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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 261 - (show annotations) (download)
Fri Sep 21 08:37:48 2007 UTC (7 years, 3 months ago) by ph10
File size: 141931 byte(s)
Final doc updates and file tidies for 7.4.

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

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