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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 236 - (show annotations) (download)
Tue Sep 11 12:57:06 2007 UTC (7 years, 2 months ago) by ph10
File size: 140336 byte(s)
<config.h> => "config.h" and also some cases of <pcre.h>.

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

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