where the number was greater than 9, were
840 not recognized as back references, even though there were sufficient
841 previous subpatterns.
842
843 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
844 versions of gcc, e.g. 2.95.4.
845
846
847 Version 6.1 21-Jun-05
848 ---------------------
849
850 1. There was one reference to the variable "posix" in pcretest.c that was not
851 surrounded by "#if !defined NOPOSIX".
852
853 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
854 the cross-check on the old pcre_info() function, for the benefit of the
855 cut-down version of PCRE that is currently imported into Exim.
856
857 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
858 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
859 patterns that start (?i) but I don't think that matters. The use of (?i) is
860 just an example; this all applies to the other options as well.
861
862 4. Since libtool seems to echo the compile commands it is issuing, the output
863 from "make" can be reduced a bit by putting "@" in front of each libtool
864 compile command.
865
866 5. Patch from the folks at Google for configure.in to be a bit more thorough
867 in checking for a suitable C++ installation before trying to compile the
868 C++ stuff. This should fix a reported problem when a compiler was present,
869 but no suitable headers.
870
871 6. The man pages all had just "PCRE" as their title. I have changed them to
872 be the relevant file name. I have also arranged that these names are
873 retained in the file doc/pcre.txt, which is a concatenation in text format
874 of all the man pages except the little individual ones for each function.
875
876 7. The NON-UNIX-USE file had not been updated for the different set of source
877 files that come with release 6. I also added a few comments about the C++
878 wrapper.
879
880
881 Version 6.0 07-Jun-05
882 ---------------------
883
884 1. Some minor internal re-organization to help with my DFA experiments.
885
886 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
887 didn't matter for the library itself when fully configured, but did matter
888 when compiling without UCP support, or within Exim, where the ucp files are
889 not imported.
890
891 3. Refactoring of the library code to split up the various functions into
892 different source modules. The addition of the new DFA matching code (see
893 below) to a single monolithic source would have made it really too
894 unwieldy, quite apart from causing all the code to be include in a
895 statically linked application, when only some functions are used. This is
896 relevant even without the DFA addition now that patterns can be compiled in
897 one application and matched in another.
898
899 The downside of splitting up is that there have to be some external
900 functions and data tables that are used internally in different modules of
901 the library but which are not part of the API. These have all had their
902 names changed to start with "_pcre_" so that they are unlikely to clash
903 with other external names.
904
905 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
906 a different (DFA) algorithm. Although it is slower than the original
907 function, it does have some advantages for certain types of matching
908 problem.
909
910 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
911 including restarting after a partial match.
912
913 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
914 defined when compiling for Windows was sent to me. I have put it into the
915 code, though I have no means of testing or verifying it.
916
917 7. Added the pcre_refcount() auxiliary function.
918
919 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
920 match before or at the first newline in the subject string. In pcretest,
921 the /f option on a pattern can be used to set this.
922
923 9. A repeated \w when used in UTF-8 mode with characters greater than 256
924 would behave wrongly. This has been present in PCRE since release 4.0.
925
926 10. A number of changes to the pcregrep command:
927
928 (a) Refactored how -x works; insert ^(...)$ instead of setting
929 PCRE_ANCHORED and checking the length, in preparation for adding
930 something similar for -w.
931
932 (b) Added the -w (match as a word) option.
933
934 (c) Refactored the way lines are read and buffered so as to have more
935 than one at a time available.
936
937 (d) Implemented a pcregrep test script.
938
939 (e) Added the -M (multiline match) option. This allows patterns to match
940 over several lines of the subject. The buffering ensures that at least
941 8K, or the rest of the document (whichever is the shorter) is available
942 for matching (and similarly the previous 8K for lookbehind assertions).
943
944 (f) Changed the --help output so that it now says
945
946 -w, --word-regex(p)
947
948 instead of two lines, one with "regex" and the other with "regexp"
949 because that confused at least one person since the short forms are the
950 same. (This required a bit of code, as the output is generated
951 automatically from a table. It wasn't just a text change.)
952
953 (g) -- can be used to terminate pcregrep options if the next thing isn't an
954 option but starts with a hyphen. Could be a pattern or a path name
955 starting with a hyphen, for instance.
956
957 (h) "-" can be given as a file name to represent stdin.
958
959 (i) When file names are being printed, "(standard input)" is used for
960 the standard input, for compatibility with GNU grep. Previously
961 "<stdin>" was used.
962
963 (j) The option --label=xxx can be used to supply a name to be used for
964 stdin when file names are being printed. There is no short form.
965
966 (k) Re-factored the options decoding logic because we are going to add
967 two more options that take data. Such options can now be given in four
968 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
969
970 (l) Added the -A, -B, and -C options for requesting that lines of context
971 around matches be printed.
972
973 (m) Added the -L option to print the names of files that do not contain
974 any matching lines, that is, the complement of -l.
975
976 (n) The return code is 2 if any file cannot be opened, but pcregrep does
977 continue to scan other files.
978
979 (o) The -s option was incorrectly implemented. For compatibility with other
980 greps, it now suppresses the error message for a non-existent or non-
981 accessible file (but not the return code). There is a new option called
982 -q that suppresses the output of matching lines, which was what -s was
983 previously doing.
984
985 (p) Added --include and --exclude options to specify files for inclusion
986 and exclusion when recursing.
987
988 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
989 Hopefully, it now does.
990
991 12. Missing cast in pcre_study().
992
993 13. Added an "uninstall" target to the makefile.
994
995 14. Replaced "extern" in the function prototypes in Makefile.in with
996 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
997 world, but is set differently for Windows.
998
999 15. Added a second compiling function called pcre_compile2(). The only
1000 difference is that it has an extra argument, which is a pointer to an
1001 integer error code. When there is a compile-time failure, this is set
1002 non-zero, in addition to the error test pointer being set to point to an
1003 error message. The new argument may be NULL if no error number is required
1004 (but then you may as well call pcre_compile(), which is now just a
1005 wrapper). This facility is provided because some applications need a
1006 numeric error indication, but it has also enabled me to tidy up the way
1007 compile-time errors are handled in the POSIX wrapper.
1008
1009 16. Added VPATH=.libs to the makefile; this should help when building with one
1010 prefix path and installing with another. (Or so I'm told by someone who
1011 knows more about this stuff than I do.)
1012
1013 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1014 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1015 match everything, including newlines. This is not POSIX-compatible, but
1016 somebody wanted the feature. From pcretest it can be activated by using
1017 both the P and the s flags.
1018
1019 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1020
1021 19. libpcre.pc was being incorrectly installed as executable.
1022
1023 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1024 it now also looks for '\r' so that it will work unmodified on Windows.
1025
1026 21. Added Google's contributed C++ wrapper to the distribution.
1027
1028 22. Added some untidy missing memory free() calls in pcretest, to keep
1029 Electric Fence happy when testing.
1030
1031
1032
1033 Version 5.0 13-Sep-04
1034 ---------------------
1035
1036 1. Internal change: literal characters are no longer packed up into items
1037 containing multiple characters in a single byte-string. Each character
1038 is now matched using a separate opcode. However, there may be more than one
1039 byte in the character in UTF-8 mode.
1040
1041 2. The pcre_callout_block structure has two new fields: pattern_position and
1042 next_item_length. These contain the offset in the pattern to the next match
1043 item, and its length, respectively.
1044
1045 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1046 insertion of callouts before each pattern item. Added the /C option to
1047 pcretest to make use of this.
1048
1049 4. On the advice of a Windows user, the lines
1050
1051 #if defined(_WIN32) || defined(WIN32)
1052 _setmode( _fileno( stdout ), 0x8000 );
1053 #endif /* defined(_WIN32) || defined(WIN32) */
1054
1055 have been added to the source of pcretest. This apparently does useful
1056 magic in relation to line terminators.
