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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 239 - (show annotations) (download)
Tue Sep 11 13:46:30 2007 UTC (6 years, 11 months ago) by ph10
File size: 140464 byte(s)
Fix line ending missing for pcregrep with --colour.

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

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