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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 232 - (show annotations) (download)
Tue Sep 11 11:18:50 2007 UTC (6 years, 7 months ago) by ph10
File size: 139879 byte(s)
doc/pcresyntax.html was missing from the distribution.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.4 10-Sep-07
5 ---------------------
6
7 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
8 means that a class such as [\s] counted as "explicit reference to CR or
9 LF". That isn't really right - the whole point of the change was to try to
10 help when there was an actual mention of one of the two characters. So now
11 the change happens only if \r or \n (or a literal CR or LF) character is
12 encountered.
13
14 2. The 32-bit options word was also used for 6 internal flags, but the numbers
15 of both had grown to the point where there were only 3 bits left.
16 Fortunately, there was spare space in the data structure, and so I have
17 moved the internal flags into a new 16-bit field to free up more option
18 bits.
19
20 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
21 but did not set the internal JCHANGED flag - either of these is enough to
22 control the way the "get" function works - but the PCRE_INFO_JCHANGED
23 facility is supposed to tell if (?J) was ever used, so now (?J) at the
24 start sets both bits.
25
26 4. Added options (at build time, compile time, exec time) to change \R from
27 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
28
29 5. doc/pcresyntax.html was missing from the distribution.
30
31
32 Version 7.3 28-Aug-07
33 ---------------------
34
35 1. In the rejigging of the build system that eventually resulted in 7.1, the
36 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
37 brackets there is not right, since it causes compilers to look for an
38 installed pcre.h, not the version that is in the source that is being
39 compiled (which of course may be different). I have changed it back to:
40
41 #include "pcre.h"
42
43 I have a vague recollection that the change was concerned with compiling in
44 different directories, but in the new build system, that is taken care of
45 by the VPATH setting the Makefile.
46
47 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
48 when the subject happened to end in the byte 0x85 (e.g. if the last
49 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
50 characters but of course it shouldn't be taken as a newline when it is part
51 of another character. The bug was that, for an unlimited repeat of . in
52 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
53 characters when looking for a newline.
54
55 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
56
57 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
58 in debug output.
59
60 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
61 long printing in the pcrecpp unittest when running under MinGW.
62
63 6. ESC_K was left out of the EBCDIC table.
64
65 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
66 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
67 limit also applies to "virtual nesting" when a pattern is recursive, and in
68 this case 1000 isn't so big. I have been able to remove this limit at the
69 expense of backing off one optimization in certain circumstances. Normally,
70 when pcre_exec() would call its internal match() function recursively and
71 immediately return the result unconditionally, it uses a "tail recursion"
72 feature to save stack. However, when a subpattern that can match an empty
73 string has an unlimited repetition quantifier, it no longer makes this
74 optimization. That gives it a stack frame in which to save the data for
75 checking that an empty string has been matched. Previously this was taken
76 from the 1000-entry workspace that had been reserved. So now there is no
77 explicit limit, but more stack is used.
78
79 8. Applied Daniel's patches to solve problems with the import/export magic
80 syntax that is required for Windows, and which was going wrong for the
81 pcreposix and pcrecpp parts of the library. These were overlooked when this
82 problem was solved for the main library.
83
84 9. There were some crude static tests to avoid integer overflow when computing
85 the size of patterns that contain repeated groups with explicit upper
86 limits. As the maximum quantifier is 65535, the maximum group length was
87 set at 30,000 so that the product of these two numbers did not overflow a
88 32-bit integer. However, it turns out that people want to use groups that
89 are longer than 30,000 bytes (though not repeat them that many times).
90 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
91 made it possible to implement the integer overflow checks in a much more
92 dynamic way, which I have now done. The artificial limitation on group
93 length has been removed - we now have only the limit on the total length of
94 the compiled pattern, which depends on the LINK_SIZE setting.
95
96 10. Fixed a bug in the documentation for get/copy named substring when
97 duplicate names are permitted. If none of the named substrings are set, the
98 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
99 empty string.
100
101 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
102 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
103 because the ] is interpreted as the first data character and the
104 terminating ] is not found. PCRE has been made compatible with Perl in this
105 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
106 cause memory overwriting.
107
108 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
109 string has been matched (to stop an infinite loop). It was not recognizing
110 a conditional subpattern that could match an empty string if that
111 subpattern was within another subpattern. For example, it looped when
112 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
113 condition was not nested. This bug has been fixed.
114
115 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
116 past the start of the subject in the presence of bytes with the top bit
117 set, for example "\x8aBCD".
118
119 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
120 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
121
122 14. Optimized (?!) to (*FAIL).
123
124 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
125 This restricts code points to be within the range 0 to 0x10FFFF, excluding
126 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
127 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
128 does: it's just the validity check that is more restrictive.
129
130 16. Inserted checks for integer overflows during escape sequence (backslash)
131 processing, and also fixed erroneous offset values for syntax errors during
132 backslash processing.
133
134 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
135 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
136
137 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
138 caused an overrun.
139
140 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
141 something other than just ASCII characters) inside a group that had an
142 unlimited repeat caused a loop at compile time (while checking to see
143 whether the group could match an empty string).
144
145 20. Debugging a pattern containing \p or \P could cause a crash. For example,
146 [\P{Any}] did so. (Error in the code for printing property names.)
147
148 21. An orphan \E inside a character class could cause a crash.
149
150 22. A repeated capturing bracket such as (A)? could cause a wild memory
151 reference during compilation.
152
153 23. There are several functions in pcre_compile() that scan along a compiled
154 expression for various reasons (e.g. to see if it's fixed length for look
155 behind). There were bugs in these functions when a repeated \p or \P was
156 present in the pattern. These operators have additional parameters compared
157 with \d, etc, and these were not being taken into account when moving along
158 the compiled data. Specifically:
159
160 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
161 length.
162
163 (b) An item such as \pL+ within a repeated group could cause crashes or
164 loops.
165
166 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
167 "reference to non-existent subpattern" error.
168
169 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
170
171 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
172 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
173
174 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
175
176 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
177 character were causing crashes (broken optimization).
178
179 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
180 \p or \P) caused a compile-time loop.
181
182 28. More problems have arisen in unanchored patterns when CRLF is a valid line
183 break. For example, the unstudied pattern [\r\n]A does not match the string
184 "\r\nA" because change 7.0/46 below moves the current point on by two
185 characters after failing to match at the start. However, the pattern \nA
186 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
187 the same is true. There doesn't seem any very clean way out of this, but
188 what I have chosen to do makes the common cases work: PCRE now takes note
189 of whether there can be an explicit match for \r or \n anywhere in the
190 pattern, and if so, 7.0/46 no longer applies. As part of this change,
191 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
192 pattern has explicit CR or LF references.
193
194 29. Added (*CR) etc for changing newline setting at start of pattern.
195
196
197 Version 7.2 19-Jun-07
198 ---------------------
199
200 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
201 which is apparently normally available under Windows.
202
203 2. Re-jig the pcregrep tests with different newline settings in an attempt
204 to make them independent of the local environment's newline setting.
205
206 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
207
208 4. Some of the "internals" tests were previously cut out when the link size
209 was not 2, because the output contained actual offsets. The recent new
210 "Z" feature of pcretest means that these can be cut out, making the tests
211 usable with all link sizes.
212
213 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
214 stack recursion. This gives a massive performance boost under BSD, but just
215 a small improvement under Linux. However, it saves one field in the frame
216 in all cases.
217
218 6. Added more features from the forthcoming Perl 5.10:
219
220 (a) (?-n) (where n is a string of digits) is a relative subroutine or
221 recursion call. It refers to the nth most recently opened parentheses.
222
223 (b) (?+n) is also a relative subroutine call; it refers to the nth next
224 to be opened parentheses.
225
226 (c) Conditions that refer to capturing parentheses can be specified
227 relatively, for example, (?(-2)... or (?(+3)...
228
229 (d) \K resets the start of the current match so that everything before
230 is not part of it.
231
232 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
233
234 (f) \g{name} is another synonym - part of Perl 5.10's unification of
235 reference syntax.
236
237 (g) (?| introduces a group in which the numbering of parentheses in each
238 alternative starts with the same number.
239
240 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
241
242 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
243 PCRE_INFO_JCHANGED.
244
245 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
246 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
247 for detecting groups that can match an empty string.
248
249 9. A pattern with a very large number of alternatives (more than several
250 hundred) was running out of internal workspace during the pre-compile
251 phase, where pcre_compile() figures out how much memory will be needed. A
252 bit of new cunning has reduced the workspace needed for groups with
253 alternatives. The 1000-alternative test pattern now uses 12 bytes of
254 workspace instead of running out of the 4096 that are available.
255
256 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
257
258 11. Applied patch from Google to remove an optimization that didn't quite work.
259 The report of the bug said:
260
261 pcrecpp::RE("a*").FullMatch("aaa") matches, while
262 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
263 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
264
265 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
266 it matched the wrong number of bytes.
267
268
269 Version 7.1 24-Apr-07
270 ---------------------
271
272 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
273 that is more "standard", making use of automake and other Autotools. There
274 is some re-arrangement of the files and adjustment of comments consequent
275 on this.
276
277 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
278 for recursive directory scanning broke on some systems because the files
279 are not scanned in any specific order and on different systems the order
280 was different. A call to "sort" has been inserted into RunGrepTest for the
281 approprate test as a short-term fix. In the longer term there may be an
282 alternative.
283
284 3. I had an email from Eric Raymond about problems translating some of PCRE's
285 man pages to HTML (despite the fact that I distribute HTML pages, some
286 people do their own conversions for various reasons). The problems
287 concerned the use of low-level troff macros .br and .in. I have therefore
288 removed all such uses from the man pages (some were redundant, some could
289 be replaced by .nf/.fi pairs). The 132html script that I use to generate
290 HTML has been updated to handle .nf/.fi and to complain if it encounters
291 .br or .in.
292
293 4. Updated comments in configure.ac that get placed in config.h.in and also
294 arranged for config.h to be included in the distribution, with the name
295 config.h.generic, for the benefit of those who have to compile without
296 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
297
298 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
299 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
300 makevp.bat for the current PCRE, using the additional files
301 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
302
303 6. A Windows user reported a minor discrepancy with test 2, which turned out
304 to be caused by a trailing space on an input line that had got lost in his
305 copy. The trailing space was an accident, so I've just removed it.
306
307 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
308 that is needed.
309
310 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
311 as "const" (a) because they are and (b) because it helps the PHP
312 maintainers who have recently made a script to detect big data structures
313 in the php code that should be moved to the .rodata section. I remembered
314 to update Builducptable as well, so it won't revert if ucptable.h is ever
315 re-created.
316
317 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
318 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
319 order to be able to cut out the UTF-8 tables in the latter when UTF-8
320 support is not required. This saves 1.5-2K of code, which is important in
321 some applications.
322
323 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
324 so as not to refer to the tables, even though these functions will never be
325 called when UTF-8 support is disabled. Otherwise there are problems with a
326 shared library.
327
328 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
329
330 (a) It was defining its arguments as char * instead of void *.
331
332 (b) It was assuming that all moves were upwards in memory; this was true
333 a long time ago when I wrote it, but is no longer the case.
334
335 The emulated memove() is provided for those environments that have neither
336 memmove() nor bcopy(). I didn't think anyone used it these days, but that
337 is clearly not the case, as these two bugs were recently reported.
338
339 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
340 and Detrail to create the HTML documentation, the .txt form of the man
341 pages, and it removes trailing spaces from listed files. It also creates
342 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
343 case, it wraps all the #defines with #ifndefs. This script should be run
344 before "make dist".
345
346 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
347 with Unicode property support.
348
349 (a) For a maximizing quantifier, if the two different cases of the
350 character were of different lengths in their UTF-8 codings (there are
351 some cases like this - I found 11), and the matching function had to
352 back up over a mixture of the two cases, it incorrectly assumed they
353 were both the same length.
354
355 (b) When PCRE was configured to use the heap rather than the stack for
356 recursion during matching, it was not correctly preserving the data for
357 the other case of a UTF-8 character when checking ahead for a match
358 while processing a minimizing repeat. If the check also involved
359 matching a wide character, but failed, corruption could cause an
360 erroneous result when trying to check for a repeat of the original
361 character.
362
363 13. Some tidying changes to the testing mechanism:
364
365 (a) The RunTest script now detects the internal link size and whether there
366 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
367 values substituted by "configure". (The RunGrepTest script already did
368 this for UTF-8.) The configure.ac script no longer substitutes the
369 relevant variables.
370
371 (b) The debugging options /B and /D in pcretest show the compiled bytecode
372 with length and offset values. This means that the output is different
373 for different internal link sizes. Test 2 is skipped for link sizes
374 other than 2 because of this, bypassing the problem. Unfortunately,
375 there was also a test in test 3 (the locale tests) that used /B and
376 failed for link sizes other than 2. Rather than cut the whole test out,
377 I have added a new /Z option to pcretest that replaces the length and
378 offset values with spaces. This is now used to make test 3 independent
379 of link size. (Test 2 will be tidied up later.)
