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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 295 - (show annotations) (download)
Mon Dec 31 17:00:24 2007 UTC (6 years, 8 months ago) by ph10
File size: 148260 byte(s)
Make POSIX character class parsing more like Perl.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.5 31-Dec-07
5 ---------------------
6
7 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
8 values in parens when parsing an RE using the C++ wrapper."
9
10 2. Negative specials like \S did not work in character classes in UTF-8 mode.
11 Characters greater than 255 were excluded from the class instead of being
12 included.
13
14 3. The same bug as (2) above applied to negated POSIX classes such as
15 [:^space:].
16
17 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
18 defined or documented. It seems to have been a typo for PCRE_STATIC, so
19 I have changed it.
20
21 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
22 first named subpattern) and a construct such as (?&a) would reference the
23 first named subpattern whose name started with "a" (in other words, the
24 length check was missing). Both these problems are fixed. "Subpattern name
25 expected" is now given for (?&) (a zero-length name), and this patch also
26 makes it give the same error for \k'' (previously it complained that that
27 was a reference to a non-existent subpattern).
28
29 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
30 this is right because (?- can be followed by option settings as well as by
31 digits. I have, however, made the messages clearer.
32
33 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
34 than the number used in the conditional) now cause a compile-time error.
35 This is actually not compatible with Perl, which accepts such patterns, but
36 treats the conditional as always being FALSE (as PCRE used to), but it
37 seems to me that giving a diagnostic is better.
38
39 8. Change "alphameric" to the more common word "alphanumeric" in comments
40 and messages.
41
42 9. Fix two occurrences of "backslash" in comments that should have been
43 "backspace".
44
45 10. Remove two redundant lines of code that can never be obeyed (their function
46 was moved elsewhere).
47
48 11. The program that makes PCRE's Unicode character property table had a bug
49 which caused it to generate incorrect table entries for sequences of
50 characters that have the same character type, but are in different scripts.
51 It amalgamated them into a single range, with the script of the first of
52 them. In other words, some characters were in the wrong script. There were
53 thirteen such cases, affecting characters in the following ranges:
54
55 U+002b0 - U+002c1
56 U+0060c - U+0060d
57 U+0061e - U+00612
58 U+0064b - U+0065e
59 U+0074d - U+0076d
60 U+01800 - U+01805
61 U+01d00 - U+01d77
62 U+01d9b - U+01dbf
63 U+0200b - U+0200f
64 U+030fc - U+030fe
65 U+03260 - U+0327f
66 U+0fb46 - U+0fbb1
67 U+10450 - U+1049d
68
69 12. The -o option (show only the matching part of a line) for pcregrep was not
70 compatible with GNU grep in that, if there was more than one match in a
71 line, it showed only the first of them. It now behaves in the same way as
72 GNU grep.
73
74 13. If the -o and -v options were combined for pcregrep, it printed a blank
75 line for every non-matching line. GNU grep prints nothing, and pcregrep now
76 does the same. The return code can be used to tell if there were any
77 non-matching lines.
78
79 14. Added --file-offsets and --line-offsets to pcregrep.
80
81 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
82 infinitely looping recursion. The bug was that positive lookaheads were not
83 being skipped when checking for a possible empty match (negative lookaheads
84 and both kinds of lookbehind were skipped).
85
86 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
87 inclusion of <windows.h> to before rather than after the definition of
88 INVALID_FILE_ATTRIBUTES (patch from David Byron).
89
90 17. Specifying a possessive quantifier with a specific limit for a Unicode
91 character property caused pcre_compile() to compile bad code, which led at
92 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
93 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
94 caused the error; without that there was no problem.
95
96 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
97
98 19. Added --enable-pcretest-libreadline.
99
100 20. In pcrecpp.cc, the variable 'count' was incremented twice in
101 RE::GlobalReplace(). As a result, the number of replacements returned was
102 double what it should be. I have removed one of the increments.
103
104 21. Several CMake things:
105
106 (1) Arranged that, when cmake is used on Unix, the libraries end up with
107 the names libpcre and libpcreposix, not just pcre and pcreposix.
108
109 (2) The above change means that pcretest and pcregrep are now correctly
110 linked with the newly-built libraries, not previously installed ones.
111
112 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
113
114 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
115 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
116 UTF-8 newline character). The key issue is that the pattern starts .*;
117 this means that the match must be either at the beginning, or after a
118 newline. The bug was in the code for advancing after a failed match and
119 checking that the new position followed a newline. It was not taking
120 account of UTF-8 characters correctly.
121
122 23. PCRE was behaving differently from Perl in the way it recognized POSIX
123 character classes. PCRE was not treating the sequence [:...:] as a
124 character class unless the ... were all letters. Perl, however, seems to
125 allow any characters between [: and :], though of course it rejects as
126 unknown any "names" that contain non-letters, because all the known class
127 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
128 for example, whereas PCRE did not - it did not recognize a POSIX character
129 class. This seemed a bit dangerous, so the code has been changed to be
130 closer to Perl. The behaviour is not identical to Perl, because PCRE will
131 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
132 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
133 Perl does, and where it didn't before.
134
135
136 Version 7.4 21-Sep-07
137 ---------------------
138
139 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
140 means that a class such as [\s] counted as "explicit reference to CR or
141 LF". That isn't really right - the whole point of the change was to try to
142 help when there was an actual mention of one of the two characters. So now
143 the change happens only if \r or \n (or a literal CR or LF) character is
144 encountered.
145
146 2. The 32-bit options word was also used for 6 internal flags, but the numbers
147 of both had grown to the point where there were only 3 bits left.
148 Fortunately, there was spare space in the data structure, and so I have
149 moved the internal flags into a new 16-bit field to free up more option
150 bits.
151
152 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
153 but did not set the internal JCHANGED flag - either of these is enough to
154 control the way the "get" function works - but the PCRE_INFO_JCHANGED
155 facility is supposed to tell if (?J) was ever used, so now (?J) at the
156 start sets both bits.
157
158 4. Added options (at build time, compile time, exec time) to change \R from
159 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
160
161 5. doc/pcresyntax.html was missing from the distribution.
162
163 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
164 compatibility, even though it is no longer used.
165
166 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
167 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
168 windows.h file is present (where different names are used). [This was
169 reversed later after testing - see 16 below.]
170
171 8. Changed all #include <config.h> to #include "config.h". There were also
172 some further <pcre.h> cases that I changed to "pcre.h".
173
174 9. When pcregrep was used with the --colour option, it missed the line ending
175 sequence off the lines that it output.
176
177 10. It was pointed out to me that arrays of string pointers cause lots of
178 relocations when a shared library is dynamically loaded. A technique of
179 using a single long string with a table of offsets can drastically reduce
180 these. I have refactored PCRE in four places to do this. The result is
181 dramatic:
182
183 Originally: 290
184 After changing UCP table: 187
185 After changing error message table: 43
186 After changing table of "verbs" 36
187 After changing table of Posix names 22
188
189 Thanks to the folks working on Gregex for glib for this insight.
190
191 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
192 unicode-properties was also set.
193
194 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
195
196 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
197 checked only for CRLF.
198
199 14. Added casts to pcretest.c to avoid compiler warnings.
200
201 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
202
203 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
204 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
205 entirely. This removes changes made in 7 above.
206
207 17. The CMake files have been updated, and there is now more information about
208 building with CMake in the NON-UNIX-USE document.
209
210
211 Version 7.3 28-Aug-07
212 ---------------------
213
214 1. In the rejigging of the build system that eventually resulted in 7.1, the
215 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
216 brackets there is not right, since it causes compilers to look for an
217 installed pcre.h, not the version that is in the source that is being
218 compiled (which of course may be different). I have changed it back to:
219
220 #include "pcre.h"
221
222 I have a vague recollection that the change was concerned with compiling in
223 different directories, but in the new build system, that is taken care of
224 by the VPATH setting the Makefile.
225
226 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
227 when the subject happened to end in the byte 0x85 (e.g. if the last
228 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
229 characters but of course it shouldn't be taken as a newline when it is part
230 of another character. The bug was that, for an unlimited repeat of . in
231 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
232 characters when looking for a newline.
233
234 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
235
236 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
237 in debug output.
238
239 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
240 long printing in the pcrecpp unittest when running under MinGW.
241
242 6. ESC_K was left out of the EBCDIC table.
243
244 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
245 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
246 limit also applies to "virtual nesting" when a pattern is recursive, and in
247 this case 1000 isn't so big. I have been able to remove this limit at the
248 expense of backing off one optimization in certain circumstances. Normally,
249 when pcre_exec() would call its internal match() function recursively and
250 immediately return the result unconditionally, it uses a "tail recursion"
251 feature to save stack. However, when a subpattern that can match an empty
252 string has an unlimited repetition quantifier, it no longer makes this
253 optimization. That gives it a stack frame in which to save the data for
254 checking that an empty string has been matched. Previously this was taken
255 from the 1000-entry workspace that had been reserved. So now there is no
256 explicit limit, but more stack is used.
257
258 8. Applied Daniel's patches to solve problems with the import/export magic
259 syntax that is required for Windows, and which was going wrong for the
260 pcreposix and pcrecpp parts of the library. These were overlooked when this
261 problem was solved for the main library.
262
263 9. There were some crude static tests to avoid integer overflow when computing
264 the size of patterns that contain repeated groups with explicit upper
265 limits. As the maximum quantifier is 65535, the maximum group length was
266 set at 30,000 so that the product of these two numbers did not overflow a
267 32-bit integer. However, it turns out that people want to use groups that
268 are longer than 30,000 bytes (though not repeat them that many times).
269 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
270 made it possible to implement the integer overflow checks in a much more
271 dynamic way, which I have now done. The artificial limitation on group
272 length has been removed - we now have only the limit on the total length of
273 the compiled pattern, which depends on the LINK_SIZE setting.
274
275 10. Fixed a bug in the documentation for get/copy named substring when
276 duplicate names are permitted. If none of the named substrings are set, the
277 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
278 empty string.
279
280 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
281 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
282 because the ] is interpreted as the first data character and the
283 terminating ] is not found. PCRE has been made compatible with Perl in this
284 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
285 cause memory overwriting.
286
287 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
288 string has been matched (to stop an infinite loop). It was not recognizing
289 a conditional subpattern that could match an empty string if that
290 subpattern was within another subpattern. For example, it looped when
291 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
292 condition was not nested. This bug has been fixed.
293
294 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
295 past the start of the subject in the presence of bytes with the top bit
296 set, for example "\x8aBCD".
297
298 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
299 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
300
301 14. Optimized (?!) to (*FAIL).
302
303 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
304 This restricts code points to be within the range 0 to 0x10FFFF, excluding
305 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
306 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
307 does: it's just the validity check that is more restrictive.
308
309 16. Inserted checks for integer overflows during escape sequence (backslash)
310 processing, and also fixed erroneous offset values for syntax errors during
311 backslash processing.
312
313 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
314 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
315
316 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
317 caused an overrun.
318
319 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
320 something other than just ASCII characters) inside a group that had an
321 unlimited repeat caused a loop at compile time (while checking to see
322 whether the group could match an empty string).
323
324 20. Debugging a pattern containing \p or \P could cause a crash. For example,
325 [\P{Any}] did so. (Error in the code for printing property names.)
326
327 21. An orphan \E inside a character class could cause a crash.
328
329 22. A repeated capturing bracket such as (A)? could cause a wild memory
330 reference during compilation.
331
332 23. There are several functions in pcre_compile() that scan along a compiled
333 expression for various reasons (e.g. to see if it's fixed length for look
334 behind). There were bugs in these functions when a repeated \p or \P was
335 present in the pattern. These operators have additional parameters compared
336 with \d, etc, and these were not being taken into account when moving along
337 the compiled data. Specifically:
338
339 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
340 length.
341
342 (b) An item such as \pL+ within a repeated group could cause crashes or
343 loops.
344
345 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
346 "reference to non-existent subpattern" error.
347
348 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
349
350 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
351 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
352
353 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
354
355 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
356 character were causing crashes (broken optimization).
357
358 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
359 \p or \P) caused a compile-time loop.
360
361 28. More problems have arisen in unanchored patterns when CRLF is a valid line
362 break. For example, the unstudied pattern [\r\n]A does not match the string
363 "\r\nA" because change 7.0/46 below moves the current point on by two
364 characters after failing to match at the start. However, the pattern \nA
365 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
366 the same is true. There doesn't seem any very clean way out of this, but
367 what I have chosen to do makes the common cases work: PCRE now takes note
368 of whether there can be an explicit match for \r or \n anywhere in the
369 pattern, and if so, 7.0/46 no longer applies. As part of this change,
370 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
371 pattern has explicit CR or LF references.
372
373 29. Added (*CR) etc for changing newline setting at start of pattern.
374
375
376 Version 7.2 19-Jun-07
377 ---------------------
378
379 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
380 which is apparently normally available under Windows.
381
382 2. Re-jig the pcregrep tests with different newline settings in an attempt
383 to make them independent of the local environment's newline setting.
384
385 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
386
387 4. Some of the "internals" tests were previously cut out when the link size
388 was not 2, because the output contained actual offsets. The recent new
389 "Z" feature of pcretest means that these can be cut out, making the tests
390 usable with all link sizes.
391
392 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
393 stack recursion. This gives a massive performance boost under BSD, but just
394 a small improvement under Linux. However, it saves one field in the frame
395 in all cases.
396
397 6. Added more features from the forthcoming Perl 5.10:
398
399 (a) (?-n) (where n is a string of digits) is a relative subroutine or
400 recursion call. It refers to the nth most recently opened parentheses.
