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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 313 - (show annotations) (download)
Wed Jan 23 18:05:06 2008 UTC (6 years, 11 months ago) by ph10
File size: 150383 byte(s)
Tidies for 7.6-RC2.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.6 21-Jan-08
5 ---------------------
6
7 1. A character class containing a very large number of characters with
8 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
9 overflow.
10
11 2. Patch to cut out the "long long" test in pcrecpp_unittest when
12 HAVE_LONG_LONG is not defined.
13
14 3. Applied Christian Ehrlicher's patch to update the CMake build files to
15 bring them up to date and include new features. This patch includes:
16
17 - Fixed PH's badly added libz and libbz2 support.
18 - Fixed a problem with static linking.
19 - Added pcredemo. [But later removed - see 7 below.]
20 - Fixed dftables problem and added an option.
21 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
22 HAVE_LONG_LONG.
23 - Added readline support for pcretest.
24 - Added an listing of the option settings after cmake has run.
25
26 4. A user submitted a patch to Makefile that makes it easy to created a dll
27 under mingw. I added stuff to Makefile.am that cause it to include this
28 special target, without affecting anything else.
29
30 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
31 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
32 exported in the Windows port". It has not yet been confirmed that the patch
33 solves the problem, but it does no harm.
34
35 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
36 NON_STANDARD_LIB_SUFFIX and correct the comment about stack recursion.
37
38 7. Remove the automatic building of pcredemo from the ./configure system and
39 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
40 of a program that users should build themselves after PCRE is installed, so
41 building it automatically is not really right. What is more, it gave
42 trouble in some build environments.
43
44
45 Version 7.5 10-Jan-08
46 ---------------------
47
48 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
49 values in parens when parsing an RE using the C++ wrapper."
50
51 2. Negative specials like \S did not work in character classes in UTF-8 mode.
52 Characters greater than 255 were excluded from the class instead of being
53 included.
54
55 3. The same bug as (2) above applied to negated POSIX classes such as
56 [:^space:].
57
58 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
59 defined or documented. It seems to have been a typo for PCRE_STATIC, so
60 I have changed it.
61
62 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
63 first named subpattern) and a construct such as (?&a) would reference the
64 first named subpattern whose name started with "a" (in other words, the
65 length check was missing). Both these problems are fixed. "Subpattern name
66 expected" is now given for (?&) (a zero-length name), and this patch also
67 makes it give the same error for \k'' (previously it complained that that
68 was a reference to a non-existent subpattern).
69
70 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
71 this is right because (?- can be followed by option settings as well as by
72 digits. I have, however, made the messages clearer.
73
74 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
75 than the number used in the conditional) now cause a compile-time error.
76 This is actually not compatible with Perl, which accepts such patterns, but
77 treats the conditional as always being FALSE (as PCRE used to), but it
78 seems to me that giving a diagnostic is better.
79
80 8. Change "alphameric" to the more common word "alphanumeric" in comments
81 and messages.
82
83 9. Fix two occurrences of "backslash" in comments that should have been
84 "backspace".
85
86 10. Remove two redundant lines of code that can never be obeyed (their function
87 was moved elsewhere).
88
89 11. The program that makes PCRE's Unicode character property table had a bug
90 which caused it to generate incorrect table entries for sequences of
91 characters that have the same character type, but are in different scripts.
92 It amalgamated them into a single range, with the script of the first of
93 them. In other words, some characters were in the wrong script. There were
94 thirteen such cases, affecting characters in the following ranges:
95
96 U+002b0 - U+002c1
97 U+0060c - U+0060d
98 U+0061e - U+00612
99 U+0064b - U+0065e
100 U+0074d - U+0076d
101 U+01800 - U+01805
102 U+01d00 - U+01d77
103 U+01d9b - U+01dbf
104 U+0200b - U+0200f
105 U+030fc - U+030fe
106 U+03260 - U+0327f
107 U+0fb46 - U+0fbb1
108 U+10450 - U+1049d
109
110 12. The -o option (show only the matching part of a line) for pcregrep was not
111 compatible with GNU grep in that, if there was more than one match in a
112 line, it showed only the first of them. It now behaves in the same way as
113 GNU grep.
114
115 13. If the -o and -v options were combined for pcregrep, it printed a blank
116 line for every non-matching line. GNU grep prints nothing, and pcregrep now
117 does the same. The return code can be used to tell if there were any
118 non-matching lines.
119
120 14. Added --file-offsets and --line-offsets to pcregrep.
121
122 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
123 infinitely looping recursion. The bug was that positive lookaheads were not
124 being skipped when checking for a possible empty match (negative lookaheads
125 and both kinds of lookbehind were skipped).
126
127 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
128 inclusion of <windows.h> to before rather than after the definition of
129 INVALID_FILE_ATTRIBUTES (patch from David Byron).
130
131 17. Specifying a possessive quantifier with a specific limit for a Unicode
132 character property caused pcre_compile() to compile bad code, which led at
133 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
134 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
135 caused the error; without that there was no problem.
136
137 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
138
139 19. Added --enable-pcretest-libreadline.
140
141 20. In pcrecpp.cc, the variable 'count' was incremented twice in
142 RE::GlobalReplace(). As a result, the number of replacements returned was
143 double what it should be. I removed one of the increments, but Craig sent a
144 later patch that removed the other one (the right fix) and added unit tests
145 that check the return values (which was not done before).
146
147 21. Several CMake things:
148
149 (1) Arranged that, when cmake is used on Unix, the libraries end up with
150 the names libpcre and libpcreposix, not just pcre and pcreposix.
151
152 (2) The above change means that pcretest and pcregrep are now correctly
153 linked with the newly-built libraries, not previously installed ones.
154
155 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
156
157 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
158 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
159 UTF-8 newline character). The key issue is that the pattern starts .*;
160 this means that the match must be either at the beginning, or after a
161 newline. The bug was in the code for advancing after a failed match and
162 checking that the new position followed a newline. It was not taking
163 account of UTF-8 characters correctly.
164
165 23. PCRE was behaving differently from Perl in the way it recognized POSIX
166 character classes. PCRE was not treating the sequence [:...:] as a
167 character class unless the ... were all letters. Perl, however, seems to
168 allow any characters between [: and :], though of course it rejects as
169 unknown any "names" that contain non-letters, because all the known class
170 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
171 for example, whereas PCRE did not - it did not recognize a POSIX character
172 class. This seemed a bit dangerous, so the code has been changed to be
173 closer to Perl. The behaviour is not identical to Perl, because PCRE will
174 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
175 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
176 Perl does, and where it didn't before.
177
178 24. Rewrite so as to remove the single use of %n from pcregrep because in some
179 Windows environments %n is disabled by default.
180
181
182 Version 7.4 21-Sep-07
183 ---------------------
184
185 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
186 means that a class such as [\s] counted as "explicit reference to CR or
187 LF". That isn't really right - the whole point of the change was to try to
188 help when there was an actual mention of one of the two characters. So now
189 the change happens only if \r or \n (or a literal CR or LF) character is
190 encountered.
191
192 2. The 32-bit options word was also used for 6 internal flags, but the numbers
193 of both had grown to the point where there were only 3 bits left.
194 Fortunately, there was spare space in the data structure, and so I have
195 moved the internal flags into a new 16-bit field to free up more option
196 bits.
197
198 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
199 but did not set the internal JCHANGED flag - either of these is enough to
200 control the way the "get" function works - but the PCRE_INFO_JCHANGED
201 facility is supposed to tell if (?J) was ever used, so now (?J) at the
202 start sets both bits.
203
204 4. Added options (at build time, compile time, exec time) to change \R from
205 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
206
207 5. doc/pcresyntax.html was missing from the distribution.
208
209 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
210 compatibility, even though it is no longer used.
211
212 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
213 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
214 windows.h file is present (where different names are used). [This was
215 reversed later after testing - see 16 below.]
216
217 8. Changed all #include <config.h> to #include "config.h". There were also
218 some further <pcre.h> cases that I changed to "pcre.h".
219
220 9. When pcregrep was used with the --colour option, it missed the line ending
221 sequence off the lines that it output.
222
223 10. It was pointed out to me that arrays of string pointers cause lots of
224 relocations when a shared library is dynamically loaded. A technique of
225 using a single long string with a table of offsets can drastically reduce
226 these. I have refactored PCRE in four places to do this. The result is
227 dramatic:
228
229 Originally: 290
230 After changing UCP table: 187
231 After changing error message table: 43
232 After changing table of "verbs" 36
233 After changing table of Posix names 22
234
235 Thanks to the folks working on Gregex for glib for this insight.
236
237 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
238 unicode-properties was also set.
239
240 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
241
242 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
243 checked only for CRLF.
244
245 14. Added casts to pcretest.c to avoid compiler warnings.
246
247 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
248
249 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
250 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
251 entirely. This removes changes made in 7 above.
252
253 17. The CMake files have been updated, and there is now more information about
254 building with CMake in the NON-UNIX-USE document.
255
256
257 Version 7.3 28-Aug-07
258 ---------------------
259
260 1. In the rejigging of the build system that eventually resulted in 7.1, the
261 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
262 brackets there is not right, since it causes compilers to look for an
263 installed pcre.h, not the version that is in the source that is being
264 compiled (which of course may be different). I have changed it back to:
265
266 #include "pcre.h"
267
268 I have a vague recollection that the change was concerned with compiling in
269 different directories, but in the new build system, that is taken care of
270 by the VPATH setting the Makefile.
271
272 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
273 when the subject happened to end in the byte 0x85 (e.g. if the last
274 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
275 characters but of course it shouldn't be taken as a newline when it is part
276 of another character. The bug was that, for an unlimited repeat of . in
277 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
278 characters when looking for a newline.
279
280 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
281
282 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
283 in debug output.
284
285 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
286 long printing in the pcrecpp unittest when running under MinGW.
287
288 6. ESC_K was left out of the EBCDIC table.
289
290 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
291 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
292 limit also applies to "virtual nesting" when a pattern is recursive, and in
293 this case 1000 isn't so big. I have been able to remove this limit at the
294 expense of backing off one optimization in certain circumstances. Normally,
295 when pcre_exec() would call its internal match() function recursively and
296 immediately return the result unconditionally, it uses a "tail recursion"
297 feature to save stack. However, when a subpattern that can match an empty
298 string has an unlimited repetition quantifier, it no longer makes this
299 optimization. That gives it a stack frame in which to save the data for
300 checking that an empty string has been matched. Previously this was taken
301 from the 1000-entry workspace that had been reserved. So now there is no
302 explicit limit, but more stack is used.
303
304 8. Applied Daniel's patches to solve problems with the import/export magic
305 syntax that is required for Windows, and which was going wrong for the
306 pcreposix and pcrecpp parts of the library. These were overlooked when this
307 problem was solved for the main library.
308
309 9. There were some crude static tests to avoid integer overflow when computing
310 the size of patterns that contain repeated groups with explicit upper
311 limits. As the maximum quantifier is 65535, the maximum group length was
312 set at 30,000 so that the product of these two numbers did not overflow a
313 32-bit integer. However, it turns out that people want to use groups that
314 are longer than 30,000 bytes (though not repeat them that many times).
315 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
316 made it possible to implement the integer overflow checks in a much more
317 dynamic way, which I have now done. The artificial limitation on group
318 length has been removed - we now have only the limit on the total length of
319 the compiled pattern, which depends on the LINK_SIZE setting.
320
321 10. Fixed a bug in the documentation for get/copy named substring when
322 duplicate names are permitted. If none of the named substrings are set, the
323 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
324 empty string.
325
326 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
327 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
328 because the ] is interpreted as the first data character and the
329 terminating ] is not found. PCRE has been made compatible with Perl in this
330 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
331 cause memory overwriting.
332
333 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
334 string has been matched (to stop an infinite loop). It was not recognizing
335 a conditional subpattern that could match an empty string if that
336 subpattern was within another subpattern. For example, it looped when
337 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
338 condition was not nested. This bug has been fixed.
339
340 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
341 past the start of the subject in the presence of bytes with the top bit
342 set, for example "\x8aBCD".
343
344 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
345 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
346
347 14. Optimized (?!) to (*FAIL).
348
349 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
350 This restricts code points to be within the range 0 to 0x10FFFF, excluding
351 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
352 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
353 does: it's just the validity check that is more restrictive.
354
355 16. Inserted checks for integer overflows during escape sequence (backslash)
356 processing, and also fixed erroneous offset values for syntax errors during
357 backslash processing.
358
359 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
360 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
361
362 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
363 caused an overrun.
364
365 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
366 something other than just ASCII characters) inside a group that had an
367 unlimited repeat caused a loop at compile time (while checking to see
368 whether the group could match an empty string).
369
370 20. Debugging a pattern containing \p or \P could cause a crash. For example,
371 [\P{Any}] did so. (Error in the code for printing property names.)
372
373 21. An orphan \E inside a character class could cause a crash.
374
375 22. A repeated capturing bracket such as (A)? could cause a wild memory
376 reference during compilation.
377
378 23. There are several functions in pcre_compile() that scan along a compiled
379 expression for various reasons (e.g. to see if it's fixed length for look
380 behind). There were bugs in these functions when a repeated \p or \P was
381 present in the pattern. These operators have additional parameters compared
382 with \d, etc, and these were not being taken into account when moving along
383 the compiled data. Specifically:
384
385 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
386 length.
387
388 (b) An item such as \pL+ within a repeated group could cause crashes or
389 loops.
390
391 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
392 "reference to non-existent subpattern" error.
393
394 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
395
396 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
397 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
398
399 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
400
401 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
402 character were causing crashes (broken optimization).
403
404 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
405 \p or \P) caused a compile-time loop.
