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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 267 - (show annotations) (download)
Wed Nov 14 11:48:49 2007 UTC (6 years, 5 months ago) by ph10
File size: 142567 byte(s)
Fixed reference to PCRECPP_STATIC that should have been PCRE_STATIC.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.5 12-Nov-07
5 ---------------------
6
7 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
8 values in parens when parsing an RE using the C++ wrapper."
9
10 2. Negative specials like \S did not work in character classes in UTF-8 mode.
11 Characters greater than 255 were excluded from the class instead of being
12 included.
13
14 3. The same bug as (2) above applied to negated POSIX classes such as
15 [:^space:].
16
17 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
18 defined or documented. It seems to have been a typo for PCRE_STATIC, so
19 I have changed it.
20
21
22 Version 7.4 21-Sep-07
23 ---------------------
24
25 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
26 means that a class such as [\s] counted as "explicit reference to CR or
27 LF". That isn't really right - the whole point of the change was to try to
28 help when there was an actual mention of one of the two characters. So now
29 the change happens only if \r or \n (or a literal CR or LF) character is
30 encountered.
31
32 2. The 32-bit options word was also used for 6 internal flags, but the numbers
33 of both had grown to the point where there were only 3 bits left.
34 Fortunately, there was spare space in the data structure, and so I have
35 moved the internal flags into a new 16-bit field to free up more option
36 bits.
37
38 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
39 but did not set the internal JCHANGED flag - either of these is enough to
40 control the way the "get" function works - but the PCRE_INFO_JCHANGED
41 facility is supposed to tell if (?J) was ever used, so now (?J) at the
42 start sets both bits.
43
44 4. Added options (at build time, compile time, exec time) to change \R from
45 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
46
47 5. doc/pcresyntax.html was missing from the distribution.
48
49 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
50 compatibility, even though it is no longer used.
51
52 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
53 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
54 windows.h file is present (where different names are used). [This was
55 reversed later after testing - see 16 below.]
56
57 8. Changed all #include <config.h> to #include "config.h". There were also
58 some further <pcre.h> cases that I changed to "pcre.h".
59
60 9. When pcregrep was used with the --colour option, it missed the line ending
61 sequence off the lines that it output.
62
63 10. It was pointed out to me that arrays of string pointers cause lots of
64 relocations when a shared library is dynamically loaded. A technique of
65 using a single long string with a table of offsets can drastically reduce
66 these. I have refactored PCRE in four places to do this. The result is
67 dramatic:
68
69 Originally: 290
70 After changing UCP table: 187
71 After changing error message table: 43
72 After changing table of "verbs" 36
73 After changing table of Posix names 22
74
75 Thanks to the folks working on Gregex for glib for this insight.
76
77 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
78 unicode-properties was also set.
79
80 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
81
82 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
83 checked only for CRLF.
84
85 14. Added casts to pcretest.c to avoid compiler warnings.
86
87 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
88
89 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
90 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
91 entirely. This removes changes made in 7 above.
92
93 17. The CMake files have been updated, and there is now more information about
94 building with CMake in the NON-UNIX-USE document.
95
96
97 Version 7.3 28-Aug-07
98 ---------------------
99
100 1. In the rejigging of the build system that eventually resulted in 7.1, the
101 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
102 brackets there is not right, since it causes compilers to look for an
103 installed pcre.h, not the version that is in the source that is being
104 compiled (which of course may be different). I have changed it back to:
105
106 #include "pcre.h"
107
108 I have a vague recollection that the change was concerned with compiling in
109 different directories, but in the new build system, that is taken care of
110 by the VPATH setting the Makefile.
111
112 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
113 when the subject happened to end in the byte 0x85 (e.g. if the last
114 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
115 characters but of course it shouldn't be taken as a newline when it is part
116 of another character. The bug was that, for an unlimited repeat of . in
117 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
118 characters when looking for a newline.
119
120 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
121
122 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
123 in debug output.
124
125 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
126 long printing in the pcrecpp unittest when running under MinGW.
127
128 6. ESC_K was left out of the EBCDIC table.
129
130 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
131 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
132 limit also applies to "virtual nesting" when a pattern is recursive, and in
133 this case 1000 isn't so big. I have been able to remove this limit at the
134 expense of backing off one optimization in certain circumstances. Normally,
135 when pcre_exec() would call its internal match() function recursively and
136 immediately return the result unconditionally, it uses a "tail recursion"
137 feature to save stack. However, when a subpattern that can match an empty
138 string has an unlimited repetition quantifier, it no longer makes this
139 optimization. That gives it a stack frame in which to save the data for
140 checking that an empty string has been matched. Previously this was taken
141 from the 1000-entry workspace that had been reserved. So now there is no
142 explicit limit, but more stack is used.
143
144 8. Applied Daniel's patches to solve problems with the import/export magic
145 syntax that is required for Windows, and which was going wrong for the
146 pcreposix and pcrecpp parts of the library. These were overlooked when this
147 problem was solved for the main library.
148
149 9. There were some crude static tests to avoid integer overflow when computing
150 the size of patterns that contain repeated groups with explicit upper
151 limits. As the maximum quantifier is 65535, the maximum group length was
152 set at 30,000 so that the product of these two numbers did not overflow a
153 32-bit integer. However, it turns out that people want to use groups that
154 are longer than 30,000 bytes (though not repeat them that many times).
155 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
156 made it possible to implement the integer overflow checks in a much more
157 dynamic way, which I have now done. The artificial limitation on group
158 length has been removed - we now have only the limit on the total length of
159 the compiled pattern, which depends on the LINK_SIZE setting.
160
161 10. Fixed a bug in the documentation for get/copy named substring when
162 duplicate names are permitted. If none of the named substrings are set, the
163 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
164 empty string.
165
166 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
167 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
168 because the ] is interpreted as the first data character and the
169 terminating ] is not found. PCRE has been made compatible with Perl in this
170 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
171 cause memory overwriting.
172
173 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
174 string has been matched (to stop an infinite loop). It was not recognizing
175 a conditional subpattern that could match an empty string if that
176 subpattern was within another subpattern. For example, it looped when
177 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
178 condition was not nested. This bug has been fixed.
179
180 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
181 past the start of the subject in the presence of bytes with the top bit
182 set, for example "\x8aBCD".
183
184 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
185 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
186
187 14. Optimized (?!) to (*FAIL).
188
189 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
190 This restricts code points to be within the range 0 to 0x10FFFF, excluding
191 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
192 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
193 does: it's just the validity check that is more restrictive.
194
195 16. Inserted checks for integer overflows during escape sequence (backslash)
196 processing, and also fixed erroneous offset values for syntax errors during
197 backslash processing.
198
199 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
200 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
201
202 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
203 caused an overrun.
204
205 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
206 something other than just ASCII characters) inside a group that had an
207 unlimited repeat caused a loop at compile time (while checking to see
208 whether the group could match an empty string).
209
210 20. Debugging a pattern containing \p or \P could cause a crash. For example,
211 [\P{Any}] did so. (Error in the code for printing property names.)
212
213 21. An orphan \E inside a character class could cause a crash.
214
215 22. A repeated capturing bracket such as (A)? could cause a wild memory
216 reference during compilation.
217
218 23. There are several functions in pcre_compile() that scan along a compiled
219 expression for various reasons (e.g. to see if it's fixed length for look
220 behind). There were bugs in these functions when a repeated \p or \P was
221 present in the pattern. These operators have additional parameters compared
222 with \d, etc, and these were not being taken into account when moving along
223 the compiled data. Specifically:
224
225 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
226 length.
227
228 (b) An item such as \pL+ within a repeated group could cause crashes or
229 loops.
230
231 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
232 "reference to non-existent subpattern" error.
233
234 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
235
236 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
237 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
238
239 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
240
241 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
242 character were causing crashes (broken optimization).
243
244 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
245 \p or \P) caused a compile-time loop.
246
247 28. More problems have arisen in unanchored patterns when CRLF is a valid line
248 break. For example, the unstudied pattern [\r\n]A does not match the string
249 "\r\nA" because change 7.0/46 below moves the current point on by two
250 characters after failing to match at the start. However, the pattern \nA
251 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
252 the same is true. There doesn't seem any very clean way out of this, but
253 what I have chosen to do makes the common cases work: PCRE now takes note
254 of whether there can be an explicit match for \r or \n anywhere in the
255 pattern, and if so, 7.0/46 no longer applies. As part of this change,
256 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
257 pattern has explicit CR or LF references.
258
259 29. Added (*CR) etc for changing newline setting at start of pattern.
260
261
262 Version 7.2 19-Jun-07
263 ---------------------
264
265 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
266 which is apparently normally available under Windows.
267
268 2. Re-jig the pcregrep tests with different newline settings in an attempt
269 to make them independent of the local environment's newline setting.
270
271 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
272
273 4. Some of the "internals" tests were previously cut out when the link size
274 was not 2, because the output contained actual offsets. The recent new
275 "Z" feature of pcretest means that these can be cut out, making the tests
276 usable with all link sizes.
277
278 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
279 stack recursion. This gives a massive performance boost under BSD, but just
280 a small improvement under Linux. However, it saves one field in the frame
281 in all cases.
282
283 6. Added more features from the forthcoming Perl 5.10:
284
285 (a) (?-n) (where n is a string of digits) is a relative subroutine or
286 recursion call. It refers to the nth most recently opened parentheses.
287
288 (b) (?+n) is also a relative subroutine call; it refers to the nth next
289 to be opened parentheses.
290
291 (c) Conditions that refer to capturing parentheses can be specified
292 relatively, for example, (?(-2)... or (?(+3)...
293
294 (d) \K resets the start of the current match so that everything before
295 is not part of it.
296
297 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
298
299 (f) \g{name} is another synonym - part of Perl 5.10's unification of
300 reference syntax.
301
302 (g) (?| introduces a group in which the numbering of parentheses in each
303 alternative starts with the same number.
304
305 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
306
307 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
308 PCRE_INFO_JCHANGED.
309
310 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
311 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
312 for detecting groups that can match an empty string.
313
314 9. A pattern with a very large number of alternatives (more than several
315 hundred) was running out of internal workspace during the pre-compile
316 phase, where pcre_compile() figures out how much memory will be needed. A
317 bit of new cunning has reduced the workspace needed for groups with
318 alternatives. The 1000-alternative test pattern now uses 12 bytes of
319 workspace instead of running out of the 4096 that are available.
320
321 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
322
323 11. Applied patch from Google to remove an optimization that didn't quite work.
324 The report of the bug said:
325
326 pcrecpp::RE("a*").FullMatch("aaa") matches, while
327 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
328 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
329
330 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
331 it matched the wrong number of bytes.
332
333
334 Version 7.1 24-Apr-07
335 ---------------------
336
337 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
338 that is more "standard", making use of automake and other Autotools. There
339 is some re-arrangement of the files and adjustment of comments consequent
340 on this.
341
342 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
343 for recursive directory scanning broke on some systems because the files
344 are not scanned in any specific order and on different systems the order
345 was different. A call to "sort" has been inserted into RunGrepTest for the
346 approprate test as a short-term fix. In the longer term there may be an
347 alternative.
348
349 3. I had an email from Eric Raymond about problems translating some of PCRE's
350 man pages to HTML (despite the fact that I distribute HTML pages, some
351 people do their own conversions for various reasons). The problems
352 concerned the use of low-level troff macros .br and .in. I have therefore
353 removed all such uses from the man pages (some were redundant, some could
354 be replaced by .nf/.fi pairs). The 132html script that I use to generate
355 HTML has been updated to handle .nf/.fi and to complain if it encounters
356 .br or .in.
357
358 4. Updated comments in configure.ac that get placed in config.h.in and also
359 arranged for config.h to be included in the distribution, with the name
360 config.h.generic, for the benefit of those who have to compile without
361 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
362
363 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
364 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
365 makevp.bat for the current PCRE, using the additional files
366 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
367
368 6. A Windows user reported a minor discrepancy with test 2, which turned out
369 to be caused by a trailing space on an input line that had got lost in his
370 copy. The trailing space was an accident, so I've just removed it.
371
372 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
373 that is needed.
374
375 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
376 as "const" (a) because they are and (b) because it helps the PHP
377 maintainers who have recently made a script to detect big data structures
378 in the php code that should be moved to the .rodata section. I remembered
379 to update Builducptable as well, so it won't revert if ucptable.h is ever
380 re-created.
381
382 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
383 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
384 order to be able to cut out the UTF-8 tables in the latter when UTF-8
385 support is not required. This saves 1.5-2K of code, which is important in
386 some applications.
387
388 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
389 so as not to refer to the tables, even though these functions will never be
390 called when UTF-8 support is disabled. Otherwise there are problems with a
391 shared library.
392
393 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
394
395 (a) It was defining its arguments as char * instead of void *.
396
397 (b) It was assuming that all moves were upwards in memory; this was true
398 a long time ago when I wrote it, but is no longer the case.
399
400 The emulated memove() is provided for those environments that have neither
401 memmove() nor bcopy(). I didn't think anyone used it these days, but that
402 is clearly not the case, as these two bugs were recently reported.