401
402 (b) (?+n) is also a relative subroutine call; it refers to the nth next
403 to be opened parentheses.
404
405 (c) Conditions that refer to capturing parentheses can be specified
406 relatively, for example, (?(-2)... or (?(+3)...
407
408 (d) \K resets the start of the current match so that everything before
409 is not part of it.
410
411 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
412
413 (f) \g{name} is another synonym - part of Perl 5.10's unification of
414 reference syntax.
415
416 (g) (?| introduces a group in which the numbering of parentheses in each
417 alternative starts with the same number.
418
419 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
420
421 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
422 PCRE_INFO_JCHANGED.
423
424 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
425 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
426 for detecting groups that can match an empty string.
427
428 9. A pattern with a very large number of alternatives (more than several
429 hundred) was running out of internal workspace during the pre-compile
430 phase, where pcre_compile() figures out how much memory will be needed. A
431 bit of new cunning has reduced the workspace needed for groups with
432 alternatives. The 1000-alternative test pattern now uses 12 bytes of
433 workspace instead of running out of the 4096 that are available.
434
435 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
436
437 11. Applied patch from Google to remove an optimization that didn't quite work.
438 The report of the bug said:
439
440 pcrecpp::RE("a*").FullMatch("aaa") matches, while
441 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
442 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
443
444 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
445 it matched the wrong number of bytes.
446
447
448 Version 7.1 24-Apr-07
449 ---------------------
450
451 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
452 that is more "standard", making use of automake and other Autotools. There
453 is some re-arrangement of the files and adjustment of comments consequent
454 on this.
455
456 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
457 for recursive directory scanning broke on some systems because the files
458 are not scanned in any specific order and on different systems the order
459 was different. A call to "sort" has been inserted into RunGrepTest for the
460 approprate test as a short-term fix. In the longer term there may be an
461 alternative.
462
463 3. I had an email from Eric Raymond about problems translating some of PCRE's
464 man pages to HTML (despite the fact that I distribute HTML pages, some
465 people do their own conversions for various reasons). The problems
466 concerned the use of low-level troff macros .br and .in. I have therefore
467 removed all such uses from the man pages (some were redundant, some could
468 be replaced by .nf/.fi pairs). The 132html script that I use to generate
469 HTML has been updated to handle .nf/.fi and to complain if it encounters
470 .br or .in.
471
472 4. Updated comments in configure.ac that get placed in config.h.in and also
473 arranged for config.h to be included in the distribution, with the name
474 config.h.generic, for the benefit of those who have to compile without
475 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
476
477 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
478 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
479 makevp.bat for the current PCRE, using the additional files
480 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
481
482 6. A Windows user reported a minor discrepancy with test 2, which turned out
483 to be caused by a trailing space on an input line that had got lost in his
484 copy. The trailing space was an accident, so I've just removed it.
485
486 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
487 that is needed.
488
489 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
490 as "const" (a) because they are and (b) because it helps the PHP
491 maintainers who have recently made a script to detect big data structures
492 in the php code that should be moved to the .rodata section. I remembered
493 to update Builducptable as well, so it won't revert if ucptable.h is ever
494 re-created.
495
496 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
497 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
498 order to be able to cut out the UTF-8 tables in the latter when UTF-8
499 support is not required. This saves 1.5-2K of code, which is important in
500 some applications.
501
502 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
503 so as not to refer to the tables, even though these functions will never be
504 called when UTF-8 support is disabled. Otherwise there are problems with a
505 shared library.
506
507 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
508
509 (a) It was defining its arguments as char * instead of void *.
510
511 (b) It was assuming that all moves were upwards in memory; this was true
512 a long time ago when I wrote it, but is no longer the case.
513
514 The emulated memove() is provided for those environments that have neither
515 memmove() nor bcopy(). I didn't think anyone used it these days, but that
516 is clearly not the case, as these two bugs were recently reported.
517
518 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
519 and Detrail to create the HTML documentation, the .txt form of the man
520 pages, and it removes trailing spaces from listed files. It also creates
521 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
522 case, it wraps all the #defines with #ifndefs. This script should be run
523 before "make dist".
524
525 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
526 with Unicode property support.
527
528 (a) For a maximizing quantifier, if the two different cases of the
529 character were of different lengths in their UTF-8 codings (there are
530 some cases like this - I found 11), and the matching function had to
531 back up over a mixture of the two cases, it incorrectly assumed they
532 were both the same length.
533
534 (b) When PCRE was configured to use the heap rather than the stack for
535 recursion during matching, it was not correctly preserving the data for
536 the other case of a UTF-8 character when checking ahead for a match
537 while processing a minimizing repeat. If the check also involved
538 matching a wide character, but failed, corruption could cause an
539 erroneous result when trying to check for a repeat of the original
540 character.
541
542 13. Some tidying changes to the testing mechanism:
543
544 (a) The RunTest script now detects the internal link size and whether there
545 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
546 values substituted by "configure". (The RunGrepTest script already did
547 this for UTF-8.) The configure.ac script no longer substitutes the
548 relevant variables.
549
550 (b) The debugging options /B and /D in pcretest show the compiled bytecode
551 with length and offset values. This means that the output is different
552 for different internal link sizes. Test 2 is skipped for link sizes
553 other than 2 because of this, bypassing the problem. Unfortunately,
554 there was also a test in test 3 (the locale tests) that used /B and
555 failed for link sizes other than 2. Rather than cut the whole test out,
556 I have added a new /Z option to pcretest that replaces the length and
557 offset values with spaces. This is now used to make test 3 independent
558 of link size. (Test 2 will be tidied up later.)
559
560 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
561 segmentation fault instead of returning the appropriate error message.
562
563 15. In multiline mode when the newline sequence was set to "any", the pattern
564 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
565 This doesn't seem right; it now treats the CRLF combination as the line
566 ending, and so does not match in that case. It's only a pattern such as ^$
567 that would hit this one: something like ^ABC$ would have failed after \r
568 and then tried again after \r\n.
569
570 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
571 in an attempt to make files that differ only in their line terminators
572 compare equal. This works on Linux.
573
574 17. Under certain error circumstances pcregrep might try to free random memory
575 as it exited. This is now fixed, thanks to valgrind.
576
577 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
578 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
579 was because its rules for how to advance for /g after matching an empty
580 string at the end of a line did not allow for this case. They now check for
581 it specially.
582
583 20. pcretest is supposed to handle patterns and data of any length, by
584 extending its buffers when necessary. It was getting this wrong when the
585 buffer for a data line had to be extended.
586
587 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
588 CRLF as a newline sequence.
589
590 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
591 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
592 I have nevertheless tidied it up.
593
594 23. Added some casts to kill warnings from HP-UX ia64 compiler.
595
596 24. Added a man page for pcre-config.
597
598
599 Version 7.0 19-Dec-06
600 ---------------------
601
602 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
603 moving to gcc 4.1.1.
604
605 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
606 sys/time.h, which is documented as needed for this function. It doesn't
607 seem to matter on Linux, but it showed up on some releases of OS X.
608
609 3. It seems that there are systems where bytes whose values are greater than
610 127 match isprint() in the "C" locale. The "C" locale should be the
611 default when a C program starts up. In most systems, only ASCII printing
612 characters match isprint(). This difference caused the output from pcretest
613 to vary, making some of the tests fail. I have changed pcretest so that:
614
615 (a) When it is outputting text in the compiled version of a pattern, bytes
616 other than 32-126 are always shown as hex escapes.
617
618 (b) When it is outputting text that is a matched part of a subject string,
619 it does the same, unless a different locale has been set for the match
620 (using the /L modifier). In this case, it uses isprint() to decide.
621
622 4. Fixed a major bug that caused incorrect computation of the amount of memory
623 required for a compiled pattern when options that changed within the
624 pattern affected the logic of the preliminary scan that determines the
625 length. The relevant options are -x, and -i in UTF-8 mode. The result was
626 that the computed length was too small. The symptoms of this bug were
627 either the PCRE error "internal error: code overflow" from pcre_compile(),
628 or a glibc crash with a message such as "pcretest: free(): invalid next
629 size (fast)". Examples of patterns that provoked this bug (shown in
630 pcretest format) are:
631
632 /(?-x: )/x
633 /(?x)(?-x: \s*#\s*)/
634 /((?i)[\x{c0}])/8
635 /(?i:[\x{c0}])/8
636
637 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
638 is now done differently.
639
640 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
641 wrapper classes; (b) implement a new function in the C++ scanner that is
642 more efficient than the old way of doing things because it avoids levels of
643 recursion in the regex matching; (c) add a paragraph to the documentation
644 for the FullMatch() function.
645
646 6. The escape sequence \n was being treated as whatever was defined as
647 "newline". Not only was this contrary to the documentation, which states
648 that \n is character 10 (hex 0A), but it also went horribly wrong when
649 "newline" was defined as CRLF. This has been fixed.
650
651 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
652 was being set to -1 for the "end of line" case (supposedly a value that no
653 character can have). Though this value is never used (the check for end of
654 line is "zero bytes in current character"), it caused compiler complaints.
655 I've changed it to 0xffffffff.
656
657 8. In pcre_version.c, the version string was being built by a sequence of
658 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
659 string (as it is for production releases) called a macro with an empty
660 argument. The C standard says the result of this is undefined. The gcc
661 compiler treats it as an empty string (which was what was wanted) but it is
662 reported that Visual C gives an error. The source has been hacked around to
663 avoid this problem.
664
665 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
666 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
667 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
668 of them did).
669
670 10. Originally, pcretest opened its input and output without "b"; then I was
671 told that "b" was needed in some environments, so it was added for release
672 5.0 to both the input and output. (It makes no difference on Unix-like
673 systems.) Later I was told that it is wrong for the input on Windows. I've
674 now abstracted the modes into two macros, to make it easier to fiddle with
675 them, and removed "b" from the input mode under Windows.
676
677 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
678
679 12. Added -help and --help to pcretest as an official way of being reminded
680 of the options.
681
682 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
683 and pcrecpp.cc because they annoy compilers at high warning levels.
684
685 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
686
687 15. Fixed an occurrence of == in configure.ac that should have been = (shell
688 scripts are not C programs :-) and which was not noticed because it works
689 on Linux.
690
691 16. pcretest is supposed to handle any length of pattern and data line (as one
692 line or as a continued sequence of lines) by extending its input buffer if
693 necessary. This feature was broken for very long pattern lines, leading to
694 a string of junk being passed to pcre_compile() if the pattern was longer
695 than about 50K.
696
697 17. I have done a major re-factoring of the way pcre_compile() computes the
698 amount of memory needed for a compiled pattern. Previously, there was code
699 that made a preliminary scan of the pattern in order to do this. That was
700 OK when PCRE was new, but as the facilities have expanded, it has become
701 harder and harder to keep it in step with the real compile phase, and there
702 have been a number of bugs (see for example, 4 above). I have now found a
703 cunning way of running the real compile function in a "fake" mode that
704 enables it to compute how much memory it would need, while actually only
705 ever using a few hundred bytes of working memory and without too many
706 tests of the mode. This should make future maintenance and development
707 easier. A side effect of this work is that the limit of 200 on the nesting
708 depth of parentheses has been removed (though this was never a serious
709 limitation, I suspect). However, there is a downside: pcre_compile() now
710 runs more slowly than before (30% or more, depending on the pattern). I
711 hope this isn't a big issue. There is no effect on runtime performance.
712
713 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
714 newline (only possible for the last line of a file) and it was a
715 pattern that set a locale (followed by /Lsomething), pcretest crashed.
716
717 19. Added additional timing features to pcretest. (1) The -tm option now times
718 matching only, not compiling. (2) Both -t and -tm can be followed, as a
719 separate command line item, by a number that specifies the number of
720 repeats to use when timing. The default is 50000; this gives better
721 precision, but takes uncomfortably long for very large patterns.
722
723 20. Extended pcre_study() to be more clever in cases where a branch of a
724 subpattern has no definite first character. For example, (a*|b*)[cd] would
725 previously give no result from pcre_study(). Now it recognizes that the
726 first character must be a, b, c, or d.
727
728 21. There was an incorrect error "recursive call could loop indefinitely" if
729 a subpattern (or the entire pattern) that was being tested for matching an
730 empty string contained only one non-empty item after a nested subpattern.
731 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
732 incorrectly, because the \d was being skipped in the check.
733
734 22. The pcretest program now has a new pattern option /B and a command line
735 option -b, which is equivalent to adding /B to every pattern. This causes
736 it to show the compiled bytecode, without the additional information that
737 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
738 is the same as /B/I).
739
740 23. A new optimization is now able automatically to treat some sequences such
741 as a*b as a*+b. More specifically, if something simple (such as a character
742 or a simple class like \d) has an unlimited quantifier, and is followed by
743 something that cannot possibly match the quantified thing, the quantifier
744 is automatically "possessified".
745
746 24. A recursive reference to a subpattern whose number was greater than 39
747 went wrong under certain circumstances in UTF-8 mode. This bug could also
748 have affected the operation of pcre_study().
749
750 25. Realized that a little bit of performance could be had by replacing
751 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
752
753 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
754
755 27. Possessive quantifiers such as a++ were previously implemented by turning
756 them into atomic groups such as ($>a+). Now they have their own opcodes,
757 which improves performance. This includes the automatically created ones
758 from 23 above.
759
760 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
761 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
762 the first matched character to be a colon. This applied both to named and
763 numbered groups.
764
765 29. The ucpinternal.h header file was missing its idempotency #ifdef.