380
381 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
382 segmentation fault instead of returning the appropriate error message.
383
384 15. In multiline mode when the newline sequence was set to "any", the pattern
385 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
386 This doesn't seem right; it now treats the CRLF combination as the line
387 ending, and so does not match in that case. It's only a pattern such as ^$
388 that would hit this one: something like ^ABC$ would have failed after \r
389 and then tried again after \r\n.
390
391 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
392 in an attempt to make files that differ only in their line terminators
393 compare equal. This works on Linux.
394
395 17. Under certain error circumstances pcregrep might try to free random memory
396 as it exited. This is now fixed, thanks to valgrind.
397
398 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
399 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
400 was because its rules for how to advance for /g after matching an empty
401 string at the end of a line did not allow for this case. They now check for
402 it specially.
403
404 20. pcretest is supposed to handle patterns and data of any length, by
405 extending its buffers when necessary. It was getting this wrong when the
406 buffer for a data line had to be extended.
407
408 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
409 CRLF as a newline sequence.
410
411 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
412 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
413 I have nevertheless tidied it up.
414
415 23. Added some casts to kill warnings from HP-UX ia64 compiler.
416
417 24. Added a man page for pcre-config.
418
419
420 Version 7.0 19-Dec-06
421 ---------------------
422
423 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
424 moving to gcc 4.1.1.
425
426 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
427 sys/time.h, which is documented as needed for this function. It doesn't
428 seem to matter on Linux, but it showed up on some releases of OS X.
429
430 3. It seems that there are systems where bytes whose values are greater than
431 127 match isprint() in the "C" locale. The "C" locale should be the
432 default when a C program starts up. In most systems, only ASCII printing
433 characters match isprint(). This difference caused the output from pcretest
434 to vary, making some of the tests fail. I have changed pcretest so that:
435
436 (a) When it is outputting text in the compiled version of a pattern, bytes
437 other than 32-126 are always shown as hex escapes.
438
439 (b) When it is outputting text that is a matched part of a subject string,
440 it does the same, unless a different locale has been set for the match
441 (using the /L modifier). In this case, it uses isprint() to decide.
442
443 4. Fixed a major bug that caused incorrect computation of the amount of memory
444 required for a compiled pattern when options that changed within the
445 pattern affected the logic of the preliminary scan that determines the
446 length. The relevant options are -x, and -i in UTF-8 mode. The result was
447 that the computed length was too small. The symptoms of this bug were
448 either the PCRE error "internal error: code overflow" from pcre_compile(),
449 or a glibc crash with a message such as "pcretest: free(): invalid next
450 size (fast)". Examples of patterns that provoked this bug (shown in
451 pcretest format) are:
452
453 /(?-x: )/x
454 /(?x)(?-x: \s*#\s*)/
455 /((?i)[\x{c0}])/8
456 /(?i:[\x{c0}])/8
457
458 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
459 is now done differently.
460
461 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
462 wrapper classes; (b) implement a new function in the C++ scanner that is
463 more efficient than the old way of doing things because it avoids levels of
464 recursion in the regex matching; (c) add a paragraph to the documentation
465 for the FullMatch() function.
466
467 6. The escape sequence \n was being treated as whatever was defined as
468 "newline". Not only was this contrary to the documentation, which states
469 that \n is character 10 (hex 0A), but it also went horribly wrong when
470 "newline" was defined as CRLF. This has been fixed.
471
472 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
473 was being set to -1 for the "end of line" case (supposedly a value that no
474 character can have). Though this value is never used (the check for end of
475 line is "zero bytes in current character"), it caused compiler complaints.
476 I've changed it to 0xffffffff.
477
478 8. In pcre_version.c, the version string was being built by a sequence of
479 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
480 string (as it is for production releases) called a macro with an empty
481 argument. The C standard says the result of this is undefined. The gcc
482 compiler treats it as an empty string (which was what was wanted) but it is
483 reported that Visual C gives an error. The source has been hacked around to
484 avoid this problem.
485
486 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
487 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
488 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
489 of them did).
490
491 10. Originally, pcretest opened its input and output without "b"; then I was
492 told that "b" was needed in some environments, so it was added for release
493 5.0 to both the input and output. (It makes no difference on Unix-like
494 systems.) Later I was told that it is wrong for the input on Windows. I've
495 now abstracted the modes into two macros, to make it easier to fiddle with
496 them, and removed "b" from the input mode under Windows.
497
498 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
499
500 12. Added -help and --help to pcretest as an official way of being reminded
501 of the options.
502
503 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
504 and pcrecpp.cc because they annoy compilers at high warning levels.
505
506 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
507
508 15. Fixed an occurrence of == in configure.ac that should have been = (shell
509 scripts are not C programs :-) and which was not noticed because it works
510 on Linux.
511
512 16. pcretest is supposed to handle any length of pattern and data line (as one
513 line or as a continued sequence of lines) by extending its input buffer if
514 necessary. This feature was broken for very long pattern lines, leading to
515 a string of junk being passed to pcre_compile() if the pattern was longer
516 than about 50K.
517
518 17. I have done a major re-factoring of the way pcre_compile() computes the
519 amount of memory needed for a compiled pattern. Previously, there was code
520 that made a preliminary scan of the pattern in order to do this. That was
521 OK when PCRE was new, but as the facilities have expanded, it has become
522 harder and harder to keep it in step with the real compile phase, and there
523 have been a number of bugs (see for example, 4 above). I have now found a
524 cunning way of running the real compile function in a "fake" mode that
525 enables it to compute how much memory it would need, while actually only
526 ever using a few hundred bytes of working memory and without too many
527 tests of the mode. This should make future maintenance and development
528 easier. A side effect of this work is that the limit of 200 on the nesting
529 depth of parentheses has been removed (though this was never a serious
530 limitation, I suspect). However, there is a downside: pcre_compile() now
531 runs more slowly than before (30% or more, depending on the pattern). I
532 hope this isn't a big issue. There is no effect on runtime performance.
533
534 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
535 newline (only possible for the last line of a file) and it was a
536 pattern that set a locale (followed by /Lsomething), pcretest crashed.
537
538 19. Added additional timing features to pcretest. (1) The -tm option now times
539 matching only, not compiling. (2) Both -t and -tm can be followed, as a
540 separate command line item, by a number that specifies the number of
541 repeats to use when timing. The default is 50000; this gives better
542 precision, but takes uncomfortably long for very large patterns.
543
544 20. Extended pcre_study() to be more clever in cases where a branch of a
545 subpattern has no definite first character. For example, (a*|b*)[cd] would
546 previously give no result from pcre_study(). Now it recognizes that the
547 first character must be a, b, c, or d.
548
549 21. There was an incorrect error "recursive call could loop indefinitely" if
550 a subpattern (or the entire pattern) that was being tested for matching an
551 empty string contained only one non-empty item after a nested subpattern.
552 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
553 incorrectly, because the \d was being skipped in the check.
554
555 22. The pcretest program now has a new pattern option /B and a command line
556 option -b, which is equivalent to adding /B to every pattern. This causes
557 it to show the compiled bytecode, without the additional information that
558 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
559 is the same as /B/I).
560
561 23. A new optimization is now able automatically to treat some sequences such
562 as a*b as a*+b. More specifically, if something simple (such as a character
563 or a simple class like \d) has an unlimited quantifier, and is followed by
564 something that cannot possibly match the quantified thing, the quantifier
565 is automatically "possessified".
566
567 24. A recursive reference to a subpattern whose number was greater than 39
568 went wrong under certain circumstances in UTF-8 mode. This bug could also
569 have affected the operation of pcre_study().
570
571 25. Realized that a little bit of performance could be had by replacing
572 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
573
574 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
575
576 27. Possessive quantifiers such as a++ were previously implemented by turning
577 them into atomic groups such as ($>a+). Now they have their own opcodes,
578 which improves performance. This includes the automatically created ones
579 from 23 above.
580
581 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
582 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
583 the first matched character to be a colon. This applied both to named and
584 numbered groups.
585
586 29. The ucpinternal.h header file was missing its idempotency #ifdef.
587
588 30. I was sent a "project" file called libpcre.a.dev which I understand makes
589 building PCRE on Windows easier, so I have included it in the distribution.
590
591 31. There is now a check in pcretest against a ridiculously large number being
592 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
593 loop, the loop is abandoned.
594
595 32. Forward references to subpatterns in conditions such as (?(2)...) where
596 subpattern 2 is defined later cause pcre_compile() to search forwards in
597 the pattern for the relevant set of parentheses. This search went wrong
598 when there were unescaped parentheses in a character class, parentheses
599 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
600
601 33. "Subroutine" calls and backreferences were previously restricted to
602 referencing subpatterns earlier in the regex. This restriction has now
603 been removed.
604
605 34. Added a number of extra features that are going to be in Perl 5.10. On the
606 whole, these are just syntactic alternatives for features that PCRE had
607 previously implemented using the Python syntax or my own invention. The
608 other formats are all retained for compatibility.
609
610 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
611 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
612 also .NET compatible.
613
614 (b) A recursion or subroutine call to a named group can now be defined as
615 (?&name) as well as (?P>name).
616
617 (c) A backreference to a named group can now be defined as \k<name> or
618 \k'name' as well as (?P=name). The new forms, as well as being in Perl
619 5.10, are also .NET compatible.
620
621 (d) A conditional reference to a named group can now use the syntax
622 (?(<name>) or (?('name') as well as (?(name).
623
624 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
625 groups (named and numbered) that are never evaluated inline, but can be
626 called as "subroutines" from elsewhere. In effect, the DEFINE condition
627 is always false. There may be only one alternative in such a group.
628
629 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
630 as the simple (?(R). The condition is true only if the most recent
631 recursion is that of the given number or name. It does not search out
632 through the entire recursion stack.
633
634 (g) The escape \gN or \g{N} has been added, where N is a positive or
635 negative number, specifying an absolute or relative reference.
636
637 35. Tidied to get rid of some further signed/unsigned compiler warnings and
638 some "unreachable code" warnings.
639
640 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
641 things, this adds five new scripts.
642
643 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
644 There were also incompatibilities regarding the handling of \Q..\E inside
645 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
646 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
647
648 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
649 matches an empty string, and forcibly breaks the loop. There were bugs in
650 this code in non-simple cases. For a pattern such as ^(a()*)* matched
651 against aaaa the result was just "a" rather than "aaaa", for example. Two
652 separate and independent bugs (that affected different cases) have been
653 fixed.
654
655 39. Refactored the code to abolish the use of different opcodes for small
656 capturing bracket numbers. This is a tidy that I avoided doing when I
657 removed the limit on the number of capturing brackets for 3.5 back in 2001.
658 The new approach is not only tidier, it makes it possible to reduce the
659 memory needed to fix the previous bug (38).
660
661 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
662 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
663 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
664 mode.
665
666 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
667 report.
668
669 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
670 copy construction and assignment in the C++ wrapper.
671
672 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
673 couple of bugs that could have given wrong results in the "--newline=crlf"
674 case.
675
676 44. Added a number of casts and did some reorganization of signed/unsigned int
677 variables following suggestions from Dair Grant. Also renamed the variable
678 "this" as "item" because it is a C++ keyword.
679
680 45. Arranged for dftables to add
681
682 #include "pcre_internal.h"
683
684 to pcre_chartables.c because without it, gcc 4.x may remove the array
685 definition from the final binary if PCRE is built into a static library and
686 dead code stripping is activated.
687
688 46. For an unanchored pattern, if a match attempt fails at the start of a
689 newline sequence, and the newline setting is CRLF or ANY, and the next two
690 characters are CRLF, advance by two characters instead of one.
691
692
693 Version 6.7 04-Jul-06
694 ---------------------
695
696 1. In order to handle tests when input lines are enormously long, pcretest has
697 been re-factored so that it automatically extends its buffers when
698 necessary. The code is crude, but this _is_ just a test program. The
699 default size has been increased from 32K to 50K.
700
701 2. The code in pcre_study() was using the value of the re argument before
702 testing it for NULL. (Of course, in any sensible call of the function, it
703 won't be NULL.)
704
705 3. The memmove() emulation function in pcre_internal.h, which is used on
706 systems that lack both memmove() and bcopy() - that is, hardly ever -
707 was missing a "static" storage class specifier.
708
709 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
710 containing an extended class (one that cannot be represented by a bitmap
711 because it contains high-valued characters or Unicode property items, e.g.
712 [\pZ]). Almost always one would set UTF-8 mode when processing such a
713 pattern, but PCRE should not loop if you do not (it no longer does).