406
407 28. More problems have arisen in unanchored patterns when CRLF is a valid line
408 break. For example, the unstudied pattern [\r\n]A does not match the string
409 "\r\nA" because change 7.0/46 below moves the current point on by two
410 characters after failing to match at the start. However, the pattern \nA
411 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
412 the same is true. There doesn't seem any very clean way out of this, but
413 what I have chosen to do makes the common cases work: PCRE now takes note
414 of whether there can be an explicit match for \r or \n anywhere in the
415 pattern, and if so, 7.0/46 no longer applies. As part of this change,
416 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
417 pattern has explicit CR or LF references.
418
419 29. Added (*CR) etc for changing newline setting at start of pattern.
420
421
422 Version 7.2 19-Jun-07
423 ---------------------
424
425 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
426 which is apparently normally available under Windows.
427
428 2. Re-jig the pcregrep tests with different newline settings in an attempt
429 to make them independent of the local environment's newline setting.
430
431 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
432
433 4. Some of the "internals" tests were previously cut out when the link size
434 was not 2, because the output contained actual offsets. The recent new
435 "Z" feature of pcretest means that these can be cut out, making the tests
436 usable with all link sizes.
437
438 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
439 stack recursion. This gives a massive performance boost under BSD, but just
440 a small improvement under Linux. However, it saves one field in the frame
441 in all cases.
442
443 6. Added more features from the forthcoming Perl 5.10:
444
445 (a) (?-n) (where n is a string of digits) is a relative subroutine or
446 recursion call. It refers to the nth most recently opened parentheses.
447
448 (b) (?+n) is also a relative subroutine call; it refers to the nth next
449 to be opened parentheses.
450
451 (c) Conditions that refer to capturing parentheses can be specified
452 relatively, for example, (?(-2)... or (?(+3)...
453
454 (d) \K resets the start of the current match so that everything before
455 is not part of it.
456
457 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
458
459 (f) \g{name} is another synonym - part of Perl 5.10's unification of
460 reference syntax.
461
462 (g) (?| introduces a group in which the numbering of parentheses in each
463 alternative starts with the same number.
464
465 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
466
467 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
468 PCRE_INFO_JCHANGED.
469
470 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
471 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
472 for detecting groups that can match an empty string.
473
474 9. A pattern with a very large number of alternatives (more than several
475 hundred) was running out of internal workspace during the pre-compile
476 phase, where pcre_compile() figures out how much memory will be needed. A
477 bit of new cunning has reduced the workspace needed for groups with
478 alternatives. The 1000-alternative test pattern now uses 12 bytes of
479 workspace instead of running out of the 4096 that are available.
480
481 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
482
483 11. Applied patch from Google to remove an optimization that didn't quite work.
484 The report of the bug said:
485
486 pcrecpp::RE("a*").FullMatch("aaa") matches, while
487 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
488 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
489
490 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
491 it matched the wrong number of bytes.
492
493
494 Version 7.1 24-Apr-07
495 ---------------------
496
497 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
498 that is more "standard", making use of automake and other Autotools. There
499 is some re-arrangement of the files and adjustment of comments consequent
500 on this.
501
502 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
503 for recursive directory scanning broke on some systems because the files
504 are not scanned in any specific order and on different systems the order
505 was different. A call to "sort" has been inserted into RunGrepTest for the
506 approprate test as a short-term fix. In the longer term there may be an
507 alternative.
508
509 3. I had an email from Eric Raymond about problems translating some of PCRE's
510 man pages to HTML (despite the fact that I distribute HTML pages, some
511 people do their own conversions for various reasons). The problems
512 concerned the use of low-level troff macros .br and .in. I have therefore
513 removed all such uses from the man pages (some were redundant, some could
514 be replaced by .nf/.fi pairs). The 132html script that I use to generate
515 HTML has been updated to handle .nf/.fi and to complain if it encounters
516 .br or .in.
517
518 4. Updated comments in configure.ac that get placed in config.h.in and also
519 arranged for config.h to be included in the distribution, with the name
520 config.h.generic, for the benefit of those who have to compile without
521 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
522
523 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
524 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
525 makevp.bat for the current PCRE, using the additional files
526 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
527
528 6. A Windows user reported a minor discrepancy with test 2, which turned out
529 to be caused by a trailing space on an input line that had got lost in his
530 copy. The trailing space was an accident, so I've just removed it.
531
532 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
533 that is needed.
534
535 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
536 as "const" (a) because they are and (b) because it helps the PHP
537 maintainers who have recently made a script to detect big data structures
538 in the php code that should be moved to the .rodata section. I remembered
539 to update Builducptable as well, so it won't revert if ucptable.h is ever
540 re-created.
541
542 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
543 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
544 order to be able to cut out the UTF-8 tables in the latter when UTF-8
545 support is not required. This saves 1.5-2K of code, which is important in
546 some applications.
547
548 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
549 so as not to refer to the tables, even though these functions will never be
550 called when UTF-8 support is disabled. Otherwise there are problems with a
551 shared library.
552
553 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
554
555 (a) It was defining its arguments as char * instead of void *.
556
557 (b) It was assuming that all moves were upwards in memory; this was true
558 a long time ago when I wrote it, but is no longer the case.
559
560 The emulated memove() is provided for those environments that have neither
561 memmove() nor bcopy(). I didn't think anyone used it these days, but that
562 is clearly not the case, as these two bugs were recently reported.
563
564 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
565 and Detrail to create the HTML documentation, the .txt form of the man
566 pages, and it removes trailing spaces from listed files. It also creates
567 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
568 case, it wraps all the #defines with #ifndefs. This script should be run
569 before "make dist".
570
571 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
572 with Unicode property support.
573
574 (a) For a maximizing quantifier, if the two different cases of the
575 character were of different lengths in their UTF-8 codings (there are
576 some cases like this - I found 11), and the matching function had to
577 back up over a mixture of the two cases, it incorrectly assumed they
578 were both the same length.
579
580 (b) When PCRE was configured to use the heap rather than the stack for
581 recursion during matching, it was not correctly preserving the data for
582 the other case of a UTF-8 character when checking ahead for a match
583 while processing a minimizing repeat. If the check also involved
584 matching a wide character, but failed, corruption could cause an
585 erroneous result when trying to check for a repeat of the original
586 character.
587
588 13. Some tidying changes to the testing mechanism:
589
590 (a) The RunTest script now detects the internal link size and whether there
591 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
592 values substituted by "configure". (The RunGrepTest script already did
593 this for UTF-8.) The configure.ac script no longer substitutes the
594 relevant variables.
595
596 (b) The debugging options /B and /D in pcretest show the compiled bytecode
597 with length and offset values. This means that the output is different
598 for different internal link sizes. Test 2 is skipped for link sizes
599 other than 2 because of this, bypassing the problem. Unfortunately,
600 there was also a test in test 3 (the locale tests) that used /B and
601 failed for link sizes other than 2. Rather than cut the whole test out,
602 I have added a new /Z option to pcretest that replaces the length and
603 offset values with spaces. This is now used to make test 3 independent
604 of link size. (Test 2 will be tidied up later.)
605
606 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
607 segmentation fault instead of returning the appropriate error message.
608
609 15. In multiline mode when the newline sequence was set to "any", the pattern
610 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
611 This doesn't seem right; it now treats the CRLF combination as the line
612 ending, and so does not match in that case. It's only a pattern such as ^$
613 that would hit this one: something like ^ABC$ would have failed after \r
614 and then tried again after \r\n.
615
616 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
617 in an attempt to make files that differ only in their line terminators
618 compare equal. This works on Linux.
619
620 17. Under certain error circumstances pcregrep might try to free random memory
621 as it exited. This is now fixed, thanks to valgrind.
622
623 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
624 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
625 was because its rules for how to advance for /g after matching an empty
626 string at the end of a line did not allow for this case. They now check for
627 it specially.
628
629 20. pcretest is supposed to handle patterns and data of any length, by
630 extending its buffers when necessary. It was getting this wrong when the
631 buffer for a data line had to be extended.
632
633 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
634 CRLF as a newline sequence.
635
636 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
637 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
638 I have nevertheless tidied it up.
639
640 23. Added some casts to kill warnings from HP-UX ia64 compiler.
641
642 24. Added a man page for pcre-config.
643
644
645 Version 7.0 19-Dec-06
646 ---------------------
647
648 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
649 moving to gcc 4.1.1.
650
651 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
652 sys/time.h, which is documented as needed for this function. It doesn't
653 seem to matter on Linux, but it showed up on some releases of OS X.
654
655 3. It seems that there are systems where bytes whose values are greater than
656 127 match isprint() in the "C" locale. The "C" locale should be the
657 default when a C program starts up. In most systems, only ASCII printing
658 characters match isprint(). This difference caused the output from pcretest
659 to vary, making some of the tests fail. I have changed pcretest so that:
660
661 (a) When it is outputting text in the compiled version of a pattern, bytes
662 other than 32-126 are always shown as hex escapes.
663
664 (b) When it is outputting text that is a matched part of a subject string,
665 it does the same, unless a different locale has been set for the match
666 (using the /L modifier). In this case, it uses isprint() to decide.
667
668 4. Fixed a major bug that caused incorrect computation of the amount of memory
669 required for a compiled pattern when options that changed within the
670 pattern affected the logic of the preliminary scan that determines the
671 length. The relevant options are -x, and -i in UTF-8 mode. The result was
672 that the computed length was too small. The symptoms of this bug were
673 either the PCRE error "internal error: code overflow" from pcre_compile(),
674 or a glibc crash with a message such as "pcretest: free(): invalid next
675 size (fast)". Examples of patterns that provoked this bug (shown in
676 pcretest format) are:
677
678 /(?-x: )/x
679 /(?x)(?-x: \s*#\s*)/
680 /((?i)[\x{c0}])/8
681 /(?i:[\x{c0}])/8
682
683 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
684 is now done differently.
685
686 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
687 wrapper classes; (b) implement a new function in the C++ scanner that is
688 more efficient than the old way of doing things because it avoids levels of
689 recursion in the regex matching; (c) add a paragraph to the documentation
690 for the FullMatch() function.
691
692 6. The escape sequence \n was being treated as whatever was defined as
693 "newline". Not only was this contrary to the documentation, which states
694 that \n is character 10 (hex 0A), but it also went horribly wrong when
695 "newline" was defined as CRLF. This has been fixed.
696
697 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
698 was being set to -1 for the "end of line" case (supposedly a value that no
699 character can have). Though this value is never used (the check for end of
700 line is "zero bytes in current character"), it caused compiler complaints.
701 I've changed it to 0xffffffff.
702
703 8. In pcre_version.c, the version string was being built by a sequence of
704 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
705 string (as it is for production releases) called a macro with an empty
706 argument. The C standard says the result of this is undefined. The gcc
707 compiler treats it as an empty string (which was what was wanted) but it is
708 reported that Visual C gives an error. The source has been hacked around to
709 avoid this problem.
710
711 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
712 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
713 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
714 of them did).
715
716 10. Originally, pcretest opened its input and output without "b"; then I was
717 told that "b" was needed in some environments, so it was added for release
718 5.0 to both the input and output. (It makes no difference on Unix-like
719 systems.) Later I was told that it is wrong for the input on Windows. I've
720 now abstracted the modes into two macros, to make it easier to fiddle with
721 them, and removed "b" from the input mode under Windows.
722
723 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
724
725 12. Added -help and --help to pcretest as an official way of being reminded
726 of the options.
727
728 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
729 and pcrecpp.cc because they annoy compilers at high warning levels.
730
731 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
732
733 15. Fixed an occurrence of == in configure.ac that should have been = (shell
734 scripts are not C programs :-) and which was not noticed because it works
735 on Linux.
736
737 16. pcretest is supposed to handle any length of pattern and data line (as one
738 line or as a continued sequence of lines) by extending its input buffer if
739 necessary. This feature was broken for very long pattern lines, leading to
740 a string of junk being passed to pcre_compile() if the pattern was longer
741 than about 50K.
742
743 17. I have done a major re-factoring of the way pcre_compile() computes the
744 amount of memory needed for a compiled pattern. Previously, there was code
745 that made a preliminary scan of the pattern in order to do this. That was
746 OK when PCRE was new, but as the facilities have expanded, it has become
747 harder and harder to keep it in step with the real compile phase, and there
748 have been a number of bugs (see for example, 4 above). I have now found a
749 cunning way of running the real compile function in a "fake" mode that
750 enables it to compute how much memory it would need, while actually only
751 ever using a few hundred bytes of working memory and without too many
752 tests of the mode. This should make future maintenance and development
753 easier. A side effect of this work is that the limit of 200 on the nesting
754 depth of parentheses has been removed (though this was never a serious
755 limitation, I suspect). However, there is a downside: pcre_compile() now
756 runs more slowly than before (30% or more, depending on the pattern). I
757 hope this isn't a big issue. There is no effect on runtime performance.
758
759 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
760 newline (only possible for the last line of a file) and it was a
761 pattern that set a locale (followed by /Lsomething), pcretest crashed.
762
763 19. Added additional timing features to pcretest. (1) The -tm option now times
764 matching only, not compiling. (2) Both -t and -tm can be followed, as a
765 separate command line item, by a number that specifies the number of
766 repeats to use when timing. The default is 50000; this gives better
767 precision, but takes uncomfortably long for very large patterns.
768
769 20. Extended pcre_study() to be more clever in cases where a branch of a
770 subpattern has no definite first character. For example, (a*|b*)[cd] would
771 previously give no result from pcre_study(). Now it recognizes that the
772 first character must be a, b, c, or d.
773
774 21. There was an incorrect error "recursive call could loop indefinitely" if
775 a subpattern (or the entire pattern) that was being tested for matching an
776 empty string contained only one non-empty item after a nested subpattern.