766
767 30. I was sent a "project" file called libpcre.a.dev which I understand makes
768 building PCRE on Windows easier, so I have included it in the distribution.
769
770 31. There is now a check in pcretest against a ridiculously large number being
771 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
772 loop, the loop is abandoned.
773
774 32. Forward references to subpatterns in conditions such as (?(2)...) where
775 subpattern 2 is defined later cause pcre_compile() to search forwards in
776 the pattern for the relevant set of parentheses. This search went wrong
777 when there were unescaped parentheses in a character class, parentheses
778 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
779
780 33. "Subroutine" calls and backreferences were previously restricted to
781 referencing subpatterns earlier in the regex. This restriction has now
782 been removed.
783
784 34. Added a number of extra features that are going to be in Perl 5.10. On the
785 whole, these are just syntactic alternatives for features that PCRE had
786 previously implemented using the Python syntax or my own invention. The
787 other formats are all retained for compatibility.
788
789 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
790 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
791 also .NET compatible.
792
793 (b) A recursion or subroutine call to a named group can now be defined as
794 (?&name) as well as (?P>name).
795
796 (c) A backreference to a named group can now be defined as \k<name> or
797 \k'name' as well as (?P=name). The new forms, as well as being in Perl
798 5.10, are also .NET compatible.
799
800 (d) A conditional reference to a named group can now use the syntax
801 (?(<name>) or (?('name') as well as (?(name).
802
803 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
804 groups (named and numbered) that are never evaluated inline, but can be
805 called as "subroutines" from elsewhere. In effect, the DEFINE condition
806 is always false. There may be only one alternative in such a group.
807
808 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
809 as the simple (?(R). The condition is true only if the most recent
810 recursion is that of the given number or name. It does not search out
811 through the entire recursion stack.
812
813 (g) The escape \gN or \g{N} has been added, where N is a positive or
814 negative number, specifying an absolute or relative reference.
815
816 35. Tidied to get rid of some further signed/unsigned compiler warnings and
817 some "unreachable code" warnings.
818
819 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
820 things, this adds five new scripts.
821
822 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
823 There were also incompatibilities regarding the handling of \Q..\E inside
824 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
825 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
826
827 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
828 matches an empty string, and forcibly breaks the loop. There were bugs in
829 this code in non-simple cases. For a pattern such as ^(a()*)* matched
830 against aaaa the result was just "a" rather than "aaaa", for example. Two
831 separate and independent bugs (that affected different cases) have been
832 fixed.
833
834 39. Refactored the code to abolish the use of different opcodes for small
835 capturing bracket numbers. This is a tidy that I avoided doing when I
836 removed the limit on the number of capturing brackets for 3.5 back in 2001.
837 The new approach is not only tidier, it makes it possible to reduce the
838 memory needed to fix the previous bug (38).
839
840 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
841 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
842 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
843 mode.
844
845 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
846 report.
847
848 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
849 copy construction and assignment in the C++ wrapper.
850
851 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
852 couple of bugs that could have given wrong results in the "--newline=crlf"
853 case.
854
855 44. Added a number of casts and did some reorganization of signed/unsigned int
856 variables following suggestions from Dair Grant. Also renamed the variable
857 "this" as "item" because it is a C++ keyword.
858
859 45. Arranged for dftables to add
860
861 #include "pcre_internal.h"
862
863 to pcre_chartables.c because without it, gcc 4.x may remove the array
864 definition from the final binary if PCRE is built into a static library and
865 dead code stripping is activated.
866
867 46. For an unanchored pattern, if a match attempt fails at the start of a
868 newline sequence, and the newline setting is CRLF or ANY, and the next two
869 characters are CRLF, advance by two characters instead of one.
870
871
872 Version 6.7 04-Jul-06
873 ---------------------
874
875 1. In order to handle tests when input lines are enormously long, pcretest has
876 been re-factored so that it automatically extends its buffers when
877 necessary. The code is crude, but this _is_ just a test program. The
878 default size has been increased from 32K to 50K.
879
880 2. The code in pcre_study() was using the value of the re argument before
881 testing it for NULL. (Of course, in any sensible call of the function, it
882 won't be NULL.)
883
884 3. The memmove() emulation function in pcre_internal.h, which is used on
885 systems that lack both memmove() and bcopy() - that is, hardly ever -
886 was missing a "static" storage class specifier.
887
888 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
889 containing an extended class (one that cannot be represented by a bitmap
890 because it contains high-valued characters or Unicode property items, e.g.
891 [\pZ]). Almost always one would set UTF-8 mode when processing such a
892 pattern, but PCRE should not loop if you do not (it no longer does).
893 [Detail: two cases were found: (a) a repeated subpattern containing an
894 extended class; (b) a recursive reference to a subpattern that followed a
895 previous extended class. It wasn't skipping over the extended class
896 correctly when UTF-8 mode was not set.]
897
898 5. A negated single-character class was not being recognized as fixed-length
899 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
900 compile error "lookbehind assertion is not fixed length".
901
902 6. The RunPerlTest auxiliary script was showing an unexpected difference
903 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
904 write a Perl script that can interpret lines of an input file either as
905 byte characters or as UTF-8, which is what "perltest" was being required to
906 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
907 can't do is switch easily at run time between having the "use utf8;" pragma
908 or not. In the end, I fudged it by using the RunPerlTest script to insert
909 "use utf8;" explicitly for the UTF-8 tests.
910
911 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
912 the end of the subject string, contrary to the documentation and to what
913 Perl does. This was true of both matching functions. Now it matches only at
914 the start of the subject and immediately after *internal* newlines.
915
916 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
917 a pointer to an int instead of a pointer to an unsigned long int. This
918 caused problems on 64-bit systems.
919
920 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
921 instance of the 'standard' template library not being so standard".
922
923 10. There was no check on the number of named subpatterns nor the maximum
924 length of a subpattern name. The product of these values is used to compute
925 the size of the memory block for a compiled pattern. By supplying a very
926 long subpattern name and a large number of named subpatterns, the size
927 computation could be caused to overflow. This is now prevented by limiting
928 the length of names to 32 characters, and the number of named subpatterns
929 to 10,000.
930
931 11. Subpatterns that are repeated with specific counts have to be replicated in
932 the compiled pattern. The size of memory for this was computed from the
933 length of the subpattern and the repeat count. The latter is limited to
934 65535, but there was no limit on the former, meaning that integer overflow
935 could in principle occur. The compiled length of a repeated subpattern is
936 now limited to 30,000 bytes in order to prevent this.
937
938 12. Added the optional facility to have named substrings with the same name.
939
940 13. Added the ability to use a named substring as a condition, using the
941 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
942 are numbers (not recommended). Forward references are permitted.
943
944 14. Added forward references in named backreferences (if you see what I mean).
945
946 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
947 pattern could run off the end of the subject. For example, the pattern
948 "(?s)(.{1,5})"8 did this with the subject "ab".
949
950 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
951 PCRE_CASELESS was set when matching characters that were quantified with ?
952 or *.
953
954 17. A character class other than a single negated character that had a minimum
955 but no maximum quantifier - for example [ab]{6,} - was not handled
956 correctly by pce_dfa_exec(). It would match only one character.
957
958 18. A valid (though odd) pattern that looked like a POSIX character
959 class but used an invalid character after [ (for example [[,abc,]]) caused
960 pcre_compile() to give the error "Failed: internal error: code overflow" or
961 in some cases to crash with a glibc free() error. This could even happen if
962 the pattern terminated after [[ but there just happened to be a sequence of
963 letters, a binary zero, and a closing ] in the memory that followed.
964
965 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
966 over the years. Originally (before any Unicode support), just the bottom 8
967 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
968 output from "man perlunicode" includes this:
969
970 The regular expression compiler produces polymorphic opcodes. That
971 is, the pattern adapts to the data and automatically switches to
972 the Unicode character scheme when presented with Unicode data--or
973 instead uses a traditional byte scheme when presented with byte
974 data.
975
976 Sadly, a wide octal escape does not cause a switch, and in a string with
977 no other multibyte characters, these octal escapes are treated as before.
978 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
979 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
980 Unicode string.
981
982 I have not perpetrated such confusion in PCRE. Up till now, it took just
983 the bottom 8 bits, as in old Perl. I have now made octal escapes with
984 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
985 translate to the appropriate multibyte character.
986
987 29. Applied some refactoring to reduce the number of warnings from Microsoft
988 and Borland compilers. This has included removing the fudge introduced
989 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
990 a warning about an unused variable.
991
992 21. PCRE has not included VT (character 0x0b) in the set of whitespace
993 characters since release 4.0, because Perl (from release 5.004) does not.
994 [Or at least, is documented not to: some releases seem to be in conflict
995 with the documentation.] However, when a pattern was studied with
996 pcre_study() and all its branches started with \s, PCRE still included VT
997 as a possible starting character. Of course, this did no harm; it just
998 caused an unnecessary match attempt.
999
1000 22. Removed a now-redundant internal flag bit that recorded the fact that case
1001 dependency changed within the pattern. This was once needed for "required
1002 byte" processing, but is no longer used. This recovers a now-scarce options
1003 bit. Also moved the least significant internal flag bit to the most-
1004 significant bit of the word, which was not previously used (hangover from
1005 the days when it was an int rather than a uint) to free up another bit for
1006 the future.
1007
1008 23. Added support for CRLF line endings as well as CR and LF. As well as the
1009 default being selectable at build time, it can now be changed at runtime
1010 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1011 specify that it is scanning data with non-default line endings.
1012
1013 24. Changed the definition of CXXLINK to make it agree with the definition of
1014 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1015
1016 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1017 recursions. This makes a big different to stack usage for some patterns.
1018
1019 26. If a subpattern containing a named recursion or subroutine reference such
1020 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1021 the space required for the compiled pattern went wrong and gave too small a
1022 value. Depending on the environment, this could lead to "Failed: internal
1023 error: code overflow at offset 49" or "glibc detected double free or
1024 corruption" errors.
1025
1026 27. Applied patches from Google (a) to support the new newline modes and (b) to
1027 advance over multibyte UTF-8 characters in GlobalReplace.
1028
1029 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1030 difference for some implementation of PCRE in some Windows version.
1031
1032 29. Added some extra testing facilities to pcretest:
1033
1034 \q<number> in a data line sets the "match limit" value
1035 \Q<number> in a data line sets the "match recursion limt" value
1036 -S <number> sets the stack size, where <number> is in megabytes
1037
1038 The -S option isn't available for Windows.
1039
1040
1041 Version 6.6 06-Feb-06
1042 ---------------------
1043
1044 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1045 in pcreposix.h. I have copied the definition from pcre.h.
1046
1047 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1048 because pcre.h is no longer a built file.
1049
1050 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1051 not normally included in the compiled code.
1052
1053
1054 Version 6.5 01-Feb-06
1055 ---------------------
1056
1057 1. When using the partial match feature with pcre_dfa_exec(), it was not
1058 anchoring the second and subsequent partial matches at the new starting
1059 point. This could lead to incorrect results. For example, with the pattern
1060 /1234/, partially matching against "123" and then "a4" gave a match.
1061
1062 2. Changes to pcregrep:
1063
1064 (a) All non-match returns from pcre_exec() were being treated as failures
1065 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1066 error message is output. Some extra information is given for the
1067 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1068 probably the only errors that are likely to be caused by users (by
1069 specifying a regex that has nested indefinite repeats, for instance).
1070 If there are more than 20 of these errors, pcregrep is abandoned.
1071
1072 (b) A binary zero was treated as data while matching, but terminated the
1073 output line if it was written out. This has been fixed: binary zeroes
1074 are now no different to any other data bytes.
1075
1076 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1077 used to set a locale for matching. The --locale=xxxx long option has
1078 been added (no short equivalent) to specify a locale explicitly on the
1079 pcregrep command, overriding the environment variables.
1080
1081 (d) When -B was used with -n, some line numbers in the output were one less
1082 than they should have been.
1083
1084 (e) Added the -o (--only-matching) option.
1085
1086 (f) If -A or -C was used with -c (count only), some lines of context were
1087 accidentally printed for the final match.
1088
1089 (g) Added the -H (--with-filename) option.
1090
1091 (h) The combination of options -rh failed to suppress file names for files
1092 that were found from directory arguments.
1093
1094 (i) Added the -D (--devices) and -d (--directories) options.
1095
1096 (j) Added the -F (--fixed-strings) option.
1097
1098 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1099
1100 (l) Added the --colo(u)r option.
1101
1102 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1103 is not present by default.
1104
1105 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1106 items such as (?R) or (?1), when the recursion could match a number of
1107 alternatives. If it matched one of the alternatives, but subsequently,
1108 outside the recursion, there was a failure, the code tried to back up into
1109 the recursion. However, because of the way PCRE is implemented, this is not
1110 possible, and the result was an incorrect result from the match.
1111
1112 In order to prevent this happening, the specification of recursion has
1113 been changed so that all such subpatterns are automatically treated as
1114 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1115
1116 4. I had overlooked the fact that, in some locales, there are characters for
1117 which isalpha() is true but neither isupper() nor islower() are true. In
1118 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1119 and ordfeminine) are like this. This affected the treatment of \w and \W
1120 when they appeared in character classes, but not when they appeared outside
1121 a character class. The bit map for "word" characters is now created
1122 separately from the results of isalnum() instead of just taking it from the
1123 upper, lower, and digit maps. (Plus the underscore character, of course.)