714 [Detail: two cases were found: (a) a repeated subpattern containing an
715 extended class; (b) a recursive reference to a subpattern that followed a
716 previous extended class. It wasn't skipping over the extended class
717 correctly when UTF-8 mode was not set.]
718
719 5. A negated single-character class was not being recognized as fixed-length
720 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
721 compile error "lookbehind assertion is not fixed length".
722
723 6. The RunPerlTest auxiliary script was showing an unexpected difference
724 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
725 write a Perl script that can interpret lines of an input file either as
726 byte characters or as UTF-8, which is what "perltest" was being required to
727 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
728 can't do is switch easily at run time between having the "use utf8;" pragma
729 or not. In the end, I fudged it by using the RunPerlTest script to insert
730 "use utf8;" explicitly for the UTF-8 tests.
731
732 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
733 the end of the subject string, contrary to the documentation and to what
734 Perl does. This was true of both matching functions. Now it matches only at
735 the start of the subject and immediately after *internal* newlines.
736
737 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
738 a pointer to an int instead of a pointer to an unsigned long int. This
739 caused problems on 64-bit systems.
740
741 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
742 instance of the 'standard' template library not being so standard".
743
744 10. There was no check on the number of named subpatterns nor the maximum
745 length of a subpattern name. The product of these values is used to compute
746 the size of the memory block for a compiled pattern. By supplying a very
747 long subpattern name and a large number of named subpatterns, the size
748 computation could be caused to overflow. This is now prevented by limiting
749 the length of names to 32 characters, and the number of named subpatterns
750 to 10,000.
751
752 11. Subpatterns that are repeated with specific counts have to be replicated in
753 the compiled pattern. The size of memory for this was computed from the
754 length of the subpattern and the repeat count. The latter is limited to
755 65535, but there was no limit on the former, meaning that integer overflow
756 could in principle occur. The compiled length of a repeated subpattern is
757 now limited to 30,000 bytes in order to prevent this.
758
759 12. Added the optional facility to have named substrings with the same name.
760
761 13. Added the ability to use a named substring as a condition, using the
762 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
763 are numbers (not recommended). Forward references are permitted.
764
765 14. Added forward references in named backreferences (if you see what I mean).
766
767 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
768 pattern could run off the end of the subject. For example, the pattern
769 "(?s)(.{1,5})"8 did this with the subject "ab".
770
771 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
772 PCRE_CASELESS was set when matching characters that were quantified with ?
773 or *.
774
775 17. A character class other than a single negated character that had a minimum
776 but no maximum quantifier - for example [ab]{6,} - was not handled
777 correctly by pce_dfa_exec(). It would match only one character.
778
779 18. A valid (though odd) pattern that looked like a POSIX character
780 class but used an invalid character after [ (for example [[,abc,]]) caused
781 pcre_compile() to give the error "Failed: internal error: code overflow" or
782 in some cases to crash with a glibc free() error. This could even happen if
783 the pattern terminated after [[ but there just happened to be a sequence of
784 letters, a binary zero, and a closing ] in the memory that followed.
785
786 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
787 over the years. Originally (before any Unicode support), just the bottom 8
788 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
789 output from "man perlunicode" includes this:
790
791 The regular expression compiler produces polymorphic opcodes. That
792 is, the pattern adapts to the data and automatically switches to
793 the Unicode character scheme when presented with Unicode data--or
794 instead uses a traditional byte scheme when presented with byte
795 data.
796
797 Sadly, a wide octal escape does not cause a switch, and in a string with
798 no other multibyte characters, these octal escapes are treated as before.
799 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
800 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
801 Unicode string.
802
803 I have not perpetrated such confusion in PCRE. Up till now, it took just
804 the bottom 8 bits, as in old Perl. I have now made octal escapes with
805 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
806 translate to the appropriate multibyte character.
807
808 29. Applied some refactoring to reduce the number of warnings from Microsoft
809 and Borland compilers. This has included removing the fudge introduced
810 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
811 a warning about an unused variable.
812
813 21. PCRE has not included VT (character 0x0b) in the set of whitespace
814 characters since release 4.0, because Perl (from release 5.004) does not.
815 [Or at least, is documented not to: some releases seem to be in conflict
816 with the documentation.] However, when a pattern was studied with
817 pcre_study() and all its branches started with \s, PCRE still included VT
818 as a possible starting character. Of course, this did no harm; it just
819 caused an unnecessary match attempt.
820
821 22. Removed a now-redundant internal flag bit that recorded the fact that case
822 dependency changed within the pattern. This was once needed for "required
823 byte" processing, but is no longer used. This recovers a now-scarce options
824 bit. Also moved the least significant internal flag bit to the most-
825 significant bit of the word, which was not previously used (hangover from
826 the days when it was an int rather than a uint) to free up another bit for
827 the future.
828
829 23. Added support for CRLF line endings as well as CR and LF. As well as the
830 default being selectable at build time, it can now be changed at runtime
831 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
832 specify that it is scanning data with non-default line endings.
833
834 24. Changed the definition of CXXLINK to make it agree with the definition of
835 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
836
837 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
838 recursions. This makes a big different to stack usage for some patterns.
839
840 26. If a subpattern containing a named recursion or subroutine reference such
841 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
842 the space required for the compiled pattern went wrong and gave too small a
843 value. Depending on the environment, this could lead to "Failed: internal
844 error: code overflow at offset 49" or "glibc detected double free or
845 corruption" errors.
846
847 27. Applied patches from Google (a) to support the new newline modes and (b) to
848 advance over multibyte UTF-8 characters in GlobalReplace.
849
850 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
851 difference for some implementation of PCRE in some Windows version.
852
853 29. Added some extra testing facilities to pcretest:
854
855 \q<number> in a data line sets the "match limit" value
856 \Q<number> in a data line sets the "match recursion limt" value
857 -S <number> sets the stack size, where <number> is in megabytes
858
859 The -S option isn't available for Windows.
860
861
862 Version 6.6 06-Feb-06
863 ---------------------
864
865 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
866 in pcreposix.h. I have copied the definition from pcre.h.
867
868 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
869 because pcre.h is no longer a built file.
870
871 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
872 not normally included in the compiled code.
873
874
875 Version 6.5 01-Feb-06
876 ---------------------
877
878 1. When using the partial match feature with pcre_dfa_exec(), it was not
879 anchoring the second and subsequent partial matches at the new starting
880 point. This could lead to incorrect results. For example, with the pattern
881 /1234/, partially matching against "123" and then "a4" gave a match.
882
883 2. Changes to pcregrep:
884
885 (a) All non-match returns from pcre_exec() were being treated as failures
886 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
887 error message is output. Some extra information is given for the
888 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
889 probably the only errors that are likely to be caused by users (by
890 specifying a regex that has nested indefinite repeats, for instance).
891 If there are more than 20 of these errors, pcregrep is abandoned.
892
893 (b) A binary zero was treated as data while matching, but terminated the
894 output line if it was written out. This has been fixed: binary zeroes
895 are now no different to any other data bytes.
896
897 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
898 used to set a locale for matching. The --locale=xxxx long option has
899 been added (no short equivalent) to specify a locale explicitly on the
900 pcregrep command, overriding the environment variables.
901
902 (d) When -B was used with -n, some line numbers in the output were one less
903 than they should have been.
904
905 (e) Added the -o (--only-matching) option.
906
907 (f) If -A or -C was used with -c (count only), some lines of context were
908 accidentally printed for the final match.
909
910 (g) Added the -H (--with-filename) option.
911
912 (h) The combination of options -rh failed to suppress file names for files
913 that were found from directory arguments.
914
915 (i) Added the -D (--devices) and -d (--directories) options.
916
917 (j) Added the -F (--fixed-strings) option.
918
919 (k) Allow "-" to be used as a file name for -f as well as for a data file.
920
921 (l) Added the --colo(u)r option.
922
923 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
924 is not present by default.
925
926 3. A nasty bug was discovered in the handling of recursive patterns, that is,
927 items such as (?R) or (?1), when the recursion could match a number of
928 alternatives. If it matched one of the alternatives, but subsequently,
929 outside the recursion, there was a failure, the code tried to back up into
930 the recursion. However, because of the way PCRE is implemented, this is not
931 possible, and the result was an incorrect result from the match.
932
933 In order to prevent this happening, the specification of recursion has
934 been changed so that all such subpatterns are automatically treated as
935 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
936
937 4. I had overlooked the fact that, in some locales, there are characters for
938 which isalpha() is true but neither isupper() nor islower() are true. In
939 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
940 and ordfeminine) are like this. This affected the treatment of \w and \W
941 when they appeared in character classes, but not when they appeared outside
942 a character class. The bit map for "word" characters is now created
943 separately from the results of isalnum() instead of just taking it from the
944 upper, lower, and digit maps. (Plus the underscore character, of course.)
945
946 5. The above bug also affected the handling of POSIX character classes such as
947 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
948 permanent tables. Instead, the bit maps for such a class were previously
949 created as the appropriate unions of the upper, lower, and digit bitmaps.
950 Now they are created by subtraction from the [[:word:]] class, which has
951 its own bitmap.
952
953 6. The [[:blank:]] character class matches horizontal, but not vertical space.
954 It is created by subtracting the vertical space characters (\x09, \x0a,
955 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
956 subtraction was done in the overall bitmap for a character class, meaning
957 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
958 be recognized. This bug has been fixed.
959
960 7. Patches from the folks at Google:
961
962 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
963 real life, but is still worth protecting against".
964
965 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
966 regular expressions".
967
968 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
969 have it.
970
971 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
972 "configure" and the latter not, in order to fix a problem somebody had
973 with compiling the Arg class on HP-UX.
974
975 (e) Improve the error-handling of the C++ wrapper a little bit.
976
977 (f) New tests for checking recursion limiting.
978
979 8. The pcre_memmove() function, which is used only if the environment does not
980 have a standard memmove() function (and is therefore rarely compiled),
981 contained two bugs: (a) use of int instead of size_t, and (b) it was not
982 returning a result (though PCRE never actually uses the result).
983
984 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
985 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
986 returned instead of calling malloc() with an overflowing number that would
987 most likely cause subsequent chaos.
988
989 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
990
991 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
992 with this option is matched, the nmatch and pmatch options of regexec() are
993 ignored.
994
995 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
996 provided in case anyone wants to the the POSIX interface with UTF-8
997 strings.
998
999 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1000 C++ linking (needed for some HP-UX environments).
1001
1002 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1003 (unused parameter) and in the pcre_printint() function (omitted "default"
1004 switch label when the default is to do nothing).
1005
1006 15. Added some code to make it possible, when PCRE is compiled as a C++
1007 library, to replace subject pointers for pcre_exec() with a smart pointer
1008 class, thus making it possible to process discontinuous strings.
1009
1010 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1011 much the same function. They were added by different people who were trying
1012 to make PCRE easy to compile on non-Unix systems. It has been suggested
1013 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1014 for compiling on Windows systems. I have therefore replaced it with
1015 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1016 defaults to "extern" for C or "extern C" for C++, which works fine on
1017 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1018 SCOPE with something explicit in config.h. In addition:
1019
1020 (a) pcreposix.h still had just "extern" instead of either of these macros;
1021 I have replaced it with PCRE_DATA_SCOPE.
1022
1023 (b) Functions such as _pcre_xclass(), which are internal to the library,
1024 but external in the C sense, all had PCRE_EXPORT in their definitions.
1025 This is apparently wrong for the Windows case, so I have removed it.
1026 (It makes no difference on Unix-like systems.)
1027
1028 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1029 of recursive calls to match(). This is different to MATCH_LIMIT because
1030 that limits the total number of calls to match(), not all of which increase
1031 the depth of recursion. Limiting the recursion depth limits the amount of
1032 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1033 when PCRE is compiled, and changed at run time. A patch from Google adds
1034 this functionality to the C++ interface.
1035
1036 18. Changes to the handling of Unicode character properties:
1037
1038 (a) Updated the table to Unicode 4.1.0.
1039
1040 (b) Recognize characters that are not in the table as "Cn" (undefined).
1041
1042 (c) I revised the way the table is implemented to a much improved format
1043 which includes recognition of ranges. It now supports the ranges that
1044 are defined in UnicodeData.txt, and it also amalgamates other
1045 characters into ranges. This has reduced the number of entries in the
1046 table from around 16,000 to around 3,000, thus reducing its size
1047 considerably. I realized I did not need to use a tree structure after
1048 all - a binary chop search is just as efficient. Having reduced the
1049 number of entries, I extended their size from 6 bytes to 8 bytes to
1050 allow for more data.
1051
1052 (d) Added support for Unicode script names via properties such as \p{Han}.
1053
1054 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1055 matching that character.