1057
1058 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1059 for the benefit of those environments where the "b" makes a difference.
1060
1061 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1062 to know about it. I have put a hack into configure.in that adds in code
1063 to set GCC=yes if CC=icc. This seems to end up at a point in the
1064 generated configure script that is early enough to affect the setting of
1065 compiler options, which is what is needed, but I have no means of testing
1066 whether it really works. (The user who reported this had patched the
1067 generated configure script, which of course I cannot do.)
1068
1069 LATER: After change 22 below (new libtool files), the configure script
1070 seems to know about icc (and also ecc). Therefore, I have commented out
1071 this hack in configure.in.
1072
1073 7. Added support for pkg-config (2 patches were sent in).
1074
1075 8. Negated POSIX character classes that used a combination of internal tables
1076 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1077 [[:^ascii]]. Typically, they would match almost any characters. The other
1078 POSIX classes were not broken in this way.
1079
1080 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1081 to find the match, as PCRE was deluded into thinking that the match had to
1082 start at the start point or following a newline. The same bug applied to
1083 patterns with negative forward assertions or any backward assertions
1084 preceding ".*" at the start, unless the pattern required a fixed first
1085 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1086
1087 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1088 starting at the last subject character, bytes beyond the end of the subject
1089 string were read.
1090
1091 11. Renamed the variable "class" as "classbits" to make life easier for C++
1092 users. (Previously there was a macro definition, but it apparently wasn't
1093 enough.)
1094
1095 12. Added the new field "tables" to the extra data so that tables can be passed
1096 in at exec time, or the internal tables can be re-selected. This allows
1097 a compiled regex to be saved and re-used at a later time by a different
1098 program that might have everything at different addresses.
1099
1100 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1101 -R library as well as a -L library.
1102
1103 14. The debugging options of pcretest (-d on the command line or D on a
1104 pattern) showed incorrect output for anything following an extended class
1105 that contained multibyte characters and which was followed by a quantifier.
1106
1107 15. Added optional support for general category Unicode character properties
1108 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1109 support. It adds about 90K to the size of the library. The meanings of the
1110 inbuilt class escapes such as \d and \s have NOT been changed.
1111
1112 16. Updated pcredemo.c to include calls to free() to release the memory for the
1113 compiled pattern.
1114
1115 17. The generated file chartables.c was being created in the source directory
1116 instead of in the building directory. This caused the build to fail if the
1117 source directory was different from the building directory, and was
1118 read-only.
1119
1120 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1121 file. No doubt somebody will tell me if they don't make sense... Also added
1122 Dan Mooney's comments about building on OpenVMS.
1123
1124 19. Added support for partial matching via the PCRE_PARTIAL option for
1125 pcre_exec() and the \P data escape in pcretest.
1126
1127 20. Extended pcretest with 3 new pattern features:
1128
1129 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1130 write the compiled pattern to the file whose name is "rest-of-line".
1131 This is a straight binary dump of the data, with the saved pointer to
1132 the character tables forced to be NULL. The study data, if any, is
1133 written too. After writing, pcretest reads a new pattern.
1134
1135 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1136 compiled pattern from the given file. There must not be any
1137 occurrences of "<" in the file name (pretty unlikely); if there are,
1138 pcretest will instead treat the initial "<" as a pattern delimiter.
1139 After reading in the pattern, pcretest goes on to read data lines as
1140 usual.
1141
1142 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1143 and 16-bit fields in a compiled pattern, to simulate a pattern that
1144 was compiled on a host of opposite endianness.
1145
1146 21. The pcre-exec() function can now cope with patterns that were compiled on
1147 hosts of opposite endianness, with this restriction:
1148
1149 As for any compiled expression that is saved and used later, the tables
1150 pointer field cannot be preserved; the extra_data field in the arguments
1151 to pcre_exec() should be used to pass in a tables address if a value
1152 other than the default internal tables were used at compile time.
1153
1154 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1155 now diagnosed as an error. Previously, most of the time, a negative number
1156 would have been treated as zero, but if in addition "ovector" was passed as
1157 NULL, a crash could occur.
1158
1159 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1160 new versions from the libtool 1.5 distribution (the last one is a copy of
1161 a file called libtool.m4). This seems to have fixed the need to patch
1162 "configure" to support Darwin 1.3 (which I used to do). However, I still
1163 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1164 workstation).
1165
1166 24. Changed the PCRE licence to be the more standard "BSD" licence.
1167
1168
1169 Version 4.5 01-Dec-03
1170 ---------------------
1171
1172 1. There has been some re-arrangement of the code for the match() function so
1173 that it can be compiled in a version that does not call itself recursively.
1174 Instead, it keeps those local variables that need separate instances for
1175 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1176 needs to "recurse". Keeping track of where control must go is done by means
1177 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1178 hide most of the details from the main code, and operates only if
1179 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1180 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1181 operating.
1182
1183 To make it easier for callers to provide specially tailored get/free
1184 functions for this usage, two new functions, pcre_stack_malloc, and
1185 pcre_stack_free, are used. They are always called in strict stacking order,
1186 and the size of block requested is always the same.
1187
1188 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1189 PCRE has been compiled to use the stack or the heap for recursion. The
1190 -C option of pcretest uses this to show which version is compiled.
1191
1192 A new data escape \S, is added to pcretest; it causes the amounts of store
1193 obtained and freed by both kinds of malloc/free at match time to be added
1194 to the output.
1195
1196 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1197 what's available on my current Linux desktop machine.
1198
1199 3. When matching a UTF-8 string, the test for a valid string at the start has
1200 been extended. If start_offset is not zero, PCRE now checks that it points
1201 to a byte that is the start of a UTF-8 character. If not, it returns
1202 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1203 this is necessary because there may be backward assertions in the pattern.
1204 When matching the same subject several times, it may save resources to use
1205 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1206
1207 4. The code for checking the validity of UTF-8 strings has been tightened so
1208 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1209 containing "overlong sequences".
1210
1211 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1212 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1213 should have been "&", but it just so happened that all the cases this let
1214 through by mistake were picked up later in the function.
1215
1216 6. I had used a variable called "isblank" - this is a C99 function, causing
1217 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1218
1219 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1220 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1221 so the version is shown; (c) stop "make test" echoing "./RunTest".
1222
1223 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1224
1225 9. The prototype for memmove() for systems that don't have it was using
1226 size_t, but the inclusion of the header that defines size_t was later. I've
1227 moved the #includes for the C headers earlier to avoid this.
1228
1229 10. Added some adjustments to the code to make it easier to compiler on certain
1230 special systems:
1231
1232 (a) Some "const" qualifiers were missing.
1233 (b) Added the macro EXPORT before all exported functions; by default this
1234 is defined to be empty.
1235 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1236 that it reads its output file name as an argument instead of writing
1237 to the standard output and assuming this can be redirected.
1238
1239 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1240 class containing characters with values greater than 255, PCRE compilation
1241 went into a loop.
1242
1243 12. A recursive reference to a subpattern that was within another subpattern
1244 that had a minimum quantifier of zero caused PCRE to crash. For example,
1245 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1246 recursion. If the recursively-called subpattern itself had a zero repeat,
1247 that was OK.
1248
1249 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1250 buffer into which it was copied (for escape processing) was still set at
1251 1024, so long lines caused crashes.
1252
1253 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1254 "internal error: code overflow...". This applied to any character class
1255 that was followed by a possessive quantifier.