777 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
778 incorrectly, because the \d was being skipped in the check.
779
780 22. The pcretest program now has a new pattern option /B and a command line
781 option -b, which is equivalent to adding /B to every pattern. This causes
782 it to show the compiled bytecode, without the additional information that
783 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
784 is the same as /B/I).
785
786 23. A new optimization is now able automatically to treat some sequences such
787 as a*b as a*+b. More specifically, if something simple (such as a character
788 or a simple class like \d) has an unlimited quantifier, and is followed by
789 something that cannot possibly match the quantified thing, the quantifier
790 is automatically "possessified".
791
792 24. A recursive reference to a subpattern whose number was greater than 39
793 went wrong under certain circumstances in UTF-8 mode. This bug could also
794 have affected the operation of pcre_study().
795
796 25. Realized that a little bit of performance could be had by replacing
797 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
798
799 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
800
801 27. Possessive quantifiers such as a++ were previously implemented by turning
802 them into atomic groups such as ($>a+). Now they have their own opcodes,
803 which improves performance. This includes the automatically created ones
804 from 23 above.
805
806 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
807 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
808 the first matched character to be a colon. This applied both to named and
809 numbered groups.
810
811 29. The ucpinternal.h header file was missing its idempotency #ifdef.
812
813 30. I was sent a "project" file called libpcre.a.dev which I understand makes
814 building PCRE on Windows easier, so I have included it in the distribution.
815
816 31. There is now a check in pcretest against a ridiculously large number being
817 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
818 loop, the loop is abandoned.
819
820 32. Forward references to subpatterns in conditions such as (?(2)...) where
821 subpattern 2 is defined later cause pcre_compile() to search forwards in
822 the pattern for the relevant set of parentheses. This search went wrong
823 when there were unescaped parentheses in a character class, parentheses
824 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
825
826 33. "Subroutine" calls and backreferences were previously restricted to
827 referencing subpatterns earlier in the regex. This restriction has now
828 been removed.
829
830 34. Added a number of extra features that are going to be in Perl 5.10. On the
831 whole, these are just syntactic alternatives for features that PCRE had
832 previously implemented using the Python syntax or my own invention. The
833 other formats are all retained for compatibility.
834
835 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
836 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
837 also .NET compatible.
838
839 (b) A recursion or subroutine call to a named group can now be defined as
840 (?&name) as well as (?P>name).
841
842 (c) A backreference to a named group can now be defined as \k<name> or
843 \k'name' as well as (?P=name). The new forms, as well as being in Perl
844 5.10, are also .NET compatible.
845
846 (d) A conditional reference to a named group can now use the syntax
847 (?(<name>) or (?('name') as well as (?(name).
848
849 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
850 groups (named and numbered) that are never evaluated inline, but can be
851 called as "subroutines" from elsewhere. In effect, the DEFINE condition
852 is always false. There may be only one alternative in such a group.
853
854 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
855 as the simple (?(R). The condition is true only if the most recent
856 recursion is that of the given number or name. It does not search out
857 through the entire recursion stack.
858
859 (g) The escape \gN or \g{N} has been added, where N is a positive or
860 negative number, specifying an absolute or relative reference.
861
862 35. Tidied to get rid of some further signed/unsigned compiler warnings and
863 some "unreachable code" warnings.
864
865 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
866 things, this adds five new scripts.
867
868 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
869 There were also incompatibilities regarding the handling of \Q..\E inside
870 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
871 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
872
873 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
874 matches an empty string, and forcibly breaks the loop. There were bugs in
875 this code in non-simple cases. For a pattern such as ^(a()*)* matched
876 against aaaa the result was just "a" rather than "aaaa", for example. Two
877 separate and independent bugs (that affected different cases) have been
878 fixed.
879
880 39. Refactored the code to abolish the use of different opcodes for small
881 capturing bracket numbers. This is a tidy that I avoided doing when I
882 removed the limit on the number of capturing brackets for 3.5 back in 2001.
883 The new approach is not only tidier, it makes it possible to reduce the
884 memory needed to fix the previous bug (38).
885
886 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
887 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
888 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
889 mode.
890
891 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
892 report.
893
894 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
895 copy construction and assignment in the C++ wrapper.
896
897 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
898 couple of bugs that could have given wrong results in the "--newline=crlf"
899 case.
900
901 44. Added a number of casts and did some reorganization of signed/unsigned int
902 variables following suggestions from Dair Grant. Also renamed the variable
903 "this" as "item" because it is a C++ keyword.
904
905 45. Arranged for dftables to add
906
907 #include "pcre_internal.h"
908
909 to pcre_chartables.c because without it, gcc 4.x may remove the array
910 definition from the final binary if PCRE is built into a static library and
911 dead code stripping is activated.
912
913 46. For an unanchored pattern, if a match attempt fails at the start of a
914 newline sequence, and the newline setting is CRLF or ANY, and the next two
915 characters are CRLF, advance by two characters instead of one.
916
917
918 Version 6.7 04-Jul-06
919 ---------------------
920
921 1. In order to handle tests when input lines are enormously long, pcretest has
922 been re-factored so that it automatically extends its buffers when
923 necessary. The code is crude, but this _is_ just a test program. The
924 default size has been increased from 32K to 50K.
925
926 2. The code in pcre_study() was using the value of the re argument before
927 testing it for NULL. (Of course, in any sensible call of the function, it
928 won't be NULL.)
929
930 3. The memmove() emulation function in pcre_internal.h, which is used on
931 systems that lack both memmove() and bcopy() - that is, hardly ever -
932 was missing a "static" storage class specifier.
933
934 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
935 containing an extended class (one that cannot be represented by a bitmap
936 because it contains high-valued characters or Unicode property items, e.g.
937 [\pZ]). Almost always one would set UTF-8 mode when processing such a
938 pattern, but PCRE should not loop if you do not (it no longer does).
939 [Detail: two cases were found: (a) a repeated subpattern containing an
940 extended class; (b) a recursive reference to a subpattern that followed a
941 previous extended class. It wasn't skipping over the extended class
942 correctly when UTF-8 mode was not set.]
943
944 5. A negated single-character class was not being recognized as fixed-length
945 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
946 compile error "lookbehind assertion is not fixed length".
947
948 6. The RunPerlTest auxiliary script was showing an unexpected difference
949 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
950 write a Perl script that can interpret lines of an input file either as
951 byte characters or as UTF-8, which is what "perltest" was being required to
952 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
953 can't do is switch easily at run time between having the "use utf8;" pragma
954 or not. In the end, I fudged it by using the RunPerlTest script to insert
955 "use utf8;" explicitly for the UTF-8 tests.
956
957 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
958 the end of the subject string, contrary to the documentation and to what
959 Perl does. This was true of both matching functions. Now it matches only at
960 the start of the subject and immediately after *internal* newlines.
961
962 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
963 a pointer to an int instead of a pointer to an unsigned long int. This
964 caused problems on 64-bit systems.
965
966 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
967 instance of the 'standard' template library not being so standard".
968
969 10. There was no check on the number of named subpatterns nor the maximum
970 length of a subpattern name. The product of these values is used to compute
971 the size of the memory block for a compiled pattern. By supplying a very
972 long subpattern name and a large number of named subpatterns, the size
973 computation could be caused to overflow. This is now prevented by limiting
974 the length of names to 32 characters, and the number of named subpatterns
975 to 10,000.
976
977 11. Subpatterns that are repeated with specific counts have to be replicated in
978 the compiled pattern. The size of memory for this was computed from the
979 length of the subpattern and the repeat count. The latter is limited to
980 65535, but there was no limit on the former, meaning that integer overflow
981 could in principle occur. The compiled length of a repeated subpattern is
982 now limited to 30,000 bytes in order to prevent this.
983
984 12. Added the optional facility to have named substrings with the same name.
985
986 13. Added the ability to use a named substring as a condition, using the
987 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
988 are numbers (not recommended). Forward references are permitted.
989
990 14. Added forward references in named backreferences (if you see what I mean).
991
992 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
993 pattern could run off the end of the subject. For example, the pattern
994 "(?s)(.{1,5})"8 did this with the subject "ab".
995
996 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
997 PCRE_CASELESS was set when matching characters that were quantified with ?
998 or *.
999
1000 17. A character class other than a single negated character that had a minimum
1001 but no maximum quantifier - for example [ab]{6,} - was not handled
1002 correctly by pce_dfa_exec(). It would match only one character.
1003
1004 18. A valid (though odd) pattern that looked like a POSIX character
1005 class but used an invalid character after [ (for example [[,abc,]]) caused
1006 pcre_compile() to give the error "Failed: internal error: code overflow" or
1007 in some cases to crash with a glibc free() error. This could even happen if
1008 the pattern terminated after [[ but there just happened to be a sequence of
1009 letters, a binary zero, and a closing ] in the memory that followed.
1010
1011 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1012 over the years. Originally (before any Unicode support), just the bottom 8
1013 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1014 output from "man perlunicode" includes this:
1015
1016 The regular expression compiler produces polymorphic opcodes. That
1017 is, the pattern adapts to the data and automatically switches to
1018 the Unicode character scheme when presented with Unicode data--or
1019 instead uses a traditional byte scheme when presented with byte
1020 data.
1021
1022 Sadly, a wide octal escape does not cause a switch, and in a string with
1023 no other multibyte characters, these octal escapes are treated as before.
1024 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1025 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1026 Unicode string.
1027
1028 I have not perpetrated such confusion in PCRE. Up till now, it took just
1029 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1030 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1031 translate to the appropriate multibyte character.
1032
1033 29. Applied some refactoring to reduce the number of warnings from Microsoft
1034 and Borland compilers. This has included removing the fudge introduced
1035 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1036 a warning about an unused variable.
1037
1038 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1039 characters since release 4.0, because Perl (from release 5.004) does not.
1040 [Or at least, is documented not to: some releases seem to be in conflict
1041 with the documentation.] However, when a pattern was studied with
1042 pcre_study() and all its branches started with \s, PCRE still included VT
1043 as a possible starting character. Of course, this did no harm; it just
1044 caused an unnecessary match attempt.
1045
1046 22. Removed a now-redundant internal flag bit that recorded the fact that case
1047 dependency changed within the pattern. This was once needed for "required
1048 byte" processing, but is no longer used. This recovers a now-scarce options
1049 bit. Also moved the least significant internal flag bit to the most-
1050 significant bit of the word, which was not previously used (hangover from
1051 the days when it was an int rather than a uint) to free up another bit for
1052 the future.
1053
1054 23. Added support for CRLF line endings as well as CR and LF. As well as the
1055 default being selectable at build time, it can now be changed at runtime
1056 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1057 specify that it is scanning data with non-default line endings.
1058
1059 24. Changed the definition of CXXLINK to make it agree with the definition of
1060 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1061
1062 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1063 recursions. This makes a big different to stack usage for some patterns.
1064
1065 26. If a subpattern containing a named recursion or subroutine reference such
1066 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1067 the space required for the compiled pattern went wrong and gave too small a
1068 value. Depending on the environment, this could lead to "Failed: internal
1069 error: code overflow at offset 49" or "glibc detected double free or
1070 corruption" errors.
1071
1072 27. Applied patches from Google (a) to support the new newline modes and (b) to
1073 advance over multibyte UTF-8 characters in GlobalReplace.
1074
1075 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1076 difference for some implementation of PCRE in some Windows version.
1077
1078 29. Added some extra testing facilities to pcretest:
1079
1080 \q<number> in a data line sets the "match limit" value
1081 \Q<number> in a data line sets the "match recursion limt" value
1082 -S <number> sets the stack size, where <number> is in megabytes
1083
1084 The -S option isn't available for Windows.
1085
1086
1087 Version 6.6 06-Feb-06
1088 ---------------------
1089
1090 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1091 in pcreposix.h. I have copied the definition from pcre.h.
1092
1093 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1094 because pcre.h is no longer a built file.
1095
1096 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1097 not normally included in the compiled code.
1098
1099
1100 Version 6.5 01-Feb-06
1101 ---------------------
1102
1103 1. When using the partial match feature with pcre_dfa_exec(), it was not
1104 anchoring the second and subsequent partial matches at the new starting
1105 point. This could lead to incorrect results. For example, with the pattern
1106 /1234/, partially matching against "123" and then "a4" gave a match.
1107
1108 2. Changes to pcregrep:
1109
1110 (a) All non-match returns from pcre_exec() were being treated as failures
1111 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1112 error message is output. Some extra information is given for the
1113 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1114 probably the only errors that are likely to be caused by users (by
1115 specifying a regex that has nested indefinite repeats, for instance).
1116 If there are more than 20 of these errors, pcregrep is abandoned.
1117
1118 (b) A binary zero was treated as data while matching, but terminated the
1119 output line if it was written out. This has been fixed: binary zeroes
1120 are now no different to any other data bytes.
1121
1122 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1123 used to set a locale for matching. The --locale=xxxx long option has
1124 been added (no short equivalent) to specify a locale explicitly on the
1125 pcregrep command, overriding the environment variables.
1126
1127 (d) When -B was used with -n, some line numbers in the output were one less
1128 than they should have been.
1129
1130 (e) Added the -o (--only-matching) option.
1131
1132 (f) If -A or -C was used with -c (count only), some lines of context were
1133 accidentally printed for the final match.
1134
1135 (g) Added the -H (--with-filename) option.
1136
1137 (h) The combination of options -rh failed to suppress file names for files
1138 that were found from directory arguments.
1139
1140 (i) Added the -D (--devices) and -d (--directories) options.
1141
1142 (j) Added the -F (--fixed-strings) option.
1143
1144 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1145
1146 (l) Added the --colo(u)r option.
1147
1148 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1149 is not present by default.