403
404 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
405 and Detrail to create the HTML documentation, the .txt form of the man
406 pages, and it removes trailing spaces from listed files. It also creates
407 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
408 case, it wraps all the #defines with #ifndefs. This script should be run
409 before "make dist".
410
411 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
412 with Unicode property support.
413
414 (a) For a maximizing quantifier, if the two different cases of the
415 character were of different lengths in their UTF-8 codings (there are
416 some cases like this - I found 11), and the matching function had to
417 back up over a mixture of the two cases, it incorrectly assumed they
418 were both the same length.
419
420 (b) When PCRE was configured to use the heap rather than the stack for
421 recursion during matching, it was not correctly preserving the data for
422 the other case of a UTF-8 character when checking ahead for a match
423 while processing a minimizing repeat. If the check also involved
424 matching a wide character, but failed, corruption could cause an
425 erroneous result when trying to check for a repeat of the original
426 character.
427
428 13. Some tidying changes to the testing mechanism:
429
430 (a) The RunTest script now detects the internal link size and whether there
431 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
432 values substituted by "configure". (The RunGrepTest script already did
433 this for UTF-8.) The configure.ac script no longer substitutes the
434 relevant variables.
435
436 (b) The debugging options /B and /D in pcretest show the compiled bytecode
437 with length and offset values. This means that the output is different
438 for different internal link sizes. Test 2 is skipped for link sizes
439 other than 2 because of this, bypassing the problem. Unfortunately,
440 there was also a test in test 3 (the locale tests) that used /B and
441 failed for link sizes other than 2. Rather than cut the whole test out,
442 I have added a new /Z option to pcretest that replaces the length and
443 offset values with spaces. This is now used to make test 3 independent
444 of link size. (Test 2 will be tidied up later.)
445
446 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
447 segmentation fault instead of returning the appropriate error message.
448
449 15. In multiline mode when the newline sequence was set to "any", the pattern
450 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
451 This doesn't seem right; it now treats the CRLF combination as the line
452 ending, and so does not match in that case. It's only a pattern such as ^$
453 that would hit this one: something like ^ABC$ would have failed after \r
454 and then tried again after \r\n.
455
456 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
457 in an attempt to make files that differ only in their line terminators
458 compare equal. This works on Linux.
459
460 17. Under certain error circumstances pcregrep might try to free random memory
461 as it exited. This is now fixed, thanks to valgrind.
462
463 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
464 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
465 was because its rules for how to advance for /g after matching an empty
466 string at the end of a line did not allow for this case. They now check for
467 it specially.
468
469 20. pcretest is supposed to handle patterns and data of any length, by
470 extending its buffers when necessary. It was getting this wrong when the
471 buffer for a data line had to be extended.
472
473 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
474 CRLF as a newline sequence.
475
476 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
477 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
478 I have nevertheless tidied it up.
479
480 23. Added some casts to kill warnings from HP-UX ia64 compiler.
481
482 24. Added a man page for pcre-config.
483
484
485 Version 7.0 19-Dec-06
486 ---------------------
487
488 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
489 moving to gcc 4.1.1.
490
491 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
492 sys/time.h, which is documented as needed for this function. It doesn't
493 seem to matter on Linux, but it showed up on some releases of OS X.
494
495 3. It seems that there are systems where bytes whose values are greater than
496 127 match isprint() in the "C" locale. The "C" locale should be the
497 default when a C program starts up. In most systems, only ASCII printing
498 characters match isprint(). This difference caused the output from pcretest
499 to vary, making some of the tests fail. I have changed pcretest so that:
500
501 (a) When it is outputting text in the compiled version of a pattern, bytes
502 other than 32-126 are always shown as hex escapes.
503
504 (b) When it is outputting text that is a matched part of a subject string,
505 it does the same, unless a different locale has been set for the match
506 (using the /L modifier). In this case, it uses isprint() to decide.
507
508 4. Fixed a major bug that caused incorrect computation of the amount of memory
509 required for a compiled pattern when options that changed within the
510 pattern affected the logic of the preliminary scan that determines the
511 length. The relevant options are -x, and -i in UTF-8 mode. The result was
512 that the computed length was too small. The symptoms of this bug were
513 either the PCRE error "internal error: code overflow" from pcre_compile(),
514 or a glibc crash with a message such as "pcretest: free(): invalid next
515 size (fast)". Examples of patterns that provoked this bug (shown in
516 pcretest format) are:
517
518 /(?-x: )/x
519 /(?x)(?-x: \s*#\s*)/
520 /((?i)[\x{c0}])/8
521 /(?i:[\x{c0}])/8
522
523 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
524 is now done differently.
525
526 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
527 wrapper classes; (b) implement a new function in the C++ scanner that is
528 more efficient than the old way of doing things because it avoids levels of
529 recursion in the regex matching; (c) add a paragraph to the documentation
530 for the FullMatch() function.
531
532 6. The escape sequence \n was being treated as whatever was defined as
533 "newline". Not only was this contrary to the documentation, which states
534 that \n is character 10 (hex 0A), but it also went horribly wrong when
535 "newline" was defined as CRLF. This has been fixed.
536
537 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
538 was being set to -1 for the "end of line" case (supposedly a value that no
539 character can have). Though this value is never used (the check for end of
540 line is "zero bytes in current character"), it caused compiler complaints.
541 I've changed it to 0xffffffff.
542
543 8. In pcre_version.c, the version string was being built by a sequence of
544 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
545 string (as it is for production releases) called a macro with an empty
546 argument. The C standard says the result of this is undefined. The gcc
547 compiler treats it as an empty string (which was what was wanted) but it is
548 reported that Visual C gives an error. The source has been hacked around to
549 avoid this problem.
550
551 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
552 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
553 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
554 of them did).
555
556 10. Originally, pcretest opened its input and output without "b"; then I was
557 told that "b" was needed in some environments, so it was added for release
558 5.0 to both the input and output. (It makes no difference on Unix-like
559 systems.) Later I was told that it is wrong for the input on Windows. I've
560 now abstracted the modes into two macros, to make it easier to fiddle with
561 them, and removed "b" from the input mode under Windows.
562
563 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
564
565 12. Added -help and --help to pcretest as an official way of being reminded
566 of the options.
567
568 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
569 and pcrecpp.cc because they annoy compilers at high warning levels.
570
571 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
572
573 15. Fixed an occurrence of == in configure.ac that should have been = (shell
574 scripts are not C programs :-) and which was not noticed because it works
575 on Linux.
576
577 16. pcretest is supposed to handle any length of pattern and data line (as one
578 line or as a continued sequence of lines) by extending its input buffer if
579 necessary. This feature was broken for very long pattern lines, leading to
580 a string of junk being passed to pcre_compile() if the pattern was longer
581 than about 50K.
582
583 17. I have done a major re-factoring of the way pcre_compile() computes the
584 amount of memory needed for a compiled pattern. Previously, there was code
585 that made a preliminary scan of the pattern in order to do this. That was
586 OK when PCRE was new, but as the facilities have expanded, it has become
587 harder and harder to keep it in step with the real compile phase, and there
588 have been a number of bugs (see for example, 4 above). I have now found a
589 cunning way of running the real compile function in a "fake" mode that
590 enables it to compute how much memory it would need, while actually only
591 ever using a few hundred bytes of working memory and without too many
592 tests of the mode. This should make future maintenance and development
593 easier. A side effect of this work is that the limit of 200 on the nesting
594 depth of parentheses has been removed (though this was never a serious
595 limitation, I suspect). However, there is a downside: pcre_compile() now
596 runs more slowly than before (30% or more, depending on the pattern). I
597 hope this isn't a big issue. There is no effect on runtime performance.
598
599 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
600 newline (only possible for the last line of a file) and it was a
601 pattern that set a locale (followed by /Lsomething), pcretest crashed.
602
603 19. Added additional timing features to pcretest. (1) The -tm option now times
604 matching only, not compiling. (2) Both -t and -tm can be followed, as a
605 separate command line item, by a number that specifies the number of
606 repeats to use when timing. The default is 50000; this gives better
607 precision, but takes uncomfortably long for very large patterns.
608
609 20. Extended pcre_study() to be more clever in cases where a branch of a
610 subpattern has no definite first character. For example, (a*|b*)[cd] would
611 previously give no result from pcre_study(). Now it recognizes that the
612 first character must be a, b, c, or d.
613
614 21. There was an incorrect error "recursive call could loop indefinitely" if
615 a subpattern (or the entire pattern) that was being tested for matching an
616 empty string contained only one non-empty item after a nested subpattern.
617 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
618 incorrectly, because the \d was being skipped in the check.
619
620 22. The pcretest program now has a new pattern option /B and a command line
621 option -b, which is equivalent to adding /B to every pattern. This causes
622 it to show the compiled bytecode, without the additional information that
623 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
624 is the same as /B/I).
625
626 23. A new optimization is now able automatically to treat some sequences such
627 as a*b as a*+b. More specifically, if something simple (such as a character
628 or a simple class like \d) has an unlimited quantifier, and is followed by
629 something that cannot possibly match the quantified thing, the quantifier
630 is automatically "possessified".
631
632 24. A recursive reference to a subpattern whose number was greater than 39
633 went wrong under certain circumstances in UTF-8 mode. This bug could also
634 have affected the operation of pcre_study().
635
636 25. Realized that a little bit of performance could be had by replacing
637 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
638
639 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
640
641 27. Possessive quantifiers such as a++ were previously implemented by turning
642 them into atomic groups such as ($>a+). Now they have their own opcodes,
643 which improves performance. This includes the automatically created ones
644 from 23 above.
645
646 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
647 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
648 the first matched character to be a colon. This applied both to named and
649 numbered groups.
650
651 29. The ucpinternal.h header file was missing its idempotency #ifdef.
652
653 30. I was sent a "project" file called libpcre.a.dev which I understand makes
654 building PCRE on Windows easier, so I have included it in the distribution.
655
656 31. There is now a check in pcretest against a ridiculously large number being
657 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
658 loop, the loop is abandoned.
659
660 32. Forward references to subpatterns in conditions such as (?(2)...) where
661 subpattern 2 is defined later cause pcre_compile() to search forwards in
662 the pattern for the relevant set of parentheses. This search went wrong
663 when there were unescaped parentheses in a character class, parentheses
664 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
665
666 33. "Subroutine" calls and backreferences were previously restricted to
667 referencing subpatterns earlier in the regex. This restriction has now
668 been removed.
669
670 34. Added a number of extra features that are going to be in Perl 5.10. On the
671 whole, these are just syntactic alternatives for features that PCRE had
672 previously implemented using the Python syntax or my own invention. The
673 other formats are all retained for compatibility.
674
675 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
676 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
677 also .NET compatible.
678
679 (b) A recursion or subroutine call to a named group can now be defined as
680 (?&name) as well as (?P>name).
681
682 (c) A backreference to a named group can now be defined as \k<name> or
683 \k'name' as well as (?P=name). The new forms, as well as being in Perl
684 5.10, are also .NET compatible.
685
686 (d) A conditional reference to a named group can now use the syntax
687 (?(<name>) or (?('name') as well as (?(name).
688
689 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
690 groups (named and numbered) that are never evaluated inline, but can be
691 called as "subroutines" from elsewhere. In effect, the DEFINE condition
692 is always false. There may be only one alternative in such a group.
693
694 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
695 as the simple (?(R). The condition is true only if the most recent
696 recursion is that of the given number or name. It does not search out
697 through the entire recursion stack.
698
699 (g) The escape \gN or \g{N} has been added, where N is a positive or
700 negative number, specifying an absolute or relative reference.
701
702 35. Tidied to get rid of some further signed/unsigned compiler warnings and
703 some "unreachable code" warnings.
704
705 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
706 things, this adds five new scripts.
707
708 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
709 There were also incompatibilities regarding the handling of \Q..\E inside
710 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
711 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
712
713 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
714 matches an empty string, and forcibly breaks the loop. There were bugs in
715 this code in non-simple cases. For a pattern such as ^(a()*)* matched
716 against aaaa the result was just "a" rather than "aaaa", for example. Two
717 separate and independent bugs (that affected different cases) have been
718 fixed.
719
720 39. Refactored the code to abolish the use of different opcodes for small
721 capturing bracket numbers. This is a tidy that I avoided doing when I
722 removed the limit on the number of capturing brackets for 3.5 back in 2001.
723 The new approach is not only tidier, it makes it possible to reduce the
724 memory needed to fix the previous bug (38).
725
726 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
727 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
728 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
729 mode.
730
731 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
732 report.
733
734 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
735 copy construction and assignment in the C++ wrapper.
736
737 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
738 couple of bugs that could have given wrong results in the "--newline=crlf"
739 case.
740
741 44. Added a number of casts and did some reorganization of signed/unsigned int
742 variables following suggestions from Dair Grant. Also renamed the variable
743 "this" as "item" because it is a C++ keyword.