1124
1125 5. The above bug also affected the handling of POSIX character classes such as
1126 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1127 permanent tables. Instead, the bit maps for such a class were previously
1128 created as the appropriate unions of the upper, lower, and digit bitmaps.
1129 Now they are created by subtraction from the [[:word:]] class, which has
1130 its own bitmap.
1131
1132 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1133 It is created by subtracting the vertical space characters (\x09, \x0a,
1134 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1135 subtraction was done in the overall bitmap for a character class, meaning
1136 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1137 be recognized. This bug has been fixed.
1138
1139 7. Patches from the folks at Google:
1140
1141 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1142 real life, but is still worth protecting against".
1143
1144 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1145 regular expressions".
1146
1147 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1148 have it.
1149
1150 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1151 "configure" and the latter not, in order to fix a problem somebody had
1152 with compiling the Arg class on HP-UX.
1153
1154 (e) Improve the error-handling of the C++ wrapper a little bit.
1155
1156 (f) New tests for checking recursion limiting.
1157
1158 8. The pcre_memmove() function, which is used only if the environment does not
1159 have a standard memmove() function (and is therefore rarely compiled),
1160 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1161 returning a result (though PCRE never actually uses the result).
1162
1163 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1164 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1165 returned instead of calling malloc() with an overflowing number that would
1166 most likely cause subsequent chaos.
1167
1168 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1169
1170 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1171 with this option is matched, the nmatch and pmatch options of regexec() are
1172 ignored.
1173
1174 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1175 provided in case anyone wants to the the POSIX interface with UTF-8
1176 strings.
1177
1178 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1179 C++ linking (needed for some HP-UX environments).
1180
1181 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1182 (unused parameter) and in the pcre_printint() function (omitted "default"
1183 switch label when the default is to do nothing).
1184
1185 15. Added some code to make it possible, when PCRE is compiled as a C++
1186 library, to replace subject pointers for pcre_exec() with a smart pointer
1187 class, thus making it possible to process discontinuous strings.
1188
1189 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1190 much the same function. They were added by different people who were trying
1191 to make PCRE easy to compile on non-Unix systems. It has been suggested
1192 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1193 for compiling on Windows systems. I have therefore replaced it with
1194 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1195 defaults to "extern" for C or "extern C" for C++, which works fine on
1196 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1197 SCOPE with something explicit in config.h. In addition:
1198
1199 (a) pcreposix.h still had just "extern" instead of either of these macros;
1200 I have replaced it with PCRE_DATA_SCOPE.
1201
1202 (b) Functions such as _pcre_xclass(), which are internal to the library,
1203 but external in the C sense, all had PCRE_EXPORT in their definitions.
1204 This is apparently wrong for the Windows case, so I have removed it.
1205 (It makes no difference on Unix-like systems.)
1206
1207 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1208 of recursive calls to match(). This is different to MATCH_LIMIT because
1209 that limits the total number of calls to match(), not all of which increase
1210 the depth of recursion. Limiting the recursion depth limits the amount of
1211 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1212 when PCRE is compiled, and changed at run time. A patch from Google adds
1213 this functionality to the C++ interface.
1214
1215 18. Changes to the handling of Unicode character properties:
1216
1217 (a) Updated the table to Unicode 4.1.0.
1218
1219 (b) Recognize characters that are not in the table as "Cn" (undefined).
1220
1221 (c) I revised the way the table is implemented to a much improved format
1222 which includes recognition of ranges. It now supports the ranges that
1223 are defined in UnicodeData.txt, and it also amalgamates other
1224 characters into ranges. This has reduced the number of entries in the
1225 table from around 16,000 to around 3,000, thus reducing its size
1226 considerably. I realized I did not need to use a tree structure after
1227 all - a binary chop search is just as efficient. Having reduced the
1228 number of entries, I extended their size from 6 bytes to 8 bytes to
1229 allow for more data.
1230
1231 (d) Added support for Unicode script names via properties such as \p{Han}.
1232
1233 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1234 matching that character.
1235
1236 20. When matching a repeated Unicode property with a minimum greater than zero,
1237 (for example \pL{2,}), PCRE could look past the end of the subject if it
1238 reached it while seeking the minimum number of characters. This could
1239 happen only if some of the characters were more than one byte long, because
1240 there is a check for at least the minimum number of bytes.
1241
1242 21. Refactored the implementation of \p and \P so as to be more general, to
1243 allow for more different types of property in future. This has changed the
1244 compiled form incompatibly. Anybody with saved compiled patterns that use
1245 \p or \P will have to recompile them.
1246
1247 22. Added "Any" and "L&" to the supported property types.
1248
1249 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1250 but give a compile time error if the value is greater than 0xff.
1251
1252 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1253 accidentally not being installed or uninstalled.
1254
1255 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1256 made were to insert the current release number. This seemed silly, because
1257 it made things harder for people building PCRE on systems that don't run
1258 "configure". I have turned pcre.h into a distributed file, no longer built
1259 by "configure", with the version identification directly included. There is
1260 no longer a pcre.h.in file.
1261
1262 However, this change necessitated a change to the pcre-config script as
1263 well. It is built from pcre-config.in, and one of the substitutions was the
1264 release number. I have updated configure.ac so that ./configure now finds
1265 the release number by grepping pcre.h.
1266
1267 26. Added the ability to run the tests under valgrind.
1268
1269
1270 Version 6.4 05-Sep-05
1271 ---------------------
1272
1273 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1274 "--" to be printed when multiple files were scanned, even when none of the
1275 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1276 consider it to be a bug, and have restored the previous behaviour.
1277
1278 2. A couple of code tidies to get rid of compiler warnings.
1279
1280 3. The pcretest program used to cheat by referring to symbols in the library
1281 whose names begin with _pcre_. These are internal symbols that are not
1282 really supposed to be visible externally, and in some environments it is
1283 possible to suppress them. The cheating is now confined to including
1284 certain files from the library's source, which is a bit cleaner.
1285
1286 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1287 file's purpose clearer.
1288
1289 5. Reorganized pcre_ucp_findchar().
1290
1291
1292 Version 6.3 15-Aug-05
1293 ---------------------
1294
1295 1. The file libpcre.pc.in did not have general read permission in the tarball.
1296
1297 2. There were some problems when building without C++ support:
1298
1299 (a) If C++ support was not built, "make install" and "make test" still
1300 tried to test it.
1301
1302 (b) There were problems when the value of CXX was explicitly set. Some
1303 changes have been made to try to fix these, and ...
1304
1305 (c) --disable-cpp can now be used to explicitly disable C++ support.
1306
1307 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1308 backslash in a target when C++ was disabled. This confuses some
1309 versions of "make", apparently. Using an intermediate variable solves
1310 this. (Same for CPP_LOBJ.)
1311
1312 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1313 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1314 necessary on certain architectures.
1315
1316 4. Added a setting of -export-symbols-regex to the link command to remove
1317 those symbols that are exported in the C sense, but actually are local
1318 within the library, and not documented. Their names all begin with
1319 "_pcre_". This is not a perfect job, because (a) we have to except some
1320 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1321 available (and never for static libraries). I have made a note to try to
1322 find a way round (a) in the future.
1323
1324
1325 Version 6.2 01-Aug-05
1326 ---------------------
1327
1328 1. There was no test for integer overflow of quantifier values. A construction
1329 such as {1111111111111111} would give undefined results. What is worse, if
1330 a minimum quantifier for a parenthesized subpattern overflowed and became
1331 negative, the calculation of the memory size went wrong. This could have
1332 led to memory overwriting.
1333
1334 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1335
1336 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1337 operating environments where this matters.
1338
1339 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1340 PCRE options from within the C++ wrapper.
1341
1342 5. Named capturing subpatterns were not being correctly counted when a pattern
1343 was compiled. This caused two problems: (a) If there were more than 100
1344 such subpatterns, the calculation of the memory needed for the whole
1345 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1346 back references of the form \12, where the number was greater than 9, were
1347 not recognized as back references, even though there were sufficient
1348 previous subpatterns.
1349
1350 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1351 versions of gcc, e.g. 2.95.4.
1352
1353
1354 Version 6.1 21-Jun-05
1355 ---------------------
1356
1357 1. There was one reference to the variable "posix" in pcretest.c that was not
1358 surrounded by "#if !defined NOPOSIX".
1359
1360 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1361 the cross-check on the old pcre_info() function, for the benefit of the
1362 cut-down version of PCRE that is currently imported into Exim.
1363
1364 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1365 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1366 patterns that start (?i) but I don't think that matters. The use of (?i) is
1367 just an example; this all applies to the other options as well.
1368
1369 4. Since libtool seems to echo the compile commands it is issuing, the output
1370 from "make" can be reduced a bit by putting "@" in front of each libtool
1371 compile command.
1372
1373 5. Patch from the folks at Google for configure.in to be a bit more thorough
1374 in checking for a suitable C++ installation before trying to compile the
1375 C++ stuff. This should fix a reported problem when a compiler was present,
1376 but no suitable headers.
1377
1378 6. The man pages all had just "PCRE" as their title. I have changed them to
1379 be the relevant file name. I have also arranged that these names are
1380 retained in the file doc/pcre.txt, which is a concatenation in text format
1381 of all the man pages except the little individual ones for each function.
1382
1383 7. The NON-UNIX-USE file had not been updated for the different set of source
1384 files that come with release 6. I also added a few comments about the C++
1385 wrapper.
1386
1387
1388 Version 6.0 07-Jun-05
1389 ---------------------
1390
1391 1. Some minor internal re-organization to help with my DFA experiments.
1392
1393 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1394 didn't matter for the library itself when fully configured, but did matter
1395 when compiling without UCP support, or within Exim, where the ucp files are
1396 not imported.
1397
1398 3. Refactoring of the library code to split up the various functions into
1399 different source modules. The addition of the new DFA matching code (see
1400 below) to a single monolithic source would have made it really too
1401 unwieldy, quite apart from causing all the code to be include in a
1402 statically linked application, when only some functions are used. This is
1403 relevant even without the DFA addition now that patterns can be compiled in
1404 one application and matched in another.
1405
1406 The downside of splitting up is that there have to be some external
1407 functions and data tables that are used internally in different modules of
1408 the library but which are not part of the API. These have all had their
1409 names changed to start with "_pcre_" so that they are unlikely to clash
1410 with other external names.
1411
1412 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1413 a different (DFA) algorithm. Although it is slower than the original
1414 function, it does have some advantages for certain types of matching
1415 problem.
1416
1417 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1418 including restarting after a partial match.
1419
1420 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1421 defined when compiling for Windows was sent to me. I have put it into the
1422 code, though I have no means of testing or verifying it.
1423
1424 7. Added the pcre_refcount() auxiliary function.
1425
1426 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1427 match before or at the first newline in the subject string. In pcretest,
1428 the /f option on a pattern can be used to set this.
1429
1430 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1431 would behave wrongly. This has been present in PCRE since release 4.0.
1432
1433 10. A number of changes to the pcregrep command:
1434
1435 (a) Refactored how -x works; insert ^(...)$ instead of setting
1436 PCRE_ANCHORED and checking the length, in preparation for adding
1437 something similar for -w.
1438
1439 (b) Added the -w (match as a word) option.
1440
1441 (c) Refactored the way lines are read and buffered so as to have more
1442 than one at a time available.
1443
1444 (d) Implemented a pcregrep test script.
1445
1446 (e) Added the -M (multiline match) option. This allows patterns to match
1447 over several lines of the subject. The buffering ensures that at least
1448 8K, or the rest of the document (whichever is the shorter) is available
1449 for matching (and similarly the previous 8K for lookbehind assertions).
1450
1451 (f) Changed the --help output so that it now says
1452
1453 -w, --word-regex(p)
1454
1455 instead of two lines, one with "regex" and the other with "regexp"
1456 because that confused at least one person since the short forms are the
1457 same. (This required a bit of code, as the output is generated
1458 automatically from a table. It wasn't just a text change.)
1459
1460 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1461 option but starts with a hyphen. Could be a pattern or a path name
1462 starting with a hyphen, for instance.
1463
1464 (h) "-" can be given as a file name to represent stdin.
1465
1466 (i) When file names are being printed, "(standard input)" is used for
1467 the standard input, for compatibility with GNU grep. Previously
1468 "<stdin>" was used.
1469
1470 (j) The option --label=xxx can be used to supply a name to be used for
1471 stdin when file names are being printed. There is no short form.
1472
1473 (k) Re-factored the options decoding logic because we are going to add
1474 two more options that take data. Such options can now be given in four
1475 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1476
1477 (l) Added the -A, -B, and -C options for requesting that lines of context
1478 around matches be printed.
1479
1480 (m) Added the -L option to print the names of files that do not contain
1481 any matching lines, that is, the complement of -l.
1482
1483 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1484 continue to scan other files.
1485
1486 (o) The -s option was incorrectly implemented. For compatibility with other
1487 greps, it now suppresses the error message for a non-existent or non-
1488 accessible file (but not the return code). There is a new option called
1489 -q that suppresses the output of matching lines, which was what -s was
1490 previously doing.
1491
1492 (p) Added --include and --exclude options to specify files for inclusion
1493 and exclusion when recursing.
1494
1495 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1496 Hopefully, it now does.
1497
1498 12. Missing cast in pcre_study().
1499
1500 13. Added an "uninstall" target to the makefile.
1501
1502 14. Replaced "extern" in the function prototypes in Makefile.in with
1503 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1504 world, but is set differently for Windows.
