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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 102 - (show annotations) (download)
Tue Mar 6 15:50:45 2007 UTC (7 years, 6 months ago) by ph10
File size: 120960 byte(s)
Fix test problem caused by training space in the data.

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

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