744
745 45. Arranged for dftables to add
746
747 #include "pcre_internal.h"
748
749 to pcre_chartables.c because without it, gcc 4.x may remove the array
750 definition from the final binary if PCRE is built into a static library and
751 dead code stripping is activated.
752
753 46. For an unanchored pattern, if a match attempt fails at the start of a
754 newline sequence, and the newline setting is CRLF or ANY, and the next two
755 characters are CRLF, advance by two characters instead of one.
756
757
758 Version 6.7 04-Jul-06
759 ---------------------
760
761 1. In order to handle tests when input lines are enormously long, pcretest has
762 been re-factored so that it automatically extends its buffers when
763 necessary. The code is crude, but this _is_ just a test program. The
764 default size has been increased from 32K to 50K.
765
766 2. The code in pcre_study() was using the value of the re argument before
767 testing it for NULL. (Of course, in any sensible call of the function, it
768 won't be NULL.)
769
770 3. The memmove() emulation function in pcre_internal.h, which is used on
771 systems that lack both memmove() and bcopy() - that is, hardly ever -
772 was missing a "static" storage class specifier.
773
774 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
775 containing an extended class (one that cannot be represented by a bitmap
776 because it contains high-valued characters or Unicode property items, e.g.
777 [\pZ]). Almost always one would set UTF-8 mode when processing such a
778 pattern, but PCRE should not loop if you do not (it no longer does).
779 [Detail: two cases were found: (a) a repeated subpattern containing an
780 extended class; (b) a recursive reference to a subpattern that followed a
781 previous extended class. It wasn't skipping over the extended class
782 correctly when UTF-8 mode was not set.]
783
784 5. A negated single-character class was not being recognized as fixed-length
785 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
786 compile error "lookbehind assertion is not fixed length".
787
788 6. The RunPerlTest auxiliary script was showing an unexpected difference
789 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
790 write a Perl script that can interpret lines of an input file either as
791 byte characters or as UTF-8, which is what "perltest" was being required to
792 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
793 can't do is switch easily at run time between having the "use utf8;" pragma
794 or not. In the end, I fudged it by using the RunPerlTest script to insert
795 "use utf8;" explicitly for the UTF-8 tests.
796
797 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
798 the end of the subject string, contrary to the documentation and to what
799 Perl does. This was true of both matching functions. Now it matches only at
800 the start of the subject and immediately after *internal* newlines.
801
802 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
803 a pointer to an int instead of a pointer to an unsigned long int. This
804 caused problems on 64-bit systems.
805
806 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
807 instance of the 'standard' template library not being so standard".
808
809 10. There was no check on the number of named subpatterns nor the maximum
810 length of a subpattern name. The product of these values is used to compute
811 the size of the memory block for a compiled pattern. By supplying a very
812 long subpattern name and a large number of named subpatterns, the size
813 computation could be caused to overflow. This is now prevented by limiting
814 the length of names to 32 characters, and the number of named subpatterns
815 to 10,000.
816
817 11. Subpatterns that are repeated with specific counts have to be replicated in
818 the compiled pattern. The size of memory for this was computed from the
819 length of the subpattern and the repeat count. The latter is limited to
820 65535, but there was no limit on the former, meaning that integer overflow
821 could in principle occur. The compiled length of a repeated subpattern is
822 now limited to 30,000 bytes in order to prevent this.
823
824 12. Added the optional facility to have named substrings with the same name.
825
826 13. Added the ability to use a named substring as a condition, using the
827 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
828 are numbers (not recommended). Forward references are permitted.
829
830 14. Added forward references in named backreferences (if you see what I mean).
831
832 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
833 pattern could run off the end of the subject. For example, the pattern
834 "(?s)(.{1,5})"8 did this with the subject "ab".
835
836 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
837 PCRE_CASELESS was set when matching characters that were quantified with ?
838 or *.
839
840 17. A character class other than a single negated character that had a minimum
841 but no maximum quantifier - for example [ab]{6,} - was not handled
842 correctly by pce_dfa_exec(). It would match only one character.
843
844 18. A valid (though odd) pattern that looked like a POSIX character
845 class but used an invalid character after [ (for example [[,abc,]]) caused
846 pcre_compile() to give the error "Failed: internal error: code overflow" or
847 in some cases to crash with a glibc free() error. This could even happen if
848 the pattern terminated after [[ but there just happened to be a sequence of
849 letters, a binary zero, and a closing ] in the memory that followed.
850
851 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
852 over the years. Originally (before any Unicode support), just the bottom 8
853 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
854 output from "man perlunicode" includes this:
855
856 The regular expression compiler produces polymorphic opcodes. That
857 is, the pattern adapts to the data and automatically switches to
858 the Unicode character scheme when presented with Unicode data--or
859 instead uses a traditional byte scheme when presented with byte
860 data.
861
862 Sadly, a wide octal escape does not cause a switch, and in a string with
863 no other multibyte characters, these octal escapes are treated as before.
864 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
865 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
866 Unicode string.
867
868 I have not perpetrated such confusion in PCRE. Up till now, it took just
869 the bottom 8 bits, as in old Perl. I have now made octal escapes with
870 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
871 translate to the appropriate multibyte character.
872
873 29. Applied some refactoring to reduce the number of warnings from Microsoft
874 and Borland compilers. This has included removing the fudge introduced
875 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
876 a warning about an unused variable.
877
878 21. PCRE has not included VT (character 0x0b) in the set of whitespace
879 characters since release 4.0, because Perl (from release 5.004) does not.
880 [Or at least, is documented not to: some releases seem to be in conflict
881 with the documentation.] However, when a pattern was studied with
882 pcre_study() and all its branches started with \s, PCRE still included VT
883 as a possible starting character. Of course, this did no harm; it just
884 caused an unnecessary match attempt.
885
886 22. Removed a now-redundant internal flag bit that recorded the fact that case
887 dependency changed within the pattern. This was once needed for "required
888 byte" processing, but is no longer used. This recovers a now-scarce options
889 bit. Also moved the least significant internal flag bit to the most-
890 significant bit of the word, which was not previously used (hangover from
891 the days when it was an int rather than a uint) to free up another bit for
892 the future.
893
894 23. Added support for CRLF line endings as well as CR and LF. As well as the
895 default being selectable at build time, it can now be changed at runtime
896 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
897 specify that it is scanning data with non-default line endings.
898
899 24. Changed the definition of CXXLINK to make it agree with the definition of
900 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
901
902 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
903 recursions. This makes a big different to stack usage for some patterns.
904
905 26. If a subpattern containing a named recursion or subroutine reference such
906 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
907 the space required for the compiled pattern went wrong and gave too small a
908 value. Depending on the environment, this could lead to "Failed: internal
909 error: code overflow at offset 49" or "glibc detected double free or
910 corruption" errors.
911
912 27. Applied patches from Google (a) to support the new newline modes and (b) to
913 advance over multibyte UTF-8 characters in GlobalReplace.
914
915 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
916 difference for some implementation of PCRE in some Windows version.
917
918 29. Added some extra testing facilities to pcretest:
919
920 \q<number> in a data line sets the "match limit" value
921 \Q<number> in a data line sets the "match recursion limt" value
922 -S <number> sets the stack size, where <number> is in megabytes
923
924 The -S option isn't available for Windows.
925
926
927 Version 6.6 06-Feb-06
928 ---------------------
929
930 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
931 in pcreposix.h. I have copied the definition from pcre.h.
932
933 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
934 because pcre.h is no longer a built file.
935
936 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
937 not normally included in the compiled code.
938
939
940 Version 6.5 01-Feb-06
941 ---------------------
942
943 1. When using the partial match feature with pcre_dfa_exec(), it was not
944 anchoring the second and subsequent partial matches at the new starting
945 point. This could lead to incorrect results. For example, with the pattern
946 /1234/, partially matching against "123" and then "a4" gave a match.
947
948 2. Changes to pcregrep:
949
950 (a) All non-match returns from pcre_exec() were being treated as failures
951 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
952 error message is output. Some extra information is given for the
953 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
954 probably the only errors that are likely to be caused by users (by
955 specifying a regex that has nested indefinite repeats, for instance).
956 If there are more than 20 of these errors, pcregrep is abandoned.
957
958 (b) A binary zero was treated as data while matching, but terminated the
959 output line if it was written out. This has been fixed: binary zeroes
960 are now no different to any other data bytes.
961
962 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
963 used to set a locale for matching. The --locale=xxxx long option has
964 been added (no short equivalent) to specify a locale explicitly on the
965 pcregrep command, overriding the environment variables.
966
967 (d) When -B was used with -n, some line numbers in the output were one less
968 than they should have been.
969
970 (e) Added the -o (--only-matching) option.
971
972 (f) If -A or -C was used with -c (count only), some lines of context were
973 accidentally printed for the final match.
974
975 (g) Added the -H (--with-filename) option.
976
977 (h) The combination of options -rh failed to suppress file names for files
978 that were found from directory arguments.
979
980 (i) Added the -D (--devices) and -d (--directories) options.
981
982 (j) Added the -F (--fixed-strings) option.
983
984 (k) Allow "-" to be used as a file name for -f as well as for a data file.
985
986 (l) Added the --colo(u)r option.
987
988 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
989 is not present by default.
990
991 3. A nasty bug was discovered in the handling of recursive patterns, that is,
992 items such as (?R) or (?1), when the recursion could match a number of
993 alternatives. If it matched one of the alternatives, but subsequently,
994 outside the recursion, there was a failure, the code tried to back up into
995 the recursion. However, because of the way PCRE is implemented, this is not
996 possible, and the result was an incorrect result from the match.
997
998 In order to prevent this happening, the specification of recursion has
999 been changed so that all such subpatterns are automatically treated as
1000 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1001
1002 4. I had overlooked the fact that, in some locales, there are characters for
1003 which isalpha() is true but neither isupper() nor islower() are true. In
1004 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1005 and ordfeminine) are like this. This affected the treatment of \w and \W
1006 when they appeared in character classes, but not when they appeared outside
1007 a character class. The bit map for "word" characters is now created
1008 separately from the results of isalnum() instead of just taking it from the
1009 upper, lower, and digit maps. (Plus the underscore character, of course.)
1010
1011 5. The above bug also affected the handling of POSIX character classes such as
1012 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1013 permanent tables. Instead, the bit maps for such a class were previously
1014 created as the appropriate unions of the upper, lower, and digit bitmaps.
1015 Now they are created by subtraction from the [[:word:]] class, which has
1016 its own bitmap.
1017
1018 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1019 It is created by subtracting the vertical space characters (\x09, \x0a,
1020 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1021 subtraction was done in the overall bitmap for a character class, meaning
1022 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1023 be recognized. This bug has been fixed.
1024
1025 7. Patches from the folks at Google:
1026
1027 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1028 real life, but is still worth protecting against".
1029
1030 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1031 regular expressions".
1032
1033 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1034 have it.
1035
1036 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1037 "configure" and the latter not, in order to fix a problem somebody had
1038 with compiling the Arg class on HP-UX.
1039
1040 (e) Improve the error-handling of the C++ wrapper a little bit.
1041
1042 (f) New tests for checking recursion limiting.
1043
1044 8. The pcre_memmove() function, which is used only if the environment does not
1045 have a standard memmove() function (and is therefore rarely compiled),
1046 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1047 returning a result (though PCRE never actually uses the result).
1048
1049 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1050 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1051 returned instead of calling malloc() with an overflowing number that would
1052 most likely cause subsequent chaos.
1053
1054 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1055
1056 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1057 with this option is matched, the nmatch and pmatch options of regexec() are
1058 ignored.
1059
1060 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1061 provided in case anyone wants to the the POSIX interface with UTF-8
1062 strings.
1063
1064 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1065 C++ linking (needed for some HP-UX environments).
1066
1067 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1068 (unused parameter) and in the pcre_printint() function (omitted "default"
1069 switch label when the default is to do nothing).
1070
1071 15. Added some code to make it possible, when PCRE is compiled as a C++
1072 library, to replace subject pointers for pcre_exec() with a smart pointer
1073 class, thus making it possible to process discontinuous strings.
1074
1075 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1076 much the same function. They were added by different people who were trying
1077 to make PCRE easy to compile on non-Unix systems. It has been suggested
1078 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1079 for compiling on Windows systems. I have therefore replaced it with
1080 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1081 defaults to "extern" for C or "extern C" for C++, which works fine on
1082 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1083 SCOPE with something explicit in config.h. In addition:
1084
1085 (a) pcreposix.h still had just "extern" instead of either of these macros;
1086 I have replaced it with PCRE_DATA_SCOPE.
1087
1088 (b) Functions such as _pcre_xclass(), which are internal to the library,
1089 but external in the C sense, all had PCRE_EXPORT in their definitions.
1090 This is apparently wrong for the Windows case, so I have removed it.
1091 (It makes no difference on Unix-like systems.)
1092
1093 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1094 of recursive calls to match(). This is different to MATCH_LIMIT because
1095 that limits the total number of calls to match(), not all of which increase
1096 the depth of recursion. Limiting the recursion depth limits the amount of
1097 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1098 when PCRE is compiled, and changed at run time. A patch from Google adds
1099 this functionality to the C++ interface.