1056
1057 20. When matching a repeated Unicode property with a minimum greater than zero,
1058 (for example \pL{2,}), PCRE could look past the end of the subject if it
1059 reached it while seeking the minimum number of characters. This could
1060 happen only if some of the characters were more than one byte long, because
1061 there is a check for at least the minimum number of bytes.
1062
1063 21. Refactored the implementation of \p and \P so as to be more general, to
1064 allow for more different types of property in future. This has changed the
1065 compiled form incompatibly. Anybody with saved compiled patterns that use
1066 \p or \P will have to recompile them.
1067
1068 22. Added "Any" and "L&" to the supported property types.
1069
1070 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1071 but give a compile time error if the value is greater than 0xff.
1072
1073 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1074 accidentally not being installed or uninstalled.
1075
1076 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1077 made were to insert the current release number. This seemed silly, because
1078 it made things harder for people building PCRE on systems that don't run
1079 "configure". I have turned pcre.h into a distributed file, no longer built
1080 by "configure", with the version identification directly included. There is
1081 no longer a pcre.h.in file.
1082
1083 However, this change necessitated a change to the pcre-config script as
1084 well. It is built from pcre-config.in, and one of the substitutions was the
1085 release number. I have updated configure.ac so that ./configure now finds
1086 the release number by grepping pcre.h.
1087
1088 26. Added the ability to run the tests under valgrind.
1089
1090
1091 Version 6.4 05-Sep-05
1092 ---------------------
1093
1094 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1095 "--" to be printed when multiple files were scanned, even when none of the
1096 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1097 consider it to be a bug, and have restored the previous behaviour.
1098
1099 2. A couple of code tidies to get rid of compiler warnings.
1100
1101 3. The pcretest program used to cheat by referring to symbols in the library
1102 whose names begin with _pcre_. These are internal symbols that are not
1103 really supposed to be visible externally, and in some environments it is
1104 possible to suppress them. The cheating is now confined to including
1105 certain files from the library's source, which is a bit cleaner.
1106
1107 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1108 file's purpose clearer.
1109
1110 5. Reorganized pcre_ucp_findchar().
1111
1112
1113 Version 6.3 15-Aug-05
1114 ---------------------
1115
1116 1. The file libpcre.pc.in did not have general read permission in the tarball.
1117
1118 2. There were some problems when building without C++ support:
1119
1120 (a) If C++ support was not built, "make install" and "make test" still
1121 tried to test it.
1122
1123 (b) There were problems when the value of CXX was explicitly set. Some
1124 changes have been made to try to fix these, and ...
1125
1126 (c) --disable-cpp can now be used to explicitly disable C++ support.
1127
1128 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1129 backslash in a target when C++ was disabled. This confuses some
1130 versions of "make", apparently. Using an intermediate variable solves
1131 this. (Same for CPP_LOBJ.)
1132
1133 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1134 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1135 necessary on certain architectures.
1136
1137 4. Added a setting of -export-symbols-regex to the link command to remove
1138 those symbols that are exported in the C sense, but actually are local
1139 within the library, and not documented. Their names all begin with
1140 "_pcre_". This is not a perfect job, because (a) we have to except some
1141 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1142 available (and never for static libraries). I have made a note to try to
1143 find a way round (a) in the future.
1144
1145
1146 Version 6.2 01-Aug-05
1147 ---------------------
1148
1149 1. There was no test for integer overflow of quantifier values. A construction
1150 such as {1111111111111111} would give undefined results. What is worse, if
1151 a minimum quantifier for a parenthesized subpattern overflowed and became
1152 negative, the calculation of the memory size went wrong. This could have
1153 led to memory overwriting.
1154
1155 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1156
1157 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1158 operating environments where this matters.
1159
1160 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1161 PCRE options from within the C++ wrapper.
1162
1163 5. Named capturing subpatterns were not being correctly counted when a pattern
1164 was compiled. This caused two problems: (a) If there were more than 100
1165 such subpatterns, the calculation of the memory needed for the whole
1166 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1167 back references of the form \12, where the number was greater than 9, were
1168 not recognized as back references, even though there were sufficient
1169 previous subpatterns.
1170
1171 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1172 versions of gcc, e.g. 2.95.4.
1173
1174
1175 Version 6.1 21-Jun-05
1176 ---------------------
1177
1178 1. There was one reference to the variable "posix" in pcretest.c that was not
1179 surrounded by "#if !defined NOPOSIX".
1180
1181 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1182 the cross-check on the old pcre_info() function, for the benefit of the
1183 cut-down version of PCRE that is currently imported into Exim.
1184
1185 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1186 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1187 patterns that start (?i) but I don't think that matters. The use of (?i) is
1188 just an example; this all applies to the other options as well.
1189
1190 4. Since libtool seems to echo the compile commands it is issuing, the output
1191 from "make" can be reduced a bit by putting "@" in front of each libtool
1192 compile command.
1193
1194 5. Patch from the folks at Google for configure.in to be a bit more thorough
1195 in checking for a suitable C++ installation before trying to compile the
1196 C++ stuff. This should fix a reported problem when a compiler was present,
1197 but no suitable headers.
1198
1199 6. The man pages all had just "PCRE" as their title. I have changed them to
1200 be the relevant file name. I have also arranged that these names are
1201 retained in the file doc/pcre.txt, which is a concatenation in text format
1202 of all the man pages except the little individual ones for each function.
1203
1204 7. The NON-UNIX-USE file had not been updated for the different set of source
1205 files that come with release 6. I also added a few comments about the C++
1206 wrapper.
1207
1208
1209 Version 6.0 07-Jun-05
1210 ---------------------
1211
1212 1. Some minor internal re-organization to help with my DFA experiments.
1213
1214 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1215 didn't matter for the library itself when fully configured, but did matter
1216 when compiling without UCP support, or within Exim, where the ucp files are
1217 not imported.
1218
1219 3. Refactoring of the library code to split up the various functions into
1220 different source modules. The addition of the new DFA matching code (see
1221 below) to a single monolithic source would have made it really too
1222 unwieldy, quite apart from causing all the code to be include in a
1223 statically linked application, when only some functions are used. This is
1224 relevant even without the DFA addition now that patterns can be compiled in
1225 one application and matched in another.
1226
1227 The downside of splitting up is that there have to be some external
1228 functions and data tables that are used internally in different modules of
1229 the library but which are not part of the API. These have all had their
1230 names changed to start with "_pcre_" so that they are unlikely to clash
1231 with other external names.
1232
1233 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1234 a different (DFA) algorithm. Although it is slower than the original
1235 function, it does have some advantages for certain types of matching
1236 problem.
1237
1238 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1239 including restarting after a partial match.
1240
1241 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1242 defined when compiling for Windows was sent to me. I have put it into the
1243 code, though I have no means of testing or verifying it.
1244
1245 7. Added the pcre_refcount() auxiliary function.
1246
1247 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1248 match before or at the first newline in the subject string. In pcretest,
1249 the /f option on a pattern can be used to set this.
1250
1251 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1252 would behave wrongly. This has been present in PCRE since release 4.0.
1253
1254 10. A number of changes to the pcregrep command:
1255
1256 (a) Refactored how -x works; insert ^(...)$ instead of setting
1257 PCRE_ANCHORED and checking the length, in preparation for adding
1258 something similar for -w.
1259
1260 (b) Added the -w (match as a word) option.
1261
1262 (c) Refactored the way lines are read and buffered so as to have more
1263 than one at a time available.
1264
1265 (d) Implemented a pcregrep test script.
1266
1267 (e) Added the -M (multiline match) option. This allows patterns to match
1268 over several lines of the subject. The buffering ensures that at least
1269 8K, or the rest of the document (whichever is the shorter) is available
1270 for matching (and similarly the previous 8K for lookbehind assertions).
1271
1272 (f) Changed the --help output so that it now says
1273
1274 -w, --word-regex(p)
1275
1276 instead of two lines, one with "regex" and the other with "regexp"
1277 because that confused at least one person since the short forms are the
1278 same. (This required a bit of code, as the output is generated
1279 automatically from a table. It wasn't just a text change.)
1280
1281 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1282 option but starts with a hyphen. Could be a pattern or a path name
1283 starting with a hyphen, for instance.
1284
1285 (h) "-" can be given as a file name to represent stdin.
1286
1287 (i) When file names are being printed, "(standard input)" is used for
1288 the standard input, for compatibility with GNU grep. Previously
1289 "<stdin>" was used.
1290
1291 (j) The option --label=xxx can be used to supply a name to be used for
1292 stdin when file names are being printed. There is no short form.
1293
1294 (k) Re-factored the options decoding logic because we are going to add
1295 two more options that take data. Such options can now be given in four
1296 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1297
1298 (l) Added the -A, -B, and -C options for requesting that lines of context
1299 around matches be printed.
1300
1301 (m) Added the -L option to print the names of files that do not contain
1302 any matching lines, that is, the complement of -l.
1303
1304 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1305 continue to scan other files.
1306
1307 (o) The -s option was incorrectly implemented. For compatibility with other
1308 greps, it now suppresses the error message for a non-existent or non-
1309 accessible file (but not the return code). There is a new option called
1310 -q that suppresses the output of matching lines, which was what -s was
1311 previously doing.
1312
1313 (p) Added --include and --exclude options to specify files for inclusion
1314 and exclusion when recursing.
1315
1316 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1317 Hopefully, it now does.
1318
1319 12. Missing cast in pcre_study().
1320
1321 13. Added an "uninstall" target to the makefile.
1322
1323 14. Replaced "extern" in the function prototypes in Makefile.in with
1324 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1325 world, but is set differently for Windows.
1326
1327 15. Added a second compiling function called pcre_compile2(). The only
1328 difference is that it has an extra argument, which is a pointer to an
1329 integer error code. When there is a compile-time failure, this is set
1330 non-zero, in addition to the error test pointer being set to point to an
1331 error message. The new argument may be NULL if no error number is required
1332 (but then you may as well call pcre_compile(), which is now just a
1333 wrapper). This facility is provided because some applications need a
1334 numeric error indication, but it has also enabled me to tidy up the way
1335 compile-time errors are handled in the POSIX wrapper.
1336
1337 16. Added VPATH=.libs to the makefile; this should help when building with one
1338 prefix path and installing with another. (Or so I'm told by someone who
1339 knows more about this stuff than I do.)
1340
1341 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1342 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1343 match everything, including newlines. This is not POSIX-compatible, but
1344 somebody wanted the feature. From pcretest it can be activated by using
1345 both the P and the s flags.
1346
1347 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1348
1349 19. libpcre.pc was being incorrectly installed as executable.
1350
1351 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1352 it now also looks for '\r' so that it will work unmodified on Windows.
1353
1354 21. Added Google's contributed C++ wrapper to the distribution.
1355
1356 22. Added some untidy missing memory free() calls in pcretest, to keep
1357 Electric Fence happy when testing.
1358
1359
1360
1361 Version 5.0 13-Sep-04
1362 ---------------------
1363
1364 1. Internal change: literal characters are no longer packed up into items
1365 containing multiple characters in a single byte-string. Each character
1366 is now matched using a separate opcode. However, there may be more than one
1367 byte in the character in UTF-8 mode.
1368
1369 2. The pcre_callout_block structure has two new fields: pattern_position and
1370 next_item_length. These contain the offset in the pattern to the next match
1371 item, and its length, respectively.
1372
1373 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1374 insertion of callouts before each pattern item. Added the /C option to
1375 pcretest to make use of this.
1376
1377 4. On the advice of a Windows user, the lines
1378
1379 #if defined(_WIN32) || defined(WIN32)
1380 _setmode( _fileno( stdout ), 0x8000 );
1381 #endif /* defined(_WIN32) || defined(WIN32) */
1382
1383 have been added to the source of pcretest. This apparently does useful
1384 magic in relation to line terminators.
1385
1386 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1387 for the benefit of those environments where the "b" makes a difference.
1388
1389 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1390 to know about it. I have put a hack into configure.in that adds in code
1391 to set GCC=yes if CC=icc. This seems to end up at a point in the
1392 generated configure script that is early enough to affect the setting of
1393 compiler options, which is what is needed, but I have no means of testing
1394 whether it really works. (The user who reported this had patched the
1395 generated configure script, which of course I cannot do.)
1396
1397 LATER: After change 22 below (new libtool files), the configure script
1398 seems to know about icc (and also ecc). Therefore, I have commented out
1399 this hack in configure.in.
1400
1401 7. Added support for pkg-config (2 patches were sent in).
1402
1403 8. Negated POSIX character classes that used a combination of internal tables
1404 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1405 [[:^ascii]]. Typically, they would match almost any characters. The other
1406 POSIX classes were not broken in this way.
1407
1408 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1409 to find the match, as PCRE was deluded into thinking that the match had to
1410 start at the start point or following a newline. The same bug applied to
1411 patterns with negative forward assertions or any backward assertions
1412 preceding ".*" at the start, unless the pattern required a fixed first
1413 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1414
1415 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1416 starting at the last subject character, bytes beyond the end of the subject
1417 string were read.
