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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 231 - (show annotations) (download)
Tue Sep 11 11:15:33 2007 UTC (6 years, 10 months ago) by ph10
File size: 139814 byte(s)
Add facility to make \R match only CR, LF, or CRLF.

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

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