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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 240 - (show annotations) (download)
Tue Sep 11 15:47:20 2007 UTC (6 years, 11 months ago) by ph10
File size: 141107 byte(s)
Refactoring to reduce the number of relocations in a shared library.

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

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