1418
1419 11. Renamed the variable "class" as "classbits" to make life easier for C++
1420 users. (Previously there was a macro definition, but it apparently wasn't
1421 enough.)
1422
1423 12. Added the new field "tables" to the extra data so that tables can be passed
1424 in at exec time, or the internal tables can be re-selected. This allows
1425 a compiled regex to be saved and re-used at a later time by a different
1426 program that might have everything at different addresses.
1427
1428 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1429 -R library as well as a -L library.
1430
1431 14. The debugging options of pcretest (-d on the command line or D on a
1432 pattern) showed incorrect output for anything following an extended class
1433 that contained multibyte characters and which was followed by a quantifier.
1434
1435 15. Added optional support for general category Unicode character properties
1436 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1437 support. It adds about 90K to the size of the library. The meanings of the
1438 inbuilt class escapes such as \d and \s have NOT been changed.
1439
1440 16. Updated pcredemo.c to include calls to free() to release the memory for the
1441 compiled pattern.
1442
1443 17. The generated file chartables.c was being created in the source directory
1444 instead of in the building directory. This caused the build to fail if the
1445 source directory was different from the building directory, and was
1446 read-only.
1447
1448 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1449 file. No doubt somebody will tell me if they don't make sense... Also added
1450 Dan Mooney's comments about building on OpenVMS.
1451
1452 19. Added support for partial matching via the PCRE_PARTIAL option for
1453 pcre_exec() and the \P data escape in pcretest.
1454
1455 20. Extended pcretest with 3 new pattern features:
1456
1457 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1458 write the compiled pattern to the file whose name is "rest-of-line".
1459 This is a straight binary dump of the data, with the saved pointer to
1460 the character tables forced to be NULL. The study data, if any, is
1461 written too. After writing, pcretest reads a new pattern.
1462
1463 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1464 compiled pattern from the given file. There must not be any
1465 occurrences of "<" in the file name (pretty unlikely); if there are,
1466 pcretest will instead treat the initial "<" as a pattern delimiter.
1467 After reading in the pattern, pcretest goes on to read data lines as
1468 usual.
1469
1470 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1471 and 16-bit fields in a compiled pattern, to simulate a pattern that
1472 was compiled on a host of opposite endianness.
1473
1474 21. The pcre-exec() function can now cope with patterns that were compiled on
1475 hosts of opposite endianness, with this restriction:
1476
1477 As for any compiled expression that is saved and used later, the tables
1478 pointer field cannot be preserved; the extra_data field in the arguments
1479 to pcre_exec() should be used to pass in a tables address if a value
1480 other than the default internal tables were used at compile time.
1481
1482 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1483 now diagnosed as an error. Previously, most of the time, a negative number
1484 would have been treated as zero, but if in addition "ovector" was passed as
1485 NULL, a crash could occur.
1486
1487 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1488 new versions from the libtool 1.5 distribution (the last one is a copy of
1489 a file called libtool.m4). This seems to have fixed the need to patch
1490 "configure" to support Darwin 1.3 (which I used to do). However, I still
1491 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1492 workstation).
1493
1494 24. Changed the PCRE licence to be the more standard "BSD" licence.
1495
1496
1497 Version 4.5 01-Dec-03
1498 ---------------------
1499
1500 1. There has been some re-arrangement of the code for the match() function so
1501 that it can be compiled in a version that does not call itself recursively.
1502 Instead, it keeps those local variables that need separate instances for
1503 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1504 needs to "recurse". Keeping track of where control must go is done by means
1505 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1506 hide most of the details from the main code, and operates only if
1507 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1508 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1509 operating.
1510
1511 To make it easier for callers to provide specially tailored get/free
1512 functions for this usage, two new functions, pcre_stack_malloc, and
1513 pcre_stack_free, are used. They are always called in strict stacking order,
1514 and the size of block requested is always the same.
1515
1516 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1517 PCRE has been compiled to use the stack or the heap for recursion. The
1518 -C option of pcretest uses this to show which version is compiled.
1519
1520 A new data escape \S, is added to pcretest; it causes the amounts of store
1521 obtained and freed by both kinds of malloc/free at match time to be added
1522 to the output.
1523
1524 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1525 what's available on my current Linux desktop machine.
1526
1527 3. When matching a UTF-8 string, the test for a valid string at the start has
1528 been extended. If start_offset is not zero, PCRE now checks that it points
1529 to a byte that is the start of a UTF-8 character. If not, it returns
1530 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1531 this is necessary because there may be backward assertions in the pattern.
1532 When matching the same subject several times, it may save resources to use
1533 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1534
1535 4. The code for checking the validity of UTF-8 strings has been tightened so
1536 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1537 containing "overlong sequences".
1538
1539 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1540 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1541 should have been "&", but it just so happened that all the cases this let
1542 through by mistake were picked up later in the function.
1543
1544 6. I had used a variable called "isblank" - this is a C99 function, causing
1545 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1546
1547 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1548 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1549 so the version is shown; (c) stop "make test" echoing "./RunTest".
1550
1551 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1552
1553 9. The prototype for memmove() for systems that don't have it was using
1554 size_t, but the inclusion of the header that defines size_t was later. I've
1555 moved the #includes for the C headers earlier to avoid this.
1556
1557 10. Added some adjustments to the code to make it easier to compiler on certain
1558 special systems:
1559
1560 (a) Some "const" qualifiers were missing.
1561 (b) Added the macro EXPORT before all exported functions; by default this
1562 is defined to be empty.
1563 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1564 that it reads its output file name as an argument instead of writing
1565 to the standard output and assuming this can be redirected.
1566
1567 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1568 class containing characters with values greater than 255, PCRE compilation
1569 went into a loop.
1570
1571 12. A recursive reference to a subpattern that was within another subpattern
1572 that had a minimum quantifier of zero caused PCRE to crash. For example,
1573 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1574 recursion. If the recursively-called subpattern itself had a zero repeat,
1575 that was OK.
1576
1577 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1578 buffer into which it was copied (for escape processing) was still set at
1579 1024, so long lines caused crashes.
1580
1581 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1582 "internal error: code overflow...". This applied to any character class
1583 that was followed by a possessive quantifier.
1584
1585 15. Modified the Makefile to add libpcre.la as a prerequisite for
1586 libpcreposix.la because I was told this is needed for a parallel build to
1587 work.
1588
1589 16. If a pattern that contained .* following optional items at the start was
1590 studied, the wrong optimizing data was generated, leading to matching
1591 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1592 matching string must start with a or b or c. The correct conclusion for
1593 this pattern is that a match can start with any character.
1594
1595
1596 Version 4.4 13-Aug-03
1597 ---------------------
1598
1599 1. In UTF-8 mode, a character class containing characters with values between
1600 127 and 255 was not handled correctly if the compiled pattern was studied.
1601 In fixing this, I have also improved the studying algorithm for such
1602 classes (slightly).
1603
1604 2. Three internal functions had redundant arguments passed to them. Removal
1605 might give a very teeny performance improvement.
1606
1607 3. Documentation bug: the value of the capture_top field in a callout is *one
1608 more than* the number of the hightest numbered captured substring.
1609
1610 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1611 in incorrectly linking with a previously installed version. They now link
1612 explicitly with libpcre.la.
1613
1614 5. configure.in no longer needs to recognize Cygwin specially.
1615
1616 6. A problem in pcre.in for Windows platforms is fixed.
1617
1618 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1619 pcretest, it used to include the size of the study block as part of its
1620 output. Unfortunately, the structure contains a field that has a different
1621 size on different hardware architectures. This meant that the tests that
1622 showed this size failed. As the block is currently always of a fixed size,
1623 this information isn't actually particularly useful in pcretest output, so
1624 I have just removed it.
1625
1626 8. Three pre-processor statements accidentally did not start in column 1.
1627 Sadly, there are *still* compilers around that complain, even though
1628 standard C has not required this for well over a decade. Sigh.
1629
1630 9. In pcretest, the code for checking callouts passed small integers in the
1631 callout_data field, which is a void * field. However, some picky compilers
1632 complained about the casts involved for this on 64-bit systems. Now
1633 pcretest passes the address of the small integer instead, which should get
1634 rid of the warnings.
1635
1636 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1637 both compile and run time, and gives an error if an invalid UTF-8 sequence
1638 is found. There is a option for disabling this check in cases where the
1639 string is known to be correct and/or the maximum performance is wanted.
1640
1641 11. In response to a bug report, I changed one line in Makefile.in from
1642
1643 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1644 to
1645 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1646
1647 to look similar to other lines, but I have no way of telling whether this
1648 is the right thing to do, as I do not use Windows. No doubt I'll get told
1649 if it's wrong...
1650
1651
1652 Version 4.3 21-May-03
1653 ---------------------
1654
1655 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1656 Makefile.
1657
1658 2. Some refactoring to improve the quality of the code:
1659
1660 (i) The utf8_table... variables are now declared "const".
1661
1662 (ii) The code for \cx, which used the "case flipping" table to upper case
1663 lower case letters, now just substracts 32. This is ASCII-specific,
1664 but the whole concept of \cx is ASCII-specific, so it seems
1665 reasonable.
1666
1667 (iii) PCRE was using its character types table to recognize decimal and
1668 hexadecimal digits in the pattern. This is silly, because it handles
1669 only 0-9, a-f, and A-F, but the character types table is locale-
1670 specific, which means strange things might happen. A private
1671 table is now used for this - though it costs 256 bytes, a table is
1672 much faster than multiple explicit tests. Of course, the standard
1673 character types table is still used for matching digits in subject
1674 strings against \d.
1675
1676 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1677 ending in _t are). So I've renamed it as ESC_tee.
1678
1679 3. The first argument for regexec() in the POSIX wrapper should have been
1680 defined as "const".
1681
1682 4. Changed pcretest to use malloc() for its buffers so that they can be
1683 Electric Fenced for debugging.
1684
1685 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1686 to read one or more bytes before the start of the subject string. Often this
1687 had no effect on PCRE's behaviour, but in some circumstances it could
1688 provoke a segmentation fault.
1689
1690 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1691 to try to read one or more bytes before the start of the subject string.
1692
1693 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1694 UTF-8 support could misbehave in various ways if the subject string
1695 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1696 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1697 back over UTF-8 characters.)
1698
1699
1700 Version 4.2 14-Apr-03
1701 ---------------------
1702
1703 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1704
1705 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1706 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1707 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1708 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1709 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1710 and BUILD_EXEEXT
1711 Note: automatic setting of the BUILD variables is not yet working
1712 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1713 compile-time but not at link-time
1714 [LINK]: use for linking executables only
1715 make different versions for Windows and non-Windows
1716 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1717 libraries
1718 [LINK_FOR_BUILD]: new variable
1719 [OBJEXT]: use throughout
1720 [EXEEXT]: use throughout
1721 <winshared>: new target
1722 <wininstall>: new target
1723 <dftables.o>: use native compiler
1724 <dftables>: use native linker
1725 <install>: handle Windows platform correctly
1726 <clean>: ditto
1727 <check>: ditto
1728 copy DLL to top builddir before testing
1729
1730 As part of these changes, -no-undefined was removed again. This was reported
1731 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1732 in any case.
1733
1734 3. Some tidies to get rid of compiler warnings:
1735
1736 . In the match_data structure, match_limit was an unsigned long int, whereas
1737 match_call_count was an int. I've made them both unsigned long ints.
1738
1739 . In pcretest the fact that a const uschar * doesn't automatically cast to
1740 a void * provoked a warning.
1741
1742 . Turning on some more compiler warnings threw up some "shadow" variables
1743 and a few more missing casts.
1744
1745 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1746 option, a class that contained a single character with a value between 128
1747 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1748
1749 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1750 option, a class that contained several characters, but with at least one
1751 whose value was between 128 and 255 caused PCRE to crash.
1752
1753
1754 Version 4.1 12-Mar-03
1755 ---------------------
1756
1757 1. Compiling with gcc -pedantic found a couple of places where casts were
1758 needed, and a string in dftables.c that was longer than standard compilers are
1759 required to support.
1760
1761 2. Compiling with Sun's compiler found a few more places where the code could
1762 be tidied up in order to avoid warnings.
1763
1764 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1765 first of these names is deprecated in the latest Autoconf in favour of the name
1766 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1767 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1768 analogy I have changed it to CFLAGS_FOR_BUILD.
1769
1770 4. Added -no-undefined to the linking command in the Makefile, because this is
1771 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1772 linking step for the pcreposix library.
1773
1774 5. PCRE was failing to diagnose the case of two named groups with the same
1775 name.