1100
1101 18. Changes to the handling of Unicode character properties:
1102
1103 (a) Updated the table to Unicode 4.1.0.
1104
1105 (b) Recognize characters that are not in the table as "Cn" (undefined).
1106
1107 (c) I revised the way the table is implemented to a much improved format
1108 which includes recognition of ranges. It now supports the ranges that
1109 are defined in UnicodeData.txt, and it also amalgamates other
1110 characters into ranges. This has reduced the number of entries in the
1111 table from around 16,000 to around 3,000, thus reducing its size
1112 considerably. I realized I did not need to use a tree structure after
1113 all - a binary chop search is just as efficient. Having reduced the
1114 number of entries, I extended their size from 6 bytes to 8 bytes to
1115 allow for more data.
1116
1117 (d) Added support for Unicode script names via properties such as \p{Han}.
1118
1119 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1120 matching that character.
1121
1122 20. When matching a repeated Unicode property with a minimum greater than zero,
1123 (for example \pL{2,}), PCRE could look past the end of the subject if it
1124 reached it while seeking the minimum number of characters. This could
1125 happen only if some of the characters were more than one byte long, because
1126 there is a check for at least the minimum number of bytes.
1127
1128 21. Refactored the implementation of \p and \P so as to be more general, to
1129 allow for more different types of property in future. This has changed the
1130 compiled form incompatibly. Anybody with saved compiled patterns that use
1131 \p or \P will have to recompile them.
1132
1133 22. Added "Any" and "L&" to the supported property types.
1134
1135 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1136 but give a compile time error if the value is greater than 0xff.
1137
1138 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1139 accidentally not being installed or uninstalled.
1140
1141 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1142 made were to insert the current release number. This seemed silly, because
1143 it made things harder for people building PCRE on systems that don't run
1144 "configure". I have turned pcre.h into a distributed file, no longer built
1145 by "configure", with the version identification directly included. There is
1146 no longer a pcre.h.in file.
1147
1148 However, this change necessitated a change to the pcre-config script as
1149 well. It is built from pcre-config.in, and one of the substitutions was the
1150 release number. I have updated configure.ac so that ./configure now finds
1151 the release number by grepping pcre.h.
1152
1153 26. Added the ability to run the tests under valgrind.
1154
1155
1156 Version 6.4 05-Sep-05
1157 ---------------------
1158
1159 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1160 "--" to be printed when multiple files were scanned, even when none of the
1161 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1162 consider it to be a bug, and have restored the previous behaviour.
1163
1164 2. A couple of code tidies to get rid of compiler warnings.
1165
1166 3. The pcretest program used to cheat by referring to symbols in the library
1167 whose names begin with _pcre_. These are internal symbols that are not
1168 really supposed to be visible externally, and in some environments it is
1169 possible to suppress them. The cheating is now confined to including
1170 certain files from the library's source, which is a bit cleaner.
1171
1172 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1173 file's purpose clearer.
1174
1175 5. Reorganized pcre_ucp_findchar().
1176
1177
1178 Version 6.3 15-Aug-05
1179 ---------------------
1180
1181 1. The file libpcre.pc.in did not have general read permission in the tarball.
1182
1183 2. There were some problems when building without C++ support:
1184
1185 (a) If C++ support was not built, "make install" and "make test" still
1186 tried to test it.
1187
1188 (b) There were problems when the value of CXX was explicitly set. Some
1189 changes have been made to try to fix these, and ...
1190
1191 (c) --disable-cpp can now be used to explicitly disable C++ support.
1192
1193 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1194 backslash in a target when C++ was disabled. This confuses some
1195 versions of "make", apparently. Using an intermediate variable solves
1196 this. (Same for CPP_LOBJ.)
1197
1198 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1199 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1200 necessary on certain architectures.
1201
1202 4. Added a setting of -export-symbols-regex to the link command to remove
1203 those symbols that are exported in the C sense, but actually are local
1204 within the library, and not documented. Their names all begin with
1205 "_pcre_". This is not a perfect job, because (a) we have to except some
1206 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1207 available (and never for static libraries). I have made a note to try to
1208 find a way round (a) in the future.
1209
1210
1211 Version 6.2 01-Aug-05
1212 ---------------------
1213
1214 1. There was no test for integer overflow of quantifier values. A construction
1215 such as {1111111111111111} would give undefined results. What is worse, if
1216 a minimum quantifier for a parenthesized subpattern overflowed and became
1217 negative, the calculation of the memory size went wrong. This could have
1218 led to memory overwriting.
1219
1220 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1221
1222 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1223 operating environments where this matters.
1224
1225 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1226 PCRE options from within the C++ wrapper.
1227
1228 5. Named capturing subpatterns were not being correctly counted when a pattern
1229 was compiled. This caused two problems: (a) If there were more than 100
1230 such subpatterns, the calculation of the memory needed for the whole
1231 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1232 back references of the form \12, where the number was greater than 9, were
1233 not recognized as back references, even though there were sufficient
1234 previous subpatterns.
1235
1236 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1237 versions of gcc, e.g. 2.95.4.
1238
1239
1240 Version 6.1 21-Jun-05
1241 ---------------------
1242
1243 1. There was one reference to the variable "posix" in pcretest.c that was not
1244 surrounded by "#if !defined NOPOSIX".
1245
1246 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1247 the cross-check on the old pcre_info() function, for the benefit of the
1248 cut-down version of PCRE that is currently imported into Exim.
1249
1250 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1251 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1252 patterns that start (?i) but I don't think that matters. The use of (?i) is
1253 just an example; this all applies to the other options as well.
1254
1255 4. Since libtool seems to echo the compile commands it is issuing, the output
1256 from "make" can be reduced a bit by putting "@" in front of each libtool
1257 compile command.
1258
1259 5. Patch from the folks at Google for configure.in to be a bit more thorough
1260 in checking for a suitable C++ installation before trying to compile the
1261 C++ stuff. This should fix a reported problem when a compiler was present,
1262 but no suitable headers.
1263
1264 6. The man pages all had just "PCRE" as their title. I have changed them to
1265 be the relevant file name. I have also arranged that these names are
1266 retained in the file doc/pcre.txt, which is a concatenation in text format
1267 of all the man pages except the little individual ones for each function.
1268
1269 7. The NON-UNIX-USE file had not been updated for the different set of source
1270 files that come with release 6. I also added a few comments about the C++
1271 wrapper.
1272
1273
1274 Version 6.0 07-Jun-05
1275 ---------------------
1276
1277 1. Some minor internal re-organization to help with my DFA experiments.
1278
1279 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1280 didn't matter for the library itself when fully configured, but did matter
1281 when compiling without UCP support, or within Exim, where the ucp files are
1282 not imported.
1283
1284 3. Refactoring of the library code to split up the various functions into
1285 different source modules. The addition of the new DFA matching code (see
1286 below) to a single monolithic source would have made it really too
1287 unwieldy, quite apart from causing all the code to be include in a
1288 statically linked application, when only some functions are used. This is
1289 relevant even without the DFA addition now that patterns can be compiled in
1290 one application and matched in another.
1291
1292 The downside of splitting up is that there have to be some external
1293 functions and data tables that are used internally in different modules of
1294 the library but which are not part of the API. These have all had their
1295 names changed to start with "_pcre_" so that they are unlikely to clash
1296 with other external names.
1297
1298 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1299 a different (DFA) algorithm. Although it is slower than the original
1300 function, it does have some advantages for certain types of matching
1301 problem.
1302
1303 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1304 including restarting after a partial match.
1305
1306 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1307 defined when compiling for Windows was sent to me. I have put it into the
1308 code, though I have no means of testing or verifying it.
1309
1310 7. Added the pcre_refcount() auxiliary function.
1311
1312 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1313 match before or at the first newline in the subject string. In pcretest,
1314 the /f option on a pattern can be used to set this.
1315
1316 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1317 would behave wrongly. This has been present in PCRE since release 4.0.
1318
1319 10. A number of changes to the pcregrep command:
1320
1321 (a) Refactored how -x works; insert ^(...)$ instead of setting
1322 PCRE_ANCHORED and checking the length, in preparation for adding
1323 something similar for -w.
1324
1325 (b) Added the -w (match as a word) option.
1326
1327 (c) Refactored the way lines are read and buffered so as to have more
1328 than one at a time available.
1329
1330 (d) Implemented a pcregrep test script.
1331
1332 (e) Added the -M (multiline match) option. This allows patterns to match
1333 over several lines of the subject. The buffering ensures that at least
1334 8K, or the rest of the document (whichever is the shorter) is available
1335 for matching (and similarly the previous 8K for lookbehind assertions).
1336
1337 (f) Changed the --help output so that it now says
1338
1339 -w, --word-regex(p)
1340
1341 instead of two lines, one with "regex" and the other with "regexp"
1342 because that confused at least one person since the short forms are the
1343 same. (This required a bit of code, as the output is generated
1344 automatically from a table. It wasn't just a text change.)
1345
1346 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1347 option but starts with a hyphen. Could be a pattern or a path name
1348 starting with a hyphen, for instance.
1349
1350 (h) "-" can be given as a file name to represent stdin.
1351
1352 (i) When file names are being printed, "(standard input)" is used for
1353 the standard input, for compatibility with GNU grep. Previously
1354 "<stdin>" was used.
1355
1356 (j) The option --label=xxx can be used to supply a name to be used for
1357 stdin when file names are being printed. There is no short form.
1358
1359 (k) Re-factored the options decoding logic because we are going to add
1360 two more options that take data. Such options can now be given in four
1361 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1362
1363 (l) Added the -A, -B, and -C options for requesting that lines of context
1364 around matches be printed.
1365
1366 (m) Added the -L option to print the names of files that do not contain
1367 any matching lines, that is, the complement of -l.
1368
1369 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1370 continue to scan other files.
1371
1372 (o) The -s option was incorrectly implemented. For compatibility with other
1373 greps, it now suppresses the error message for a non-existent or non-
1374 accessible file (but not the return code). There is a new option called
1375 -q that suppresses the output of matching lines, which was what -s was
1376 previously doing.
1377
1378 (p) Added --include and --exclude options to specify files for inclusion
1379 and exclusion when recursing.
1380
1381 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1382 Hopefully, it now does.
1383
1384 12. Missing cast in pcre_study().
1385
1386 13. Added an "uninstall" target to the makefile.
1387
1388 14. Replaced "extern" in the function prototypes in Makefile.in with
1389 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1390 world, but is set differently for Windows.
1391
1392 15. Added a second compiling function called pcre_compile2(). The only
1393 difference is that it has an extra argument, which is a pointer to an
1394 integer error code. When there is a compile-time failure, this is set
1395 non-zero, in addition to the error test pointer being set to point to an
1396 error message. The new argument may be NULL if no error number is required
1397 (but then you may as well call pcre_compile(), which is now just a
1398 wrapper). This facility is provided because some applications need a
1399 numeric error indication, but it has also enabled me to tidy up the way
1400 compile-time errors are handled in the POSIX wrapper.
1401
1402 16. Added VPATH=.libs to the makefile; this should help when building with one
1403 prefix path and installing with another. (Or so I'm told by someone who
1404 knows more about this stuff than I do.)
1405
1406 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1407 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1408 match everything, including newlines. This is not POSIX-compatible, but
1409 somebody wanted the feature. From pcretest it can be activated by using
1410 both the P and the s flags.
1411
1412 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1413
1414 19. libpcre.pc was being incorrectly installed as executable.
1415
1416 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1417 it now also looks for '\r' so that it will work unmodified on Windows.
1418
1419 21. Added Google's contributed C++ wrapper to the distribution.
1420
1421 22. Added some untidy missing memory free() calls in pcretest, to keep
1422 Electric Fence happy when testing.
1423
1424
1425
1426 Version 5.0 13-Sep-04
1427 ---------------------
1428
1429 1. Internal change: literal characters are no longer packed up into items
1430 containing multiple characters in a single byte-string. Each character
1431 is now matched using a separate opcode. However, there may be more than one
1432 byte in the character in UTF-8 mode.
1433
1434 2. The pcre_callout_block structure has two new fields: pattern_position and
1435 next_item_length. These contain the offset in the pattern to the next match
1436 item, and its length, respectively.
1437
1438 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1439 insertion of callouts before each pattern item. Added the /C option to
1440 pcretest to make use of this.
1441
1442 4. On the advice of a Windows user, the lines
1443
1444 #if defined(_WIN32) || defined(WIN32)
1445 _setmode( _fileno( stdout ), 0x8000 );
1446 #endif /* defined(_WIN32) || defined(WIN32) */
1447
1448 have been added to the source of pcretest. This apparently does useful
1449 magic in relation to line terminators.
1450
1451 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1452 for the benefit of those environments where the "b" makes a difference.