1150
1151 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1152 items such as (?R) or (?1), when the recursion could match a number of
1153 alternatives. If it matched one of the alternatives, but subsequently,
1154 outside the recursion, there was a failure, the code tried to back up into
1155 the recursion. However, because of the way PCRE is implemented, this is not
1156 possible, and the result was an incorrect result from the match.
1157
1158 In order to prevent this happening, the specification of recursion has
1159 been changed so that all such subpatterns are automatically treated as
1160 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1161
1162 4. I had overlooked the fact that, in some locales, there are characters for
1163 which isalpha() is true but neither isupper() nor islower() are true. In
1164 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1165 and ordfeminine) are like this. This affected the treatment of \w and \W
1166 when they appeared in character classes, but not when they appeared outside
1167 a character class. The bit map for "word" characters is now created
1168 separately from the results of isalnum() instead of just taking it from the
1169 upper, lower, and digit maps. (Plus the underscore character, of course.)
1170
1171 5. The above bug also affected the handling of POSIX character classes such as
1172 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1173 permanent tables. Instead, the bit maps for such a class were previously
1174 created as the appropriate unions of the upper, lower, and digit bitmaps.
1175 Now they are created by subtraction from the [[:word:]] class, which has
1176 its own bitmap.
1177
1178 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1179 It is created by subtracting the vertical space characters (\x09, \x0a,
1180 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1181 subtraction was done in the overall bitmap for a character class, meaning
1182 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1183 be recognized. This bug has been fixed.
1184
1185 7. Patches from the folks at Google:
1186
1187 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1188 real life, but is still worth protecting against".
1189
1190 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1191 regular expressions".
1192
1193 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1194 have it.
1195
1196 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1197 "configure" and the latter not, in order to fix a problem somebody had
1198 with compiling the Arg class on HP-UX.
1199
1200 (e) Improve the error-handling of the C++ wrapper a little bit.
1201
1202 (f) New tests for checking recursion limiting.
1203
1204 8. The pcre_memmove() function, which is used only if the environment does not
1205 have a standard memmove() function (and is therefore rarely compiled),
1206 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1207 returning a result (though PCRE never actually uses the result).
1208
1209 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1210 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1211 returned instead of calling malloc() with an overflowing number that would
1212 most likely cause subsequent chaos.
1213
1214 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1215
1216 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1217 with this option is matched, the nmatch and pmatch options of regexec() are
1218 ignored.
1219
1220 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1221 provided in case anyone wants to the the POSIX interface with UTF-8
1222 strings.
1223
1224 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1225 C++ linking (needed for some HP-UX environments).
1226
1227 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1228 (unused parameter) and in the pcre_printint() function (omitted "default"
1229 switch label when the default is to do nothing).
1230
1231 15. Added some code to make it possible, when PCRE is compiled as a C++
1232 library, to replace subject pointers for pcre_exec() with a smart pointer
1233 class, thus making it possible to process discontinuous strings.
1234
1235 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1236 much the same function. They were added by different people who were trying
1237 to make PCRE easy to compile on non-Unix systems. It has been suggested
1238 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1239 for compiling on Windows systems. I have therefore replaced it with
1240 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1241 defaults to "extern" for C or "extern C" for C++, which works fine on
1242 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1243 SCOPE with something explicit in config.h. In addition:
1244
1245 (a) pcreposix.h still had just "extern" instead of either of these macros;
1246 I have replaced it with PCRE_DATA_SCOPE.
1247
1248 (b) Functions such as _pcre_xclass(), which are internal to the library,
1249 but external in the C sense, all had PCRE_EXPORT in their definitions.
1250 This is apparently wrong for the Windows case, so I have removed it.
1251 (It makes no difference on Unix-like systems.)
1252
1253 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1254 of recursive calls to match(). This is different to MATCH_LIMIT because
1255 that limits the total number of calls to match(), not all of which increase
1256 the depth of recursion. Limiting the recursion depth limits the amount of
1257 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1258 when PCRE is compiled, and changed at run time. A patch from Google adds
1259 this functionality to the C++ interface.
1260
1261 18. Changes to the handling of Unicode character properties:
1262
1263 (a) Updated the table to Unicode 4.1.0.
1264
1265 (b) Recognize characters that are not in the table as "Cn" (undefined).
1266
1267 (c) I revised the way the table is implemented to a much improved format
1268 which includes recognition of ranges. It now supports the ranges that
1269 are defined in UnicodeData.txt, and it also amalgamates other
1270 characters into ranges. This has reduced the number of entries in the
1271 table from around 16,000 to around 3,000, thus reducing its size
1272 considerably. I realized I did not need to use a tree structure after
1273 all - a binary chop search is just as efficient. Having reduced the
1274 number of entries, I extended their size from 6 bytes to 8 bytes to
1275 allow for more data.
1276
1277 (d) Added support for Unicode script names via properties such as \p{Han}.
1278
1279 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1280 matching that character.
1281
1282 20. When matching a repeated Unicode property with a minimum greater than zero,
1283 (for example \pL{2,}), PCRE could look past the end of the subject if it
1284 reached it while seeking the minimum number of characters. This could
1285 happen only if some of the characters were more than one byte long, because
1286 there is a check for at least the minimum number of bytes.
1287
1288 21. Refactored the implementation of \p and \P so as to be more general, to
1289 allow for more different types of property in future. This has changed the
1290 compiled form incompatibly. Anybody with saved compiled patterns that use
1291 \p or \P will have to recompile them.
1292
1293 22. Added "Any" and "L&" to the supported property types.
1294
1295 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1296 but give a compile time error if the value is greater than 0xff.
1297
1298 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1299 accidentally not being installed or uninstalled.
1300
1301 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1302 made were to insert the current release number. This seemed silly, because
1303 it made things harder for people building PCRE on systems that don't run
1304 "configure". I have turned pcre.h into a distributed file, no longer built
1305 by "configure", with the version identification directly included. There is
1306 no longer a pcre.h.in file.
1307
1308 However, this change necessitated a change to the pcre-config script as
1309 well. It is built from pcre-config.in, and one of the substitutions was the
1310 release number. I have updated configure.ac so that ./configure now finds
1311 the release number by grepping pcre.h.
1312
1313 26. Added the ability to run the tests under valgrind.
1314
1315
1316 Version 6.4 05-Sep-05
1317 ---------------------
1318
1319 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1320 "--" to be printed when multiple files were scanned, even when none of the
1321 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1322 consider it to be a bug, and have restored the previous behaviour.
1323
1324 2. A couple of code tidies to get rid of compiler warnings.
1325
1326 3. The pcretest program used to cheat by referring to symbols in the library
1327 whose names begin with _pcre_. These are internal symbols that are not
1328 really supposed to be visible externally, and in some environments it is
1329 possible to suppress them. The cheating is now confined to including
1330 certain files from the library's source, which is a bit cleaner.
1331
1332 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1333 file's purpose clearer.
1334
1335 5. Reorganized pcre_ucp_findchar().
1336
1337
1338 Version 6.3 15-Aug-05
1339 ---------------------
1340
1341 1. The file libpcre.pc.in did not have general read permission in the tarball.
1342
1343 2. There were some problems when building without C++ support:
1344
1345 (a) If C++ support was not built, "make install" and "make test" still
1346 tried to test it.
1347
1348 (b) There were problems when the value of CXX was explicitly set. Some
1349 changes have been made to try to fix these, and ...
1350
1351 (c) --disable-cpp can now be used to explicitly disable C++ support.
1352
1353 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1354 backslash in a target when C++ was disabled. This confuses some
1355 versions of "make", apparently. Using an intermediate variable solves
1356 this. (Same for CPP_LOBJ.)
1357
1358 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1359 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1360 necessary on certain architectures.
1361
1362 4. Added a setting of -export-symbols-regex to the link command to remove
1363 those symbols that are exported in the C sense, but actually are local
1364 within the library, and not documented. Their names all begin with
1365 "_pcre_". This is not a perfect job, because (a) we have to except some
1366 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1367 available (and never for static libraries). I have made a note to try to
1368 find a way round (a) in the future.
1369
1370
1371 Version 6.2 01-Aug-05
1372 ---------------------
1373
1374 1. There was no test for integer overflow of quantifier values. A construction
1375 such as {1111111111111111} would give undefined results. What is worse, if
1376 a minimum quantifier for a parenthesized subpattern overflowed and became
1377 negative, the calculation of the memory size went wrong. This could have
1378 led to memory overwriting.
1379
1380 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1381
1382 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1383 operating environments where this matters.
1384
1385 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1386 PCRE options from within the C++ wrapper.
1387
1388 5. Named capturing subpatterns were not being correctly counted when a pattern
1389 was compiled. This caused two problems: (a) If there were more than 100
1390 such subpatterns, the calculation of the memory needed for the whole
1391 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1392 back references of the form \12, where the number was greater than 9, were
1393 not recognized as back references, even though there were sufficient
1394 previous subpatterns.
1395
1396 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1397 versions of gcc, e.g. 2.95.4.
1398
1399
1400 Version 6.1 21-Jun-05
1401 ---------------------
1402
1403 1. There was one reference to the variable "posix" in pcretest.c that was not
1404 surrounded by "#if !defined NOPOSIX".
1405
1406 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1407 the cross-check on the old pcre_info() function, for the benefit of the
1408 cut-down version of PCRE that is currently imported into Exim.
1409
1410 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1411 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1412 patterns that start (?i) but I don't think that matters. The use of (?i) is
1413 just an example; this all applies to the other options as well.
1414
1415 4. Since libtool seems to echo the compile commands it is issuing, the output
1416 from "make" can be reduced a bit by putting "@" in front of each libtool
1417 compile command.
1418
1419 5. Patch from the folks at Google for configure.in to be a bit more thorough
1420 in checking for a suitable C++ installation before trying to compile the
1421 C++ stuff. This should fix a reported problem when a compiler was present,
1422 but no suitable headers.
1423
1424 6. The man pages all had just "PCRE" as their title. I have changed them to
1425 be the relevant file name. I have also arranged that these names are
1426 retained in the file doc/pcre.txt, which is a concatenation in text format
1427 of all the man pages except the little individual ones for each function.
1428
1429 7. The NON-UNIX-USE file had not been updated for the different set of source
1430 files that come with release 6. I also added a few comments about the C++
1431 wrapper.
1432
1433
1434 Version 6.0 07-Jun-05
1435 ---------------------
1436
1437 1. Some minor internal re-organization to help with my DFA experiments.
1438
1439 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1440 didn't matter for the library itself when fully configured, but did matter
1441 when compiling without UCP support, or within Exim, where the ucp files are
1442 not imported.
1443
1444 3. Refactoring of the library code to split up the various functions into
1445 different source modules. The addition of the new DFA matching code (see
1446 below) to a single monolithic source would have made it really too
1447 unwieldy, quite apart from causing all the code to be include in a
1448 statically linked application, when only some functions are used. This is
1449 relevant even without the DFA addition now that patterns can be compiled in
1450 one application and matched in another.
1451
1452 The downside of splitting up is that there have to be some external
1453 functions and data tables that are used internally in different modules of
1454 the library but which are not part of the API. These have all had their
1455 names changed to start with "_pcre_" so that they are unlikely to clash
1456 with other external names.
1457
1458 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1459 a different (DFA) algorithm. Although it is slower than the original
1460 function, it does have some advantages for certain types of matching
1461 problem.
1462
1463 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1464 including restarting after a partial match.
1465
1466 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1467 defined when compiling for Windows was sent to me. I have put it into the
1468 code, though I have no means of testing or verifying it.
1469
1470 7. Added the pcre_refcount() auxiliary function.
1471
1472 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1473 match before or at the first newline in the subject string. In pcretest,
1474 the /f option on a pattern can be used to set this.
1475
1476 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1477 would behave wrongly. This has been present in PCRE since release 4.0.
1478
1479 10. A number of changes to the pcregrep command:
1480
1481 (a) Refactored how -x works; insert ^(...)$ instead of setting
1482 PCRE_ANCHORED and checking the length, in preparation for adding
1483 something similar for -w.
1484
1485 (b) Added the -w (match as a word) option.
1486
1487 (c) Refactored the way lines are read and buffered so as to have more
1488 than one at a time available.
1489
1490 (d) Implemented a pcregrep test script.
1491
1492 (e) Added the -M (multiline match) option. This allows patterns to match
1493 over several lines of the subject. The buffering ensures that at least
1494 8K, or the rest of the document (whichever is the shorter) is available
1495 for matching (and similarly the previous 8K for lookbehind assertions).
1496
1497 (f) Changed the --help output so that it now says
1498
1499 -w, --word-regex(p)
1500
1501 instead of two lines, one with "regex" and the other with "regexp"
1502 because that confused at least one person since the short forms are the
1503 same. (This required a bit of code, as the output is generated
1504 automatically from a table. It wasn't just a text change.)
1505
1506 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1507 option but starts with a hyphen. Could be a pattern or a path name
1508 starting with a hyphen, for instance.
1509
1510 (h) "-" can be given as a file name to represent stdin.
1511
1512 (i) When file names are being printed, "(standard input)" is used for
1513 the standard input, for compatibility with GNU grep. Previously
1514 "<stdin>" was used.
1515
1516 (j) The option --label=xxx can be used to supply a name to be used for
1517 stdin when file names are being printed. There is no short form.
1518
1519 (k) Re-factored the options decoding logic because we are going to add
1520 two more options that take data. Such options can now be given in four
1521 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1522
1523 (l) Added the -A, -B, and -C options for requesting that lines of context
1524 around matches be printed.
1525
1526 (m) Added the -L option to print the names of files that do not contain
1527 any matching lines, that is, the complement of -l.
1528
1529 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1530 continue to scan other files.