1256
1257 15. Modified the Makefile to add libpcre.la as a prerequisite for
1258 libpcreposix.la because I was told this is needed for a parallel build to
1259 work.
1260
1261 16. If a pattern that contained .* following optional items at the start was
1262 studied, the wrong optimizing data was generated, leading to matching
1263 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1264 matching string must start with a or b or c. The correct conclusion for
1265 this pattern is that a match can start with any character.
1266
1267
1268 Version 4.4 13-Aug-03
1269 ---------------------
1270
1271 1. In UTF-8 mode, a character class containing characters with values between
1272 127 and 255 was not handled correctly if the compiled pattern was studied.
1273 In fixing this, I have also improved the studying algorithm for such
1274 classes (slightly).
1275
1276 2. Three internal functions had redundant arguments passed to them. Removal
1277 might give a very teeny performance improvement.
1278
1279 3. Documentation bug: the value of the capture_top field in a callout is *one
1280 more than* the number of the hightest numbered captured substring.
1281
1282 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1283 in incorrectly linking with a previously installed version. They now link
1284 explicitly with libpcre.la.
1285
1286 5. configure.in no longer needs to recognize Cygwin specially.
1287
1288 6. A problem in pcre.in for Windows platforms is fixed.
1289
1290 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1291 pcretest, it used to include the size of the study block as part of its
1292 output. Unfortunately, the structure contains a field that has a different
1293 size on different hardware architectures. This meant that the tests that
1294 showed this size failed. As the block is currently always of a fixed size,
1295 this information isn't actually particularly useful in pcretest output, so
1296 I have just removed it.
1297
1298 8. Three pre-processor statements accidentally did not start in column 1.
1299 Sadly, there are *still* compilers around that complain, even though
1300 standard C has not required this for well over a decade. Sigh.
1301
1302 9. In pcretest, the code for checking callouts passed small integers in the
1303 callout_data field, which is a void * field. However, some picky compilers
1304 complained about the casts involved for this on 64-bit systems. Now
1305 pcretest passes the address of the small integer instead, which should get
1306 rid of the warnings.
1307
1308 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1309 both compile and run time, and gives an error if an invalid UTF-8 sequence
1310 is found. There is a option for disabling this check in cases where the
1311 string is known to be correct and/or the maximum performance is wanted.
1312
1313 11. In response to a bug report, I changed one line in Makefile.in from
1314
1315 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1316 to
1317 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1318
1319 to look similar to other lines, but I have no way of telling whether this
1320 is the right thing to do, as I do not use Windows. No doubt I'll get told
1321 if it's wrong...
1322
1323
1324 Version 4.3 21-May-03
1325 ---------------------
1326
1327 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1328 Makefile.
1329
1330 2. Some refactoring to improve the quality of the code:
1331
1332 (i) The utf8_table... variables are now declared "const".
1333
1334 (ii) The code for \cx, which used the "case flipping" table to upper case
1335 lower case letters, now just substracts 32. This is ASCII-specific,
1336 but the whole concept of \cx is ASCII-specific, so it seems
1337 reasonable.
1338
1339 (iii) PCRE was using its character types table to recognize decimal and
1340 hexadecimal digits in the pattern. This is silly, because it handles
1341 only 0-9, a-f, and A-F, but the character types table is locale-
1342 specific, which means strange things might happen. A private
1343 table is now used for this - though it costs 256 bytes, a table is
1344 much faster than multiple explicit tests. Of course, the standard
1345 character types table is still used for matching digits in subject
1346 strings against \d.
1347
1348 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1349 ending in _t are). So I've renamed it as ESC_tee.
1350
1351 3. The first argument for regexec() in the POSIX wrapper should have been
1352 defined as "const".
1353
1354 4. Changed pcretest to use malloc() for its buffers so that they can be
1355 Electric Fenced for debugging.
1356
1357 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1358 to read one or more bytes before the start of the subject string. Often this
1359 had no effect on PCRE's behaviour, but in some circumstances it could
1360 provoke a segmentation fault.
1361
1362 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1363 to try to read one or more bytes before the start of the subject string.
1364
1365 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1366 UTF-8 support could misbehave in various ways if the subject string
1367 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1368 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1369 back over UTF-8 characters.)
1370
1371
1372 Version 4.2 14-Apr-03
1373 ---------------------
1374
1375 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1376
1377 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1378 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1379 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1380 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1381 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1382 and BUILD_EXEEXT
1383 Note: automatic setting of the BUILD variables is not yet working
1384 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1385 compile-time but not at link-time
1386 [LINK]: use for linking executables only
1387 make different versions for Windows and non-Windows
1388 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1389 libraries
1390 [LINK_FOR_BUILD]: new variable
1391 [OBJEXT]: use throughout
1392 [EXEEXT]: use throughout
1393 <winshared>: new target
1394 <wininstall>: new target
1395 <dftables.o>: use native compiler
1396 <dftables>: use native linker
1397 <install>: handle Windows platform correctly
1398 <clean>: ditto
1399 <check>: ditto
1400 copy DLL to top builddir before testing
1401
1402 As part of these changes, -no-undefined was removed again. This was reported
1403 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1404 in any case.
1405
1406 3. Some tidies to get rid of compiler warnings:
1407
1408 . In the match_data structure, match_limit was an unsigned long int, whereas
1409 match_call_count was an int. I've made them both unsigned long ints.
1410
1411 . In pcretest the fact that a const uschar * doesn't automatically cast to
1412 a void * provoked a warning.
1413
1414 . Turning on some more compiler warnings threw up some "shadow" variables
1415 and a few more missing casts.
1416
1417 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1418 option, a class that contained a single character with a value between 128
1419 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1420
1421 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1422 option, a class that contained several characters, but with at least one
1423 whose value was between 128 and 255 caused PCRE to crash.
1424
1425
1426 Version 4.1 12-Mar-03
1427 ---------------------
1428
1429 1. Compiling with gcc -pedantic found a couple of places where casts were
1430 needed, and a string in dftables.c that was longer than standard compilers are
1431 required to support.
1432
1433 2. Compiling with Sun's compiler found a few more places where the code could
1434 be tidied up in order to avoid warnings.
1435
1436 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1437 first of these names is deprecated in the latest Autoconf in favour of the name
1438 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1439 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1440 analogy I have changed it to CFLAGS_FOR_BUILD.
1441
1442 4. Added -no-undefined to the linking command in the Makefile, because this is
1443 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1444 linking step for the pcreposix library.
1445
1446 5. PCRE was failing to diagnose the case of two named groups with the same
1447 name.
1448
1449 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1450 literal character that is needed in the subject for a match, and scans along to
1451 ensure that it is present before embarking on the full matching process. This
1452 saves time in cases of nested unlimited repeats that are never going to match.
1453 Problem: the scan can take a lot of time if the subject is very long (e.g.
1454 megabytes), thus penalizing straightforward matches. It is now done only if the
1455 amount of subject to be scanned is less than 1000 bytes.
1456
1457 7. A lesser problem with the same optimization is that it was recording the
1458 first character of an anchored pattern as "needed", thus provoking a search
1459 right along the subject, even when the first match of the pattern was going to
1460 fail. The "needed" character is now not set for anchored patterns, unless it
1461 follows something in the pattern that is of non-fixed length. Thus, it still
1462 fulfils its original purpose of finding quick non-matches in cases of nested
1463 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1464
1465
1466 Version 4.0 17-Feb-03
1467 ---------------------
1468
1469 1. If a comment in an extended regex that started immediately after a meta-item
1470 extended to the end of string, PCRE compiled incorrect data. This could lead to
1471 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1472
1473 2. Moved to autoconf 2.53 and libtool 1.4.2.