1453
1454 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1455 to know about it. I have put a hack into configure.in that adds in code
1456 to set GCC=yes if CC=icc. This seems to end up at a point in the
1457 generated configure script that is early enough to affect the setting of
1458 compiler options, which is what is needed, but I have no means of testing
1459 whether it really works. (The user who reported this had patched the
1460 generated configure script, which of course I cannot do.)
1461
1462 LATER: After change 22 below (new libtool files), the configure script
1463 seems to know about icc (and also ecc). Therefore, I have commented out
1464 this hack in configure.in.
1465
1466 7. Added support for pkg-config (2 patches were sent in).
1467
1468 8. Negated POSIX character classes that used a combination of internal tables
1469 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1470 [[:^ascii]]. Typically, they would match almost any characters. The other
1471 POSIX classes were not broken in this way.
1472
1473 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1474 to find the match, as PCRE was deluded into thinking that the match had to
1475 start at the start point or following a newline. The same bug applied to
1476 patterns with negative forward assertions or any backward assertions
1477 preceding ".*" at the start, unless the pattern required a fixed first
1478 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1479
1480 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1481 starting at the last subject character, bytes beyond the end of the subject
1482 string were read.
1483
1484 11. Renamed the variable "class" as "classbits" to make life easier for C++
1485 users. (Previously there was a macro definition, but it apparently wasn't
1486 enough.)
1487
1488 12. Added the new field "tables" to the extra data so that tables can be passed
1489 in at exec time, or the internal tables can be re-selected. This allows
1490 a compiled regex to be saved and re-used at a later time by a different
1491 program that might have everything at different addresses.
1492
1493 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1494 -R library as well as a -L library.
1495
1496 14. The debugging options of pcretest (-d on the command line or D on a
1497 pattern) showed incorrect output for anything following an extended class
1498 that contained multibyte characters and which was followed by a quantifier.
1499
1500 15. Added optional support for general category Unicode character properties
1501 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1502 support. It adds about 90K to the size of the library. The meanings of the
1503 inbuilt class escapes such as \d and \s have NOT been changed.
1504
1505 16. Updated pcredemo.c to include calls to free() to release the memory for the
1506 compiled pattern.
1507
1508 17. The generated file chartables.c was being created in the source directory
1509 instead of in the building directory. This caused the build to fail if the
1510 source directory was different from the building directory, and was
1511 read-only.
1512
1513 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1514 file. No doubt somebody will tell me if they don't make sense... Also added
1515 Dan Mooney's comments about building on OpenVMS.
1516
1517 19. Added support for partial matching via the PCRE_PARTIAL option for
1518 pcre_exec() and the \P data escape in pcretest.
1519
1520 20. Extended pcretest with 3 new pattern features:
1521
1522 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1523 write the compiled pattern to the file whose name is "rest-of-line".
1524 This is a straight binary dump of the data, with the saved pointer to
1525 the character tables forced to be NULL. The study data, if any, is
1526 written too. After writing, pcretest reads a new pattern.
1527
1528 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1529 compiled pattern from the given file. There must not be any
1530 occurrences of "<" in the file name (pretty unlikely); if there are,
1531 pcretest will instead treat the initial "<" as a pattern delimiter.
1532 After reading in the pattern, pcretest goes on to read data lines as
1533 usual.
1534
1535 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1536 and 16-bit fields in a compiled pattern, to simulate a pattern that
1537 was compiled on a host of opposite endianness.
1538
1539 21. The pcre-exec() function can now cope with patterns that were compiled on
1540 hosts of opposite endianness, with this restriction:
1541
1542 As for any compiled expression that is saved and used later, the tables
1543 pointer field cannot be preserved; the extra_data field in the arguments
1544 to pcre_exec() should be used to pass in a tables address if a value
1545 other than the default internal tables were used at compile time.
1546
1547 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1548 now diagnosed as an error. Previously, most of the time, a negative number
1549 would have been treated as zero, but if in addition "ovector" was passed as
1550 NULL, a crash could occur.
1551
1552 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1553 new versions from the libtool 1.5 distribution (the last one is a copy of
1554 a file called libtool.m4). This seems to have fixed the need to patch
1555 "configure" to support Darwin 1.3 (which I used to do). However, I still
1556 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1557 workstation).
1558
1559 24. Changed the PCRE licence to be the more standard "BSD" licence.
1560
1561
1562 Version 4.5 01-Dec-03
1563 ---------------------
1564
1565 1. There has been some re-arrangement of the code for the match() function so
1566 that it can be compiled in a version that does not call itself recursively.
1567 Instead, it keeps those local variables that need separate instances for
1568 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1569 needs to "recurse". Keeping track of where control must go is done by means
1570 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1571 hide most of the details from the main code, and operates only if
1572 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1573 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1574 operating.
1575
1576 To make it easier for callers to provide specially tailored get/free
1577 functions for this usage, two new functions, pcre_stack_malloc, and
1578 pcre_stack_free, are used. They are always called in strict stacking order,
1579 and the size of block requested is always the same.
1580
1581 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1582 PCRE has been compiled to use the stack or the heap for recursion. The
1583 -C option of pcretest uses this to show which version is compiled.
1584
1585 A new data escape \S, is added to pcretest; it causes the amounts of store
1586 obtained and freed by both kinds of malloc/free at match time to be added
1587 to the output.
1588
1589 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1590 what's available on my current Linux desktop machine.
1591
1592 3. When matching a UTF-8 string, the test for a valid string at the start has
1593 been extended. If start_offset is not zero, PCRE now checks that it points
1594 to a byte that is the start of a UTF-8 character. If not, it returns
1595 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1596 this is necessary because there may be backward assertions in the pattern.
1597 When matching the same subject several times, it may save resources to use
1598 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1599
1600 4. The code for checking the validity of UTF-8 strings has been tightened so
1601 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1602 containing "overlong sequences".
1603
1604 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1605 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1606 should have been "&", but it just so happened that all the cases this let
1607 through by mistake were picked up later in the function.
1608
1609 6. I had used a variable called "isblank" - this is a C99 function, causing
1610 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1611
1612 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1613 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1614 so the version is shown; (c) stop "make test" echoing "./RunTest".
1615
1616 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1617
1618 9. The prototype for memmove() for systems that don't have it was using
1619 size_t, but the inclusion of the header that defines size_t was later. I've
1620 moved the #includes for the C headers earlier to avoid this.
1621
1622 10. Added some adjustments to the code to make it easier to compiler on certain
1623 special systems:
1624
1625 (a) Some "const" qualifiers were missing.
1626 (b) Added the macro EXPORT before all exported functions; by default this
1627 is defined to be empty.
1628 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1629 that it reads its output file name as an argument instead of writing
1630 to the standard output and assuming this can be redirected.
1631
1632 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1633 class containing characters with values greater than 255, PCRE compilation
1634 went into a loop.
1635
1636 12. A recursive reference to a subpattern that was within another subpattern
1637 that had a minimum quantifier of zero caused PCRE to crash. For example,
1638 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1639 recursion. If the recursively-called subpattern itself had a zero repeat,
1640 that was OK.
1641
1642 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1643 buffer into which it was copied (for escape processing) was still set at
1644 1024, so long lines caused crashes.
1645
1646 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1647 "internal error: code overflow...". This applied to any character class
1648 that was followed by a possessive quantifier.
1649
1650 15. Modified the Makefile to add libpcre.la as a prerequisite for
1651 libpcreposix.la because I was told this is needed for a parallel build to
1652 work.
1653
1654 16. If a pattern that contained .* following optional items at the start was
1655 studied, the wrong optimizing data was generated, leading to matching
1656 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1657 matching string must start with a or b or c. The correct conclusion for
1658 this pattern is that a match can start with any character.
1659
1660
1661 Version 4.4 13-Aug-03
1662 ---------------------
1663
1664 1. In UTF-8 mode, a character class containing characters with values between
1665 127 and 255 was not handled correctly if the compiled pattern was studied.
1666 In fixing this, I have also improved the studying algorithm for such
1667 classes (slightly).
1668
1669 2. Three internal functions had redundant arguments passed to them. Removal
1670 might give a very teeny performance improvement.
1671
1672 3. Documentation bug: the value of the capture_top field in a callout is *one
1673 more than* the number of the hightest numbered captured substring.
1674
1675 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1676 in incorrectly linking with a previously installed version. They now link
1677 explicitly with libpcre.la.
1678
1679 5. configure.in no longer needs to recognize Cygwin specially.
1680
1681 6. A problem in pcre.in for Windows platforms is fixed.
1682
1683 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1684 pcretest, it used to include the size of the study block as part of its
1685 output. Unfortunately, the structure contains a field that has a different
1686 size on different hardware architectures. This meant that the tests that
1687 showed this size failed. As the block is currently always of a fixed size,
1688 this information isn't actually particularly useful in pcretest output, so
1689 I have just removed it.
1690
1691 8. Three pre-processor statements accidentally did not start in column 1.
1692 Sadly, there are *still* compilers around that complain, even though
1693 standard C has not required this for well over a decade. Sigh.
1694
1695 9. In pcretest, the code for checking callouts passed small integers in the
1696 callout_data field, which is a void * field. However, some picky compilers
1697 complained about the casts involved for this on 64-bit systems. Now
1698 pcretest passes the address of the small integer instead, which should get
1699 rid of the warnings.
1700
1701 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1702 both compile and run time, and gives an error if an invalid UTF-8 sequence
1703 is found. There is a option for disabling this check in cases where the
1704 string is known to be correct and/or the maximum performance is wanted.
1705
1706 11. In response to a bug report, I changed one line in Makefile.in from
1707
1708 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1709 to
1710 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1711
1712 to look similar to other lines, but I have no way of telling whether this
1713 is the right thing to do, as I do not use Windows. No doubt I'll get told
1714 if it's wrong...
1715
1716
1717 Version 4.3 21-May-03
1718 ---------------------
1719
1720 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1721 Makefile.
1722
1723 2. Some refactoring to improve the quality of the code:
1724
1725 (i) The utf8_table... variables are now declared "const".
1726
1727 (ii) The code for \cx, which used the "case flipping" table to upper case
1728 lower case letters, now just substracts 32. This is ASCII-specific,
1729 but the whole concept of \cx is ASCII-specific, so it seems
1730 reasonable.
1731
1732 (iii) PCRE was using its character types table to recognize decimal and
1733 hexadecimal digits in the pattern. This is silly, because it handles
1734 only 0-9, a-f, and A-F, but the character types table is locale-
1735 specific, which means strange things might happen. A private
1736 table is now used for this - though it costs 256 bytes, a table is
1737 much faster than multiple explicit tests. Of course, the standard
1738 character types table is still used for matching digits in subject
1739 strings against \d.
1740
1741 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1742 ending in _t are). So I've renamed it as ESC_tee.
1743
1744 3. The first argument for regexec() in the POSIX wrapper should have been
1745 defined as "const".
1746
1747 4. Changed pcretest to use malloc() for its buffers so that they can be
1748 Electric Fenced for debugging.
1749
1750 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1751 to read one or more bytes before the start of the subject string. Often this
1752 had no effect on PCRE's behaviour, but in some circumstances it could
1753 provoke a segmentation fault.
1754
1755 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1756 to try to read one or more bytes before the start of the subject string.
1757
1758 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1759 UTF-8 support could misbehave in various ways if the subject string
1760 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1761 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1762 back over UTF-8 characters.)
1763
1764
1765 Version 4.2 14-Apr-03
1766 ---------------------
1767
1768 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1769
1770 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1771 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1772 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1773 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1774 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1775 and BUILD_EXEEXT
1776 Note: automatic setting of the BUILD variables is not yet working
1777 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1778 compile-time but not at link-time
1779 [LINK]: use for linking executables only
1780 make different versions for Windows and non-Windows
1781 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1782 libraries
1783 [LINK_FOR_BUILD]: new variable
1784 [OBJEXT]: use throughout
1785 [EXEEXT]: use throughout
1786 <winshared>: new target
1787 <wininstall>: new target
1788 <dftables.o>: use native compiler
1789 <dftables>: use native linker
1790 <install>: handle Windows platform correctly
1791 <clean>: ditto
1792 <check>: ditto
1793 copy DLL to top builddir before testing
1794
1795 As part of these changes, -no-undefined was removed again. This was reported
1796 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1797 in any case.
1798
1799 3. Some tidies to get rid of compiler warnings:
1800
1801 . In the match_data structure, match_limit was an unsigned long int, whereas
1802 match_call_count was an int. I've made them both unsigned long ints.
1803
1804 . In pcretest the fact that a const uschar * doesn't automatically cast to
1805 a void * provoked a warning.
1806
1807 . Turning on some more compiler warnings threw up some "shadow" variables
1808 and a few more missing casts.
1809
1810 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1811 option, a class that contained a single character with a value between 128
1812 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1813
1814 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1815 option, a class that contained several characters, but with at least one
1816 whose value was between 128 and 255 caused PCRE to crash.
1817
1818
1819 Version 4.1 12-Mar-03
1820 ---------------------
1821
1822 1. Compiling with gcc -pedantic found a couple of places where casts were
1823 needed, and a string in dftables.c that was longer than standard compilers are
1824 required to support.