1531
1532 (o) The -s option was incorrectly implemented. For compatibility with other
1533 greps, it now suppresses the error message for a non-existent or non-
1534 accessible file (but not the return code). There is a new option called
1535 -q that suppresses the output of matching lines, which was what -s was
1536 previously doing.
1537
1538 (p) Added --include and --exclude options to specify files for inclusion
1539 and exclusion when recursing.
1540
1541 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1542 Hopefully, it now does.
1543
1544 12. Missing cast in pcre_study().
1545
1546 13. Added an "uninstall" target to the makefile.
1547
1548 14. Replaced "extern" in the function prototypes in Makefile.in with
1549 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1550 world, but is set differently for Windows.
1551
1552 15. Added a second compiling function called pcre_compile2(). The only
1553 difference is that it has an extra argument, which is a pointer to an
1554 integer error code. When there is a compile-time failure, this is set
1555 non-zero, in addition to the error test pointer being set to point to an
1556 error message. The new argument may be NULL if no error number is required
1557 (but then you may as well call pcre_compile(), which is now just a
1558 wrapper). This facility is provided because some applications need a
1559 numeric error indication, but it has also enabled me to tidy up the way
1560 compile-time errors are handled in the POSIX wrapper.
1561
1562 16. Added VPATH=.libs to the makefile; this should help when building with one
1563 prefix path and installing with another. (Or so I'm told by someone who
1564 knows more about this stuff than I do.)
1565
1566 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1567 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1568 match everything, including newlines. This is not POSIX-compatible, but
1569 somebody wanted the feature. From pcretest it can be activated by using
1570 both the P and the s flags.
1571
1572 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1573
1574 19. libpcre.pc was being incorrectly installed as executable.
1575
1576 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1577 it now also looks for '\r' so that it will work unmodified on Windows.
1578
1579 21. Added Google's contributed C++ wrapper to the distribution.
1580
1581 22. Added some untidy missing memory free() calls in pcretest, to keep
1582 Electric Fence happy when testing.
1583
1584
1585
1586 Version 5.0 13-Sep-04
1587 ---------------------
1588
1589 1. Internal change: literal characters are no longer packed up into items
1590 containing multiple characters in a single byte-string. Each character
1591 is now matched using a separate opcode. However, there may be more than one
1592 byte in the character in UTF-8 mode.
1593
1594 2. The pcre_callout_block structure has two new fields: pattern_position and
1595 next_item_length. These contain the offset in the pattern to the next match
1596 item, and its length, respectively.
1597
1598 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1599 insertion of callouts before each pattern item. Added the /C option to
1600 pcretest to make use of this.
1601
1602 4. On the advice of a Windows user, the lines
1603
1604 #if defined(_WIN32) || defined(WIN32)
1605 _setmode( _fileno( stdout ), 0x8000 );
1606 #endif /* defined(_WIN32) || defined(WIN32) */
1607
1608 have been added to the source of pcretest. This apparently does useful
1609 magic in relation to line terminators.
1610
1611 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1612 for the benefit of those environments where the "b" makes a difference.
1613
1614 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1615 to know about it. I have put a hack into configure.in that adds in code
1616 to set GCC=yes if CC=icc. This seems to end up at a point in the
1617 generated configure script that is early enough to affect the setting of
1618 compiler options, which is what is needed, but I have no means of testing
1619 whether it really works. (The user who reported this had patched the
1620 generated configure script, which of course I cannot do.)
1621
1622 LATER: After change 22 below (new libtool files), the configure script
1623 seems to know about icc (and also ecc). Therefore, I have commented out
1624 this hack in configure.in.
1625
1626 7. Added support for pkg-config (2 patches were sent in).
1627
1628 8. Negated POSIX character classes that used a combination of internal tables
1629 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1630 [[:^ascii]]. Typically, they would match almost any characters. The other
1631 POSIX classes were not broken in this way.
1632
1633 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1634 to find the match, as PCRE was deluded into thinking that the match had to
1635 start at the start point or following a newline. The same bug applied to
1636 patterns with negative forward assertions or any backward assertions
1637 preceding ".*" at the start, unless the pattern required a fixed first
1638 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1639
1640 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1641 starting at the last subject character, bytes beyond the end of the subject
1642 string were read.
1643
1644 11. Renamed the variable "class" as "classbits" to make life easier for C++
1645 users. (Previously there was a macro definition, but it apparently wasn't
1646 enough.)
1647
1648 12. Added the new field "tables" to the extra data so that tables can be passed
1649 in at exec time, or the internal tables can be re-selected. This allows
1650 a compiled regex to be saved and re-used at a later time by a different
1651 program that might have everything at different addresses.
1652
1653 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1654 -R library as well as a -L library.
1655
1656 14. The debugging options of pcretest (-d on the command line or D on a
1657 pattern) showed incorrect output for anything following an extended class
1658 that contained multibyte characters and which was followed by a quantifier.
1659
1660 15. Added optional support for general category Unicode character properties
1661 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1662 support. It adds about 90K to the size of the library. The meanings of the
1663 inbuilt class escapes such as \d and \s have NOT been changed.
1664
1665 16. Updated pcredemo.c to include calls to free() to release the memory for the
1666 compiled pattern.
1667
1668 17. The generated file chartables.c was being created in the source directory
1669 instead of in the building directory. This caused the build to fail if the
1670 source directory was different from the building directory, and was
1671 read-only.
1672
1673 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1674 file. No doubt somebody will tell me if they don't make sense... Also added
1675 Dan Mooney's comments about building on OpenVMS.
1676
1677 19. Added support for partial matching via the PCRE_PARTIAL option for
1678 pcre_exec() and the \P data escape in pcretest.
1679
1680 20. Extended pcretest with 3 new pattern features:
1681
1682 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1683 write the compiled pattern to the file whose name is "rest-of-line".
1684 This is a straight binary dump of the data, with the saved pointer to
1685 the character tables forced to be NULL. The study data, if any, is
1686 written too. After writing, pcretest reads a new pattern.
1687
1688 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1689 compiled pattern from the given file. There must not be any
1690 occurrences of "<" in the file name (pretty unlikely); if there are,
1691 pcretest will instead treat the initial "<" as a pattern delimiter.
1692 After reading in the pattern, pcretest goes on to read data lines as
1693 usual.
1694
1695 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1696 and 16-bit fields in a compiled pattern, to simulate a pattern that
1697 was compiled on a host of opposite endianness.
1698
1699 21. The pcre-exec() function can now cope with patterns that were compiled on
1700 hosts of opposite endianness, with this restriction:
1701
1702 As for any compiled expression that is saved and used later, the tables
1703 pointer field cannot be preserved; the extra_data field in the arguments
1704 to pcre_exec() should be used to pass in a tables address if a value
1705 other than the default internal tables were used at compile time.
1706
1707 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1708 now diagnosed as an error. Previously, most of the time, a negative number
1709 would have been treated as zero, but if in addition "ovector" was passed as
1710 NULL, a crash could occur.
1711
1712 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1713 new versions from the libtool 1.5 distribution (the last one is a copy of
1714 a file called libtool.m4). This seems to have fixed the need to patch
1715 "configure" to support Darwin 1.3 (which I used to do). However, I still
1716 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1717 workstation).
1718
1719 24. Changed the PCRE licence to be the more standard "BSD" licence.
1720
1721
1722 Version 4.5 01-Dec-03
1723 ---------------------
1724
1725 1. There has been some re-arrangement of the code for the match() function so
1726 that it can be compiled in a version that does not call itself recursively.
1727 Instead, it keeps those local variables that need separate instances for
1728 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1729 needs to "recurse". Keeping track of where control must go is done by means
1730 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1731 hide most of the details from the main code, and operates only if
1732 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1733 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1734 operating.
1735
1736 To make it easier for callers to provide specially tailored get/free
1737 functions for this usage, two new functions, pcre_stack_malloc, and
1738 pcre_stack_free, are used. They are always called in strict stacking order,
1739 and the size of block requested is always the same.
1740
1741 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1742 PCRE has been compiled to use the stack or the heap for recursion. The
1743 -C option of pcretest uses this to show which version is compiled.
1744
1745 A new data escape \S, is added to pcretest; it causes the amounts of store
1746 obtained and freed by both kinds of malloc/free at match time to be added
1747 to the output.
1748
1749 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1750 what's available on my current Linux desktop machine.
1751
1752 3. When matching a UTF-8 string, the test for a valid string at the start has
1753 been extended. If start_offset is not zero, PCRE now checks that it points
1754 to a byte that is the start of a UTF-8 character. If not, it returns
1755 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1756 this is necessary because there may be backward assertions in the pattern.
1757 When matching the same subject several times, it may save resources to use
1758 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1759
1760 4. The code for checking the validity of UTF-8 strings has been tightened so
1761 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1762 containing "overlong sequences".
1763
1764 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1765 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1766 should have been "&", but it just so happened that all the cases this let
1767 through by mistake were picked up later in the function.
1768
1769 6. I had used a variable called "isblank" - this is a C99 function, causing
1770 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1771
1772 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1773 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1774 so the version is shown; (c) stop "make test" echoing "./RunTest".
1775
1776 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1777
1778 9. The prototype for memmove() for systems that don't have it was using
1779 size_t, but the inclusion of the header that defines size_t was later. I've
1780 moved the #includes for the C headers earlier to avoid this.
1781
1782 10. Added some adjustments to the code to make it easier to compiler on certain
1783 special systems:
1784
1785 (a) Some "const" qualifiers were missing.
1786 (b) Added the macro EXPORT before all exported functions; by default this
1787 is defined to be empty.
1788 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1789 that it reads its output file name as an argument instead of writing
1790 to the standard output and assuming this can be redirected.
1791
1792 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1793 class containing characters with values greater than 255, PCRE compilation
1794 went into a loop.
1795
1796 12. A recursive reference to a subpattern that was within another subpattern
1797 that had a minimum quantifier of zero caused PCRE to crash. For example,
1798 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1799 recursion. If the recursively-called subpattern itself had a zero repeat,
1800 that was OK.
1801
1802 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1803 buffer into which it was copied (for escape processing) was still set at
1804 1024, so long lines caused crashes.
1805
1806 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1807 "internal error: code overflow...". This applied to any character class
1808 that was followed by a possessive quantifier.
1809
1810 15. Modified the Makefile to add libpcre.la as a prerequisite for
1811 libpcreposix.la because I was told this is needed for a parallel build to
1812 work.
1813
1814 16. If a pattern that contained .* following optional items at the start was
1815 studied, the wrong optimizing data was generated, leading to matching
1816 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1817 matching string must start with a or b or c. The correct conclusion for
1818 this pattern is that a match can start with any character.
1819
1820
1821 Version 4.4 13-Aug-03
1822 ---------------------
1823
1824 1. In UTF-8 mode, a character class containing characters with values between
1825 127 and 255 was not handled correctly if the compiled pattern was studied.
1826 In fixing this, I have also improved the studying algorithm for such
1827 classes (slightly).
1828
1829 2. Three internal functions had redundant arguments passed to them. Removal
1830 might give a very teeny performance improvement.
1831
1832 3. Documentation bug: the value of the capture_top field in a callout is *one
1833 more than* the number of the hightest numbered captured substring.
1834
1835 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1836 in incorrectly linking with a previously installed version. They now link
1837 explicitly with libpcre.la.
1838
1839 5. configure.in no longer needs to recognize Cygwin specially.
1840
1841 6. A problem in pcre.in for Windows platforms is fixed.
1842
1843 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1844 pcretest, it used to include the size of the study block as part of its
1845 output. Unfortunately, the structure contains a field that has a different
1846 size on different hardware architectures. This meant that the tests that
1847 showed this size failed. As the block is currently always of a fixed size,
1848 this information isn't actually particularly useful in pcretest output, so
1849 I have just removed it.
1850
1851 8. Three pre-processor statements accidentally did not start in column 1.
1852 Sadly, there are *still* compilers around that complain, even though
1853 standard C has not required this for well over a decade. Sigh.
1854
1855 9. In pcretest, the code for checking callouts passed small integers in the
1856 callout_data field, which is a void * field. However, some picky compilers
1857 complained about the casts involved for this on 64-bit systems. Now
1858 pcretest passes the address of the small integer instead, which should get
1859 rid of the warnings.
1860
1861 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1862 both compile and run time, and gives an error if an invalid UTF-8 sequence
1863 is found. There is a option for disabling this check in cases where the
1864 string is known to be correct and/or the maximum performance is wanted.
1865
1866 11. In response to a bug report, I changed one line in Makefile.in from
1867
1868 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1869 to
1870 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1871
1872 to look similar to other lines, but I have no way of telling whether this
1873 is the right thing to do, as I do not use Windows. No doubt I'll get told
1874 if it's wrong...
1875
1876
1877 Version 4.3 21-May-03
1878 ---------------------
1879
1880 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1881 Makefile.
1882
1883 2. Some refactoring to improve the quality of the code:
1884
1885 (i) The utf8_table... variables are now declared "const".
1886
1887 (ii) The code for \cx, which used the "case flipping" table to upper case
1888 lower case letters, now just substracts 32. This is ASCII-specific,
1889 but the whole concept of \cx is ASCII-specific, so it seems
1890 reasonable.
1891
1892 (iii) PCRE was using its character types table to recognize decimal and
1893 hexadecimal digits in the pattern. This is silly, because it handles
1894 only 0-9, a-f, and A-F, but the character types table is locale-
1895 specific, which means strange things might happen. A private
1896 table is now used for this - though it costs 256 bytes, a table is
1897 much faster than multiple explicit tests. Of course, the standard
1898 character types table is still used for matching digits in subject
1899 strings against \d.
1900
1901 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1902 ending in _t are). So I've renamed it as ESC_tee.