1474
1475 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1476 the special perltest8 script is no longer needed - all the tests can be run
1477 from a single perltest script.
1478
1479 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1480 by \s. It has now been removed in PCRE. This means it isn't recognized as
1481 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1482 class [:space:] *does* include VT, thereby creating a mess.
1483
1484 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1485 space and tab.
1486
1487 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1488 its new features into the main test script, reducing the number of scripts.
1489
1490 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1491 were backward compatible, and made the (?i) apply to the whole pattern, as if
1492 /i were given. Now it behaves more logically, and applies the option setting
1493 only to what follows. PCRE has been changed to follow suit. However, if it
1494 finds options settings right at the start of the pattern, it extracts them into
1495 the global options, as before. Thus, they show up in the info data.
1496
1497 8. Added support for the \Q...\E escape sequence. Characters in between are
1498 treated as literals. This is slightly different from Perl in that $ and @ are
1499 also handled as literals inside the quotes. In Perl, they will cause variable
1500 interpolation. Note the following examples:
1501
1502 Pattern PCRE matches Perl matches
1503
1504 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1505 \Qabc\$xyz\E abc\$xyz abc\$xyz
1506 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1507
1508 For compatibility with Perl, \Q...\E sequences are recognized inside character
1509 classes as well as outside them.
1510
1511 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1512 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1513 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1514 signed/unsigned warnings.
1515
1516 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1517 option for pcretest, so I've replaced it by a simple function that does just
1518 that job.
1519
1520 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1521 "pcregrep -".
1522
1523 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1524 Java package. This provides some syntactic sugar for simple cases of what my
1525 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1526 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1527 item, you can use this simplified notation. Note that only makes sense with
1528 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1529 greediness, whatever the setting of the PCRE_UNGREEDY option.
1530
1531 13. A change of greediness default within a pattern was not taking effect at
1532 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1533 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1534 was abstracted outside.
1535
1536 14. PCRE now supports the \G assertion. It is true when the current matching
1537 position is at the start point of the match. This differs from \A when the
1538 starting offset is non-zero. Used with the /g option of pcretest (or similar
1539 code), it works in the same way as it does for Perl's /g option. If all
1540 alternatives of a regex begin with \G, the expression is anchored to the start
1541 match position, and the "anchored" flag is set in the compiled expression.
1542
1543 15. Some bugs concerning the handling of certain option changes within patterns
1544 have been fixed. These applied to options other than (?ims). For example,
1545 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1546 been the other way round. Some of this was related to change 7 above.
1547
1548 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1549 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1550 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1551 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1552
1553 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1554 mode. Unlike ".", it always matches newline, whatever the setting of
1555 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1556 assertions. Perl allows it, but it doesn't (in general) work because it can't
1557 calculate the length of the lookbehind. At least, that's the case for Perl
1558 5.8.0 - I've been told they are going to document that it doesn't work in
1559 future.
1560
1561 18. Added an error diagnosis for escapes that PCRE does not support: these are
1562 \L, \l, \N, \P, \p, \U, \u, and \X.
1563
1564 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1565 reading past the end of the pattern in cases such as /[abcd/.
1566
1567 20. PCRE was getting more memory than necessary for patterns with classes that
1568 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1569
1570 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1571 compiling PCRE for use with Virtual Pascal.
1572
1573 22. Small fix to the Makefile to make it work properly if the build is done
1574 outside the source tree.
1575
1576 23. Added a new extension: a condition to go with recursion. If a conditional
1577 subpattern starts with (?(R) the "true" branch is used if recursion has
1578 happened, whereas the "false" branch is used only at the top level.
1579
1580 24. When there was a very long string of literal characters (over 255 bytes
1581 without UTF support, over 250 bytes with UTF support), the computation of how
1582 much memory was required could be incorrect, leading to segfaults or other
1583 strange effects.
1584
1585 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1586 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1587 there was a subsequent back reference to those brackets. This meant that, for
1588 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1589 possible to check for precisely this case. All we can do is abandon the
1590 optimization if .* occurs inside capturing brackets when there are any back
1591 references whatsoever. (See below for a better fix that came later.)
1592
1593 26. The handling of the optimization for finding the first character of a
1594 non-anchored pattern, and for finding a character that is required later in the
1595 match were failing in some cases. This didn't break the matching; it just
1596 failed to optimize when it could. The way this is done has been re-implemented.
1597
1598 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1599
1600 28. Added a new feature that provides some of the functionality that Perl
1601 provides with (?{...}). The facility is termed a "callout". The way it is done
1602 in PCRE is for the caller to provide an optional function, by setting
1603 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1604 global variable. By default it is unset, which disables all calling out. To get
1605 the function called, the regex must include (?C) at appropriate points. This
1606 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1607 This provides a means of identifying different callout points. When PCRE
1608 reaches such a point in the regex, if pcre_callout has been set, the external
1609 function is called. It is provided with data in a structure called
1610 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1611 matching continues; if it returns a non-zero value, the match at the current
1612 point fails. However, backtracking will occur if possible. [This was changed
1613 later and other features added - see item 49 below.]
1614
1615 29. pcretest is upgraded to test the callout functionality. It provides a
1616 callout function that displays information. By default, it shows the start of
1617 the match and the current position in the text. There are some new data escapes
1618 to vary what happens:
1619
1620 \C+ in addition, show current contents of captured substrings
1621 \C- do not supply a callout function
1622 \C!n return 1 when callout number n is reached
1623 \C!n!m return 1 when callout number n is reached for the mth time
1624
1625 30. If pcregrep was called with the -l option and just a single file name, it
1626 output "<stdin>" if a match was found, instead of the file name.
1627
1628 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1629 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1630 pcre_exec(). This saves a malloc/free per call. The default value of
1631 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1632 when configuring.
1633
1634 32. The default maximum size of a compiled pattern is 64K. There have been a
1635 few cases of people hitting this limit. The code now uses macros to handle the
1636 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1637 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1638 configuring. Tests 2 and 5 work only with 2-byte links because they output
1639 debugging information about compiled patterns.
1640
1641 33. Internal code re-arrangements:
1642
1643 (a) Moved the debugging function for printing out a compiled regex into
1644 its own source file (printint.c) and used #include to pull it into
1645 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1646 separate copies.
1647
1648 (b) Defined the list of op-code names for debugging as a macro in
1649 internal.h so that it is next to the definition of the opcodes.
1650
1651 (c) Defined a table of op-code lengths for simpler skipping along compiled
1652 code. This is again a macro in internal.h so that it is next to the
1653 definition of the opcodes.
1654
1655 34. Added support for recursive calls to individual subpatterns, along the
1656 lines of Robin Houston's patch (but implemented somewhat differently).
1657
1658 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1659 allow it to read and process whole directories in Win32. This code was
1660 contributed by Lionel Fourquaux; it has not been tested by me.
1661
1662 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1663 used to name a group. Names consist of alphanumerics and underscores, and must
1664 be unique. Back references use the syntax (?P=name) and recursive calls use
1665 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1666 numbers. The function pcre_fullinfo() can be used after compilation to extract
1667 a name/number map. There are three relevant calls:
1668
1669 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1670 PCRE_INFO_NAMECOUNT yields the number of entries
1671 PCRE_INFO_NAMETABLE yields a pointer to the map.
1672
1673 The map is a vector of fixed-size entries. The size of each entry depends on
1674 the length of the longest name used. The first two bytes of each entry are the
1675 group number, most significant byte first. There follows the corresponding
1676 name, zero terminated. The names are in alphabetical order.
1677
1678 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1679 case instead of 255. Making it the same both with and without UTF-8 support
1680 means that the same test output works with both.