1825
1826 2. Compiling with Sun's compiler found a few more places where the code could
1827 be tidied up in order to avoid warnings.
1828
1829 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1830 first of these names is deprecated in the latest Autoconf in favour of the name
1831 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1832 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1833 analogy I have changed it to CFLAGS_FOR_BUILD.
1834
1835 4. Added -no-undefined to the linking command in the Makefile, because this is
1836 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1837 linking step for the pcreposix library.
1838
1839 5. PCRE was failing to diagnose the case of two named groups with the same
1840 name.
1841
1842 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1843 literal character that is needed in the subject for a match, and scans along to
1844 ensure that it is present before embarking on the full matching process. This
1845 saves time in cases of nested unlimited repeats that are never going to match.
1846 Problem: the scan can take a lot of time if the subject is very long (e.g.
1847 megabytes), thus penalizing straightforward matches. It is now done only if the
1848 amount of subject to be scanned is less than 1000 bytes.
1849
1850 7. A lesser problem with the same optimization is that it was recording the
1851 first character of an anchored pattern as "needed", thus provoking a search
1852 right along the subject, even when the first match of the pattern was going to
1853 fail. The "needed" character is now not set for anchored patterns, unless it
1854 follows something in the pattern that is of non-fixed length. Thus, it still
1855 fulfils its original purpose of finding quick non-matches in cases of nested
1856 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1857
1858
1859 Version 4.0 17-Feb-03
1860 ---------------------
1861
1862 1. If a comment in an extended regex that started immediately after a meta-item
1863 extended to the end of string, PCRE compiled incorrect data. This could lead to
1864 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1865
1866 2. Moved to autoconf 2.53 and libtool 1.4.2.
1867
1868 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1869 the special perltest8 script is no longer needed - all the tests can be run
1870 from a single perltest script.
1871
1872 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1873 by \s. It has now been removed in PCRE. This means it isn't recognized as
1874 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1875 class [:space:] *does* include VT, thereby creating a mess.
1876
1877 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1878 space and tab.
1879
1880 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1881 its new features into the main test script, reducing the number of scripts.
1882
1883 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1884 were backward compatible, and made the (?i) apply to the whole pattern, as if
1885 /i were given. Now it behaves more logically, and applies the option setting
1886 only to what follows. PCRE has been changed to follow suit. However, if it
1887 finds options settings right at the start of the pattern, it extracts them into
1888 the global options, as before. Thus, they show up in the info data.
1889
1890 8. Added support for the \Q...\E escape sequence. Characters in between are
1891 treated as literals. This is slightly different from Perl in that $ and @ are
1892 also handled as literals inside the quotes. In Perl, they will cause variable
1893 interpolation. Note the following examples:
1894
1895 Pattern PCRE matches Perl matches
1896
1897 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1898 \Qabc\$xyz\E abc\$xyz abc\$xyz
1899 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1900
1901 For compatibility with Perl, \Q...\E sequences are recognized inside character
1902 classes as well as outside them.
1903
1904 9. Re-organized 3 code statements in pcretest to avoid "overflow in
1905 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1906 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1907 signed/unsigned warnings.
1908
1909 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1910 option for pcretest, so I've replaced it by a simple function that does just
1911 that job.
1912
1913 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1914 "pcregrep -".
1915
1916 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1917 Java package. This provides some syntactic sugar for simple cases of what my
1918 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1919 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1920 item, you can use this simplified notation. Note that only makes sense with
1921 greedy quantifiers. Consequently, the use of the possessive quantifier forces
1922 greediness, whatever the setting of the PCRE_UNGREEDY option.
1923
1924 13. A change of greediness default within a pattern was not taking effect at
1925 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1926 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1927 was abstracted outside.
1928
1929 14. PCRE now supports the \G assertion. It is true when the current matching
1930 position is at the start point of the match. This differs from \A when the
1931 starting offset is non-zero. Used with the /g option of pcretest (or similar
1932 code), it works in the same way as it does for Perl's /g option. If all
1933 alternatives of a regex begin with \G, the expression is anchored to the start
1934 match position, and the "anchored" flag is set in the compiled expression.
1935
1936 15. Some bugs concerning the handling of certain option changes within patterns
1937 have been fixed. These applied to options other than (?ims). For example,
1938 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1939 been the other way round. Some of this was related to change 7 above.
1940
1941 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1942 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1943 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1944 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1945
1946 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1947 mode. Unlike ".", it always matches newline, whatever the setting of
1948 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1949 assertions. Perl allows it, but it doesn't (in general) work because it can't
1950 calculate the length of the lookbehind. At least, that's the case for Perl
1951 5.8.0 - I've been told they are going to document that it doesn't work in
1952 future.
1953
1954 18. Added an error diagnosis for escapes that PCRE does not support: these are
1955 \L, \l, \N, \P, \p, \U, \u, and \X.
1956
1957 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1958 reading past the end of the pattern in cases such as /[abcd/.
1959
1960 20. PCRE was getting more memory than necessary for patterns with classes that
1961 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1962
1963 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1964 compiling PCRE for use with Virtual Pascal.
1965
1966 22. Small fix to the Makefile to make it work properly if the build is done
1967 outside the source tree.
1968
1969 23. Added a new extension: a condition to go with recursion. If a conditional
1970 subpattern starts with (?(R) the "true" branch is used if recursion has
1971 happened, whereas the "false" branch is used only at the top level.
1972
1973 24. When there was a very long string of literal characters (over 255 bytes
1974 without UTF support, over 250 bytes with UTF support), the computation of how
1975 much memory was required could be incorrect, leading to segfaults or other
1976 strange effects.
1977
1978 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1979 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1980 there was a subsequent back reference to those brackets. This meant that, for
1981 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1982 possible to check for precisely this case. All we can do is abandon the
1983 optimization if .* occurs inside capturing brackets when there are any back
1984 references whatsoever. (See below for a better fix that came later.)
1985
1986 26. The handling of the optimization for finding the first character of a
1987 non-anchored pattern, and for finding a character that is required later in the
1988 match were failing in some cases. This didn't break the matching; it just
1989 failed to optimize when it could. The way this is done has been re-implemented.
1990
1991 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1992
1993 28. Added a new feature that provides some of the functionality that Perl
1994 provides with (?{...}). The facility is termed a "callout". The way it is done
1995 in PCRE is for the caller to provide an optional function, by setting
1996 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1997 global variable. By default it is unset, which disables all calling out. To get
1998 the function called, the regex must include (?C) at appropriate points. This
1999 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2000 This provides a means of identifying different callout points. When PCRE
2001 reaches such a point in the regex, if pcre_callout has been set, the external
2002 function is called. It is provided with data in a structure called
2003 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2004 matching continues; if it returns a non-zero value, the match at the current
2005 point fails. However, backtracking will occur if possible. [This was changed
2006 later and other features added - see item 49 below.]
2007
2008 29. pcretest is upgraded to test the callout functionality. It provides a
2009 callout function that displays information. By default, it shows the start of
2010 the match and the current position in the text. There are some new data escapes
2011 to vary what happens:
2012
2013 \C+ in addition, show current contents of captured substrings
2014 \C- do not supply a callout function
2015 \C!n return 1 when callout number n is reached
2016 \C!n!m return 1 when callout number n is reached for the mth time
2017
2018 30. If pcregrep was called with the -l option and just a single file name, it
2019 output "<stdin>" if a match was found, instead of the file name.
2020
2021 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2022 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2023 pcre_exec(). This saves a malloc/free per call. The default value of
2024 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2025 when configuring.
2026
2027 32. The default maximum size of a compiled pattern is 64K. There have been a
2028 few cases of people hitting this limit. The code now uses macros to handle the
2029 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2030 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2031 configuring. Tests 2 and 5 work only with 2-byte links because they output
2032 debugging information about compiled patterns.
2033
2034 33. Internal code re-arrangements:
2035
2036 (a) Moved the debugging function for printing out a compiled regex into
2037 its own source file (printint.c) and used #include to pull it into
2038 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2039 separate copies.
2040
2041 (b) Defined the list of op-code names for debugging as a macro in
2042 internal.h so that it is next to the definition of the opcodes.
2043
2044 (c) Defined a table of op-code lengths for simpler skipping along compiled
2045 code. This is again a macro in internal.h so that it is next to the
2046 definition of the opcodes.
2047
2048 34. Added support for recursive calls to individual subpatterns, along the
2049 lines of Robin Houston's patch (but implemented somewhat differently).
2050
2051 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2052 allow it to read and process whole directories in Win32. This code was
2053 contributed by Lionel Fourquaux; it has not been tested by me.
2054
2055 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2056 used to name a group. Names consist of alphanumerics and underscores, and must
2057 be unique. Back references use the syntax (?P=name) and recursive calls use
2058 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2059 numbers. The function pcre_fullinfo() can be used after compilation to extract
2060 a name/number map. There are three relevant calls:
2061
2062 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2063 PCRE_INFO_NAMECOUNT yields the number of entries
2064 PCRE_INFO_NAMETABLE yields a pointer to the map.
2065
2066 The map is a vector of fixed-size entries. The size of each entry depends on
2067 the length of the longest name used. The first two bytes of each entry are the
2068 group number, most significant byte first. There follows the corresponding
2069 name, zero terminated. The names are in alphabetical order.
2070
2071 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2072 case instead of 255. Making it the same both with and without UTF-8 support
2073 means that the same test output works with both.
2074
2075 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2076 calling malloc() with a zero argument.
2077
2078 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2079 optimization. I've improved things by keeping a bitmap of backreferences with
2080 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2081 fact referenced, the optimization can be applied. It is unlikely that a
2082 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2083 the match to follow \n) will appear inside brackets with a number greater than
2084 31, but if it does, any back reference > 31 suppresses the optimization.
2085
2086 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2087 of disabling numbered capturing parentheses. Any opening parenthesis that is
2088 not followed by ? behaves as if it were followed by ?: but named parentheses
2089 can still be used for capturing (and they will acquire numbers in the usual
2090 way).
2091
2092 41. Redesigned the return codes from the match() function into yes/no/error so
2093 that errors can be passed back from deep inside the nested calls. A malloc
2094 failure while inside a recursive subpattern call now causes the
2095 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2096
2097 42. It is now possible to set a limit on the number of times the match()
2098 function is called in a call to pcre_exec(). This facility makes it possible to
2099 limit the amount of recursion and backtracking, though not in a directly
2100 obvious way, because the match() function is used in a number of different
2101 circumstances. The count starts from zero for each position in the subject
2102 string (for non-anchored patterns). The default limit is, for compatibility, a
2103 large number, namely 10 000 000. You can change this in two ways:
2104
2105 (a) When configuring PCRE before making, you can use --with-match-limit=n
2106 to set a default value for the compiled library.
2107
2108 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2109 a different value is set. See 45 below.
2110
2111 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2112
2113 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2114 of things that can be changed at compile time. The first argument specifies
2115 what is wanted and the second points to where the information is to be placed.
2116 The current list of available information is:
2117
2118 PCRE_CONFIG_UTF8
2119
2120 The output is an integer that is set to one if UTF-8 support is available;
2121 otherwise it is set to zero.
2122
2123 PCRE_CONFIG_NEWLINE
2124
2125 The output is an integer that it set to the value of the code that is used for
2126 newline. It is either LF (10) or CR (13).
2127
2128 PCRE_CONFIG_LINK_SIZE
2129
2130 The output is an integer that contains the number of bytes used for internal
2131 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2132
2133 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2134
2135 The output is an integer that contains the threshold above which the POSIX
2136 interface uses malloc() for output vectors. See item 31 above.
2137
2138 PCRE_CONFIG_MATCH_LIMIT
2139
2140 The output is an unsigned integer that contains the default limit of the number
2141 of match() calls in a pcre_exec() execution. See 42 above.
2142
2143 44. pcretest has been upgraded by the addition of the -C option. This causes it
2144 to extract all the available output from the new pcre_config() function, and to
2145 output it. The program then exits immediately.
2146
2147 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2148 order to support additional features. One way would have been to define
2149 pcre_exec2() (for example) with extra arguments, but this would not have been
2150 extensible, and would also have required all calls to the original function to
2151 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2152 is used for passing in "extra" data from pcre_study().
2153
2154 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2155 contains the following fields:
2156
2157 flags a bitmap indicating which of the following fields are set
2158 study_data opaque data from pcre_study()
2159 match_limit a way of specifying a limit on match() calls for a specific
2160 call to pcre_exec()
2161 callout_data data for callouts (see 49 below)
2162
2163 The flag bits are also defined in pcre.h, and are
2164
2165 PCRE_EXTRA_STUDY_DATA
2166 PCRE_EXTRA_MATCH_LIMIT
2167 PCRE_EXTRA_CALLOUT_DATA
2168
2169 The pcre_study() function now returns one of these new pcre_extra blocks, with
2170 the actual study data pointed to by the study_data field, and the
2171 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2172 before. That is, this change is entirely upwards-compatible and requires no
2173 change to existing code.