1776
1777 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1778 literal character that is needed in the subject for a match, and scans along to
1779 ensure that it is present before embarking on the full matching process. This
1780 saves time in cases of nested unlimited repeats that are never going to match.
1781 Problem: the scan can take a lot of time if the subject is very long (e.g.
1782 megabytes), thus penalizing straightforward matches. It is now done only if the
1783 amount of subject to be scanned is less than 1000 bytes.
1784
1785 7. A lesser problem with the same optimization is that it was recording the
1786 first character of an anchored pattern as "needed", thus provoking a search
1787 right along the subject, even when the first match of the pattern was going to
1788 fail. The "needed" character is now not set for anchored patterns, unless it
1789 follows something in the pattern that is of non-fixed length. Thus, it still
1790 fulfils its original purpose of finding quick non-matches in cases of nested
1791 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1792
1793
1794 Version 4.0 17-Feb-03
1795 ---------------------
1796
1797 1. If a comment in an extended regex that started immediately after a meta-item
1798 extended to the end of string, PCRE compiled incorrect data. This could lead to
1799 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1800
1801 2. Moved to autoconf 2.53 and libtool 1.4.2.
1802
1803 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1804 the special perltest8 script is no longer needed - all the tests can be run
1805 from a single perltest script.
1806
1807 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1808 by \s. It has now been removed in PCRE. This means it isn't recognized as
1809 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1810 class [:space:] *does* include VT, thereby creating a mess.
1811
1812 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1813 space and tab.
1814
1815 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1816 its new features into the main test script, reducing the number of scripts.
1817
1818 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1819 were backward compatible, and made the (?i) apply to the whole pattern, as if
1820 /i were given. Now it behaves more logically, and applies the option setting
1821 only to what follows. PCRE has been changed to follow suit. However, if it
1822 finds options settings right at the start of the pattern, it extracts them into
1823 the global options, as before. Thus, they show up in the info data.
1824
1825 8. Added support for the \Q...\E escape sequence. Characters in between are
1826 treated as literals. This is slightly different from Perl in that $ and @ are
1827 also handled as literals inside the quotes. In Perl, they will cause variable
1828 interpolation. Note the following examples:
1829
1830 Pattern PCRE matches Perl matches
1831
1832 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1833 \Qabc\$xyz\E abc\$xyz abc\$xyz
1834 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1835
1836 For compatibility with Perl, \Q...\E sequences are recognized inside character
1837 classes as well as outside them.
1838
1839 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1840 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1841 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1842 signed/unsigned warnings.
1843
1844 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1845 option for pcretest, so I've replaced it by a simple function that does just
1846 that job.
1847
1848 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1849 "pcregrep -".
1850
1851 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1852 Java package. This provides some syntactic sugar for simple cases of what my
1853 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1854 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1855 item, you can use this simplified notation. Note that only makes sense with
1856 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1857 greediness, whatever the setting of the PCRE_UNGREEDY option.
1858
1859 13. A change of greediness default within a pattern was not taking effect at
1860 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1861 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1862 was abstracted outside.
1863
1864 14. PCRE now supports the \G assertion. It is true when the current matching
1865 position is at the start point of the match. This differs from \A when the
1866 starting offset is non-zero. Used with the /g option of pcretest (or similar
1867 code), it works in the same way as it does for Perl's /g option. If all
1868 alternatives of a regex begin with \G, the expression is anchored to the start
1869 match position, and the "anchored" flag is set in the compiled expression.
1870
1871 15. Some bugs concerning the handling of certain option changes within patterns
1872 have been fixed. These applied to options other than (?ims). For example,
1873 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1874 been the other way round. Some of this was related to change 7 above.
1875
1876 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1877 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1878 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1879 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1880
1881 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1882 mode. Unlike ".", it always matches newline, whatever the setting of
1883 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1884 assertions. Perl allows it, but it doesn't (in general) work because it can't
1885 calculate the length of the lookbehind. At least, that's the case for Perl
1886 5.8.0 - I've been told they are going to document that it doesn't work in
1887 future.
1888
1889 18. Added an error diagnosis for escapes that PCRE does not support: these are
1890 \L, \l, \N, \P, \p, \U, \u, and \X.
1891
1892 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1893 reading past the end of the pattern in cases such as /[abcd/.
1894
1895 20. PCRE was getting more memory than necessary for patterns with classes that
1896 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1897
1898 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1899 compiling PCRE for use with Virtual Pascal.
1900
1901 22. Small fix to the Makefile to make it work properly if the build is done
1902 outside the source tree.
1903
1904 23. Added a new extension: a condition to go with recursion. If a conditional
1905 subpattern starts with (?(R) the "true" branch is used if recursion has
1906 happened, whereas the "false" branch is used only at the top level.
1907
1908 24. When there was a very long string of literal characters (over 255 bytes
1909 without UTF support, over 250 bytes with UTF support), the computation of how
1910 much memory was required could be incorrect, leading to segfaults or other
1911 strange effects.
1912
1913 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1914 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1915 there was a subsequent back reference to those brackets. This meant that, for
1916 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1917 possible to check for precisely this case. All we can do is abandon the
1918 optimization if .* occurs inside capturing brackets when there are any back
1919 references whatsoever. (See below for a better fix that came later.)
1920
1921 26. The handling of the optimization for finding the first character of a
1922 non-anchored pattern, and for finding a character that is required later in the
1923 match were failing in some cases. This didn't break the matching; it just
1924 failed to optimize when it could. The way this is done has been re-implemented.
1925
1926 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1927
1928 28. Added a new feature that provides some of the functionality that Perl
1929 provides with (?{...}). The facility is termed a "callout". The way it is done
1930 in PCRE is for the caller to provide an optional function, by setting
1931 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1932 global variable. By default it is unset, which disables all calling out. To get
1933 the function called, the regex must include (?C) at appropriate points. This
1934 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1935 This provides a means of identifying different callout points. When PCRE
1936 reaches such a point in the regex, if pcre_callout has been set, the external
1937 function is called. It is provided with data in a structure called
1938 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1939 matching continues; if it returns a non-zero value, the match at the current
1940 point fails. However, backtracking will occur if possible. [This was changed
1941 later and other features added - see item 49 below.]
1942
1943 29. pcretest is upgraded to test the callout functionality. It provides a
1944 callout function that displays information. By default, it shows the start of
1945 the match and the current position in the text. There are some new data escapes
1946 to vary what happens:
1947
1948 \C+ in addition, show current contents of captured substrings
1949 \C- do not supply a callout function
1950 \C!n return 1 when callout number n is reached
1951 \C!n!m return 1 when callout number n is reached for the mth time
1952
1953 30. If pcregrep was called with the -l option and just a single file name, it
1954 output "<stdin>" if a match was found, instead of the file name.
1955
1956 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1957 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1958 pcre_exec(). This saves a malloc/free per call. The default value of
1959 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1960 when configuring.
1961
1962 32. The default maximum size of a compiled pattern is 64K. There have been a
1963 few cases of people hitting this limit. The code now uses macros to handle the
1964 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1965 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1966 configuring. Tests 2 and 5 work only with 2-byte links because they output
1967 debugging information about compiled patterns.
1968
1969 33. Internal code re-arrangements:
1970
1971 (a) Moved the debugging function for printing out a compiled regex into
1972 its own source file (printint.c) and used #include to pull it into
1973 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1974 separate copies.
1975
1976 (b) Defined the list of op-code names for debugging as a macro in
1977 internal.h so that it is next to the definition of the opcodes.
1978
1979 (c) Defined a table of op-code lengths for simpler skipping along compiled
1980 code. This is again a macro in internal.h so that it is next to the
1981 definition of the opcodes.
1982
1983 34. Added support for recursive calls to individual subpatterns, along the
1984 lines of Robin Houston's patch (but implemented somewhat differently).
1985
1986 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1987 allow it to read and process whole directories in Win32. This code was
1988 contributed by Lionel Fourquaux; it has not been tested by me.
1989
1990 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1991 used to name a group. Names consist of alphanumerics and underscores, and must
1992 be unique. Back references use the syntax (?P=name) and recursive calls use
1993 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1994 numbers. The function pcre_fullinfo() can be used after compilation to extract
1995 a name/number map. There are three relevant calls:
1996
1997 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1998 PCRE_INFO_NAMECOUNT yields the number of entries
1999 PCRE_INFO_NAMETABLE yields a pointer to the map.
2000
2001 The map is a vector of fixed-size entries. The size of each entry depends on
2002 the length of the longest name used. The first two bytes of each entry are the
2003 group number, most significant byte first. There follows the corresponding
2004 name, zero terminated. The names are in alphabetical order.
2005
2006 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2007 case instead of 255. Making it the same both with and without UTF-8 support
2008 means that the same test output works with both.
2009
2010 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2011 calling malloc() with a zero argument.
2012
2013 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2014 optimization. I've improved things by keeping a bitmap of backreferences with
2015 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2016 fact referenced, the optimization can be applied. It is unlikely that a
2017 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2018 the match to follow \n) will appear inside brackets with a number greater than
2019 31, but if it does, any back reference > 31 suppresses the optimization.
2020
2021 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2022 of disabling numbered capturing parentheses. Any opening parenthesis that is
2023 not followed by ? behaves as if it were followed by ?: but named parentheses
2024 can still be used for capturing (and they will acquire numbers in the usual
2025 way).
2026
2027 41. Redesigned the return codes from the match() function into yes/no/error so
2028 that errors can be passed back from deep inside the nested calls. A malloc
2029 failure while inside a recursive subpattern call now causes the
2030 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2031
2032 42. It is now possible to set a limit on the number of times the match()
2033 function is called in a call to pcre_exec(). This facility makes it possible to
2034 limit the amount of recursion and backtracking, though not in a directly
2035 obvious way, because the match() function is used in a number of different
2036 circumstances. The count starts from zero for each position in the subject
2037 string (for non-anchored patterns). The default limit is, for compatibility, a
2038 large number, namely 10 000 000. You can change this in two ways:
2039
2040 (a) When configuring PCRE before making, you can use --with-match-limit=n
2041 to set a default value for the compiled library.
2042
2043 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2044 a different value is set. See 45 below.
2045
2046 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2047
2048 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2049 of things that can be changed at compile time. The first argument specifies
2050 what is wanted and the second points to where the information is to be placed.
2051 The current list of available information is:
2052
2053 PCRE_CONFIG_UTF8
2054
2055 The output is an integer that is set to one if UTF-8 support is available;
2056 otherwise it is set to zero.
2057
2058 PCRE_CONFIG_NEWLINE
2059
2060 The output is an integer that it set to the value of the code that is used for
2061 newline. It is either LF (10) or CR (13).
2062
2063 PCRE_CONFIG_LINK_SIZE
2064
2065 The output is an integer that contains the number of bytes used for internal
2066 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2067
2068 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2069
2070 The output is an integer that contains the threshold above which the POSIX
2071 interface uses malloc() for output vectors. See item 31 above.
2072
2073 PCRE_CONFIG_MATCH_LIMIT
2074
2075 The output is an unsigned integer that contains the default limit of the number
2076 of match() calls in a pcre_exec() execution. See 42 above.
2077
2078 44. pcretest has been upgraded by the addition of the -C option. This causes it
2079 to extract all the available output from the new pcre_config() function, and to
2080 output it. The program then exits immediately.
2081
2082 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2083 order to support additional features. One way would have been to define
2084 pcre_exec2() (for example) with extra arguments, but this would not have been
2085 extensible, and would also have required all calls to the original function to
2086 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2087 is used for passing in "extra" data from pcre_study().
2088
2089 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2090 contains the following fields:
2091
2092 flags a bitmap indicating which of the following fields are set
2093 study_data opaque data from pcre_study()
2094 match_limit a way of specifying a limit on match() calls for a specific
2095 call to pcre_exec()
2096 callout_data data for callouts (see 49 below)
2097
2098 The flag bits are also defined in pcre.h, and are
2099
2100 PCRE_EXTRA_STUDY_DATA
2101 PCRE_EXTRA_MATCH_LIMIT
2102 PCRE_EXTRA_CALLOUT_DATA
2103
2104 The pcre_study() function now returns one of these new pcre_extra blocks, with
2105 the actual study data pointed to by the study_data field, and the
2106 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2107 before. That is, this change is entirely upwards-compatible and requires no
2108 change to existing code.
2109
2110 If you want to pass in additional data to pcre_exec(), you can either place it
2111 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2112 block.
2113
2114 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2115 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2116 times with different match limits, until it finds the minimum value needed for
2117 pcre_exec() to complete. The value is then output. This can be instructive; for
2118 most simple matches the number is quite small, but for pathological cases it
2119 gets very large very quickly.