1681
1682 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1683 calling malloc() with a zero argument.
1684
1685 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1686 optimization. I've improved things by keeping a bitmap of backreferences with
1687 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1688 fact referenced, the optimization can be applied. It is unlikely that a
1689 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1690 the match to follow \n) will appear inside brackets with a number greater than
1691 31, but if it does, any back reference > 31 suppresses the optimization.
1692
1693 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1694 of disabling numbered capturing parentheses. Any opening parenthesis that is
1695 not followed by ? behaves as if it were followed by ?: but named parentheses
1696 can still be used for capturing (and they will acquire numbers in the usual
1697 way).
1698
1699 41. Redesigned the return codes from the match() function into yes/no/error so
1700 that errors can be passed back from deep inside the nested calls. A malloc
1701 failure while inside a recursive subpattern call now causes the
1702 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1703
1704 42. It is now possible to set a limit on the number of times the match()
1705 function is called in a call to pcre_exec(). This facility makes it possible to
1706 limit the amount of recursion and backtracking, though not in a directly
1707 obvious way, because the match() function is used in a number of different
1708 circumstances. The count starts from zero for each position in the subject
1709 string (for non-anchored patterns). The default limit is, for compatibility, a
1710 large number, namely 10 000 000. You can change this in two ways:
1711
1712 (a) When configuring PCRE before making, you can use --with-match-limit=n
1713 to set a default value for the compiled library.
1714
1715 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1716 a different value is set. See 45 below.
1717
1718 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1719
1720 43. Added a new function pcre_config(int, void *) to enable run-time extraction
1721 of things that can be changed at compile time. The first argument specifies
1722 what is wanted and the second points to where the information is to be placed.
1723 The current list of available information is:
1724
1725 PCRE_CONFIG_UTF8
1726
1727 The output is an integer that is set to one if UTF-8 support is available;
1728 otherwise it is set to zero.
1729
1730 PCRE_CONFIG_NEWLINE
1731
1732 The output is an integer that it set to the value of the code that is used for
1733 newline. It is either LF (10) or CR (13).
1734
1735 PCRE_CONFIG_LINK_SIZE
1736
1737 The output is an integer that contains the number of bytes used for internal
1738 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1739
1740 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1741
1742 The output is an integer that contains the threshold above which the POSIX
1743 interface uses malloc() for output vectors. See item 31 above.
1744
1745 PCRE_CONFIG_MATCH_LIMIT
1746
1747 The output is an unsigned integer that contains the default limit of the number
1748 of match() calls in a pcre_exec() execution. See 42 above.
1749
1750 44. pcretest has been upgraded by the addition of the -C option. This causes it
1751 to extract all the available output from the new pcre_config() function, and to
1752 output it. The program then exits immediately.
1753
1754 45. A need has arisen to pass over additional data with calls to pcre_exec() in
1755 order to support additional features. One way would have been to define
1756 pcre_exec2() (for example) with extra arguments, but this would not have been
1757 extensible, and would also have required all calls to the original function to
1758 be mapped to the new one. Instead, I have chosen to extend the mechanism that
1759 is used for passing in "extra" data from pcre_study().
1760
1761 The pcre_extra structure is now exposed and defined in pcre.h. It currently
1762 contains the following fields:
1763
1764 flags a bitmap indicating which of the following fields are set
1765 study_data opaque data from pcre_study()
1766 match_limit a way of specifying a limit on match() calls for a specific
1767 call to pcre_exec()
1768 callout_data data for callouts (see 49 below)
1769
1770 The flag bits are also defined in pcre.h, and are
1771
1772 PCRE_EXTRA_STUDY_DATA
1773 PCRE_EXTRA_MATCH_LIMIT
1774 PCRE_EXTRA_CALLOUT_DATA
1775
1776 The pcre_study() function now returns one of these new pcre_extra blocks, with
1777 the actual study data pointed to by the study_data field, and the
1778 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1779 before. That is, this change is entirely upwards-compatible and requires no
1780 change to existing code.
1781
1782 If you want to pass in additional data to pcre_exec(), you can either place it
1783 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1784 block.
1785
1786 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1787 data string contains the escape sequence \M, pcretest calls pcre_exec() several
1788 times with different match limits, until it finds the minimum value needed for
1789 pcre_exec() to complete. The value is then output. This can be instructive; for
1790 most simple matches the number is quite small, but for pathological cases it
1791 gets very large very quickly.
1792
1793 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1794 returns the size of the data block pointed to by the study_data field in a
1795 pcre_extra block, that is, the value that was passed as the argument to
1796 pcre_malloc() when PCRE was getting memory in which to place the information
1797 created by pcre_study(). The fourth argument should point to a size_t variable.
1798 pcretest has been extended so that this information is shown after a successful
1799 pcre_study() call when information about the compiled regex is being displayed.
1800
1801 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1802 because what follows is always an absolute path. (Later: it turns out that this
1803 is more than cosmetic for MinGW, because it doesn't like empty path
1804 components.)
1805
1806 49. Some changes have been made to the callout feature (see 28 above):
1807
1808 (i) A callout function now has three choices for what it returns:
1809
1810 0 => success, carry on matching
1811 > 0 => failure at this point, but backtrack if possible
1812 < 0 => serious error, return this value from pcre_exec()
1813
1814 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1815 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1816 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1817 use by callout functions. It will never be used by PCRE itself.
1818
1819 (ii) The pcre_extra structure (see 45 above) has a void * field called
1820 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1821 pcre_callout_block structure has a field of the same name. The contents of
1822 the field passed in the pcre_extra structure are passed to the callout
1823 function in the corresponding field in the callout block. This makes it
1824 easier to use the same callout-containing regex from multiple threads. For
1825 testing, the pcretest program has a new data escape
1826
1827 \C*n pass the number n (may be negative) as callout_data
1828
1829 If the callout function in pcretest receives a non-zero value as
1830 callout_data, it returns that value.
1831
1832 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1833 there were some redundant $(CFLAGS) in commands that are now specified as
1834 $(LINK), which already includes $(CFLAGS).
1835
1836 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1837 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1838 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1839 one-byte characters throughout. Note that case-insensitive matching applies
1840 only to characters whose values are less than 256. PCRE doesn't support the
1841 notion of cases for higher-valued characters.
1842
1843 (i) A character class whose characters are all within 0-255 is handled as
1844 a bit map, and the map is inverted for negative classes. Previously, a
1845 character > 255 always failed to match such a class; however it should
1846 match if the class was a negative one (e.g. [^ab]). This has been fixed.
1847
1848 (ii) A negated character class with a single character < 255 is coded as
1849 "not this character" (OP_NOT). This wasn't working properly when the test
1850 character was multibyte, either singly or repeated.
1851
1852 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1853 mode, for example: \x{100}{2,3}.
1854
1855 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1856 singly or repeated) now correctly test multibyte characters. However,
1857 PCRE doesn't recognize any characters with values greater than 255 as
1858 digits, spaces, or word characters. Such characters always match \D, \S,
1859 and \W, and never match \d, \s, or \w.
1860
1861 (v) Classes may now contain characters and character ranges with values
1862 greater than 255. For example: [ab\x{100}-\x{400}].
1863
1864 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1865 PCRE in UTF-8 mode.
1866
1867 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1868 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1869 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1870 value.)
1871
1872 53. The single man page has become too large. I have therefore split it up into
1873 a number of separate man pages. These also give rise to individual HTML pages;
1874 these are now put in a separate directory, and there is an index.html page that
1875 lists them all. Some hyperlinking between the pages has been installed.
1876
1877 54. Added convenience functions for handling named capturing parentheses.