1903
1904 3. The first argument for regexec() in the POSIX wrapper should have been
1905 defined as "const".
1906
1907 4. Changed pcretest to use malloc() for its buffers so that they can be
1908 Electric Fenced for debugging.
1909
1910 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1911 to read one or more bytes before the start of the subject string. Often this
1912 had no effect on PCRE's behaviour, but in some circumstances it could
1913 provoke a segmentation fault.
1914
1915 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1916 to try to read one or more bytes before the start of the subject string.
1917
1918 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1919 UTF-8 support could misbehave in various ways if the subject string
1920 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1921 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1922 back over UTF-8 characters.)
1923
1924
1925 Version 4.2 14-Apr-03
1926 ---------------------
1927
1928 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1929
1930 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1931 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1932 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1933 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1934 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1935 and BUILD_EXEEXT
1936 Note: automatic setting of the BUILD variables is not yet working
1937 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1938 compile-time but not at link-time
1939 [LINK]: use for linking executables only
1940 make different versions for Windows and non-Windows
1941 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1942 libraries
1943 [LINK_FOR_BUILD]: new variable
1944 [OBJEXT]: use throughout
1945 [EXEEXT]: use throughout
1946 <winshared>: new target
1947 <wininstall>: new target
1948 <dftables.o>: use native compiler
1949 <dftables>: use native linker
1950 <install>: handle Windows platform correctly
1951 <clean>: ditto
1952 <check>: ditto
1953 copy DLL to top builddir before testing
1954
1955 As part of these changes, -no-undefined was removed again. This was reported
1956 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1957 in any case.
1958
1959 3. Some tidies to get rid of compiler warnings:
1960
1961 . In the match_data structure, match_limit was an unsigned long int, whereas
1962 match_call_count was an int. I've made them both unsigned long ints.
1963
1964 . In pcretest the fact that a const uschar * doesn't automatically cast to
1965 a void * provoked a warning.
1966
1967 . Turning on some more compiler warnings threw up some "shadow" variables
1968 and a few more missing casts.
1969
1970 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1971 option, a class that contained a single character with a value between 128
1972 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1973
1974 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1975 option, a class that contained several characters, but with at least one
1976 whose value was between 128 and 255 caused PCRE to crash.
1977
1978
1979 Version 4.1 12-Mar-03
1980 ---------------------
1981
1982 1. Compiling with gcc -pedantic found a couple of places where casts were
1983 needed, and a string in dftables.c that was longer than standard compilers are
1984 required to support.
1985
1986 2. Compiling with Sun's compiler found a few more places where the code could
1987 be tidied up in order to avoid warnings.
1988
1989 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1990 first of these names is deprecated in the latest Autoconf in favour of the name
1991 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1992 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1993 analogy I have changed it to CFLAGS_FOR_BUILD.
1994
1995 4. Added -no-undefined to the linking command in the Makefile, because this is
1996 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1997 linking step for the pcreposix library.
1998
1999 5. PCRE was failing to diagnose the case of two named groups with the same
2000 name.
2001
2002 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2003 literal character that is needed in the subject for a match, and scans along to
2004 ensure that it is present before embarking on the full matching process. This
2005 saves time in cases of nested unlimited repeats that are never going to match.
2006 Problem: the scan can take a lot of time if the subject is very long (e.g.
2007 megabytes), thus penalizing straightforward matches. It is now done only if the
2008 amount of subject to be scanned is less than 1000 bytes.
2009
2010 7. A lesser problem with the same optimization is that it was recording the
2011 first character of an anchored pattern as "needed", thus provoking a search
2012 right along the subject, even when the first match of the pattern was going to
2013 fail. The "needed" character is now not set for anchored patterns, unless it
2014 follows something in the pattern that is of non-fixed length. Thus, it still
2015 fulfils its original purpose of finding quick non-matches in cases of nested
2016 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2017
2018
2019 Version 4.0 17-Feb-03
2020 ---------------------
2021
2022 1. If a comment in an extended regex that started immediately after a meta-item
2023 extended to the end of string, PCRE compiled incorrect data. This could lead to
2024 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2025
2026 2. Moved to autoconf 2.53 and libtool 1.4.2.
2027
2028 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2029 the special perltest8 script is no longer needed - all the tests can be run
2030 from a single perltest script.
2031
2032 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2033 by \s. It has now been removed in PCRE. This means it isn't recognized as
2034 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2035 class [:space:] *does* include VT, thereby creating a mess.
2036
2037 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2038 space and tab.
2039
2040 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2041 its new features into the main test script, reducing the number of scripts.
2042
2043 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2044 were backward compatible, and made the (?i) apply to the whole pattern, as if
2045 /i were given. Now it behaves more logically, and applies the option setting
2046 only to what follows. PCRE has been changed to follow suit. However, if it
2047 finds options settings right at the start of the pattern, it extracts them into
2048 the global options, as before. Thus, they show up in the info data.
2049
2050 8. Added support for the \Q...\E escape sequence. Characters in between are
2051 treated as literals. This is slightly different from Perl in that $ and @ are
2052 also handled as literals inside the quotes. In Perl, they will cause variable
2053 interpolation. Note the following examples:
2054
2055 Pattern PCRE matches Perl matches
2056
2057 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2058 \Qabc\$xyz\E abc\$xyz abc\$xyz
2059 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2060
2061 For compatibility with Perl, \Q...\E sequences are recognized inside character
2062 classes as well as outside them.
2063
2064 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2065 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2066 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2067 signed/unsigned warnings.
2068
2069 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2070 option for pcretest, so I've replaced it by a simple function that does just
2071 that job.
2072
2073 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2074 "pcregrep -".
2075
2076 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2077 Java package. This provides some syntactic sugar for simple cases of what my
2078 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2079 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2080 item, you can use this simplified notation. Note that only makes sense with
2081 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2082 greediness, whatever the setting of the PCRE_UNGREEDY option.
2083
2084 13. A change of greediness default within a pattern was not taking effect at
2085 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2086 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2087 was abstracted outside.
2088
2089 14. PCRE now supports the \G assertion. It is true when the current matching
2090 position is at the start point of the match. This differs from \A when the
2091 starting offset is non-zero. Used with the /g option of pcretest (or similar
2092 code), it works in the same way as it does for Perl's /g option. If all
2093 alternatives of a regex begin with \G, the expression is anchored to the start
2094 match position, and the "anchored" flag is set in the compiled expression.
2095
2096 15. Some bugs concerning the handling of certain option changes within patterns
2097 have been fixed. These applied to options other than (?ims). For example,
2098 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2099 been the other way round. Some of this was related to change 7 above.
2100
2101 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2102 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2103 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2104 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2105
2106 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2107 mode. Unlike ".", it always matches newline, whatever the setting of
2108 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2109 assertions. Perl allows it, but it doesn't (in general) work because it can't
2110 calculate the length of the lookbehind. At least, that's the case for Perl
2111 5.8.0 - I've been told they are going to document that it doesn't work in
2112 future.
2113
2114 18. Added an error diagnosis for escapes that PCRE does not support: these are
2115 \L, \l, \N, \P, \p, \U, \u, and \X.
2116
2117 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2118 reading past the end of the pattern in cases such as /[abcd/.
2119
2120 20. PCRE was getting more memory than necessary for patterns with classes that
2121 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2122
2123 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2124 compiling PCRE for use with Virtual Pascal.
2125
2126 22. Small fix to the Makefile to make it work properly if the build is done
2127 outside the source tree.
2128
2129 23. Added a new extension: a condition to go with recursion. If a conditional
2130 subpattern starts with (?(R) the "true" branch is used if recursion has
2131 happened, whereas the "false" branch is used only at the top level.
2132
2133 24. When there was a very long string of literal characters (over 255 bytes
2134 without UTF support, over 250 bytes with UTF support), the computation of how
2135 much memory was required could be incorrect, leading to segfaults or other
2136 strange effects.
2137
2138 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2139 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2140 there was a subsequent back reference to those brackets. This meant that, for
2141 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2142 possible to check for precisely this case. All we can do is abandon the
2143 optimization if .* occurs inside capturing brackets when there are any back
2144 references whatsoever. (See below for a better fix that came later.)
2145
2146 26. The handling of the optimization for finding the first character of a
2147 non-anchored pattern, and for finding a character that is required later in the
2148 match were failing in some cases. This didn't break the matching; it just
2149 failed to optimize when it could. The way this is done has been re-implemented.
2150
2151 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2152
2153 28. Added a new feature that provides some of the functionality that Perl
2154 provides with (?{...}). The facility is termed a "callout". The way it is done
2155 in PCRE is for the caller to provide an optional function, by setting
2156 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2157 global variable. By default it is unset, which disables all calling out. To get
2158 the function called, the regex must include (?C) at appropriate points. This
2159 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2160 This provides a means of identifying different callout points. When PCRE
2161 reaches such a point in the regex, if pcre_callout has been set, the external
2162 function is called. It is provided with data in a structure called
2163 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2164 matching continues; if it returns a non-zero value, the match at the current
2165 point fails. However, backtracking will occur if possible. [This was changed
2166 later and other features added - see item 49 below.]
2167
2168 29. pcretest is upgraded to test the callout functionality. It provides a
2169 callout function that displays information. By default, it shows the start of
2170 the match and the current position in the text. There are some new data escapes
2171 to vary what happens:
2172
2173 \C+ in addition, show current contents of captured substrings
2174 \C- do not supply a callout function
2175 \C!n return 1 when callout number n is reached
2176 \C!n!m return 1 when callout number n is reached for the mth time
2177
2178 30. If pcregrep was called with the -l option and just a single file name, it
2179 output "<stdin>" if a match was found, instead of the file name.
2180
2181 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2182 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2183 pcre_exec(). This saves a malloc/free per call. The default value of
2184 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2185 when configuring.
2186
2187 32. The default maximum size of a compiled pattern is 64K. There have been a
2188 few cases of people hitting this limit. The code now uses macros to handle the
2189 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2190 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2191 configuring. Tests 2 and 5 work only with 2-byte links because they output
2192 debugging information about compiled patterns.
2193
2194 33. Internal code re-arrangements:
2195
2196 (a) Moved the debugging function for printing out a compiled regex into
2197 its own source file (printint.c) and used #include to pull it into
2198 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2199 separate copies.
2200
2201 (b) Defined the list of op-code names for debugging as a macro in
2202 internal.h so that it is next to the definition of the opcodes.
2203
2204 (c) Defined a table of op-code lengths for simpler skipping along compiled
2205 code. This is again a macro in internal.h so that it is next to the
2206 definition of the opcodes.
2207
2208 34. Added support for recursive calls to individual subpatterns, along the
2209 lines of Robin Houston's patch (but implemented somewhat differently).
2210
2211 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2212 allow it to read and process whole directories in Win32. This code was
2213 contributed by Lionel Fourquaux; it has not been tested by me.
2214
2215 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2216 used to name a group. Names consist of alphanumerics and underscores, and must
2217 be unique. Back references use the syntax (?P=name) and recursive calls use
2218 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2219 numbers. The function pcre_fullinfo() can be used after compilation to extract
2220 a name/number map. There are three relevant calls:
2221
2222 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2223 PCRE_INFO_NAMECOUNT yields the number of entries
2224 PCRE_INFO_NAMETABLE yields a pointer to the map.
2225
2226 The map is a vector of fixed-size entries. The size of each entry depends on
2227 the length of the longest name used. The first two bytes of each entry are the
2228 group number, most significant byte first. There follows the corresponding
2229 name, zero terminated. The names are in alphabetical order.
2230
2231 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2232 case instead of 255. Making it the same both with and without UTF-8 support
2233 means that the same test output works with both.
2234
2235 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2236 calling malloc() with a zero argument.
2237
2238 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2239 optimization. I've improved things by keeping a bitmap of backreferences with
2240 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2241 fact referenced, the optimization can be applied. It is unlikely that a
2242 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2243 the match to follow \n) will appear inside brackets with a number greater than
2244 31, but if it does, any back reference > 31 suppresses the optimization.
2245
2246 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2247 of disabling numbered capturing parentheses. Any opening parenthesis that is
2248 not followed by ? behaves as if it were followed by ?: but named parentheses
2249 can still be used for capturing (and they will acquire numbers in the usual
2250 way).
2251
2252 41. Redesigned the return codes from the match() function into yes/no/error so
2253 that errors can be passed back from deep inside the nested calls. A malloc
2254 failure while inside a recursive subpattern call now causes the
2255 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2256
2257 42. It is now possible to set a limit on the number of times the match()
2258 function is called in a call to pcre_exec(). This facility makes it possible to
2259 limit the amount of recursion and backtracking, though not in a directly
2260 obvious way, because the match() function is used in a number of different
2261 circumstances. The count starts from zero for each position in the subject
2262 string (for non-anchored patterns). The default limit is, for compatibility, a
2263 large number, namely 10 000 000. You can change this in two ways:
2264
2265 (a) When configuring PCRE before making, you can use --with-match-limit=n
2266 to set a default value for the compiled library.
2267
2268 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2269 a different value is set. See 45 below.
2270
2271 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2272
2273 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2274 of things that can be changed at compile time. The first argument specifies
2275 what is wanted and the second points to where the information is to be placed.
2276 The current list of available information is:
2277
2278 PCRE_CONFIG_UTF8
2279
2280 The output is an integer that is set to one if UTF-8 support is available;
2281 otherwise it is set to zero.
2282
2283 PCRE_CONFIG_NEWLINE
2284
2285 The output is an integer that it set to the value of the code that is used for
2286 newline. It is either LF (10) or CR (13).
2287
2288 PCRE_CONFIG_LINK_SIZE
2289
2290 The output is an integer that contains the number of bytes used for internal
2291 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2292
2293 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2294
2295 The output is an integer that contains the threshold above which the POSIX
2296 interface uses malloc() for output vectors. See item 31 above.