2120
2121 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2122 returns the size of the data block pointed to by the study_data field in a
2123 pcre_extra block, that is, the value that was passed as the argument to
2124 pcre_malloc() when PCRE was getting memory in which to place the information
2125 created by pcre_study(). The fourth argument should point to a size_t variable.
2126 pcretest has been extended so that this information is shown after a successful
2127 pcre_study() call when information about the compiled regex is being displayed.
2128
2129 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2130 because what follows is always an absolute path. (Later: it turns out that this
2131 is more than cosmetic for MinGW, because it doesn't like empty path
2132 components.)
2133
2134 49. Some changes have been made to the callout feature (see 28 above):
2135
2136 (i) A callout function now has three choices for what it returns:
2137
2138 0 => success, carry on matching
2139 > 0 => failure at this point, but backtrack if possible
2140 < 0 => serious error, return this value from pcre_exec()
2141
2142 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2143 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2144 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2145 use by callout functions. It will never be used by PCRE itself.
2146
2147 (ii) The pcre_extra structure (see 45 above) has a void * field called
2148 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2149 pcre_callout_block structure has a field of the same name. The contents of
2150 the field passed in the pcre_extra structure are passed to the callout
2151 function in the corresponding field in the callout block. This makes it
2152 easier to use the same callout-containing regex from multiple threads. For
2153 testing, the pcretest program has a new data escape
2154
2155 \C*n pass the number n (may be negative) as callout_data
2156
2157 If the callout function in pcretest receives a non-zero value as
2158 callout_data, it returns that value.
2159
2160 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2161 there were some redundant $(CFLAGS) in commands that are now specified as
2162 $(LINK), which already includes $(CFLAGS).
2163
2164 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2165 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2166 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2167 one-byte characters throughout. Note that case-insensitive matching applies
2168 only to characters whose values are less than 256. PCRE doesn't support the
2169 notion of cases for higher-valued characters.
2170
2171 (i) A character class whose characters are all within 0-255 is handled as
2172 a bit map, and the map is inverted for negative classes. Previously, a
2173 character > 255 always failed to match such a class; however it should
2174 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2175
2176 (ii) A negated character class with a single character < 255 is coded as
2177 "not this character" (OP_NOT). This wasn't working properly when the test
2178 character was multibyte, either singly or repeated.
2179
2180 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2181 mode, for example: \x{100}{2,3}.
2182
2183 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2184 singly or repeated) now correctly test multibyte characters. However,
2185 PCRE doesn't recognize any characters with values greater than 255 as
2186 digits, spaces, or word characters. Such characters always match \D, \S,
2187 and \W, and never match \d, \s, or \w.
2188
2189 (v) Classes may now contain characters and character ranges with values
2190 greater than 255. For example: [ab\x{100}-\x{400}].
2191
2192 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2193 PCRE in UTF-8 mode.
2194
2195 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2196 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2197 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2198 value.)
2199
2200 53. The single man page has become too large. I have therefore split it up into
2201 a number of separate man pages. These also give rise to individual HTML pages;
2202 these are now put in a separate directory, and there is an index.html page that
2203 lists them all. Some hyperlinking between the pages has been installed.
2204
2205 54. Added convenience functions for handling named capturing parentheses.
2206
2207 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2208 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2209 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2210 are faulted.
2211
2212 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2213 calling configure. These values are used when compiling the dftables.c program
2214 which is run to generate the source of the default character tables. They
2215 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2216 you will need to set these values.
2217
2218 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2219
2220
2221 Version 3.9 02-Jan-02
2222 ---------------------
2223
2224 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2225
2226 2. If --disable-static was given, the building process failed when trying to
2227 build pcretest and pcregrep. (For some reason it was using libtool to compile
2228 them, which is not right, as they aren't part of the library.)
2229
2230
2231 Version 3.8 18-Dec-01
2232 ---------------------
2233
2234 1. The experimental UTF-8 code was completely screwed up. It was packing the
2235 bytes in the wrong order. How dumb can you get?
2236
2237
2238 Version 3.7 29-Oct-01
2239 ---------------------
2240
2241 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2242 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2243 this didn't happen under Solaris 8, where I normally test things.
2244
2245 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2246 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2247 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2248 this fix an hour or so after the initial 3.7 release.)
2249
2250
2251 Version 3.6 23-Oct-01
2252 ---------------------
2253
2254 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2255 offsets passed as NULL with zero offset count.
2256
2257 2. The config.guess and config.sub files had not been updated when I moved to
2258 the latest autoconf.
2259
2260
2261 Version 3.5 15-Aug-01
2262 ---------------------
2263
2264 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2265 had been forgotten.
2266
2267 2. By using declared but undefined structures, we can avoid using "void"
2268 definitions in pcre.h while keeping the internal definitions of the structures
2269 private.
2270
2271 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2272 user point of view, this means that both static and shared libraries are built
2273 by default, but this can be individually controlled. More of the work of
2274 handling this static/shared cases is now inside libtool instead of PCRE's make
2275 file.
2276
2277 4. The pcretest utility is now installed along with pcregrep because it is
2278 useful for users (to test regexs) and by doing this, it automatically gets
2279 relinked by libtool. The documentation has been turned into a man page, so
2280 there are now .1, .txt, and .html versions in /doc.
2281
2282 5. Upgrades to pcregrep:
2283 (i) Added long-form option names like gnu grep.
2284 (ii) Added --help to list all options with an explanatory phrase.
2285 (iii) Added -r, --recursive to recurse into sub-directories.
2286 (iv) Added -f, --file to read patterns from a file.
2287
2288 6. pcre_exec() was referring to its "code" argument before testing that
2289 argument for NULL (and giving an error if it was NULL).
2290
2291 7. Upgraded Makefile.in to allow for compiling in a different directory from
2292 the source directory.
2293
2294 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2295 options bits, the pointer it was passed was to an int instead of to an unsigned
2296 long int. This mattered only on 64-bit systems.
2297
2298 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2299 generated) instead of pcre.in, which it its source. Also made the same change
2300 in several of the .c files.
2301
2302 10. A new release of gcc defines printf() as a macro, which broke pcretest
2303 because it had an ifdef in the middle of a string argument for printf(). Fixed
2304 by using separate calls to printf().
2305
2306 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2307 script, to force use of CR or LF instead of \n in the source. On non-Unix
2308 systems, the value can be set in config.h.
2309
2310 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2311 absolute limit. Changed the text of the error message to make this clear, and
2312 likewise updated the man page.
2313
2314 13. The limit of 99 on the number of capturing subpatterns has been removed.
2315 The new limit is 65535, which I hope will not be a "real" limit.
2316
2317
2318 Version 3.4 22-Aug-00
2319 ---------------------
2320
2321 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2322
2323 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2324
2325
2326 Version 3.3 01-Aug-00
2327 ---------------------
2328
2329 1. If an octal character was given, but the value was greater than \377, it
2330 was not getting masked to the least significant bits, as documented. This could
2331 lead to crashes in some systems.
2332
2333 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2334 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2335
2336 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2337 These just pass their arguments on to (pcre_free)(), but they are provided
2338 because some uses of PCRE bind it to non-C systems that can call its functions,
2339 but cannot call free() or pcre_free() directly.
2340
2341 4. Add "make test" as a synonym for "make check". Corrected some comments in
2342 the Makefile.
2343
2344 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2345 Makefile.
2346
2347 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2348 command called pgrep for grepping around the active processes.
2349
2350 7. Added the beginnings of support for UTF-8 character strings.
2351
2352 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2353 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2354 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2355 out for the ar command.)
2356
2357
2358 Version 3.2 12-May-00
2359 ---------------------
2360
2361 This is purely a bug fixing release.
2362
2363 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2364 of ZA. This was just one example of several cases that could provoke this bug,
2365 which was introduced by change 9 of version 2.00. The code for breaking
2366 infinite loops after an iteration that matches an empty string was't working
2367 correctly.
2368
2369 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2370 when matched against abbab (for example). After matching an empty string, it
2371 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2372 caused it to match further down the string than it should.
2373
2374 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2375 was there because it doesn't seem to be needed, and it causes trouble on some
2376 systems, as it is not a Standard C header. It has been removed.
2377
2378 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2379 were reported on the Macintosh. The changes were from
2380
2381 while ((c = *(++ptr)) != 0 && c != '\n');
2382 to
2383 while ((c = *(++ptr)) != 0 && c != '\n') ;
2384
2385 Totally extraordinary, but if that's what it takes...
2386
2387 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2388 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2389 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2390 assumes the way PCRE uses memmove() (always moving upwards).
2391
2392 6. PCRE is being used in one environment where strchr() is not available. There
2393 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2394 faster code anyway.
2395
2396
2397 Version 3.1 09-Feb-00
2398 ---------------------
2399
2400 The only change in this release is the fixing of some bugs in Makefile.in for
2401 the "install" target:
2402
2403 (1) It was failing to install pcreposix.h.
2404
2405 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2406
2407
2408 Version 3.0 01-Feb-00
2409 ---------------------
2410
2411 1. Add support for the /+ modifier to perltest (to output $` like it does in
2412 pcretest).
2413
2414 2. Add support for the /g modifier to perltest.
2415
2416 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2417 matches null strings.
2418
2419 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2420 pattern. Perl treats empty patterns specially - it reuses the most recent
2421 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2422 effect.
2423
2424 5. The POSIX interface was broken in that it was just handing over the POSIX
2425 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2426 required a bigger vector, with some working space on the end. This means that
2427 the POSIX wrapper now has to get and free some memory, and copy the results.
2428
2429 6. Added some simple autoconf support, placing the test data and the
2430 documentation in separate directories, re-organizing some of the
2431 information files, and making it build pcre-config (a GNU standard). Also added
2432 libtool support for building PCRE as a shared library, which is now the
2433 default.
2434
2435 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2436 09 are not valid octal constants. Single digits will be used for minor values
2437 less than 10.
2438
2439 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2440 existing programs that set these in the POSIX interface can use PCRE without
2441 modification.
2442
2443 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2444 return all that pcre_info() returns, plus additional data. The pcre_info()
2445 function is retained for compatibility, but is considered to be obsolete.
2446
2447 10. Added experimental recursion feature (?R) to handle one common case that
2448 Perl 5.6 will be able to do with (?p{...}).
2449
2450 11. Added support for POSIX character classes like [:alpha:], which Perl is
2451 adopting.
2452
2453
2454 Version 2.08 31-Aug-99
2455 ----------------------
2456
2457 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2458 trying to match at the startoffset position, but instead was moving forward to
2459 the next newline as if a previous match had failed.
2460
2461 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2462 and could get into a loop if a null string was matched other than at the start
2463 of the subject.
2464
2465 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2466 be distinguished at compile time, and for completeness also added PCRE_DATE.
2467
2468 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2469 in GnuWin32 environments.
2470
2471
2472 Version 2.07 29-Jul-99
2473 ----------------------
2474
2475 1. The documentation is now supplied in plain text form and HTML as well as in
2476 the form of man page sources.
2477
2478 2. C++ compilers don't like assigning (void *) values to other pointer types.
2479 In particular this affects malloc(). Although there is no problem in Standard
2480 C, I've put in casts to keep C++ compilers happy.
2481
2482 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2483 should be (const char *).
2484
2485 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2486 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2487 However, I haven't made this a standard facility. The documentation doesn't
2488 mention it, and the Makefile doesn't support it.
2489
2490 5. The Makefile now contains an "install" target, with editable destinations at
2491 the top of the file. The pcretest program is not installed.
2492
2493 6. pgrep -V now gives the PCRE version number and date.
2494
2495 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2496 causing the entire string to be ignored, instead of just the last character.
2497
2498 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2499 non-matching string, it can take a very, very long time, even for strings of
2500 quite modest length, because of the nested recursion. PCRE now does better in
2501 some of these cases. It does this by remembering the last required literal
2502 character in the pattern, and pre-searching the subject to ensure it is present
2503 before running the real match. In other words, it applies a heuristic to detect
2504 some types of certain failure quickly, and in the above example, if presented
2505 with a string that has no trailing " it gives "no match" very quickly.
2506
2507 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2508 other alternatives are tried instead.
2509
2510
2511 Version 2.06 09-Jun-99
2512 ----------------------
2513
2514 1. Change pcretest's output for amount of store used to show just the code
2515 space, because the remainder (the data block) varies in size between 32-bit and
2516 64-bit systems.
2517
2518 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2519 start matching at. This allows lookbehinds to work when searching for multiple
2520 occurrences in a string.
2521
2522 3. Added additional options to pcretest for testing multiple occurrences:
2523
2524 /+ outputs the rest of the string that follows a match
2525 /g loops for multiple occurrences, using the new startoffset argument
2526 /G loops for multiple occurrences by passing an incremented pointer
2527
2528 4. PCRE wasn't doing the "first character" optimization for patterns starting
2529 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2530 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2531 the letter 'x'. On long subject strings, this gives a significant speed-up.