1878
1879 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1880 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1881 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1882 are faulted.
1883
1884 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1885 calling configure. These values are used when compiling the dftables.c program
1886 which is run to generate the source of the default character tables. They
1887 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1888 you will need to set these values.
1889
1890 57. Updated the building process for Windows DLL, as provided by Fred Cox.
1891
1892
1893 Version 3.9 02-Jan-02
1894 ---------------------
1895
1896 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1897
1898 2. If --disable-static was given, the building process failed when trying to
1899 build pcretest and pcregrep. (For some reason it was using libtool to compile
1900 them, which is not right, as they aren't part of the library.)
1901
1902
1903 Version 3.8 18-Dec-01
1904 ---------------------
1905
1906 1. The experimental UTF-8 code was completely screwed up. It was packing the
1907 bytes in the wrong order. How dumb can you get?
1908
1909
1910 Version 3.7 29-Oct-01
1911 ---------------------
1912
1913 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1914 This caused pcretest, when used on the test data, to segfault. Unfortunately,
1915 this didn't happen under Solaris 8, where I normally test things.
1916
1917 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1918 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1919 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1920 this fix an hour or so after the initial 3.7 release.)
1921
1922
1923 Version 3.6 23-Oct-01
1924 ---------------------
1925
1926 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1927 offsets passed as NULL with zero offset count.
1928
1929 2. The config.guess and config.sub files had not been updated when I moved to
1930 the latest autoconf.
1931
1932
1933 Version 3.5 15-Aug-01
1934 ---------------------
1935
1936 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1937 had been forgotten.
1938
1939 2. By using declared but undefined structures, we can avoid using "void"
1940 definitions in pcre.h while keeping the internal definitions of the structures
1941 private.
1942
1943 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1944 user point of view, this means that both static and shared libraries are built
1945 by default, but this can be individually controlled. More of the work of
1946 handling this static/shared cases is now inside libtool instead of PCRE's make
1947 file.
1948
1949 4. The pcretest utility is now installed along with pcregrep because it is
1950 useful for users (to test regexs) and by doing this, it automatically gets
1951 relinked by libtool. The documentation has been turned into a man page, so
1952 there are now .1, .txt, and .html versions in /doc.
1953
1954 5. Upgrades to pcregrep:
1955 (i) Added long-form option names like gnu grep.
1956 (ii) Added --help to list all options with an explanatory phrase.
1957 (iii) Added -r, --recursive to recurse into sub-directories.
1958 (iv) Added -f, --file to read patterns from a file.
1959
1960 6. pcre_exec() was referring to its "code" argument before testing that
1961 argument for NULL (and giving an error if it was NULL).
1962
1963 7. Upgraded Makefile.in to allow for compiling in a different directory from
1964 the source directory.
1965
1966 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1967 options bits, the pointer it was passed was to an int instead of to an unsigned
1968 long int. This mattered only on 64-bit systems.
1969
1970 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1971 generated) instead of pcre.in, which it its source. Also made the same change
1972 in several of the .c files.
1973
1974 10. A new release of gcc defines printf() as a macro, which broke pcretest
1975 because it had an ifdef in the middle of a string argument for printf(). Fixed
1976 by using separate calls to printf().
1977
1978 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1979 script, to force use of CR or LF instead of \n in the source. On non-Unix
1980 systems, the value can be set in config.h.
1981
1982 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1983 absolute limit. Changed the text of the error message to make this clear, and
1984 likewise updated the man page.
1985
1986 13. The limit of 99 on the number of capturing subpatterns has been removed.
1987 The new limit is 65535, which I hope will not be a "real" limit.
1988
1989
1990 Version 3.4 22-Aug-00
1991 ---------------------
1992
1993 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1994
1995 2. Diagnose condition (?(0) as an error instead of crashing on matching.
1996
1997
1998 Version 3.3 01-Aug-00
1999 ---------------------
2000
2001 1. If an octal character was given, but the value was greater than \377, it
2002 was not getting masked to the least significant bits, as documented. This could
2003 lead to crashes in some systems.
2004
2005 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2006 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2007
2008 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2009 These just pass their arguments on to (pcre_free)(), but they are provided
2010 because some uses of PCRE bind it to non-C systems that can call its functions,
2011 but cannot call free() or pcre_free() directly.
2012
2013 4. Add "make test" as a synonym for "make check". Corrected some comments in
2014 the Makefile.
2015
2016 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2017 Makefile.
2018
2019 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2020 command called pgrep for grepping around the active processes.
2021
2022 7. Added the beginnings of support for UTF-8 character strings.
2023
2024 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2025 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2026 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2027 out for the ar command.)
2028
2029
2030 Version 3.2 12-May-00
2031 ---------------------
2032
2033 This is purely a bug fixing release.
2034
2035 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2036 of ZA. This was just one example of several cases that could provoke this bug,
2037 which was introduced by change 9 of version 2.00. The code for breaking
2038 infinite loops after an iteration that matches an empty string was't working
2039 correctly.
2040
2041 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2042 when matched against abbab (for example). After matching an empty string, it
2043 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2044 caused it to match further down the string than it should.
2045
2046 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2047 was there because it doesn't seem to be needed, and it causes trouble on some
2048 systems, as it is not a Standard C header. It has been removed.
2049
2050 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2051 were reported on the Macintosh. The changes were from
2052
2053 while ((c = *(++ptr)) != 0 && c != '\n');
2054 to
2055 while ((c = *(++ptr)) != 0 && c != '\n') ;
2056
2057 Totally extraordinary, but if that's what it takes...
2058
2059 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2060 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2061 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2062 assumes the way PCRE uses memmove() (always moving upwards).
2063
2064 6. PCRE is being used in one environment where strchr() is not available. There
2065 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2066 faster code anyway.
2067
2068
2069 Version 3.1 09-Feb-00
2070 ---------------------
2071
2072 The only change in this release is the fixing of some bugs in Makefile.in for
2073 the "install" target:
2074
2075 (1) It was failing to install pcreposix.h.
2076
2077 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2078
2079
2080 Version 3.0 01-Feb-00
2081 ---------------------
2082
2083 1. Add support for the /+ modifier to perltest (to output $` like it does in
2084 pcretest).
2085
2086 2. Add support for the /g modifier to perltest.
2087
2088 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2089 matches null strings.
2090
2091 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2092 pattern. Perl treats empty patterns specially - it reuses the most recent
2093 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2094 effect.
2095
2096 5. The POSIX interface was broken in that it was just handing over the POSIX
2097 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2098 required a bigger vector, with some working space on the end. This means that
2099 the POSIX wrapper now has to get and free some memory, and copy the results.
2100
2101 6. Added some simple autoconf support, placing the test data and the
2102 documentation in separate directories, re-organizing some of the
2103 information files, and making it build pcre-config (a GNU standard). Also added
2104 libtool support for building PCRE as a shared library, which is now the
2105 default.
2106
2107 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2108 09 are not valid octal constants. Single digits will be used for minor values
2109 less than 10.
2110
2111 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2112 existing programs that set these in the POSIX interface can use PCRE without
2113 modification.
2114
2115 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2116 return all that pcre_info() returns, plus additional data. The pcre_info()
2117 function is retained for compatibility, but is considered to be obsolete.
2118
2119 10. Added experimental recursion feature (?R) to handle one common case that
2120 Perl 5.6 will be able to do with (?p{...}).
2121
2122 11. Added support for POSIX character classes like [:alpha:], which Perl is
2123 adopting.
2124
2125
2126 Version 2.08 31-Aug-99
2127 ----------------------
2128
2129 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2130 trying to match at the startoffset position, but instead was moving forward to
2131 the next newline as if a previous match had failed.