2174
2175 If you want to pass in additional data to pcre_exec(), you can either place it
2176 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2177 block.
2178
2179 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2180 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2181 times with different match limits, until it finds the minimum value needed for
2182 pcre_exec() to complete. The value is then output. This can be instructive; for
2183 most simple matches the number is quite small, but for pathological cases it
2184 gets very large very quickly.
2185
2186 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2187 returns the size of the data block pointed to by the study_data field in a
2188 pcre_extra block, that is, the value that was passed as the argument to
2189 pcre_malloc() when PCRE was getting memory in which to place the information
2190 created by pcre_study(). The fourth argument should point to a size_t variable.
2191 pcretest has been extended so that this information is shown after a successful
2192 pcre_study() call when information about the compiled regex is being displayed.
2193
2194 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2195 because what follows is always an absolute path. (Later: it turns out that this
2196 is more than cosmetic for MinGW, because it doesn't like empty path
2197 components.)
2198
2199 49. Some changes have been made to the callout feature (see 28 above):
2200
2201 (i) A callout function now has three choices for what it returns:
2202
2203 0 => success, carry on matching
2204 > 0 => failure at this point, but backtrack if possible
2205 < 0 => serious error, return this value from pcre_exec()
2206
2207 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2208 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2209 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2210 use by callout functions. It will never be used by PCRE itself.
2211
2212 (ii) The pcre_extra structure (see 45 above) has a void * field called
2213 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2214 pcre_callout_block structure has a field of the same name. The contents of
2215 the field passed in the pcre_extra structure are passed to the callout
2216 function in the corresponding field in the callout block. This makes it
2217 easier to use the same callout-containing regex from multiple threads. For
2218 testing, the pcretest program has a new data escape
2219
2220 \C*n pass the number n (may be negative) as callout_data
2221
2222 If the callout function in pcretest receives a non-zero value as
2223 callout_data, it returns that value.
2224
2225 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2226 there were some redundant $(CFLAGS) in commands that are now specified as
2227 $(LINK), which already includes $(CFLAGS).
2228
2229 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2230 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2231 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2232 one-byte characters throughout. Note that case-insensitive matching applies
2233 only to characters whose values are less than 256. PCRE doesn't support the
2234 notion of cases for higher-valued characters.
2235
2236 (i) A character class whose characters are all within 0-255 is handled as
2237 a bit map, and the map is inverted for negative classes. Previously, a
2238 character > 255 always failed to match such a class; however it should
2239 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2240
2241 (ii) A negated character class with a single character < 255 is coded as
2242 "not this character" (OP_NOT). This wasn't working properly when the test
2243 character was multibyte, either singly or repeated.
2244
2245 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2246 mode, for example: \x{100}{2,3}.
2247
2248 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2249 singly or repeated) now correctly test multibyte characters. However,
2250 PCRE doesn't recognize any characters with values greater than 255 as
2251 digits, spaces, or word characters. Such characters always match \D, \S,
2252 and \W, and never match \d, \s, or \w.
2253
2254 (v) Classes may now contain characters and character ranges with values
2255 greater than 255. For example: [ab\x{100}-\x{400}].
2256
2257 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2258 PCRE in UTF-8 mode.
2259
2260 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2261 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2262 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2263 value.)
2264
2265 53. The single man page has become too large. I have therefore split it up into
2266 a number of separate man pages. These also give rise to individual HTML pages;
2267 these are now put in a separate directory, and there is an index.html page that
2268 lists them all. Some hyperlinking between the pages has been installed.
2269
2270 54. Added convenience functions for handling named capturing parentheses.
2271
2272 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2273 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2274 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2275 are faulted.
2276
2277 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2278 calling configure. These values are used when compiling the dftables.c program
2279 which is run to generate the source of the default character tables. They
2280 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2281 you will need to set these values.
2282
2283 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2284
2285
2286 Version 3.9 02-Jan-02
2287 ---------------------
2288
2289 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2290
2291 2. If --disable-static was given, the building process failed when trying to
2292 build pcretest and pcregrep. (For some reason it was using libtool to compile
2293 them, which is not right, as they aren't part of the library.)
2294
2295
2296 Version 3.8 18-Dec-01
2297 ---------------------
2298
2299 1. The experimental UTF-8 code was completely screwed up. It was packing the
2300 bytes in the wrong order. How dumb can you get?
2301
2302
2303 Version 3.7 29-Oct-01
2304 ---------------------
2305
2306 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2307 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2308 this didn't happen under Solaris 8, where I normally test things.
2309
2310 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2311 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2312 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2313 this fix an hour or so after the initial 3.7 release.)
2314
2315
2316 Version 3.6 23-Oct-01
2317 ---------------------
2318
2319 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2320 offsets passed as NULL with zero offset count.
2321
2322 2. The config.guess and config.sub files had not been updated when I moved to
2323 the latest autoconf.
2324
2325
2326 Version 3.5 15-Aug-01
2327 ---------------------
2328
2329 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2330 had been forgotten.
2331
2332 2. By using declared but undefined structures, we can avoid using "void"
2333 definitions in pcre.h while keeping the internal definitions of the structures
2334 private.
2335
2336 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2337 user point of view, this means that both static and shared libraries are built
2338 by default, but this can be individually controlled. More of the work of
2339 handling this static/shared cases is now inside libtool instead of PCRE's make
2340 file.
2341
2342 4. The pcretest utility is now installed along with pcregrep because it is
2343 useful for users (to test regexs) and by doing this, it automatically gets
2344 relinked by libtool. The documentation has been turned into a man page, so
2345 there are now .1, .txt, and .html versions in /doc.
2346
2347 5. Upgrades to pcregrep:
2348 (i) Added long-form option names like gnu grep.
2349 (ii) Added --help to list all options with an explanatory phrase.
2350 (iii) Added -r, --recursive to recurse into sub-directories.
2351 (iv) Added -f, --file to read patterns from a file.
2352
2353 6. pcre_exec() was referring to its "code" argument before testing that
2354 argument for NULL (and giving an error if it was NULL).
2355
2356 7. Upgraded Makefile.in to allow for compiling in a different directory from
2357 the source directory.
2358
2359 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2360 options bits, the pointer it was passed was to an int instead of to an unsigned
2361 long int. This mattered only on 64-bit systems.
2362
2363 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2364 generated) instead of pcre.in, which it its source. Also made the same change
2365 in several of the .c files.
2366
2367 10. A new release of gcc defines printf() as a macro, which broke pcretest
2368 because it had an ifdef in the middle of a string argument for printf(). Fixed
2369 by using separate calls to printf().
2370
2371 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2372 script, to force use of CR or LF instead of \n in the source. On non-Unix
2373 systems, the value can be set in config.h.
2374
2375 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2376 absolute limit. Changed the text of the error message to make this clear, and
2377 likewise updated the man page.
2378
2379 13. The limit of 99 on the number of capturing subpatterns has been removed.
2380 The new limit is 65535, which I hope will not be a "real" limit.
2381
2382
2383 Version 3.4 22-Aug-00
2384 ---------------------
2385
2386 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2387
2388 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2389
2390
2391 Version 3.3 01-Aug-00
2392 ---------------------
2393
2394 1. If an octal character was given, but the value was greater than \377, it
2395 was not getting masked to the least significant bits, as documented. This could
2396 lead to crashes in some systems.
2397
2398 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2399 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2400
2401 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2402 These just pass their arguments on to (pcre_free)(), but they are provided
2403 because some uses of PCRE bind it to non-C systems that can call its functions,
2404 but cannot call free() or pcre_free() directly.
2405
2406 4. Add "make test" as a synonym for "make check". Corrected some comments in
2407 the Makefile.
2408
2409 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2410 Makefile.
2411
2412 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2413 command called pgrep for grepping around the active processes.
2414
2415 7. Added the beginnings of support for UTF-8 character strings.
2416
2417 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2418 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2419 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2420 out for the ar command.)
2421
2422
2423 Version 3.2 12-May-00
2424 ---------------------
2425
2426 This is purely a bug fixing release.
2427
2428 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2429 of ZA. This was just one example of several cases that could provoke this bug,
2430 which was introduced by change 9 of version 2.00. The code for breaking
2431 infinite loops after an iteration that matches an empty string was't working
2432 correctly.
2433
2434 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2435 when matched against abbab (for example). After matching an empty string, it
2436 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2437 caused it to match further down the string than it should.
2438
2439 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2440 was there because it doesn't seem to be needed, and it causes trouble on some
2441 systems, as it is not a Standard C header. It has been removed.
2442
2443 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2444 were reported on the Macintosh. The changes were from
2445
2446 while ((c = *(++ptr)) != 0 && c != '\n');
2447 to
2448 while ((c = *(++ptr)) != 0 && c != '\n') ;
2449
2450 Totally extraordinary, but if that's what it takes...
2451
2452 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2453 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2454 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2455 assumes the way PCRE uses memmove() (always moving upwards).
2456
2457 6. PCRE is being used in one environment where strchr() is not available. There
2458 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2459 faster code anyway.
2460
2461
2462 Version 3.1 09-Feb-00
2463 ---------------------
2464
2465 The only change in this release is the fixing of some bugs in Makefile.in for
2466 the "install" target:
2467
2468 (1) It was failing to install pcreposix.h.
2469
2470 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2471
2472
2473 Version 3.0 01-Feb-00
2474 ---------------------
2475
2476 1. Add support for the /+ modifier to perltest (to output $` like it does in
2477 pcretest).
2478
2479 2. Add support for the /g modifier to perltest.
2480
2481 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2482 matches null strings.
2483
2484 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2485 pattern. Perl treats empty patterns specially - it reuses the most recent
2486 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2487 effect.
2488
2489 5. The POSIX interface was broken in that it was just handing over the POSIX
2490 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2491 required a bigger vector, with some working space on the end. This means that
2492 the POSIX wrapper now has to get and free some memory, and copy the results.
2493
2494 6. Added some simple autoconf support, placing the test data and the
2495 documentation in separate directories, re-organizing some of the
2496 information files, and making it build pcre-config (a GNU standard). Also added
2497 libtool support for building PCRE as a shared library, which is now the
2498 default.
2499
2500 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2501 09 are not valid octal constants. Single digits will be used for minor values
2502 less than 10.
2503
2504 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2505 existing programs that set these in the POSIX interface can use PCRE without
2506 modification.
2507
2508 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2509 return all that pcre_info() returns, plus additional data. The pcre_info()
2510 function is retained for compatibility, but is considered to be obsolete.
2511
2512 10. Added experimental recursion feature (?R) to handle one common case that
2513 Perl 5.6 will be able to do with (?p{...}).
2514
2515 11. Added support for POSIX character classes like [:alpha:], which Perl is
2516 adopting.
2517
2518
2519 Version 2.08 31-Aug-99
2520 ----------------------
2521
2522 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2523 trying to match at the startoffset position, but instead was moving forward to
2524 the next newline as if a previous match had failed.
2525
2526 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2527 and could get into a loop if a null string was matched other than at the start
2528 of the subject.
2529
2530 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2531 be distinguished at compile time, and for completeness also added PCRE_DATE.
2532
2533 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2534 in GnuWin32 environments.
2535
2536
2537 Version 2.07 29-Jul-99
2538 ----------------------
2539
2540 1. The documentation is now supplied in plain text form and HTML as well as in
2541 the form of man page sources.
2542
2543 2. C++ compilers don't like assigning (void *) values to other pointer types.
2544 In particular this affects malloc(). Although there is no problem in Standard
2545 C, I've put in casts to keep C++ compilers happy.
2546
2547 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2548 should be (const char *).
2549
2550 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2551 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2552 However, I haven't made this a standard facility. The documentation doesn't
2553 mention it, and the Makefile doesn't support it.
2554
2555 5. The Makefile now contains an "install" target, with editable destinations at
2556 the top of the file. The pcretest program is not installed.
2557
2558 6. pgrep -V now gives the PCRE version number and date.
2559
2560 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2561 causing the entire string to be ignored, instead of just the last character.
2562
2563 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2564 non-matching string, it can take a very, very long time, even for strings of
2565 quite modest length, because of the nested recursion. PCRE now does better in
2566 some of these cases. It does this by remembering the last required literal
2567 character in the pattern, and pre-searching the subject to ensure it is present
2568 before running the real match. In other words, it applies a heuristic to detect
2569 some types of certain failure quickly, and in the above example, if presented
2570 with a string that has no trailing " it gives "no match" very quickly.
2571
2572 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2573 other alternatives are tried instead.
2574
2575
2576 Version 2.06 09-Jun-99
2577 ----------------------
2578
2579 1. Change pcretest's output for amount of store used to show just the code
2580 space, because the remainder (the data block) varies in size between 32-bit and
2581 64-bit systems.
2582
2583 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2584 start matching at. This allows lookbehinds to work when searching for multiple
2585 occurrences in a string.