1505
1506 15. Added a second compiling function called pcre_compile2(). The only
1507 difference is that it has an extra argument, which is a pointer to an
1508 integer error code. When there is a compile-time failure, this is set
1509 non-zero, in addition to the error test pointer being set to point to an
1510 error message. The new argument may be NULL if no error number is required
1511 (but then you may as well call pcre_compile(), which is now just a
1512 wrapper). This facility is provided because some applications need a
1513 numeric error indication, but it has also enabled me to tidy up the way
1514 compile-time errors are handled in the POSIX wrapper.
1515
1516 16. Added VPATH=.libs to the makefile; this should help when building with one
1517 prefix path and installing with another. (Or so I'm told by someone who
1518 knows more about this stuff than I do.)
1519
1520 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1521 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1522 match everything, including newlines. This is not POSIX-compatible, but
1523 somebody wanted the feature. From pcretest it can be activated by using
1524 both the P and the s flags.
1525
1526 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1527
1528 19. libpcre.pc was being incorrectly installed as executable.
1529
1530 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1531 it now also looks for '\r' so that it will work unmodified on Windows.
1532
1533 21. Added Google's contributed C++ wrapper to the distribution.
1534
1535 22. Added some untidy missing memory free() calls in pcretest, to keep
1536 Electric Fence happy when testing.
1537
1538
1539
1540 Version 5.0 13-Sep-04
1541 ---------------------
1542
1543 1. Internal change: literal characters are no longer packed up into items
1544 containing multiple characters in a single byte-string. Each character
1545 is now matched using a separate opcode. However, there may be more than one
1546 byte in the character in UTF-8 mode.
1547
1548 2. The pcre_callout_block structure has two new fields: pattern_position and
1549 next_item_length. These contain the offset in the pattern to the next match
1550 item, and its length, respectively.
1551
1552 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1553 insertion of callouts before each pattern item. Added the /C option to
1554 pcretest to make use of this.
1555
1556 4. On the advice of a Windows user, the lines
1557
1558 #if defined(_WIN32) || defined(WIN32)
1559 _setmode( _fileno( stdout ), 0x8000 );
1560 #endif /* defined(_WIN32) || defined(WIN32) */
1561
1562 have been added to the source of pcretest. This apparently does useful
1563 magic in relation to line terminators.
1564
1565 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1566 for the benefit of those environments where the "b" makes a difference.
1567
1568 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1569 to know about it. I have put a hack into configure.in that adds in code
1570 to set GCC=yes if CC=icc. This seems to end up at a point in the
1571 generated configure script that is early enough to affect the setting of
1572 compiler options, which is what is needed, but I have no means of testing
1573 whether it really works. (The user who reported this had patched the
1574 generated configure script, which of course I cannot do.)
1575
1576 LATER: After change 22 below (new libtool files), the configure script
1577 seems to know about icc (and also ecc). Therefore, I have commented out
1578 this hack in configure.in.
1579
1580 7. Added support for pkg-config (2 patches were sent in).
1581
1582 8. Negated POSIX character classes that used a combination of internal tables
1583 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1584 [[:^ascii]]. Typically, they would match almost any characters. The other
1585 POSIX classes were not broken in this way.
1586
1587 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1588 to find the match, as PCRE was deluded into thinking that the match had to
1589 start at the start point or following a newline. The same bug applied to
1590 patterns with negative forward assertions or any backward assertions
1591 preceding ".*" at the start, unless the pattern required a fixed first
1592 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1593
1594 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1595 starting at the last subject character, bytes beyond the end of the subject
1596 string were read.
1597
1598 11. Renamed the variable "class" as "classbits" to make life easier for C++
1599 users. (Previously there was a macro definition, but it apparently wasn't
1600 enough.)
1601
1602 12. Added the new field "tables" to the extra data so that tables can be passed
1603 in at exec time, or the internal tables can be re-selected. This allows
1604 a compiled regex to be saved and re-used at a later time by a different
1605 program that might have everything at different addresses.
1606
1607 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1608 -R library as well as a -L library.
1609
1610 14. The debugging options of pcretest (-d on the command line or D on a
1611 pattern) showed incorrect output for anything following an extended class
1612 that contained multibyte characters and which was followed by a quantifier.
1613
1614 15. Added optional support for general category Unicode character properties
1615 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1616 support. It adds about 90K to the size of the library. The meanings of the
1617 inbuilt class escapes such as \d and \s have NOT been changed.
1618
1619 16. Updated pcredemo.c to include calls to free() to release the memory for the
1620 compiled pattern.
1621
1622 17. The generated file chartables.c was being created in the source directory
1623 instead of in the building directory. This caused the build to fail if the
1624 source directory was different from the building directory, and was
1625 read-only.
1626
1627 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1628 file. No doubt somebody will tell me if they don't make sense... Also added
1629 Dan Mooney's comments about building on OpenVMS.
1630
1631 19. Added support for partial matching via the PCRE_PARTIAL option for
1632 pcre_exec() and the \P data escape in pcretest.
1633
1634 20. Extended pcretest with 3 new pattern features:
1635
1636 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1637 write the compiled pattern to the file whose name is "rest-of-line".
1638 This is a straight binary dump of the data, with the saved pointer to
1639 the character tables forced to be NULL. The study data, if any, is
1640 written too. After writing, pcretest reads a new pattern.
1641
1642 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1643 compiled pattern from the given file. There must not be any
1644 occurrences of "<" in the file name (pretty unlikely); if there are,
1645 pcretest will instead treat the initial "<" as a pattern delimiter.
1646 After reading in the pattern, pcretest goes on to read data lines as
1647 usual.
1648
1649 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1650 and 16-bit fields in a compiled pattern, to simulate a pattern that
1651 was compiled on a host of opposite endianness.
1652
1653 21. The pcre-exec() function can now cope with patterns that were compiled on
1654 hosts of opposite endianness, with this restriction:
1655
1656 As for any compiled expression that is saved and used later, the tables
1657 pointer field cannot be preserved; the extra_data field in the arguments
1658 to pcre_exec() should be used to pass in a tables address if a value
1659 other than the default internal tables were used at compile time.
1660
1661 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1662 now diagnosed as an error. Previously, most of the time, a negative number
1663 would have been treated as zero, but if in addition "ovector" was passed as
1664 NULL, a crash could occur.
1665
1666 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1667 new versions from the libtool 1.5 distribution (the last one is a copy of
1668 a file called libtool.m4). This seems to have fixed the need to patch
1669 "configure" to support Darwin 1.3 (which I used to do). However, I still
1670 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1671 workstation).
1672
1673 24. Changed the PCRE licence to be the more standard "BSD" licence.
1674
1675
1676 Version 4.5 01-Dec-03
1677 ---------------------
1678
1679 1. There has been some re-arrangement of the code for the match() function so
1680 that it can be compiled in a version that does not call itself recursively.
1681 Instead, it keeps those local variables that need separate instances for
1682 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1683 needs to "recurse". Keeping track of where control must go is done by means
1684 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1685 hide most of the details from the main code, and operates only if
1686 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1687 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1688 operating.
1689
1690 To make it easier for callers to provide specially tailored get/free
1691 functions for this usage, two new functions, pcre_stack_malloc, and
1692 pcre_stack_free, are used. They are always called in strict stacking order,
1693 and the size of block requested is always the same.
1694
1695 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1696 PCRE has been compiled to use the stack or the heap for recursion. The
1697 -C option of pcretest uses this to show which version is compiled.
1698
1699 A new data escape \S, is added to pcretest; it causes the amounts of store
1700 obtained and freed by both kinds of malloc/free at match time to be added
1701 to the output.
1702
1703 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1704 what's available on my current Linux desktop machine.
1705
1706 3. When matching a UTF-8 string, the test for a valid string at the start has
1707 been extended. If start_offset is not zero, PCRE now checks that it points
1708 to a byte that is the start of a UTF-8 character. If not, it returns
1709 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1710 this is necessary because there may be backward assertions in the pattern.
1711 When matching the same subject several times, it may save resources to use
1712 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1713
1714 4. The code for checking the validity of UTF-8 strings has been tightened so
1715 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1716 containing "overlong sequences".
1717
1718 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1719 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1720 should have been "&", but it just so happened that all the cases this let
1721 through by mistake were picked up later in the function.
1722
1723 6. I had used a variable called "isblank" - this is a C99 function, causing
1724 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1725
1726 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1727 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1728 so the version is shown; (c) stop "make test" echoing "./RunTest".
1729
1730 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1731
1732 9. The prototype for memmove() for systems that don't have it was using
1733 size_t, but the inclusion of the header that defines size_t was later. I've
1734 moved the #includes for the C headers earlier to avoid this.
1735
1736 10. Added some adjustments to the code to make it easier to compiler on certain
1737 special systems:
1738
1739 (a) Some "const" qualifiers were missing.
1740 (b) Added the macro EXPORT before all exported functions; by default this
1741 is defined to be empty.
1742 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1743 that it reads its output file name as an argument instead of writing
1744 to the standard output and assuming this can be redirected.
1745
1746 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1747 class containing characters with values greater than 255, PCRE compilation
1748 went into a loop.
1749
1750 12. A recursive reference to a subpattern that was within another subpattern
1751 that had a minimum quantifier of zero caused PCRE to crash. For example,
1752 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1753 recursion. If the recursively-called subpattern itself had a zero repeat,
1754 that was OK.
1755
1756 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1757 buffer into which it was copied (for escape processing) was still set at
1758 1024, so long lines caused crashes.
1759
1760 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1761 "internal error: code overflow...". This applied to any character class
1762 that was followed by a possessive quantifier.
1763
1764 15. Modified the Makefile to add libpcre.la as a prerequisite for
1765 libpcreposix.la because I was told this is needed for a parallel build to
1766 work.
1767
1768 16. If a pattern that contained .* following optional items at the start was
1769 studied, the wrong optimizing data was generated, leading to matching
1770 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1771 matching string must start with a or b or c. The correct conclusion for
1772 this pattern is that a match can start with any character.
1773
1774
1775 Version 4.4 13-Aug-03
1776 ---------------------
1777
1778 1. In UTF-8 mode, a character class containing characters with values between
1779 127 and 255 was not handled correctly if the compiled pattern was studied.
1780 In fixing this, I have also improved the studying algorithm for such
1781 classes (slightly).
1782
1783 2. Three internal functions had redundant arguments passed to them. Removal
1784 might give a very teeny performance improvement.
1785
1786 3. Documentation bug: the value of the capture_top field in a callout is *one
1787 more than* the number of the hightest numbered captured substring.
1788
1789 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1790 in incorrectly linking with a previously installed version. They now link
1791 explicitly with libpcre.la.
1792
1793 5. configure.in no longer needs to recognize Cygwin specially.
1794
1795 6. A problem in pcre.in for Windows platforms is fixed.
1796
1797 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1798 pcretest, it used to include the size of the study block as part of its
1799 output. Unfortunately, the structure contains a field that has a different
1800 size on different hardware architectures. This meant that the tests that
1801 showed this size failed. As the block is currently always of a fixed size,
1802 this information isn't actually particularly useful in pcretest output, so
1803 I have just removed it.
1804
1805 8. Three pre-processor statements accidentally did not start in column 1.
1806 Sadly, there are *still* compilers around that complain, even though
1807 standard C has not required this for well over a decade. Sigh.
1808
1809 9. In pcretest, the code for checking callouts passed small integers in the
1810 callout_data field, which is a void * field. However, some picky compilers
1811 complained about the casts involved for this on 64-bit systems. Now
1812 pcretest passes the address of the small integer instead, which should get
1813 rid of the warnings.
1814
1815 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1816 both compile and run time, and gives an error if an invalid UTF-8 sequence
1817 is found. There is a option for disabling this check in cases where the
1818 string is known to be correct and/or the maximum performance is wanted.
1819
1820 11. In response to a bug report, I changed one line in Makefile.in from
1821
1822 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1823 to
1824 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1825
1826 to look similar to other lines, but I have no way of telling whether this
1827 is the right thing to do, as I do not use Windows. No doubt I'll get told
1828 if it's wrong...
1829
1830
1831 Version 4.3 21-May-03
1832 ---------------------
1833
1834 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1835 Makefile.
1836
1837 2. Some refactoring to improve the quality of the code:
1838
1839 (i) The utf8_table... variables are now declared "const".
1840
1841 (ii) The code for \cx, which used the "case flipping" table to upper case
1842 lower case letters, now just substracts 32. This is ASCII-specific,
1843 but the whole concept of \cx is ASCII-specific, so it seems
1844 reasonable.
1845
1846 (iii) PCRE was using its character types table to recognize decimal and
1847 hexadecimal digits in the pattern. This is silly, because it handles
1848 only 0-9, a-f, and A-F, but the character types table is locale-
1849 specific, which means strange things might happen. A private
1850 table is now used for this - though it costs 256 bytes, a table is
1851 much faster than multiple explicit tests. Of course, the standard
1852 character types table is still used for matching digits in subject
1853 strings against \d.
1854
1855 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1856 ending in _t are). So I've renamed it as ESC_tee.
1857
1858 3. The first argument for regexec() in the POSIX wrapper should have been
1859 defined as "const".
1860
1861 4. Changed pcretest to use malloc() for its buffers so that they can be
1862 Electric Fenced for debugging.
1863
1864 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1865 to read one or more bytes before the start of the subject string. Often this
1866 had no effect on PCRE's behaviour, but in some circumstances it could
1867 provoke a segmentation fault.
1868
1869 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1870 to try to read one or more bytes before the start of the subject string.
1871
1872 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1873 UTF-8 support could misbehave in various ways if the subject string
1874 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1875 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1876 back over UTF-8 characters.)