2297
2298 PCRE_CONFIG_MATCH_LIMIT
2299
2300 The output is an unsigned integer that contains the default limit of the number
2301 of match() calls in a pcre_exec() execution. See 42 above.
2302
2303 44. pcretest has been upgraded by the addition of the -C option. This causes it
2304 to extract all the available output from the new pcre_config() function, and to
2305 output it. The program then exits immediately.
2306
2307 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2308 order to support additional features. One way would have been to define
2309 pcre_exec2() (for example) with extra arguments, but this would not have been
2310 extensible, and would also have required all calls to the original function to
2311 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2312 is used for passing in "extra" data from pcre_study().
2313
2314 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2315 contains the following fields:
2316
2317 flags a bitmap indicating which of the following fields are set
2318 study_data opaque data from pcre_study()
2319 match_limit a way of specifying a limit on match() calls for a specific
2320 call to pcre_exec()
2321 callout_data data for callouts (see 49 below)
2322
2323 The flag bits are also defined in pcre.h, and are
2324
2325 PCRE_EXTRA_STUDY_DATA
2326 PCRE_EXTRA_MATCH_LIMIT
2327 PCRE_EXTRA_CALLOUT_DATA
2328
2329 The pcre_study() function now returns one of these new pcre_extra blocks, with
2330 the actual study data pointed to by the study_data field, and the
2331 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2332 before. That is, this change is entirely upwards-compatible and requires no
2333 change to existing code.
2334
2335 If you want to pass in additional data to pcre_exec(), you can either place it
2336 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2337 block.
2338
2339 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2340 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2341 times with different match limits, until it finds the minimum value needed for
2342 pcre_exec() to complete. The value is then output. This can be instructive; for
2343 most simple matches the number is quite small, but for pathological cases it
2344 gets very large very quickly.
2345
2346 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2347 returns the size of the data block pointed to by the study_data field in a
2348 pcre_extra block, that is, the value that was passed as the argument to
2349 pcre_malloc() when PCRE was getting memory in which to place the information
2350 created by pcre_study(). The fourth argument should point to a size_t variable.
2351 pcretest has been extended so that this information is shown after a successful
2352 pcre_study() call when information about the compiled regex is being displayed.
2353
2354 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2355 because what follows is always an absolute path. (Later: it turns out that this
2356 is more than cosmetic for MinGW, because it doesn't like empty path
2357 components.)
2358
2359 49. Some changes have been made to the callout feature (see 28 above):
2360
2361 (i) A callout function now has three choices for what it returns:
2362
2363 0 => success, carry on matching
2364 > 0 => failure at this point, but backtrack if possible
2365 < 0 => serious error, return this value from pcre_exec()
2366
2367 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2368 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2369 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2370 use by callout functions. It will never be used by PCRE itself.
2371
2372 (ii) The pcre_extra structure (see 45 above) has a void * field called
2373 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2374 pcre_callout_block structure has a field of the same name. The contents of
2375 the field passed in the pcre_extra structure are passed to the callout
2376 function in the corresponding field in the callout block. This makes it
2377 easier to use the same callout-containing regex from multiple threads. For
2378 testing, the pcretest program has a new data escape
2379
2380 \C*n pass the number n (may be negative) as callout_data
2381
2382 If the callout function in pcretest receives a non-zero value as
2383 callout_data, it returns that value.
2384
2385 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2386 there were some redundant $(CFLAGS) in commands that are now specified as
2387 $(LINK), which already includes $(CFLAGS).
2388
2389 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2390 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2391 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2392 one-byte characters throughout. Note that case-insensitive matching applies
2393 only to characters whose values are less than 256. PCRE doesn't support the
2394 notion of cases for higher-valued characters.
2395
2396 (i) A character class whose characters are all within 0-255 is handled as
2397 a bit map, and the map is inverted for negative classes. Previously, a
2398 character > 255 always failed to match such a class; however it should
2399 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2400
2401 (ii) A negated character class with a single character < 255 is coded as
2402 "not this character" (OP_NOT). This wasn't working properly when the test
2403 character was multibyte, either singly or repeated.
2404
2405 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2406 mode, for example: \x{100}{2,3}.
2407
2408 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2409 singly or repeated) now correctly test multibyte characters. However,
2410 PCRE doesn't recognize any characters with values greater than 255 as
2411 digits, spaces, or word characters. Such characters always match \D, \S,
2412 and \W, and never match \d, \s, or \w.
2413
2414 (v) Classes may now contain characters and character ranges with values
2415 greater than 255. For example: [ab\x{100}-\x{400}].
2416
2417 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2418 PCRE in UTF-8 mode.
2419
2420 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2421 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2422 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2423 value.)
2424
2425 53. The single man page has become too large. I have therefore split it up into
2426 a number of separate man pages. These also give rise to individual HTML pages;
2427 these are now put in a separate directory, and there is an index.html page that
2428 lists them all. Some hyperlinking between the pages has been installed.
2429
2430 54. Added convenience functions for handling named capturing parentheses.
2431
2432 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2433 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2434 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2435 are faulted.
2436
2437 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2438 calling configure. These values are used when compiling the dftables.c program
2439 which is run to generate the source of the default character tables. They
2440 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2441 you will need to set these values.
2442
2443 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2444
2445
2446 Version 3.9 02-Jan-02
2447 ---------------------
2448
2449 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2450
2451 2. If --disable-static was given, the building process failed when trying to
2452 build pcretest and pcregrep. (For some reason it was using libtool to compile
2453 them, which is not right, as they aren't part of the library.)
2454
2455
2456 Version 3.8 18-Dec-01
2457 ---------------------
2458
2459 1. The experimental UTF-8 code was completely screwed up. It was packing the
2460 bytes in the wrong order. How dumb can you get?
2461
2462
2463 Version 3.7 29-Oct-01
2464 ---------------------
2465
2466 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2467 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2468 this didn't happen under Solaris 8, where I normally test things.
2469
2470 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2471 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2472 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2473 this fix an hour or so after the initial 3.7 release.)
2474
2475
2476 Version 3.6 23-Oct-01
2477 ---------------------
2478
2479 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2480 offsets passed as NULL with zero offset count.
2481
2482 2. The config.guess and config.sub files had not been updated when I moved to
2483 the latest autoconf.
2484
2485
2486 Version 3.5 15-Aug-01
2487 ---------------------
2488
2489 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2490 had been forgotten.
2491
2492 2. By using declared but undefined structures, we can avoid using "void"
2493 definitions in pcre.h while keeping the internal definitions of the structures
2494 private.
2495
2496 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2497 user point of view, this means that both static and shared libraries are built
2498 by default, but this can be individually controlled. More of the work of
2499 handling this static/shared cases is now inside libtool instead of PCRE's make
2500 file.
2501
2502 4. The pcretest utility is now installed along with pcregrep because it is
2503 useful for users (to test regexs) and by doing this, it automatically gets
2504 relinked by libtool. The documentation has been turned into a man page, so
2505 there are now .1, .txt, and .html versions in /doc.
2506
2507 5. Upgrades to pcregrep:
2508 (i) Added long-form option names like gnu grep.
2509 (ii) Added --help to list all options with an explanatory phrase.
2510 (iii) Added -r, --recursive to recurse into sub-directories.
2511 (iv) Added -f, --file to read patterns from a file.
2512
2513 6. pcre_exec() was referring to its "code" argument before testing that
2514 argument for NULL (and giving an error if it was NULL).
2515
2516 7. Upgraded Makefile.in to allow for compiling in a different directory from
2517 the source directory.
2518
2519 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2520 options bits, the pointer it was passed was to an int instead of to an unsigned
2521 long int. This mattered only on 64-bit systems.
2522
2523 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2524 generated) instead of pcre.in, which it its source. Also made the same change
2525 in several of the .c files.
2526
2527 10. A new release of gcc defines printf() as a macro, which broke pcretest
2528 because it had an ifdef in the middle of a string argument for printf(). Fixed
2529 by using separate calls to printf().
2530
2531 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2532 script, to force use of CR or LF instead of \n in the source. On non-Unix
2533 systems, the value can be set in config.h.
2534
2535 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2536 absolute limit. Changed the text of the error message to make this clear, and
2537 likewise updated the man page.
2538
2539 13. The limit of 99 on the number of capturing subpatterns has been removed.
2540 The new limit is 65535, which I hope will not be a "real" limit.
2541
2542
2543 Version 3.4 22-Aug-00
2544 ---------------------
2545
2546 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2547
2548 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2549
2550
2551 Version 3.3 01-Aug-00
2552 ---------------------
2553
2554 1. If an octal character was given, but the value was greater than \377, it
2555 was not getting masked to the least significant bits, as documented. This could
2556 lead to crashes in some systems.
2557
2558 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2559 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2560
2561 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2562 These just pass their arguments on to (pcre_free)(), but they are provided
2563 because some uses of PCRE bind it to non-C systems that can call its functions,
2564 but cannot call free() or pcre_free() directly.
2565
2566 4. Add "make test" as a synonym for "make check". Corrected some comments in
2567 the Makefile.
2568
2569 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2570 Makefile.
2571
2572 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2573 command called pgrep for grepping around the active processes.
2574
2575 7. Added the beginnings of support for UTF-8 character strings.
2576
2577 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2578 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2579 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2580 out for the ar command.)
2581
2582
2583 Version 3.2 12-May-00
2584 ---------------------
2585
2586 This is purely a bug fixing release.
2587
2588 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2589 of ZA. This was just one example of several cases that could provoke this bug,
2590 which was introduced by change 9 of version 2.00. The code for breaking
2591 infinite loops after an iteration that matches an empty string was't working
2592 correctly.
2593
2594 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2595 when matched against abbab (for example). After matching an empty string, it
2596 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2597 caused it to match further down the string than it should.
2598
2599 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2600 was there because it doesn't seem to be needed, and it causes trouble on some
2601 systems, as it is not a Standard C header. It has been removed.
2602
2603 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2604 were reported on the Macintosh. The changes were from
2605
2606 while ((c = *(++ptr)) != 0 && c != '\n');
2607 to
2608 while ((c = *(++ptr)) != 0 && c != '\n') ;
2609
2610 Totally extraordinary, but if that's what it takes...
2611
2612 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2613 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2614 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2615 assumes the way PCRE uses memmove() (always moving upwards).
2616
2617 6. PCRE is being used in one environment where strchr() is not available. There
2618 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2619 faster code anyway.
2620
2621
2622 Version 3.1 09-Feb-00
2623 ---------------------
2624
2625 The only change in this release is the fixing of some bugs in Makefile.in for
2626 the "install" target:
2627
2628 (1) It was failing to install pcreposix.h.
2629
2630 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2631
2632
2633 Version 3.0 01-Feb-00
2634 ---------------------
2635
2636 1. Add support for the /+ modifier to perltest (to output $` like it does in
2637 pcretest).
2638
2639 2. Add support for the /g modifier to perltest.
2640
2641 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2642 matches null strings.
2643
2644 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2645 pattern. Perl treats empty patterns specially - it reuses the most recent
2646 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2647 effect.
2648
2649 5. The POSIX interface was broken in that it was just handing over the POSIX
2650 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2651 required a bigger vector, with some working space on the end. This means that
2652 the POSIX wrapper now has to get and free some memory, and copy the results.
2653
2654 6. Added some simple autoconf support, placing the test data and the
2655 documentation in separate directories, re-organizing some of the
2656 information files, and making it build pcre-config (a GNU standard). Also added
2657 libtool support for building PCRE as a shared library, which is now the
2658 default.
2659
2660 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2661 09 are not valid octal constants. Single digits will be used for minor values
2662 less than 10.
2663
2664 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2665 existing programs that set these in the POSIX interface can use PCRE without
2666 modification.
2667
2668 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2669 return all that pcre_info() returns, plus additional data. The pcre_info()
2670 function is retained for compatibility, but is considered to be obsolete.
2671
2672 10. Added experimental recursion feature (?R) to handle one common case that
2673 Perl 5.6 will be able to do with (?p{...}).
2674
2675 11. Added support for POSIX character classes like [:alpha:], which Perl is
2676 adopting.
2677
2678
2679 Version 2.08 31-Aug-99
2680 ----------------------
2681
2682 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2683 trying to match at the startoffset position, but instead was moving forward to
2684 the next newline as if a previous match had failed.
2685
2686 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2687 and could get into a loop if a null string was matched other than at the start
2688 of the subject.
2689
2690 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2691 be distinguished at compile time, and for completeness also added PCRE_DATE.
2692
2693 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2694 in GnuWin32 environments.
2695
2696
2697 Version 2.07 29-Jul-99
2698 ----------------------
2699
2700 1. The documentation is now supplied in plain text form and HTML as well as in
2701 the form of man page sources.
2702
2703 2. C++ compilers don't like assigning (void *) values to other pointer types.
2704 In particular this affects malloc(). Although there is no problem in Standard
2705 C, I've put in casts to keep C++ compilers happy.
2706
2707 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2708 should be (const char *).
2709
2710 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2711 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2712 However, I haven't made this a standard facility. The documentation doesn't
2713 mention it, and the Makefile doesn't support it.
2714
2715 5. The Makefile now contains an "install" target, with editable destinations at
2716 the top of the file. The pcretest program is not installed.
2717
2718 6. pgrep -V now gives the PCRE version number and date.
2719
2720 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2721 causing the entire string to be ignored, instead of just the last character.