2132
2133 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2134 and could get into a loop if a null string was matched other than at the start
2135 of the subject.
2136
2137 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2138 be distinguished at compile time, and for completeness also added PCRE_DATE.
2139
2140 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2141 in GnuWin32 environments.
2142
2143
2144 Version 2.07 29-Jul-99
2145 ----------------------
2146
2147 1. The documentation is now supplied in plain text form and HTML as well as in
2148 the form of man page sources.
2149
2150 2. C++ compilers don't like assigning (void *) values to other pointer types.
2151 In particular this affects malloc(). Although there is no problem in Standard
2152 C, I've put in casts to keep C++ compilers happy.
2153
2154 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2155 should be (const char *).
2156
2157 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2158 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2159 However, I haven't made this a standard facility. The documentation doesn't
2160 mention it, and the Makefile doesn't support it.
2161
2162 5. The Makefile now contains an "install" target, with editable destinations at
2163 the top of the file. The pcretest program is not installed.
2164
2165 6. pgrep -V now gives the PCRE version number and date.
2166
2167 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2168 causing the entire string to be ignored, instead of just the last character.
2169
2170 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2171 non-matching string, it can take a very, very long time, even for strings of
2172 quite modest length, because of the nested recursion. PCRE now does better in
2173 some of these cases. It does this by remembering the last required literal
2174 character in the pattern, and pre-searching the subject to ensure it is present
2175 before running the real match. In other words, it applies a heuristic to detect
2176 some types of certain failure quickly, and in the above example, if presented
2177 with a string that has no trailing " it gives "no match" very quickly.
2178
2179 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2180 other alternatives are tried instead.
2181
2182
2183 Version 2.06 09-Jun-99
2184 ----------------------
2185
2186 1. Change pcretest's output for amount of store used to show just the code
2187 space, because the remainder (the data block) varies in size between 32-bit and
2188 64-bit systems.
2189
2190 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2191 start matching at. This allows lookbehinds to work when searching for multiple
2192 occurrences in a string.
2193
2194 3. Added additional options to pcretest for testing multiple occurrences:
2195
2196 /+ outputs the rest of the string that follows a match
2197 /g loops for multiple occurrences, using the new startoffset argument
2198 /G loops for multiple occurrences by passing an incremented pointer
2199
2200 4. PCRE wasn't doing the "first character" optimization for patterns starting
2201 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2202 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2203 the letter 'x'. On long subject strings, this gives a significant speed-up.
2204
2205
2206 Version 2.05 21-Apr-99
2207 ----------------------
2208
2209 1. Changed the type of magic_number from int to long int so that it works
2210 properly on 16-bit systems.
2211
2212 2. Fixed a bug which caused patterns starting with .* not to work correctly
2213 when the subject string contained newline characters. PCRE was assuming
2214 anchoring for such patterns in all cases, which is not correct because .* will
2215 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2216 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2217 must be retried after every newline in the subject.
2218
2219
2220 Version 2.04 18-Feb-99
2221 ----------------------
2222
2223 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2224 computation of the store needed to hold the pattern was incorrect (too large).
2225 If such patterns were nested a few deep, this could multiply and become a real
2226 problem.
2227
2228 2. Added /M option to pcretest to show the memory requirement of a specific
2229 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2230
2231 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2232 compiled in such a way that the backtracking after subsequent failure was
2233 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2234 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2235
2236
2237 Version 2.03 02-Feb-99
2238 ----------------------
2239
2240 1. Fixed typo and small mistake in man page.
2241
2242 2. Added 4th condition (GPL supersedes if conflict) and created separate
2243 LICENCE file containing the conditions.
2244
2245 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2246 Perl, that is the internal \ allows the delimiter to be included in the
2247 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2248 the final delimiter, add \ to the end of the pattern (to test the error).
2249
2250 4. Added the convenience functions for extracting substrings after a successful
2251 match. Updated pcretest to make it able to test these functions.
2252
2253
2254 Version 2.02 14-Jan-99
2255 ----------------------
2256
2257 1. Initialized the working variables associated with each extraction so that
2258 their saving and restoring doesn't refer to uninitialized store.
2259
2260 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2261 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2262 fix the problem.
2263
2264 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2265 calls, and wasn't printing the correct value for compiling calls. Increased the
2266 default value of LOOPREPEAT, and the number of significant figures in the
2267 times.
2268
2269 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2270
2271 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2272 a building problem on Windows NT with a FAT file system.
2273
2274
2275 Version 2.01 21-Oct-98
2276 ----------------------
2277
2278 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2279 to character tables built by pcre_maketables() in the current locale. If NULL
2280 is passed, the default tables are used.
2281
2282
2283 Version 2.00 24-Sep-98
2284 ----------------------
2285
2286 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2287 it any more.
2288
2289 2. Allow quantification of (?>) groups, and make it work correctly.
2290
2291 3. The first character computation wasn't working for (?>) groups.
2292
2293 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2294 end of the subject) and add 5.005's \z, which really does match only at the
2295 very end of the subject.
2296
2297 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2298
2299 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2300 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2301 localized options. All options to pcre_study() were also removed.
2302
2303 7. Add other new features from 5.005:
2304
2305 $(?<= positive lookbehind
2306 $(?<! negative lookbehind
2307 (?imsx-imsx) added the unsetting capability
2308 such a setting is global if at outer level; local otherwise
2309 (?imsx-imsx:) non-capturing groups with option setting
2310 (?(cond)re|re) conditional pattern matching
2311
2312 A backreference to itself in a repeated group matches the previous
2313 captured string.
2314
2315 8. General tidying up of studying (both automatic and via "study")
2316 consequential on the addition of new assertions.
2317
2318 9. As in 5.005, unlimited repeated groups that could match an empty substring
2319 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2320 runtime if any iteration does actually match an empty substring.
2321
2322 10. Include the RunTest script in the distribution.
2323
2324 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2325 discrepancies, some of which were old and were also with respect to 5.004. They
2326 have now been fixed.
2327
2328
2329 Version 1.09 28-Apr-98
2330 ----------------------
2331
2332 1. A negated single character class followed by a quantifier with a minimum
2333 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2334 program crashes, or just wrong answers. This did not apply to negated classes
2335 containing more than one character, or to minima other than one.
2336
2337
2338 Version 1.08 27-Mar-98
2339 ----------------------
2340
2341 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2342
2343 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2344 latter must appear before anything that relies on it in the pattern.
2345
2346
2347 Version 1.07 16-Feb-98
2348 ----------------------
2349
2350 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2351 repeat of a potentially empty string).
2352
2353
2354 Version 1.06 23-Jan-98
2355 ----------------------
2356
2357 1. Added Markus Oberhumer's little patches for C++.
2358
2359 2. Literal strings longer than 255 characters were broken.
2360
2361
2362 Version 1.05 23-Dec-97
2363 ----------------------
2364
2365 1. Negated character classes containing more than one character were failing if
2366 PCRE_CASELESS was set at run time.
2367
2368
2369 Version 1.04 19-Dec-97
2370 ----------------------
2371
2372 1. Corrected the man page, where some "const" qualifiers had been omitted.
2373
2374 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2375 input syntax.
2376
2377 3. Fixed memory leak which occurred when a regex with back references was
2378 matched with an offsets vector that wasn't big enough. The temporary memory
2379 that is used in this case wasn't being freed if the match failed.
2380
2381 4. Tidied pcretest to ensure it frees memory that it gets.
2382
2383 5. Temporary memory was being obtained in the case where the passed offsets
2384 vector was exactly big enough.