1877
1878
1879 Version 4.2 14-Apr-03
1880 ---------------------
1881
1882 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1883
1884 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1885 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1886 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1887 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1888 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1889 and BUILD_EXEEXT
1890 Note: automatic setting of the BUILD variables is not yet working
1891 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1892 compile-time but not at link-time
1893 [LINK]: use for linking executables only
1894 make different versions for Windows and non-Windows
1895 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1896 libraries
1897 [LINK_FOR_BUILD]: new variable
1898 [OBJEXT]: use throughout
1899 [EXEEXT]: use throughout
1900 <winshared>: new target
1901 <wininstall>: new target
1902 <dftables.o>: use native compiler
1903 <dftables>: use native linker
1904 <install>: handle Windows platform correctly
1905 <clean>: ditto
1906 <check>: ditto
1907 copy DLL to top builddir before testing
1908
1909 As part of these changes, -no-undefined was removed again. This was reported
1910 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1911 in any case.
1912
1913 3. Some tidies to get rid of compiler warnings:
1914
1915 . In the match_data structure, match_limit was an unsigned long int, whereas
1916 match_call_count was an int. I've made them both unsigned long ints.
1917
1918 . In pcretest the fact that a const uschar * doesn't automatically cast to
1919 a void * provoked a warning.
1920
1921 . Turning on some more compiler warnings threw up some "shadow" variables
1922 and a few more missing casts.
1923
1924 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1925 option, a class that contained a single character with a value between 128
1926 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1927
1928 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1929 option, a class that contained several characters, but with at least one
1930 whose value was between 128 and 255 caused PCRE to crash.
1931
1932
1933 Version 4.1 12-Mar-03
1934 ---------------------
1935
1936 1. Compiling with gcc -pedantic found a couple of places where casts were
1937 needed, and a string in dftables.c that was longer than standard compilers are
1938 required to support.
1939
1940 2. Compiling with Sun's compiler found a few more places where the code could
1941 be tidied up in order to avoid warnings.
1942
1943 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1944 first of these names is deprecated in the latest Autoconf in favour of the name
1945 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1946 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1947 analogy I have changed it to CFLAGS_FOR_BUILD.
1948
1949 4. Added -no-undefined to the linking command in the Makefile, because this is
1950 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1951 linking step for the pcreposix library.
1952
1953 5. PCRE was failing to diagnose the case of two named groups with the same
1954 name.
1955
1956 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1957 literal character that is needed in the subject for a match, and scans along to
1958 ensure that it is present before embarking on the full matching process. This
1959 saves time in cases of nested unlimited repeats that are never going to match.
1960 Problem: the scan can take a lot of time if the subject is very long (e.g.
1961 megabytes), thus penalizing straightforward matches. It is now done only if the
1962 amount of subject to be scanned is less than 1000 bytes.
1963
1964 7. A lesser problem with the same optimization is that it was recording the
1965 first character of an anchored pattern as "needed", thus provoking a search
1966 right along the subject, even when the first match of the pattern was going to
1967 fail. The "needed" character is now not set for anchored patterns, unless it
1968 follows something in the pattern that is of non-fixed length. Thus, it still
1969 fulfils its original purpose of finding quick non-matches in cases of nested
1970 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1971
1972
1973 Version 4.0 17-Feb-03
1974 ---------------------
1975
1976 1. If a comment in an extended regex that started immediately after a meta-item
1977 extended to the end of string, PCRE compiled incorrect data. This could lead to
1978 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1979
1980 2. Moved to autoconf 2.53 and libtool 1.4.2.
1981
1982 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1983 the special perltest8 script is no longer needed - all the tests can be run
1984 from a single perltest script.
1985
1986 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1987 by \s. It has now been removed in PCRE. This means it isn't recognized as
1988 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1989 class [:space:] *does* include VT, thereby creating a mess.
1990
1991 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1992 space and tab.
1993
1994 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1995 its new features into the main test script, reducing the number of scripts.
1996
1997 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1998 were backward compatible, and made the (?i) apply to the whole pattern, as if
1999 /i were given. Now it behaves more logically, and applies the option setting
2000 only to what follows. PCRE has been changed to follow suit. However, if it
2001 finds options settings right at the start of the pattern, it extracts them into
2002 the global options, as before. Thus, they show up in the info data.
2003
2004 8. Added support for the \Q...\E escape sequence. Characters in between are
2005 treated as literals. This is slightly different from Perl in that $ and @ are
2006 also handled as literals inside the quotes. In Perl, they will cause variable
2007 interpolation. Note the following examples:
2008
2009 Pattern PCRE matches Perl matches
2010
2011 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2012 \Qabc\$xyz\E abc\$xyz abc\$xyz
2013 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2014
2015 For compatibility with Perl, \Q...\E sequences are recognized inside character
2016 classes as well as outside them.
2017
2018 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2019 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2020 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2021 signed/unsigned warnings.
2022
2023 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2024 option for pcretest, so I've replaced it by a simple function that does just
2025 that job.
2026
2027 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2028 "pcregrep -".
2029
2030 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2031 Java package. This provides some syntactic sugar for simple cases of what my
2032 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2033 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2034 item, you can use this simplified notation. Note that only makes sense with
2035 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2036 greediness, whatever the setting of the PCRE_UNGREEDY option.
2037
2038 13. A change of greediness default within a pattern was not taking effect at
2039 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2040 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2041 was abstracted outside.
2042
2043 14. PCRE now supports the \G assertion. It is true when the current matching
2044 position is at the start point of the match. This differs from \A when the
2045 starting offset is non-zero. Used with the /g option of pcretest (or similar
2046 code), it works in the same way as it does for Perl's /g option. If all
2047 alternatives of a regex begin with \G, the expression is anchored to the start
2048 match position, and the "anchored" flag is set in the compiled expression.
2049
2050 15. Some bugs concerning the handling of certain option changes within patterns
2051 have been fixed. These applied to options other than (?ims). For example,
2052 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2053 been the other way round. Some of this was related to change 7 above.
2054
2055 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2056 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2057 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2058 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2059
2060 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2061 mode. Unlike ".", it always matches newline, whatever the setting of
2062 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2063 assertions. Perl allows it, but it doesn't (in general) work because it can't
2064 calculate the length of the lookbehind. At least, that's the case for Perl
2065 5.8.0 - I've been told they are going to document that it doesn't work in
2066 future.
2067
2068 18. Added an error diagnosis for escapes that PCRE does not support: these are
2069 \L, \l, \N, \P, \p, \U, \u, and \X.
2070
2071 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2072 reading past the end of the pattern in cases such as /[abcd/.
2073
2074 20. PCRE was getting more memory than necessary for patterns with classes that
2075 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2076
2077 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2078 compiling PCRE for use with Virtual Pascal.
2079
2080 22. Small fix to the Makefile to make it work properly if the build is done
2081 outside the source tree.
2082
2083 23. Added a new extension: a condition to go with recursion. If a conditional
2084 subpattern starts with (?(R) the "true" branch is used if recursion has
2085 happened, whereas the "false" branch is used only at the top level.
2086
2087 24. When there was a very long string of literal characters (over 255 bytes
2088 without UTF support, over 250 bytes with UTF support), the computation of how
2089 much memory was required could be incorrect, leading to segfaults or other
2090 strange effects.
2091
2092 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2093 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2094 there was a subsequent back reference to those brackets. This meant that, for
2095 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2096 possible to check for precisely this case. All we can do is abandon the
2097 optimization if .* occurs inside capturing brackets when there are any back
2098 references whatsoever. (See below for a better fix that came later.)
2099
2100 26. The handling of the optimization for finding the first character of a
2101 non-anchored pattern, and for finding a character that is required later in the
2102 match were failing in some cases. This didn't break the matching; it just
2103 failed to optimize when it could. The way this is done has been re-implemented.
2104
2105 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2106
2107 28. Added a new feature that provides some of the functionality that Perl
2108 provides with (?{...}). The facility is termed a "callout". The way it is done
2109 in PCRE is for the caller to provide an optional function, by setting
2110 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2111 global variable. By default it is unset, which disables all calling out. To get
2112 the function called, the regex must include (?C) at appropriate points. This
2113 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2114 This provides a means of identifying different callout points. When PCRE
2115 reaches such a point in the regex, if pcre_callout has been set, the external
2116 function is called. It is provided with data in a structure called
2117 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2118 matching continues; if it returns a non-zero value, the match at the current
2119 point fails. However, backtracking will occur if possible. [This was changed
2120 later and other features added - see item 49 below.]
2121
2122 29. pcretest is upgraded to test the callout functionality. It provides a
2123 callout function that displays information. By default, it shows the start of
2124 the match and the current position in the text. There are some new data escapes
2125 to vary what happens:
2126
2127 \C+ in addition, show current contents of captured substrings
2128 \C- do not supply a callout function
2129 \C!n return 1 when callout number n is reached
2130 \C!n!m return 1 when callout number n is reached for the mth time
2131
2132 30. If pcregrep was called with the -l option and just a single file name, it
2133 output "<stdin>" if a match was found, instead of the file name.
2134
2135 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2136 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2137 pcre_exec(). This saves a malloc/free per call. The default value of
2138 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2139 when configuring.
2140
2141 32. The default maximum size of a compiled pattern is 64K. There have been a
2142 few cases of people hitting this limit. The code now uses macros to handle the
2143 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2144 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2145 configuring. Tests 2 and 5 work only with 2-byte links because they output
2146 debugging information about compiled patterns.
2147
2148 33. Internal code re-arrangements:
2149
2150 (a) Moved the debugging function for printing out a compiled regex into
2151 its own source file (printint.c) and used #include to pull it into
2152 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2153 separate copies.
2154
2155 (b) Defined the list of op-code names for debugging as a macro in
2156 internal.h so that it is next to the definition of the opcodes.
2157
2158 (c) Defined a table of op-code lengths for simpler skipping along compiled
2159 code. This is again a macro in internal.h so that it is next to the
2160 definition of the opcodes.
2161
2162 34. Added support for recursive calls to individual subpatterns, along the
2163 lines of Robin Houston's patch (but implemented somewhat differently).
2164
2165 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2166 allow it to read and process whole directories in Win32. This code was
2167 contributed by Lionel Fourquaux; it has not been tested by me.
2168
2169 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2170 used to name a group. Names consist of alphanumerics and underscores, and must
2171 be unique. Back references use the syntax (?P=name) and recursive calls use
2172 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2173 numbers. The function pcre_fullinfo() can be used after compilation to extract
2174 a name/number map. There are three relevant calls:
2175
2176 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2177 PCRE_INFO_NAMECOUNT yields the number of entries
2178 PCRE_INFO_NAMETABLE yields a pointer to the map.
2179
2180 The map is a vector of fixed-size entries. The size of each entry depends on
2181 the length of the longest name used. The first two bytes of each entry are the
2182 group number, most significant byte first. There follows the corresponding
2183 name, zero terminated. The names are in alphabetical order.
2184
2185 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2186 case instead of 255. Making it the same both with and without UTF-8 support
2187 means that the same test output works with both.
2188
2189 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2190 calling malloc() with a zero argument.
2191
2192 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2193 optimization. I've improved things by keeping a bitmap of backreferences with
2194 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2195 fact referenced, the optimization can be applied. It is unlikely that a
2196 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2197 the match to follow \n) will appear inside brackets with a number greater than
2198 31, but if it does, any back reference > 31 suppresses the optimization.
2199
2200 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2201 of disabling numbered capturing parentheses. Any opening parenthesis that is
2202 not followed by ? behaves as if it were followed by ?: but named parentheses
2203 can still be used for capturing (and they will acquire numbers in the usual
2204 way).
2205
2206 41. Redesigned the return codes from the match() function into yes/no/error so
2207 that errors can be passed back from deep inside the nested calls. A malloc
2208 failure while inside a recursive subpattern call now causes the
2209 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2210
2211 42. It is now possible to set a limit on the number of times the match()
2212 function is called in a call to pcre_exec(). This facility makes it possible to
2213 limit the amount of recursion and backtracking, though not in a directly
2214 obvious way, because the match() function is used in a number of different
2215 circumstances. The count starts from zero for each position in the subject
2216 string (for non-anchored patterns). The default limit is, for compatibility, a
2217 large number, namely 10 000 000. You can change this in two ways:
2218
2219 (a) When configuring PCRE before making, you can use --with-match-limit=n
2220 to set a default value for the compiled library.
2221
2222 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2223 a different value is set. See 45 below.
2224
2225 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2226
2227 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2228 of things that can be changed at compile time. The first argument specifies
2229 what is wanted and the second points to where the information is to be placed.
2230 The current list of available information is:
2231
2232 PCRE_CONFIG_UTF8
2233
2234 The output is an integer that is set to one if UTF-8 support is available;
2235 otherwise it is set to zero.
2236
2237 PCRE_CONFIG_NEWLINE
2238
2239 The output is an integer that it set to the value of the code that is used for
2240 newline. It is either LF (10) or CR (13).
2241
2242 PCRE_CONFIG_LINK_SIZE
2243
2244 The output is an integer that contains the number of bytes used for internal
2245 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2246
2247 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2248
2249 The output is an integer that contains the threshold above which the POSIX
2250 interface uses malloc() for output vectors. See item 31 above.
2251
2252 PCRE_CONFIG_MATCH_LIMIT
2253
2254 The output is an unsigned integer that contains the default limit of the number
2255 of match() calls in a pcre_exec() execution. See 42 above.