2722
2723 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2724 non-matching string, it can take a very, very long time, even for strings of
2725 quite modest length, because of the nested recursion. PCRE now does better in
2726 some of these cases. It does this by remembering the last required literal
2727 character in the pattern, and pre-searching the subject to ensure it is present
2728 before running the real match. In other words, it applies a heuristic to detect
2729 some types of certain failure quickly, and in the above example, if presented
2730 with a string that has no trailing " it gives "no match" very quickly.
2731
2732 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2733 other alternatives are tried instead.
2734
2735
2736 Version 2.06 09-Jun-99
2737 ----------------------
2738
2739 1. Change pcretest's output for amount of store used to show just the code
2740 space, because the remainder (the data block) varies in size between 32-bit and
2741 64-bit systems.
2742
2743 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2744 start matching at. This allows lookbehinds to work when searching for multiple
2745 occurrences in a string.
2746
2747 3. Added additional options to pcretest for testing multiple occurrences:
2748
2749 /+ outputs the rest of the string that follows a match
2750 /g loops for multiple occurrences, using the new startoffset argument
2751 /G loops for multiple occurrences by passing an incremented pointer
2752
2753 4. PCRE wasn't doing the "first character" optimization for patterns starting
2754 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2755 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2756 the letter 'x'. On long subject strings, this gives a significant speed-up.
2757
2758
2759 Version 2.05 21-Apr-99
2760 ----------------------
2761
2762 1. Changed the type of magic_number from int to long int so that it works
2763 properly on 16-bit systems.
2764
2765 2. Fixed a bug which caused patterns starting with .* not to work correctly
2766 when the subject string contained newline characters. PCRE was assuming
2767 anchoring for such patterns in all cases, which is not correct because .* will
2768 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2769 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2770 must be retried after every newline in the subject.
2771
2772
2773 Version 2.04 18-Feb-99
2774 ----------------------
2775
2776 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2777 computation of the store needed to hold the pattern was incorrect (too large).
2778 If such patterns were nested a few deep, this could multiply and become a real
2779 problem.
2780
2781 2. Added /M option to pcretest to show the memory requirement of a specific
2782 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2783
2784 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2785 compiled in such a way that the backtracking after subsequent failure was
2786 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2787 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2788
2789
2790 Version 2.03 02-Feb-99
2791 ----------------------
2792
2793 1. Fixed typo and small mistake in man page.
2794
2795 2. Added 4th condition (GPL supersedes if conflict) and created separate
2796 LICENCE file containing the conditions.
2797
2798 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2799 Perl, that is the internal \ allows the delimiter to be included in the
2800 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2801 the final delimiter, add \ to the end of the pattern (to test the error).
2802
2803 4. Added the convenience functions for extracting substrings after a successful
2804 match. Updated pcretest to make it able to test these functions.
2805
2806
2807 Version 2.02 14-Jan-99
2808 ----------------------
2809
2810 1. Initialized the working variables associated with each extraction so that
2811 their saving and restoring doesn't refer to uninitialized store.
2812
2813 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2814 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2815 fix the problem.
2816
2817 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2818 calls, and wasn't printing the correct value for compiling calls. Increased the
2819 default value of LOOPREPEAT, and the number of significant figures in the
2820 times.
2821
2822 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2823
2824 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2825 a building problem on Windows NT with a FAT file system.
2826
2827
2828 Version 2.01 21-Oct-98
2829 ----------------------
2830
2831 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2832 to character tables built by pcre_maketables() in the current locale. If NULL
2833 is passed, the default tables are used.
2834
2835
2836 Version 2.00 24-Sep-98
2837 ----------------------
2838
2839 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2840 it any more.
2841
2842 2. Allow quantification of (?>) groups, and make it work correctly.
2843
2844 3. The first character computation wasn't working for (?>) groups.
2845
2846 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2847 end of the subject) and add 5.005's \z, which really does match only at the
2848 very end of the subject.
2849
2850 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2851
2852 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2853 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2854 localized options. All options to pcre_study() were also removed.
2855
2856 7. Add other new features from 5.005:
2857
2858 $(?<= positive lookbehind
2859 $(?<! negative lookbehind
2860 (?imsx-imsx) added the unsetting capability
2861 such a setting is global if at outer level; local otherwise
2862 (?imsx-imsx:) non-capturing groups with option setting
2863 (?(cond)re|re) conditional pattern matching
2864
2865 A backreference to itself in a repeated group matches the previous
2866 captured string.
2867
2868 8. General tidying up of studying (both automatic and via "study")
2869 consequential on the addition of new assertions.
2870
2871 9. As in 5.005, unlimited repeated groups that could match an empty substring
2872 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2873 runtime if any iteration does actually match an empty substring.
2874
2875 10. Include the RunTest script in the distribution.
2876
2877 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2878 discrepancies, some of which were old and were also with respect to 5.004. They
2879 have now been fixed.
2880
2881
2882 Version 1.09 28-Apr-98
2883 ----------------------
2884
2885 1. A negated single character class followed by a quantifier with a minimum
2886 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2887 program crashes, or just wrong answers. This did not apply to negated classes
2888 containing more than one character, or to minima other than one.
2889
2890
2891 Version 1.08 27-Mar-98
2892 ----------------------
2893
2894 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2895
2896 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2897 latter must appear before anything that relies on it in the pattern.
2898
2899
2900 Version 1.07 16-Feb-98
2901 ----------------------
2902
2903 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2904 repeat of a potentially empty string).
2905
2906
2907 Version 1.06 23-Jan-98
2908 ----------------------
2909
2910 1. Added Markus Oberhumer's little patches for C++.
2911
2912 2. Literal strings longer than 255 characters were broken.
2913
2914
2915 Version 1.05 23-Dec-97
2916 ----------------------
2917
2918 1. Negated character classes containing more than one character were failing if
2919 PCRE_CASELESS was set at run time.
2920
2921
2922 Version 1.04 19-Dec-97
2923 ----------------------
2924
2925 1. Corrected the man page, where some "const" qualifiers had been omitted.
2926
2927 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2928 input syntax.
2929
2930 3. Fixed memory leak which occurred when a regex with back references was
2931 matched with an offsets vector that wasn't big enough. The temporary memory
2932 that is used in this case wasn't being freed if the match failed.
2933
2934 4. Tidied pcretest to ensure it frees memory that it gets.
2935
2936 5. Temporary memory was being obtained in the case where the passed offsets
2937 vector was exactly big enough.
2938
2939 6. Corrected definition of offsetof() from change 5 below.
2940
2941 7. I had screwed up change 6 below and broken the rules for the use of
2942 setjmp(). Now fixed.
2943
2944
2945 Version 1.03 18-Dec-97
2946 ----------------------
2947
2948 1. A erroneous regex with a missing opening parenthesis was correctly
2949 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2950 on some systems.
2951
2952 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2953 it was reported that one broken compiler failed on the former because "code" is
2954 also an independent variable.
2955
2956 3. The erroneous regex a[]b caused an array overrun reference.
2957
2958 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2959 fail on data ending with that character. (It was going on too far, and checking
2960 the next character, typically a binary zero.) This was specific to the
2961 optimized code for single-character negative classes.
2962
2963 5. Added a contributed patch from the TIN world which does the following:
2964
2965 + Add an undef for memmove, in case the the system defines a macro for it.
2966
2967 + Add a definition of offsetof(), in case there isn't one. (I don't know
2968 the reason behind this - offsetof() is part of the ANSI standard - but
2969 it does no harm).
2970
2971 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2972 most of the places where whitespace preceded '#'. I have given up and
2973 allowed the remaining 2 cases to be at the margin.
2974
2975 + Rename some variables in pcre to eliminate shadowing. This seems very
2976 pedantic, but does no harm, of course.
2977
2978 6. Moved the call to setjmp() into its own function, to get rid of warnings
2979 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2980
2981 7. Constructs such as \d{8,} were compiling into the equivalent of
2982 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2983 outcome, but in this particular case used more store than had been allocated,
2984 which caused the bug to be discovered because it threw up an internal error.
2985
2986 8. The debugging code in both pcre and pcretest for outputting the compiled
2987 form of a regex was going wrong in the case of back references followed by
2988 curly-bracketed repeats.
2989
2990
2991 Version 1.02 12-Dec-97
2992 ----------------------
2993
2994 1. Typos in pcre.3 and comments in the source fixed.
2995
2996 2. Applied a contributed patch to get rid of places where it used to remove
2997 'const' from variables, and fixed some signed/unsigned and uninitialized
2998 variable warnings.
2999
3000 3. Added the "runtest" target to Makefile.
3001
3002 4. Set default compiler flag to -O2 rather than just -O.
3003
3004
3005 Version 1.01 19-Nov-97
3006 ----------------------
3007
3008 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3009 like /([ab]*)*/, that is, for classes with more than one character in them.
3010
3011 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3012 as /((?>a*))*/ (a PCRE_EXTRA facility).
3013
3014
3015 Version 1.00 18-Nov-97
3016 ----------------------
3017
3018 1. Added compile-time macros to support systems such as SunOS4 which don't have
3019 memmove() or strerror() but have other things that can be used instead.
3020
3021 2. Arranged that "make clean" removes the executables.
3022
3023
3024 Version 0.99 27-Oct-97
3025 ----------------------
3026
3027 1. Fixed bug in code for optimizing classes with only one character. It was
3028 initializing a 32-byte map regardless, which could cause it to run off the end
3029 of the memory it had got.
3030
3031 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3032
3033
3034 Version 0.98 22-Oct-97
3035 ----------------------
3036
3037 1. Fixed bug in code for handling temporary memory usage when there are more
3038 back references than supplied space in the ovector. This could cause segfaults.
3039
3040
3041 Version 0.97 21-Oct-97
3042 ----------------------
3043
3044 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3045
3046 2. Optimized negated single characters not to use a bit map.
3047
3048 3. Brought error texts together as macro definitions; clarified some of them;
3049 fixed one that was wrong - it said "range out of order" when it meant "invalid
3050 escape sequence".
3051
3052 4. Changed some char * arguments to const char *.
3053
3054 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3055
3056 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3057 pcretest.
3058
3059
3060 Version 0.96 16-Oct-97
3061 ----------------------
3062
3063 1. Added a simple "pgrep" utility to the distribution.
3064
3065 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3066 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3067 where "ddd" means "one or more decimal digits".
3068
3069 3. Fixed serious bug. If a pattern had a back reference, but the call to
3070 pcre_exec() didn't supply a large enough ovector to record the related
3071 identifying subpattern, the match always failed. PCRE now remembers the number
3072 of the largest back reference, and gets some temporary memory in which to save
3073 the offsets during matching if necessary, in order to ensure that
3074 backreferences always work.
3075
3076 4. Increased the compatibility with Perl in a number of ways:
3077
3078 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3079 to request this handling. The option can be set at compile or exec time.
3080
3081 (b) $ matches before a terminating newline by default; an option
3082 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3083 mode). The option can be set at compile or exec time.
3084
3085 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3086 the same as Perl's. If the decimal number it represents is less than 10
3087 or there aren't that many previous left capturing parentheses, an octal
3088 escape is read. Inside a character class, it's always an octal escape,
3089 even if it is a single digit.
3090
3091 (d) An escaped but undefined alphabetic character is taken as a literal,
3092 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3093 escapes.
3094
3095 (e) {0} is now permitted. (The previous item is removed from the compiled
3096 pattern).
3097
3098 5. Changed all the names of code files so that the basic parts are no longer
3099 than 10 characters, and abolished the teeny "globals.c" file.
3100
3101 6. Changed the handling of character classes; they are now done with a 32-byte
3102 bit map always.
3103
3104 7. Added the -d and /D options to pcretest to make it possible to look at the
3105 internals of compilation without having to recompile pcre.
3106
3107
3108 Version 0.95 23-Sep-97
3109 ----------------------
3110
3111 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3112 \x20 at the start of a run of normal characters. These were being treated as
3113 real characters, instead of the source characters being re-checked.
3114
3115
3116 Version 0.94 18-Sep-97
3117 ----------------------
3118
3119 1. The functions are now thread-safe, with the caveat that the global variables
3120 containing pointers to malloc() and free() or alternative functions are the
3121 same for all threads.
3122
3123 2. Get pcre_study() to generate a bitmap of initial characters for non-
3124 anchored patterns when this is possible, and use it if passed to pcre_exec().
3125
3126
3127 Version 0.93 15-Sep-97
3128 ----------------------
3129
3130 1. /(b)|(:+)/ was computing an incorrect first character.
3131
3132 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3133 but not actually doing anything yet.
3134
3135 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3136 as Perl does (e.g. [-az] or [az-]).
3137
3138 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3139 all possible positions.
3140
3141 5. Split up into different modules to avoid including unneeded functions in a
3142 compiled binary. However, compile and exec are still in one module. The "study"
3143 function is split off.
3144
3145 6. The character tables are now in a separate module whose source is generated
3146 by an auxiliary program - but can then be edited by hand if required. There are
3147 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3148 toupper() in the code.
3149
3150 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3151 make them global. Abolish the function for setting them, as the caller can now
3152 set them directly.
3153
3154
3155 Version 0.92 11-Sep-97
3156 ----------------------
3157
3158 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3159 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3160
3161 2. Caseless matching was not working in character classes if the characters in
3162 the pattern were in upper case.
3163
3164 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3165
3166 4. Make PCRE_ANCHORED public and accept as a compile option.
3167
3168 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3169 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3170 pass them.
3171
3172 6. Give an error if bad option bits passed at compile or run time.
3173
3174 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3175 pcretest to cause it to pass that flag.
3176
3177 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3178 options, and the first character, if set.
3179
3180 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3181
3182
3183 Version 0.91 10-Sep-97
3184 ----------------------
3185
3186 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3187 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3188
3189 2. PCRE was looping on encountering an indefinitely repeated back reference to
3190 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3191 Perl does - treats the match as successful.
3192
3193 ****

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