2586
2587 3. Added additional options to pcretest for testing multiple occurrences:
2588
2589 /+ outputs the rest of the string that follows a match
2590 /g loops for multiple occurrences, using the new startoffset argument
2591 /G loops for multiple occurrences by passing an incremented pointer
2592
2593 4. PCRE wasn't doing the "first character" optimization for patterns starting
2594 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2595 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2596 the letter 'x'. On long subject strings, this gives a significant speed-up.
2597
2598
2599 Version 2.05 21-Apr-99
2600 ----------------------
2601
2602 1. Changed the type of magic_number from int to long int so that it works
2603 properly on 16-bit systems.
2604
2605 2. Fixed a bug which caused patterns starting with .* not to work correctly
2606 when the subject string contained newline characters. PCRE was assuming
2607 anchoring for such patterns in all cases, which is not correct because .* will
2608 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2609 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2610 must be retried after every newline in the subject.
2611
2612
2613 Version 2.04 18-Feb-99
2614 ----------------------
2615
2616 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2617 computation of the store needed to hold the pattern was incorrect (too large).
2618 If such patterns were nested a few deep, this could multiply and become a real
2619 problem.
2620
2621 2. Added /M option to pcretest to show the memory requirement of a specific
2622 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2623
2624 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2625 compiled in such a way that the backtracking after subsequent failure was
2626 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2627 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2628
2629
2630 Version 2.03 02-Feb-99
2631 ----------------------
2632
2633 1. Fixed typo and small mistake in man page.
2634
2635 2. Added 4th condition (GPL supersedes if conflict) and created separate
2636 LICENCE file containing the conditions.
2637
2638 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2639 Perl, that is the internal \ allows the delimiter to be included in the
2640 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2641 the final delimiter, add \ to the end of the pattern (to test the error).
2642
2643 4. Added the convenience functions for extracting substrings after a successful
2644 match. Updated pcretest to make it able to test these functions.
2645
2646
2647 Version 2.02 14-Jan-99
2648 ----------------------
2649
2650 1. Initialized the working variables associated with each extraction so that
2651 their saving and restoring doesn't refer to uninitialized store.
2652
2653 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2654 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2655 fix the problem.
2656
2657 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2658 calls, and wasn't printing the correct value for compiling calls. Increased the
2659 default value of LOOPREPEAT, and the number of significant figures in the
2660 times.
2661
2662 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2663
2664 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2665 a building problem on Windows NT with a FAT file system.
2666
2667
2668 Version 2.01 21-Oct-98
2669 ----------------------
2670
2671 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2672 to character tables built by pcre_maketables() in the current locale. If NULL
2673 is passed, the default tables are used.
2674
2675
2676 Version 2.00 24-Sep-98
2677 ----------------------
2678
2679 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2680 it any more.
2681
2682 2. Allow quantification of (?>) groups, and make it work correctly.
2683
2684 3. The first character computation wasn't working for (?>) groups.
2685
2686 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2687 end of the subject) and add 5.005's \z, which really does match only at the
2688 very end of the subject.
2689
2690 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2691
2692 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2693 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2694 localized options. All options to pcre_study() were also removed.
2695
2696 7. Add other new features from 5.005:
2697
2698 $(?<= positive lookbehind
2699 $(?<! negative lookbehind
2700 (?imsx-imsx) added the unsetting capability
2701 such a setting is global if at outer level; local otherwise
2702 (?imsx-imsx:) non-capturing groups with option setting
2703 (?(cond)re|re) conditional pattern matching
2704
2705 A backreference to itself in a repeated group matches the previous
2706 captured string.
2707
2708 8. General tidying up of studying (both automatic and via "study")
2709 consequential on the addition of new assertions.
2710
2711 9. As in 5.005, unlimited repeated groups that could match an empty substring
2712 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2713 runtime if any iteration does actually match an empty substring.
2714
2715 10. Include the RunTest script in the distribution.
2716
2717 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2718 discrepancies, some of which were old and were also with respect to 5.004. They
2719 have now been fixed.
2720
2721
2722 Version 1.09 28-Apr-98
2723 ----------------------
2724
2725 1. A negated single character class followed by a quantifier with a minimum
2726 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2727 program crashes, or just wrong answers. This did not apply to negated classes
2728 containing more than one character, or to minima other than one.
2729
2730
2731 Version 1.08 27-Mar-98
2732 ----------------------
2733
2734 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2735
2736 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2737 latter must appear before anything that relies on it in the pattern.
2738
2739
2740 Version 1.07 16-Feb-98
2741 ----------------------
2742
2743 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2744 repeat of a potentially empty string).
2745
2746
2747 Version 1.06 23-Jan-98
2748 ----------------------
2749
2750 1. Added Markus Oberhumer's little patches for C++.
2751
2752 2. Literal strings longer than 255 characters were broken.
2753
2754
2755 Version 1.05 23-Dec-97
2756 ----------------------
2757
2758 1. Negated character classes containing more than one character were failing if
2759 PCRE_CASELESS was set at run time.
2760
2761
2762 Version 1.04 19-Dec-97
2763 ----------------------
2764
2765 1. Corrected the man page, where some "const" qualifiers had been omitted.
2766
2767 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2768 input syntax.
2769
2770 3. Fixed memory leak which occurred when a regex with back references was
2771 matched with an offsets vector that wasn't big enough. The temporary memory
2772 that is used in this case wasn't being freed if the match failed.
2773
2774 4. Tidied pcretest to ensure it frees memory that it gets.
2775
2776 5. Temporary memory was being obtained in the case where the passed offsets
2777 vector was exactly big enough.
2778
2779 6. Corrected definition of offsetof() from change 5 below.
2780
2781 7. I had screwed up change 6 below and broken the rules for the use of
2782 setjmp(). Now fixed.
2783
2784
2785 Version 1.03 18-Dec-97
2786 ----------------------
2787
2788 1. A erroneous regex with a missing opening parenthesis was correctly
2789 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2790 on some systems.
2791
2792 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2793 it was reported that one broken compiler failed on the former because "code" is
2794 also an independent variable.
2795
2796 3. The erroneous regex a[]b caused an array overrun reference.
2797
2798 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2799 fail on data ending with that character. (It was going on too far, and checking
2800 the next character, typically a binary zero.) This was specific to the
2801 optimized code for single-character negative classes.
2802
2803 5. Added a contributed patch from the TIN world which does the following:
2804
2805 + Add an undef for memmove, in case the the system defines a macro for it.
2806
2807 + Add a definition of offsetof(), in case there isn't one. (I don't know
2808 the reason behind this - offsetof() is part of the ANSI standard - but
2809 it does no harm).
2810
2811 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2812 most of the places where whitespace preceded '#'. I have given up and
2813 allowed the remaining 2 cases to be at the margin.
2814
2815 + Rename some variables in pcre to eliminate shadowing. This seems very
2816 pedantic, but does no harm, of course.
2817
2818 6. Moved the call to setjmp() into its own function, to get rid of warnings
2819 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2820
2821 7. Constructs such as \d{8,} were compiling into the equivalent of
2822 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2823 outcome, but in this particular case used more store than had been allocated,
2824 which caused the bug to be discovered because it threw up an internal error.
2825
2826 8. The debugging code in both pcre and pcretest for outputting the compiled
2827 form of a regex was going wrong in the case of back references followed by
2828 curly-bracketed repeats.
2829
2830
2831 Version 1.02 12-Dec-97
2832 ----------------------
2833
2834 1. Typos in pcre.3 and comments in the source fixed.
2835
2836 2. Applied a contributed patch to get rid of places where it used to remove
2837 'const' from variables, and fixed some signed/unsigned and uninitialized
2838 variable warnings.
2839
2840 3. Added the "runtest" target to Makefile.
2841
2842 4. Set default compiler flag to -O2 rather than just -O.
2843
2844
2845 Version 1.01 19-Nov-97
2846 ----------------------
2847
2848 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2849 like /([ab]*)*/, that is, for classes with more than one character in them.
2850
2851 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2852 as /((?>a*))*/ (a PCRE_EXTRA facility).
2853
2854
2855 Version 1.00 18-Nov-97
2856 ----------------------
2857
2858 1. Added compile-time macros to support systems such as SunOS4 which don't have
2859 memmove() or strerror() but have other things that can be used instead.
2860
2861 2. Arranged that "make clean" removes the executables.
2862
2863
2864 Version 0.99 27-Oct-97
2865 ----------------------
2866
2867 1. Fixed bug in code for optimizing classes with only one character. It was
2868 initializing a 32-byte map regardless, which could cause it to run off the end
2869 of the memory it had got.
2870
2871 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2872
2873
2874 Version 0.98 22-Oct-97
2875 ----------------------
2876
2877 1. Fixed bug in code for handling temporary memory usage when there are more
2878 back references than supplied space in the ovector. This could cause segfaults.
2879
2880
2881 Version 0.97 21-Oct-97
2882 ----------------------
2883
2884 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2885
2886 2. Optimized negated single characters not to use a bit map.
2887
2888 3. Brought error texts together as macro definitions; clarified some of them;
2889 fixed one that was wrong - it said "range out of order" when it meant "invalid
2890 escape sequence".
2891
2892 4. Changed some char * arguments to const char *.
2893
2894 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2895
2896 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2897 pcretest.
2898
2899
2900 Version 0.96 16-Oct-97
2901 ----------------------
2902
2903 1. Added a simple "pgrep" utility to the distribution.
2904
2905 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2906 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2907 where "ddd" means "one or more decimal digits".
2908
2909 3. Fixed serious bug. If a pattern had a back reference, but the call to
2910 pcre_exec() didn't supply a large enough ovector to record the related
2911 identifying subpattern, the match always failed. PCRE now remembers the number
2912 of the largest back reference, and gets some temporary memory in which to save
2913 the offsets during matching if necessary, in order to ensure that
2914 backreferences always work.
2915
2916 4. Increased the compatibility with Perl in a number of ways:
2917
2918 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2919 to request this handling. The option can be set at compile or exec time.
2920
2921 (b) $ matches before a terminating newline by default; an option
2922 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2923 mode). The option can be set at compile or exec time.
2924
2925 (c) The handling of \ followed by a digit other than 0 is now supposed to be
2926 the same as Perl's. If the decimal number it represents is less than 10
2927 or there aren't that many previous left capturing parentheses, an octal
2928 escape is read. Inside a character class, it's always an octal escape,
2929 even if it is a single digit.
2930
2931 (d) An escaped but undefined alphabetic character is taken as a literal,
2932 unless PCRE_EXTRA is set. Currently this just reserves the remaining
2933 escapes.
2934
2935 (e) {0} is now permitted. (The previous item is removed from the compiled
2936 pattern).
2937
2938 5. Changed all the names of code files so that the basic parts are no longer
2939 than 10 characters, and abolished the teeny "globals.c" file.
2940
2941 6. Changed the handling of character classes; they are now done with a 32-byte
2942 bit map always.
2943
2944 7. Added the -d and /D options to pcretest to make it possible to look at the
2945 internals of compilation without having to recompile pcre.
2946
2947
2948 Version 0.95 23-Sep-97
2949 ----------------------
2950
2951 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2952 \x20 at the start of a run of normal characters. These were being treated as
2953 real characters, instead of the source characters being re-checked.
2954
2955
2956 Version 0.94 18-Sep-97
2957 ----------------------
2958
2959 1. The functions are now thread-safe, with the caveat that the global variables
2960 containing pointers to malloc() and free() or alternative functions are the
2961 same for all threads.
2962
2963 2. Get pcre_study() to generate a bitmap of initial characters for non-
2964 anchored patterns when this is possible, and use it if passed to pcre_exec().
2965
2966
2967 Version 0.93 15-Sep-97
2968 ----------------------
2969
2970 1. /(b)|(:+)/ was computing an incorrect first character.
2971
2972 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2973 but not actually doing anything yet.
2974
2975 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2976 as Perl does (e.g. [-az] or [az-]).
2977
2978 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2979 all possible positions.
2980
2981 5. Split up into different modules to avoid including unneeded functions in a
2982 compiled binary. However, compile and exec are still in one module. The "study"
2983 function is split off.
2984
2985 6. The character tables are now in a separate module whose source is generated
2986 by an auxiliary program - but can then be edited by hand if required. There are
2987 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2988 toupper() in the code.
2989
2990 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2991 make them global. Abolish the function for setting them, as the caller can now
2992 set them directly.
2993
2994
2995 Version 0.92 11-Sep-97
2996 ----------------------
2997
2998 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2999 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3000
3001 2. Caseless matching was not working in character classes if the characters in
3002 the pattern were in upper case.
3003
3004 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3005
3006 4. Make PCRE_ANCHORED public and accept as a compile option.
3007
3008 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3009 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3010 pass them.
3011
3012 6. Give an error if bad option bits passed at compile or run time.
3013
3014 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3015 pcretest to cause it to pass that flag.
3016
3017 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3018 options, and the first character, if set.
3019
3020 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3021
3022
3023 Version 0.91 10-Sep-97
3024 ----------------------
3025
3026 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3027 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3028
3029 2. PCRE was looping on encountering an indefinitely repeated back reference to
3030 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3031 Perl does - treats the match as successful.
3032
3033 ****

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