2385
2386 6. Corrected definition of offsetof() from change 5 below.
2387
2388 7. I had screwed up change 6 below and broken the rules for the use of
2389 setjmp(). Now fixed.
2390
2391
2392 Version 1.03 18-Dec-97
2393 ----------------------
2394
2395 1. A erroneous regex with a missing opening parenthesis was correctly
2396 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2397 on some systems.
2398
2399 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2400 it was reported that one broken compiler failed on the former because "code" is
2401 also an independent variable.
2402
2403 3. The erroneous regex a[]b caused an array overrun reference.
2404
2405 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2406 fail on data ending with that character. (It was going on too far, and checking
2407 the next character, typically a binary zero.) This was specific to the
2408 optimized code for single-character negative classes.
2409
2410 5. Added a contributed patch from the TIN world which does the following:
2411
2412 + Add an undef for memmove, in case the the system defines a macro for it.
2413
2414 + Add a definition of offsetof(), in case there isn't one. (I don't know
2415 the reason behind this - offsetof() is part of the ANSI standard - but
2416 it does no harm).
2417
2418 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2419 most of the places where whitespace preceded '#'. I have given up and
2420 allowed the remaining 2 cases to be at the margin.
2421
2422 + Rename some variables in pcre to eliminate shadowing. This seems very
2423 pedantic, but does no harm, of course.
2424
2425 6. Moved the call to setjmp() into its own function, to get rid of warnings
2426 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2427
2428 7. Constructs such as \d{8,} were compiling into the equivalent of
2429 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2430 outcome, but in this particular case used more store than had been allocated,
2431 which caused the bug to be discovered because it threw up an internal error.
2432
2433 8. The debugging code in both pcre and pcretest for outputting the compiled
2434 form of a regex was going wrong in the case of back references followed by
2435 curly-bracketed repeats.
2436
2437
2438 Version 1.02 12-Dec-97
2439 ----------------------
2440
2441 1. Typos in pcre.3 and comments in the source fixed.
2442
2443 2. Applied a contributed patch to get rid of places where it used to remove
2444 'const' from variables, and fixed some signed/unsigned and uninitialized
2445 variable warnings.
2446
2447 3. Added the "runtest" target to Makefile.
2448
2449 4. Set default compiler flag to -O2 rather than just -O.
2450
2451
2452 Version 1.01 19-Nov-97
2453 ----------------------
2454
2455 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2456 like /([ab]*)*/, that is, for classes with more than one character in them.
2457
2458 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2459 as /((?>a*))*/ (a PCRE_EXTRA facility).
2460
2461
2462 Version 1.00 18-Nov-97
2463 ----------------------
2464
2465 1. Added compile-time macros to support systems such as SunOS4 which don't have
2466 memmove() or strerror() but have other things that can be used instead.
2467
2468 2. Arranged that "make clean" removes the executables.
2469
2470
2471 Version 0.99 27-Oct-97
2472 ----------------------
2473
2474 1. Fixed bug in code for optimizing classes with only one character. It was
2475 initializing a 32-byte map regardless, which could cause it to run off the end
2476 of the memory it had got.
2477
2478 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2479
2480
2481 Version 0.98 22-Oct-97
2482 ----------------------
2483
2484 1. Fixed bug in code for handling temporary memory usage when there are more
2485 back references than supplied space in the ovector. This could cause segfaults.
2486
2487
2488 Version 0.97 21-Oct-97
2489 ----------------------
2490
2491 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2492
2493 2. Optimized negated single characters not to use a bit map.
2494
2495 3. Brought error texts together as macro definitions; clarified some of them;
2496 fixed one that was wrong - it said "range out of order" when it meant "invalid
2497 escape sequence".
2498
2499 4. Changed some char * arguments to const char *.
2500
2501 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2502
2503 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2504 pcretest.
2505
2506
2507 Version 0.96 16-Oct-97
2508 ----------------------
2509
2510 1. Added a simple "pgrep" utility to the distribution.
2511
2512 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2513 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2514 where "ddd" means "one or more decimal digits".
2515
2516 3. Fixed serious bug. If a pattern had a back reference, but the call to
2517 pcre_exec() didn't supply a large enough ovector to record the related
2518 identifying subpattern, the match always failed. PCRE now remembers the number
2519 of the largest back reference, and gets some temporary memory in which to save
2520 the offsets during matching if necessary, in order to ensure that
2521 backreferences always work.
2522
2523 4. Increased the compatibility with Perl in a number of ways:
2524
2525 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2526 to request this handling. The option can be set at compile or exec time.
2527
2528 (b) $ matches before a terminating newline by default; an option
2529 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2530 mode). The option can be set at compile or exec time.
2531
2532 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2533 the same as Perl's. If the decimal number it represents is less than 10
2534 or there aren't that many previous left capturing parentheses, an octal
2535 escape is read. Inside a character class, it's always an octal escape,
2536 even if it is a single digit.
2537
2538 (d) An escaped but undefined alphabetic character is taken as a literal,
2539 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2540 escapes.
2541
2542 (e) {0} is now permitted. (The previous item is removed from the compiled
2543 pattern).
2544
2545 5. Changed all the names of code files so that the basic parts are no longer
2546 than 10 characters, and abolished the teeny "globals.c" file.
2547
2548 6. Changed the handling of character classes; they are now done with a 32-byte
2549 bit map always.
2550
2551 7. Added the -d and /D options to pcretest to make it possible to look at the
2552 internals of compilation without having to recompile pcre.
2553
2554
2555 Version 0.95 23-Sep-97
2556 ----------------------
2557
2558 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2559 \x20 at the start of a run of normal characters. These were being treated as
2560 real characters, instead of the source characters being re-checked.
2561
2562
2563 Version 0.94 18-Sep-97
2564 ----------------------
2565
2566 1. The functions are now thread-safe, with the caveat that the global variables
2567 containing pointers to malloc() and free() or alternative functions are the
2568 same for all threads.
2569
2570 2. Get pcre_study() to generate a bitmap of initial characters for non-
2571 anchored patterns when this is possible, and use it if passed to pcre_exec().
2572
2573
2574 Version 0.93 15-Sep-97
2575 ----------------------
2576
2577 1. /(b)|(:+)/ was computing an incorrect first character.
2578
2579 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2580 but not actually doing anything yet.
2581
2582 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2583 as Perl does (e.g. [-az] or [az-]).
2584
2585 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2586 all possible positions.
2587
2588 5. Split up into different modules to avoid including unneeded functions in a
2589 compiled binary. However, compile and exec are still in one module. The "study"
2590 function is split off.
2591
2592 6. The character tables are now in a separate module whose source is generated
2593 by an auxiliary program - but can then be edited by hand if required. There are
2594 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2595 toupper() in the code.
2596
2597 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2598 make them global. Abolish the function for setting them, as the caller can now
2599 set them directly.
2600
2601
2602 Version 0.92 11-Sep-97
2603 ----------------------
2604
2605 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2606 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2607
2608 2. Caseless matching was not working in character classes if the characters in
2609 the pattern were in upper case.
2610
2611 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2612
2613 4. Make PCRE_ANCHORED public and accept as a compile option.
2614
2615 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2616 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2617 pass them.
2618
2619 6. Give an error if bad option bits passed at compile or run time.
2620
2621 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2622 pcretest to cause it to pass that flag.
2623
2624 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2625 options, and the first character, if set.
2626
2627 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2628
2629
2630 Version 0.91 10-Sep-97
2631 ----------------------
2632
2633 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2634 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2635
2636 2. PCRE was looping on encountering an indefinitely repeated back reference to
2637 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2638 Perl does - treats the match as successful.
2639
2640 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12