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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 244 - (show annotations) (download)
Fri Sep 14 09:14:24 2007 UTC (7 years, 2 months ago) by ph10
File size: 141184 byte(s)
Fix broken --disable-stack-for-recursion without --enable-unicode-properties.

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

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