2256
2257 44. pcretest has been upgraded by the addition of the -C option. This causes it
2258 to extract all the available output from the new pcre_config() function, and to
2259 output it. The program then exits immediately.
2260
2261 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2262 order to support additional features. One way would have been to define
2263 pcre_exec2() (for example) with extra arguments, but this would not have been
2264 extensible, and would also have required all calls to the original function to
2265 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2266 is used for passing in "extra" data from pcre_study().
2267
2268 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2269 contains the following fields:
2270
2271 flags a bitmap indicating which of the following fields are set
2272 study_data opaque data from pcre_study()
2273 match_limit a way of specifying a limit on match() calls for a specific
2274 call to pcre_exec()
2275 callout_data data for callouts (see 49 below)
2276
2277 The flag bits are also defined in pcre.h, and are
2278
2279 PCRE_EXTRA_STUDY_DATA
2280 PCRE_EXTRA_MATCH_LIMIT
2281 PCRE_EXTRA_CALLOUT_DATA
2282
2283 The pcre_study() function now returns one of these new pcre_extra blocks, with
2284 the actual study data pointed to by the study_data field, and the
2285 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2286 before. That is, this change is entirely upwards-compatible and requires no
2287 change to existing code.
2288
2289 If you want to pass in additional data to pcre_exec(), you can either place it
2290 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2291 block.
2292
2293 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2294 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2295 times with different match limits, until it finds the minimum value needed for
2296 pcre_exec() to complete. The value is then output. This can be instructive; for
2297 most simple matches the number is quite small, but for pathological cases it
2298 gets very large very quickly.
2299
2300 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2301 returns the size of the data block pointed to by the study_data field in a
2302 pcre_extra block, that is, the value that was passed as the argument to
2303 pcre_malloc() when PCRE was getting memory in which to place the information
2304 created by pcre_study(). The fourth argument should point to a size_t variable.
2305 pcretest has been extended so that this information is shown after a successful
2306 pcre_study() call when information about the compiled regex is being displayed.
2307
2308 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2309 because what follows is always an absolute path. (Later: it turns out that this
2310 is more than cosmetic for MinGW, because it doesn't like empty path
2311 components.)
2312
2313 49. Some changes have been made to the callout feature (see 28 above):
2314
2315 (i) A callout function now has three choices for what it returns:
2316
2317 0 => success, carry on matching
2318 > 0 => failure at this point, but backtrack if possible
2319 < 0 => serious error, return this value from pcre_exec()
2320
2321 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2322 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2323 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2324 use by callout functions. It will never be used by PCRE itself.
2325
2326 (ii) The pcre_extra structure (see 45 above) has a void * field called
2327 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2328 pcre_callout_block structure has a field of the same name. The contents of
2329 the field passed in the pcre_extra structure are passed to the callout
2330 function in the corresponding field in the callout block. This makes it
2331 easier to use the same callout-containing regex from multiple threads. For
2332 testing, the pcretest program has a new data escape
2333
2334 \C*n pass the number n (may be negative) as callout_data
2335
2336 If the callout function in pcretest receives a non-zero value as
2337 callout_data, it returns that value.
2338
2339 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2340 there were some redundant $(CFLAGS) in commands that are now specified as
2341 $(LINK), which already includes $(CFLAGS).
2342
2343 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2344 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2345 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2346 one-byte characters throughout. Note that case-insensitive matching applies
2347 only to characters whose values are less than 256. PCRE doesn't support the
2348 notion of cases for higher-valued characters.
2349
2350 (i) A character class whose characters are all within 0-255 is handled as
2351 a bit map, and the map is inverted for negative classes. Previously, a
2352 character > 255 always failed to match such a class; however it should
2353 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2354
2355 (ii) A negated character class with a single character < 255 is coded as
2356 "not this character" (OP_NOT). This wasn't working properly when the test
2357 character was multibyte, either singly or repeated.
2358
2359 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2360 mode, for example: \x{100}{2,3}.
2361
2362 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2363 singly or repeated) now correctly test multibyte characters. However,
2364 PCRE doesn't recognize any characters with values greater than 255 as
2365 digits, spaces, or word characters. Such characters always match \D, \S,
2366 and \W, and never match \d, \s, or \w.
2367
2368 (v) Classes may now contain characters and character ranges with values
2369 greater than 255. For example: [ab\x{100}-\x{400}].
2370
2371 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2372 PCRE in UTF-8 mode.
2373
2374 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2375 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2376 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2377 value.)
2378
2379 53. The single man page has become too large. I have therefore split it up into
2380 a number of separate man pages. These also give rise to individual HTML pages;
2381 these are now put in a separate directory, and there is an index.html page that
2382 lists them all. Some hyperlinking between the pages has been installed.
2383
2384 54. Added convenience functions for handling named capturing parentheses.
2385
2386 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2387 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2388 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2389 are faulted.
2390
2391 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2392 calling configure. These values are used when compiling the dftables.c program
2393 which is run to generate the source of the default character tables. They
2394 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2395 you will need to set these values.
2396
2397 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2398
2399
2400 Version 3.9 02-Jan-02
2401 ---------------------
2402
2403 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2404
2405 2. If --disable-static was given, the building process failed when trying to
2406 build pcretest and pcregrep. (For some reason it was using libtool to compile
2407 them, which is not right, as they aren't part of the library.)
2408
2409
2410 Version 3.8 18-Dec-01
2411 ---------------------
2412
2413 1. The experimental UTF-8 code was completely screwed up. It was packing the
2414 bytes in the wrong order. How dumb can you get?
2415
2416
2417 Version 3.7 29-Oct-01
2418 ---------------------
2419
2420 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2421 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2422 this didn't happen under Solaris 8, where I normally test things.
2423
2424 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2425 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2426 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2427 this fix an hour or so after the initial 3.7 release.)
2428
2429
2430 Version 3.6 23-Oct-01
2431 ---------------------
2432
2433 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2434 offsets passed as NULL with zero offset count.
2435
2436 2. The config.guess and config.sub files had not been updated when I moved to
2437 the latest autoconf.
2438
2439
2440 Version 3.5 15-Aug-01
2441 ---------------------
2442
2443 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2444 had been forgotten.
2445
2446 2. By using declared but undefined structures, we can avoid using "void"
2447 definitions in pcre.h while keeping the internal definitions of the structures
2448 private.
2449
2450 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2451 user point of view, this means that both static and shared libraries are built
2452 by default, but this can be individually controlled. More of the work of
2453 handling this static/shared cases is now inside libtool instead of PCRE's make
2454 file.
2455
2456 4. The pcretest utility is now installed along with pcregrep because it is
2457 useful for users (to test regexs) and by doing this, it automatically gets
2458 relinked by libtool. The documentation has been turned into a man page, so
2459 there are now .1, .txt, and .html versions in /doc.
2460
2461 5. Upgrades to pcregrep:
2462 (i) Added long-form option names like gnu grep.
2463 (ii) Added --help to list all options with an explanatory phrase.
2464 (iii) Added -r, --recursive to recurse into sub-directories.
2465 (iv) Added -f, --file to read patterns from a file.
2466
2467 6. pcre_exec() was referring to its "code" argument before testing that
2468 argument for NULL (and giving an error if it was NULL).
2469
2470 7. Upgraded Makefile.in to allow for compiling in a different directory from
2471 the source directory.
2472
2473 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2474 options bits, the pointer it was passed was to an int instead of to an unsigned
2475 long int. This mattered only on 64-bit systems.
2476
2477 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2478 generated) instead of pcre.in, which it its source. Also made the same change
2479 in several of the .c files.
2480
2481 10. A new release of gcc defines printf() as a macro, which broke pcretest
2482 because it had an ifdef in the middle of a string argument for printf(). Fixed
2483 by using separate calls to printf().
2484
2485 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2486 script, to force use of CR or LF instead of \n in the source. On non-Unix
2487 systems, the value can be set in config.h.
2488
2489 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2490 absolute limit. Changed the text of the error message to make this clear, and
2491 likewise updated the man page.
2492
2493 13. The limit of 99 on the number of capturing subpatterns has been removed.
2494 The new limit is 65535, which I hope will not be a "real" limit.
2495
2496
2497 Version 3.4 22-Aug-00
2498 ---------------------
2499
2500 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2501
2502 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2503
2504
2505 Version 3.3 01-Aug-00
2506 ---------------------
2507
2508 1. If an octal character was given, but the value was greater than \377, it
2509 was not getting masked to the least significant bits, as documented. This could
2510 lead to crashes in some systems.
2511
2512 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2513 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2514
2515 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2516 These just pass their arguments on to (pcre_free)(), but they are provided
2517 because some uses of PCRE bind it to non-C systems that can call its functions,
2518 but cannot call free() or pcre_free() directly.
2519
2520 4. Add "make test" as a synonym for "make check". Corrected some comments in
2521 the Makefile.
2522
2523 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2524 Makefile.
2525
2526 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2527 command called pgrep for grepping around the active processes.
2528
2529 7. Added the beginnings of support for UTF-8 character strings.
2530
2531 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2532 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2533 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2534 out for the ar command.)
2535
2536
2537 Version 3.2 12-May-00
2538 ---------------------
2539
2540 This is purely a bug fixing release.
2541
2542 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2543 of ZA. This was just one example of several cases that could provoke this bug,
2544 which was introduced by change 9 of version 2.00. The code for breaking
2545 infinite loops after an iteration that matches an empty string was't working
2546 correctly.
2547
2548 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2549 when matched against abbab (for example). After matching an empty string, it
2550 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2551 caused it to match further down the string than it should.
2552
2553 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2554 was there because it doesn't seem to be needed, and it causes trouble on some
2555 systems, as it is not a Standard C header. It has been removed.
2556
2557 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2558 were reported on the Macintosh. The changes were from
2559
2560 while ((c = *(++ptr)) != 0 && c != '\n');
2561 to
2562 while ((c = *(++ptr)) != 0 && c != '\n') ;
2563
2564 Totally extraordinary, but if that's what it takes...
2565
2566 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2567 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2568 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2569 assumes the way PCRE uses memmove() (always moving upwards).
2570
2571 6. PCRE is being used in one environment where strchr() is not available. There
2572 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2573 faster code anyway.
2574
2575
2576 Version 3.1 09-Feb-00
2577 ---------------------
2578
2579 The only change in this release is the fixing of some bugs in Makefile.in for
2580 the "install" target:
2581
2582 (1) It was failing to install pcreposix.h.
2583
2584 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2585
2586
2587 Version 3.0 01-Feb-00
2588 ---------------------
2589
2590 1. Add support for the /+ modifier to perltest (to output $` like it does in
2591 pcretest).
2592
2593 2. Add support for the /g modifier to perltest.
2594
2595 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2596 matches null strings.
2597
2598 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2599 pattern. Perl treats empty patterns specially - it reuses the most recent
2600 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2601 effect.
2602
2603 5. The POSIX interface was broken in that it was just handing over the POSIX
2604 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2605 required a bigger vector, with some working space on the end. This means that
2606 the POSIX wrapper now has to get and free some memory, and copy the results.
2607
2608 6. Added some simple autoconf support, placing the test data and the
2609 documentation in separate directories, re-organizing some of the
2610 information files, and making it build pcre-config (a GNU standard). Also added
2611 libtool support for building PCRE as a shared library, which is now the
2612 default.
2613
2614 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2615 09 are not valid octal constants. Single digits will be used for minor values
2616 less than 10.
2617
2618 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2619 existing programs that set these in the POSIX interface can use PCRE without
2620 modification.
2621
2622 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2623 return all that pcre_info() returns, plus additional data. The pcre_info()
2624 function is retained for compatibility, but is considered to be obsolete.
2625
2626 10. Added experimental recursion feature (?R) to handle one common case that
2627 Perl 5.6 will be able to do with (?p{...}).
2628
2629 11. Added support for POSIX character classes like [:alpha:], which Perl is
2630 adopting.
2631
2632
2633 Version 2.08 31-Aug-99
2634 ----------------------
2635
2636 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2637 trying to match at the startoffset position, but instead was moving forward to
2638 the next newline as if a previous match had failed.
2639
2640 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2641 and could get into a loop if a null string was matched other than at the start
2642 of the subject.
2643
2644 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2645 be distinguished at compile time, and for completeness also added PCRE_DATE.
2646
2647 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2648 in GnuWin32 environments.
2649
2650
2651 Version 2.07 29-Jul-99
2652 ----------------------
2653
2654 1. The documentation is now supplied in plain text form and HTML as well as in
2655 the form of man page sources.
2656
2657 2. C++ compilers don't like assigning (void *) values to other pointer types.
2658 In particular this affects malloc(). Although there is no problem in Standard
2659 C, I've put in casts to keep C++ compilers happy.
2660
2661 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2662 should be (const char *).
2663
2664 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2665 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2666 However, I haven't made this a standard facility. The documentation doesn't
2667 mention it, and the Makefile doesn't support it.
2668
2669 5. The Makefile now contains an "install" target, with editable destinations at
2670 the top of the file. The pcretest program is not installed.
2671
2672 6. pgrep -V now gives the PCRE version number and date.
2673
2674 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2675 causing the entire string to be ignored, instead of just the last character.
2676
2677 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2678 non-matching string, it can take a very, very long time, even for strings of
2679 quite modest length, because of the nested recursion. PCRE now does better in
2680 some of these cases. It does this by remembering the last required literal
2681 character in the pattern, and pre-searching the subject to ensure it is present
2682 before running the real match. In other words, it applies a heuristic to detect
2683 some types of certain failure quickly, and in the above example, if presented
2684 with a string that has no trailing " it gives "no match" very quickly.