2532
2533
2534 Version 2.05 21-Apr-99
2535 ----------------------
2536
2537 1. Changed the type of magic_number from int to long int so that it works
2538 properly on 16-bit systems.
2539
2540 2. Fixed a bug which caused patterns starting with .* not to work correctly
2541 when the subject string contained newline characters. PCRE was assuming
2542 anchoring for such patterns in all cases, which is not correct because .* will
2543 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2544 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2545 must be retried after every newline in the subject.
2546
2547
2548 Version 2.04 18-Feb-99
2549 ----------------------
2550
2551 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2552 computation of the store needed to hold the pattern was incorrect (too large).
2553 If such patterns were nested a few deep, this could multiply and become a real
2554 problem.
2555
2556 2. Added /M option to pcretest to show the memory requirement of a specific
2557 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2558
2559 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2560 compiled in such a way that the backtracking after subsequent failure was
2561 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2562 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2563
2564
2565 Version 2.03 02-Feb-99
2566 ----------------------
2567
2568 1. Fixed typo and small mistake in man page.
2569
2570 2. Added 4th condition (GPL supersedes if conflict) and created separate
2571 LICENCE file containing the conditions.
2572
2573 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2574 Perl, that is the internal \ allows the delimiter to be included in the
2575 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2576 the final delimiter, add \ to the end of the pattern (to test the error).
2577
2578 4. Added the convenience functions for extracting substrings after a successful
2579 match. Updated pcretest to make it able to test these functions.
2580
2581
2582 Version 2.02 14-Jan-99
2583 ----------------------
2584
2585 1. Initialized the working variables associated with each extraction so that
2586 their saving and restoring doesn't refer to uninitialized store.
2587
2588 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2589 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2590 fix the problem.
2591
2592 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2593 calls, and wasn't printing the correct value for compiling calls. Increased the
2594 default value of LOOPREPEAT, and the number of significant figures in the
2595 times.
2596
2597 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2598
2599 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2600 a building problem on Windows NT with a FAT file system.
2601
2602
2603 Version 2.01 21-Oct-98
2604 ----------------------
2605
2606 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2607 to character tables built by pcre_maketables() in the current locale. If NULL
2608 is passed, the default tables are used.
2609
2610
2611 Version 2.00 24-Sep-98
2612 ----------------------
2613
2614 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2615 it any more.
2616
2617 2. Allow quantification of (?>) groups, and make it work correctly.
2618
2619 3. The first character computation wasn't working for (?>) groups.
2620
2621 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2622 end of the subject) and add 5.005's \z, which really does match only at the
2623 very end of the subject.
2624
2625 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2626
2627 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2628 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2629 localized options. All options to pcre_study() were also removed.
2630
2631 7. Add other new features from 5.005:
2632
2633 $(?<= positive lookbehind
2634 $(?<! negative lookbehind
2635 (?imsx-imsx) added the unsetting capability
2636 such a setting is global if at outer level; local otherwise
2637 (?imsx-imsx:) non-capturing groups with option setting
2638 (?(cond)re|re) conditional pattern matching
2639
2640 A backreference to itself in a repeated group matches the previous
2641 captured string.
2642
2643 8. General tidying up of studying (both automatic and via "study")
2644 consequential on the addition of new assertions.
2645
2646 9. As in 5.005, unlimited repeated groups that could match an empty substring
2647 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2648 runtime if any iteration does actually match an empty substring.
2649
2650 10. Include the RunTest script in the distribution.
2651
2652 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2653 discrepancies, some of which were old and were also with respect to 5.004. They
2654 have now been fixed.
2655
2656
2657 Version 1.09 28-Apr-98
2658 ----------------------
2659
2660 1. A negated single character class followed by a quantifier with a minimum
2661 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2662 program crashes, or just wrong answers. This did not apply to negated classes
2663 containing more than one character, or to minima other than one.
2664
2665
2666 Version 1.08 27-Mar-98
2667 ----------------------
2668
2669 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2670
2671 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2672 latter must appear before anything that relies on it in the pattern.
2673
2674
2675 Version 1.07 16-Feb-98
2676 ----------------------
2677
2678 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2679 repeat of a potentially empty string).
2680
2681
2682 Version 1.06 23-Jan-98
2683 ----------------------
2684
2685 1. Added Markus Oberhumer's little patches for C++.
2686
2687 2. Literal strings longer than 255 characters were broken.
2688
2689
2690 Version 1.05 23-Dec-97
2691 ----------------------
2692
2693 1. Negated character classes containing more than one character were failing if
2694 PCRE_CASELESS was set at run time.
2695
2696
2697 Version 1.04 19-Dec-97
2698 ----------------------
2699
2700 1. Corrected the man page, where some "const" qualifiers had been omitted.
2701
2702 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2703 input syntax.
2704
2705 3. Fixed memory leak which occurred when a regex with back references was
2706 matched with an offsets vector that wasn't big enough. The temporary memory
2707 that is used in this case wasn't being freed if the match failed.
2708
2709 4. Tidied pcretest to ensure it frees memory that it gets.
2710
2711 5. Temporary memory was being obtained in the case where the passed offsets
2712 vector was exactly big enough.
2713
2714 6. Corrected definition of offsetof() from change 5 below.
2715
2716 7. I had screwed up change 6 below and broken the rules for the use of
2717 setjmp(). Now fixed.
2718
2719
2720 Version 1.03 18-Dec-97
2721 ----------------------
2722
2723 1. A erroneous regex with a missing opening parenthesis was correctly
2724 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2725 on some systems.
2726
2727 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2728 it was reported that one broken compiler failed on the former because "code" is
2729 also an independent variable.
2730
2731 3. The erroneous regex a[]b caused an array overrun reference.
2732
2733 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2734 fail on data ending with that character. (It was going on too far, and checking
2735 the next character, typically a binary zero.) This was specific to the
2736 optimized code for single-character negative classes.
2737
2738 5. Added a contributed patch from the TIN world which does the following:
2739
2740 + Add an undef for memmove, in case the the system defines a macro for it.
2741
2742 + Add a definition of offsetof(), in case there isn't one. (I don't know
2743 the reason behind this - offsetof() is part of the ANSI standard - but
2744 it does no harm).
2745
2746 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2747 most of the places where whitespace preceded '#'. I have given up and
2748 allowed the remaining 2 cases to be at the margin.
2749
2750 + Rename some variables in pcre to eliminate shadowing. This seems very
2751 pedantic, but does no harm, of course.
2752
2753 6. Moved the call to setjmp() into its own function, to get rid of warnings
2754 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2755
2756 7. Constructs such as \d{8,} were compiling into the equivalent of
2757 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2758 outcome, but in this particular case used more store than had been allocated,
2759 which caused the bug to be discovered because it threw up an internal error.
2760
2761 8. The debugging code in both pcre and pcretest for outputting the compiled
2762 form of a regex was going wrong in the case of back references followed by
2763 curly-bracketed repeats.
2764
2765
2766 Version 1.02 12-Dec-97
2767 ----------------------
2768
2769 1. Typos in pcre.3 and comments in the source fixed.
2770
2771 2. Applied a contributed patch to get rid of places where it used to remove
2772 'const' from variables, and fixed some signed/unsigned and uninitialized
2773 variable warnings.
2774
2775 3. Added the "runtest" target to Makefile.
2776
2777 4. Set default compiler flag to -O2 rather than just -O.
2778
2779
2780 Version 1.01 19-Nov-97
2781 ----------------------
2782
2783 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2784 like /([ab]*)*/, that is, for classes with more than one character in them.
2785
2786 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2787 as /((?>a*))*/ (a PCRE_EXTRA facility).
2788
2789
2790 Version 1.00 18-Nov-97
2791 ----------------------
2792
2793 1. Added compile-time macros to support systems such as SunOS4 which don't have
2794 memmove() or strerror() but have other things that can be used instead.
2795
2796 2. Arranged that "make clean" removes the executables.
2797
2798
2799 Version 0.99 27-Oct-97
2800 ----------------------
2801
2802 1. Fixed bug in code for optimizing classes with only one character. It was
2803 initializing a 32-byte map regardless, which could cause it to run off the end
2804 of the memory it had got.
2805
2806 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2807
2808
2809 Version 0.98 22-Oct-97
2810 ----------------------
2811
2812 1. Fixed bug in code for handling temporary memory usage when there are more
2813 back references than supplied space in the ovector. This could cause segfaults.
2814
2815
2816 Version 0.97 21-Oct-97
2817 ----------------------
2818
2819 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2820
2821 2. Optimized negated single characters not to use a bit map.
2822
2823 3. Brought error texts together as macro definitions; clarified some of them;
2824 fixed one that was wrong - it said "range out of order" when it meant "invalid
2825 escape sequence".
2826
2827 4. Changed some char * arguments to const char *.
2828
2829 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2830
2831 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2832 pcretest.
2833
2834
2835 Version 0.96 16-Oct-97
2836 ----------------------
2837
2838 1. Added a simple "pgrep" utility to the distribution.
2839
2840 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2841 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2842 where "ddd" means "one or more decimal digits".
2843
2844 3. Fixed serious bug. If a pattern had a back reference, but the call to
2845 pcre_exec() didn't supply a large enough ovector to record the related
2846 identifying subpattern, the match always failed. PCRE now remembers the number
2847 of the largest back reference, and gets some temporary memory in which to save
2848 the offsets during matching if necessary, in order to ensure that
2849 backreferences always work.
2850
2851 4. Increased the compatibility with Perl in a number of ways:
2852
2853 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2854 to request this handling. The option can be set at compile or exec time.
2855
2856 (b) $ matches before a terminating newline by default; an option
2857 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2858 mode). The option can be set at compile or exec time.
2859
2860 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2861 the same as Perl's. If the decimal number it represents is less than 10
2862 or there aren't that many previous left capturing parentheses, an octal
2863 escape is read. Inside a character class, it's always an octal escape,
2864 even if it is a single digit.
2865
2866 (d) An escaped but undefined alphabetic character is taken as a literal,
2867 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2868 escapes.
2869
2870 (e) {0} is now permitted. (The previous item is removed from the compiled
2871 pattern).
2872
2873 5. Changed all the names of code files so that the basic parts are no longer
2874 than 10 characters, and abolished the teeny "globals.c" file.
2875
2876 6. Changed the handling of character classes; they are now done with a 32-byte
2877 bit map always.
2878
2879 7. Added the -d and /D options to pcretest to make it possible to look at the
2880 internals of compilation without having to recompile pcre.
2881
2882
2883 Version 0.95 23-Sep-97
2884 ----------------------
2885
2886 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2887 \x20 at the start of a run of normal characters. These were being treated as
2888 real characters, instead of the source characters being re-checked.
2889
2890
2891 Version 0.94 18-Sep-97
2892 ----------------------
2893
2894 1. The functions are now thread-safe, with the caveat that the global variables
2895 containing pointers to malloc() and free() or alternative functions are the
2896 same for all threads.
2897
2898 2. Get pcre_study() to generate a bitmap of initial characters for non-
2899 anchored patterns when this is possible, and use it if passed to pcre_exec().
2900
2901
2902 Version 0.93 15-Sep-97
2903 ----------------------
2904
2905 1. /(b)|(:+)/ was computing an incorrect first character.
2906
2907 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2908 but not actually doing anything yet.
2909
2910 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2911 as Perl does (e.g. [-az] or [az-]).
2912
2913 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2914 all possible positions.
2915
2916 5. Split up into different modules to avoid including unneeded functions in a
2917 compiled binary. However, compile and exec are still in one module. The "study"
2918 function is split off.
2919
2920 6. The character tables are now in a separate module whose source is generated
2921 by an auxiliary program - but can then be edited by hand if required. There are
2922 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2923 toupper() in the code.
2924
2925 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2926 make them global. Abolish the function for setting them, as the caller can now
2927 set them directly.
2928
2929
2930 Version 0.92 11-Sep-97
2931 ----------------------
2932
2933 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2934 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2935
2936 2. Caseless matching was not working in character classes if the characters in
2937 the pattern were in upper case.
2938
2939 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2940
2941 4. Make PCRE_ANCHORED public and accept as a compile option.
2942
2943 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2944 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2945 pass them.
2946
2947 6. Give an error if bad option bits passed at compile or run time.
2948
2949 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2950 pcretest to cause it to pass that flag.
2951
2952 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2953 options, and the first character, if set.
2954
2955 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2956
2957
2958 Version 0.91 10-Sep-97
2959 ----------------------
2960
2961 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2962 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2963
2964 2. PCRE was looping on encountering an indefinitely repeated back reference to
2965 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2966 Perl does - treats the match as successful.
2967
2968 ****

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