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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 101 - (show annotations) (download)
Tue Mar 6 15:19:44 2007 UTC (7 years, 9 months ago) by ph10
File size: 120725 byte(s)
Updated the support (such as it is) for Virtual Pascal, thanks to Stefan 
Weber: (1) pcre_internal.h was missing some function renames; (2) updated 
makevp.bat for the current PCRE, using the additional files !compile.txt, 
!linklib.txt, and pcregexp.pas.

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

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