2685
2686 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2687 other alternatives are tried instead.
2688
2689
2690 Version 2.06 09-Jun-99
2691 ----------------------
2692
2693 1. Change pcretest's output for amount of store used to show just the code
2694 space, because the remainder (the data block) varies in size between 32-bit and
2695 64-bit systems.
2696
2697 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2698 start matching at. This allows lookbehinds to work when searching for multiple
2699 occurrences in a string.
2700
2701 3. Added additional options to pcretest for testing multiple occurrences:
2702
2703 /+ outputs the rest of the string that follows a match
2704 /g loops for multiple occurrences, using the new startoffset argument
2705 /G loops for multiple occurrences by passing an incremented pointer
2706
2707 4. PCRE wasn't doing the "first character" optimization for patterns starting
2708 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2709 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2710 the letter 'x'. On long subject strings, this gives a significant speed-up.
2711
2712
2713 Version 2.05 21-Apr-99
2714 ----------------------
2715
2716 1. Changed the type of magic_number from int to long int so that it works
2717 properly on 16-bit systems.
2718
2719 2. Fixed a bug which caused patterns starting with .* not to work correctly
2720 when the subject string contained newline characters. PCRE was assuming
2721 anchoring for such patterns in all cases, which is not correct because .* will
2722 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2723 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2724 must be retried after every newline in the subject.
2725
2726
2727 Version 2.04 18-Feb-99
2728 ----------------------
2729
2730 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2731 computation of the store needed to hold the pattern was incorrect (too large).
2732 If such patterns were nested a few deep, this could multiply and become a real
2733 problem.
2734
2735 2. Added /M option to pcretest to show the memory requirement of a specific
2736 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2737
2738 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2739 compiled in such a way that the backtracking after subsequent failure was
2740 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2741 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2742
2743
2744 Version 2.03 02-Feb-99
2745 ----------------------
2746
2747 1. Fixed typo and small mistake in man page.
2748
2749 2. Added 4th condition (GPL supersedes if conflict) and created separate
2750 LICENCE file containing the conditions.
2751
2752 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2753 Perl, that is the internal \ allows the delimiter to be included in the
2754 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2755 the final delimiter, add \ to the end of the pattern (to test the error).
2756
2757 4. Added the convenience functions for extracting substrings after a successful
2758 match. Updated pcretest to make it able to test these functions.
2759
2760
2761 Version 2.02 14-Jan-99
2762 ----------------------
2763
2764 1. Initialized the working variables associated with each extraction so that
2765 their saving and restoring doesn't refer to uninitialized store.
2766
2767 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2768 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2769 fix the problem.
2770
2771 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2772 calls, and wasn't printing the correct value for compiling calls. Increased the
2773 default value of LOOPREPEAT, and the number of significant figures in the
2774 times.
2775
2776 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2777
2778 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2779 a building problem on Windows NT with a FAT file system.
2780
2781
2782 Version 2.01 21-Oct-98
2783 ----------------------
2784
2785 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2786 to character tables built by pcre_maketables() in the current locale. If NULL
2787 is passed, the default tables are used.
2788
2789
2790 Version 2.00 24-Sep-98
2791 ----------------------
2792
2793 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2794 it any more.
2795
2796 2. Allow quantification of (?>) groups, and make it work correctly.
2797
2798 3. The first character computation wasn't working for (?>) groups.
2799
2800 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2801 end of the subject) and add 5.005's \z, which really does match only at the
2802 very end of the subject.
2803
2804 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2805
2806 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2807 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2808 localized options. All options to pcre_study() were also removed.
2809
2810 7. Add other new features from 5.005:
2811
2812 $(?<= positive lookbehind
2813 $(?<! negative lookbehind
2814 (?imsx-imsx) added the unsetting capability
2815 such a setting is global if at outer level; local otherwise
2816 (?imsx-imsx:) non-capturing groups with option setting
2817 (?(cond)re|re) conditional pattern matching
2818
2819 A backreference to itself in a repeated group matches the previous
2820 captured string.
2821
2822 8. General tidying up of studying (both automatic and via "study")
2823 consequential on the addition of new assertions.
2824
2825 9. As in 5.005, unlimited repeated groups that could match an empty substring
2826 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2827 runtime if any iteration does actually match an empty substring.
2828
2829 10. Include the RunTest script in the distribution.
2830
2831 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2832 discrepancies, some of which were old and were also with respect to 5.004. They
2833 have now been fixed.
2834
2835
2836 Version 1.09 28-Apr-98
2837 ----------------------
2838
2839 1. A negated single character class followed by a quantifier with a minimum
2840 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2841 program crashes, or just wrong answers. This did not apply to negated classes
2842 containing more than one character, or to minima other than one.
2843
2844
2845 Version 1.08 27-Mar-98
2846 ----------------------
2847
2848 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2849
2850 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2851 latter must appear before anything that relies on it in the pattern.
2852
2853
2854 Version 1.07 16-Feb-98
2855 ----------------------
2856
2857 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2858 repeat of a potentially empty string).
2859
2860
2861 Version 1.06 23-Jan-98
2862 ----------------------
2863
2864 1. Added Markus Oberhumer's little patches for C++.
2865
2866 2. Literal strings longer than 255 characters were broken.
2867
2868
2869 Version 1.05 23-Dec-97
2870 ----------------------
2871
2872 1. Negated character classes containing more than one character were failing if
2873 PCRE_CASELESS was set at run time.
2874
2875
2876 Version 1.04 19-Dec-97
2877 ----------------------
2878
2879 1. Corrected the man page, where some "const" qualifiers had been omitted.
2880
2881 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2882 input syntax.
2883
2884 3. Fixed memory leak which occurred when a regex with back references was
2885 matched with an offsets vector that wasn't big enough. The temporary memory
2886 that is used in this case wasn't being freed if the match failed.
2887
2888 4. Tidied pcretest to ensure it frees memory that it gets.
2889
2890 5. Temporary memory was being obtained in the case where the passed offsets
2891 vector was exactly big enough.
2892
2893 6. Corrected definition of offsetof() from change 5 below.
2894
2895 7. I had screwed up change 6 below and broken the rules for the use of
2896 setjmp(). Now fixed.
2897
2898
2899 Version 1.03 18-Dec-97
2900 ----------------------
2901
2902 1. A erroneous regex with a missing opening parenthesis was correctly
2903 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2904 on some systems.
2905
2906 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2907 it was reported that one broken compiler failed on the former because "code" is
2908 also an independent variable.
2909
2910 3. The erroneous regex a[]b caused an array overrun reference.
2911
2912 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2913 fail on data ending with that character. (It was going on too far, and checking
2914 the next character, typically a binary zero.) This was specific to the
2915 optimized code for single-character negative classes.
2916
2917 5. Added a contributed patch from the TIN world which does the following:
2918
2919 + Add an undef for memmove, in case the the system defines a macro for it.
2920
2921 + Add a definition of offsetof(), in case there isn't one. (I don't know
2922 the reason behind this - offsetof() is part of the ANSI standard - but
2923 it does no harm).
2924
2925 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2926 most of the places where whitespace preceded '#'. I have given up and
2927 allowed the remaining 2 cases to be at the margin.
2928
2929 + Rename some variables in pcre to eliminate shadowing. This seems very
2930 pedantic, but does no harm, of course.
2931
2932 6. Moved the call to setjmp() into its own function, to get rid of warnings
2933 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2934
2935 7. Constructs such as \d{8,} were compiling into the equivalent of
2936 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2937 outcome, but in this particular case used more store than had been allocated,
2938 which caused the bug to be discovered because it threw up an internal error.
2939
2940 8. The debugging code in both pcre and pcretest for outputting the compiled
2941 form of a regex was going wrong in the case of back references followed by
2942 curly-bracketed repeats.
2943
2944
2945 Version 1.02 12-Dec-97
2946 ----------------------
2947
2948 1. Typos in pcre.3 and comments in the source fixed.
2949
2950 2. Applied a contributed patch to get rid of places where it used to remove
2951 'const' from variables, and fixed some signed/unsigned and uninitialized
2952 variable warnings.
2953
2954 3. Added the "runtest" target to Makefile.
2955
2956 4. Set default compiler flag to -O2 rather than just -O.
2957
2958
2959 Version 1.01 19-Nov-97
2960 ----------------------
2961
2962 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2963 like /([ab]*)*/, that is, for classes with more than one character in them.
2964
2965 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2966 as /((?>a*))*/ (a PCRE_EXTRA facility).
2967
2968
2969 Version 1.00 18-Nov-97
2970 ----------------------
2971
2972 1. Added compile-time macros to support systems such as SunOS4 which don't have
2973 memmove() or strerror() but have other things that can be used instead.
2974
2975 2. Arranged that "make clean" removes the executables.
2976
2977
2978 Version 0.99 27-Oct-97
2979 ----------------------
2980
2981 1. Fixed bug in code for optimizing classes with only one character. It was
2982 initializing a 32-byte map regardless, which could cause it to run off the end
2983 of the memory it had got.
2984
2985 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2986
2987
2988 Version 0.98 22-Oct-97
2989 ----------------------
2990
2991 1. Fixed bug in code for handling temporary memory usage when there are more
2992 back references than supplied space in the ovector. This could cause segfaults.
2993
2994
2995 Version 0.97 21-Oct-97
2996 ----------------------
2997
2998 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2999
3000 2. Optimized negated single characters not to use a bit map.
3001
3002 3. Brought error texts together as macro definitions; clarified some of them;
3003 fixed one that was wrong - it said "range out of order" when it meant "invalid
3004 escape sequence".
3005
3006 4. Changed some char * arguments to const char *.
3007
3008 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3009
3010 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3011 pcretest.
3012
3013
3014 Version 0.96 16-Oct-97
3015 ----------------------
3016
3017 1. Added a simple "pgrep" utility to the distribution.
3018
3019 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3020 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3021 where "ddd" means "one or more decimal digits".
3022
3023 3. Fixed serious bug. If a pattern had a back reference, but the call to
3024 pcre_exec() didn't supply a large enough ovector to record the related
3025 identifying subpattern, the match always failed. PCRE now remembers the number
3026 of the largest back reference, and gets some temporary memory in which to save
3027 the offsets during matching if necessary, in order to ensure that
3028 backreferences always work.
3029
3030 4. Increased the compatibility with Perl in a number of ways:
3031
3032 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3033 to request this handling. The option can be set at compile or exec time.
3034
3035 (b) $ matches before a terminating newline by default; an option
3036 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3037 mode). The option can be set at compile or exec time.
3038
3039 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3040 the same as Perl's. If the decimal number it represents is less than 10
3041 or there aren't that many previous left capturing parentheses, an octal
3042 escape is read. Inside a character class, it's always an octal escape,
3043 even if it is a single digit.
3044
3045 (d) An escaped but undefined alphabetic character is taken as a literal,
3046 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3047 escapes.
3048
3049 (e) {0} is now permitted. (The previous item is removed from the compiled
3050 pattern).
3051
3052 5. Changed all the names of code files so that the basic parts are no longer
3053 than 10 characters, and abolished the teeny "globals.c" file.
3054
3055 6. Changed the handling of character classes; they are now done with a 32-byte
3056 bit map always.
3057
3058 7. Added the -d and /D options to pcretest to make it possible to look at the
3059 internals of compilation without having to recompile pcre.
3060
3061
3062 Version 0.95 23-Sep-97
3063 ----------------------
3064
3065 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3066 \x20 at the start of a run of normal characters. These were being treated as
3067 real characters, instead of the source characters being re-checked.
3068
3069
3070 Version 0.94 18-Sep-97
3071 ----------------------
3072
3073 1. The functions are now thread-safe, with the caveat that the global variables
3074 containing pointers to malloc() and free() or alternative functions are the
3075 same for all threads.
3076
3077 2. Get pcre_study() to generate a bitmap of initial characters for non-
3078 anchored patterns when this is possible, and use it if passed to pcre_exec().
3079
3080
3081 Version 0.93 15-Sep-97
3082 ----------------------
3083
3084 1. /(b)|(:+)/ was computing an incorrect first character.
3085
3086 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3087 but not actually doing anything yet.
3088
3089 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3090 as Perl does (e.g. [-az] or [az-]).
3091
3092 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3093 all possible positions.
3094
3095 5. Split up into different modules to avoid including unneeded functions in a
3096 compiled binary. However, compile and exec are still in one module. The "study"
3097 function is split off.
3098
3099 6. The character tables are now in a separate module whose source is generated
3100 by an auxiliary program - but can then be edited by hand if required. There are
3101 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3102 toupper() in the code.
3103
3104 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3105 make them global. Abolish the function for setting them, as the caller can now
3106 set them directly.
3107
3108
3109 Version 0.92 11-Sep-97
3110 ----------------------
3111
3112 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3113 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3114
3115 2. Caseless matching was not working in character classes if the characters in
3116 the pattern were in upper case.
3117
3118 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3119
3120 4. Make PCRE_ANCHORED public and accept as a compile option.
3121
3122 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3123 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3124 pass them.
3125
3126 6. Give an error if bad option bits passed at compile or run time.
3127
3128 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3129 pcretest to cause it to pass that flag.
3130
3131 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3132 options, and the first character, if set.
3133
3134 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3135
3136
3137 Version 0.91 10-Sep-97
3138 ----------------------
3139
3140 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3141 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3142
3143 2. PCRE was looping on encountering an indefinitely repeated back reference to
3144 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3145 Perl does - treats the match as successful.
3146
3147 ****

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