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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 253 - (show annotations) (download)
Mon Sep 17 10:51:30 2007 UTC (7 years, 3 months ago) by ph10
File size: 141388 byte(s)
Added checks for ANY and ANYCRLF to pcrecpp.cc where previously it checked only 
for CRLF.

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

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