/[pcre]/code/tags/pcre-6.5/ChangeLog
ViewVC logotype

Contents of /code/tags/pcre-6.5/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 88 - (show annotations) (download)
Sat Feb 24 21:41:23 2007 UTC (7 years, 7 months ago) by nigel
File size: 95036 byte(s)
Tag code/trunk as code/tags/pcre-6.5.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 6.5 01-Feb-06
5 ---------------------
6
7 1. When using the partial match feature with pcre_dfa_exec(), it was not
8 anchoring the second and subsequent partial matches at the new starting
9 point. This could lead to incorrect results. For example, with the pattern
10 /1234/, partially matching against "123" and then "a4" gave a match.
11
12 2. Changes to pcregrep:
13
14 (a) All non-match returns from pcre_exec() were being treated as failures
15 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
16 error message is output. Some extra information is given for the
17 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
18 probably the only errors that are likely to be caused by users (by
19 specifying a regex that has nested indefinite repeats, for instance).
20 If there are more than 20 of these errors, pcregrep is abandoned.
21
22 (b) A binary zero was treated as data while matching, but terminated the
23 output line if it was written out. This has been fixed: binary zeroes
24 are now no different to any other data bytes.
25
26 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
27 used to set a locale for matching. The --locale=xxxx long option has
28 been added (no short equivalent) to specify a locale explicitly on the
29 pcregrep command, overriding the environment variables.
30
31 (d) When -B was used with -n, some line numbers in the output were one less
32 than they should have been.
33
34 (e) Added the -o (--only-matching) option.
35
36 (f) If -A or -C was used with -c (count only), some lines of context were
37 accidentally printed for the final match.
38
39 (g) Added the -H (--with-filename) option.
40
41 (h) The combination of options -rh failed to suppress file names for files
42 that were found from directory arguments.
43
44 (i) Added the -D (--devices) and -d (--directories) options.
45
46 (j) Added the -F (--fixed-strings) option.
47
48 (k) Allow "-" to be used as a file name for -f as well as for a data file.
49
50 (l) Added the --colo(u)r option.
51
52 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
53 is not present by default.
54
55 3. A nasty bug was discovered in the handling of recursive patterns, that is,
56 items such as (?R) or (?1), when the recursion could match a number of
57 alternatives. If it matched one of the alternatives, but subsequently,
58 outside the recursion, there was a failure, the code tried to back up into
59 the recursion. However, because of the way PCRE is implemented, this is not
60 possible, and the result was an incorrect result from the match.
61
62 In order to prevent this happening, the specification of recursion has
63 been changed so that all such subpatterns are automatically treated as
64 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
65
66 4. I had overlooked the fact that, in some locales, there are characters for
67 which isalpha() is true but neither isupper() nor islower() are true. In
68 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
69 and ordfeminine) are like this. This affected the treatment of \w and \W
70 when they appeared in character classes, but not when they appeared outside
71 a character class. The bit map for "word" characters is now created
72 separately from the results of isalnum() instead of just taking it from the
73 upper, lower, and digit maps. (Plus the underscore character, of course.)
74
75 5. The above bug also affected the handling of POSIX character classes such as
76 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
77 permanent tables. Instead, the bit maps for such a class were previously
78 created as the appropriate unions of the upper, lower, and digit bitmaps.
79 Now they are created by subtraction from the [[:word:]] class, which has
80 its own bitmap.
81
82 6. The [[:blank:]] character class matches horizontal, but not vertical space.
83 It is created by subtracting the vertical space characters (\x09, \x0a,
84 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
85 subtraction was done in the overall bitmap for a character class, meaning
86 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
87 be recognized. This bug has been fixed.
88
89 7. Patches from the folks at Google:
90
91 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
92 real life, but is still worth protecting against".
93
94 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
95 regular expressions".
96
97 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
98 have it.
99
100 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
101 "configure" and the latter not, in order to fix a problem somebody had
102 with compiling the Arg class on HP-UX.
103
104 (e) Improve the error-handling of the C++ wrapper a little bit.
105
106 (f) New tests for checking recursion limiting.
107
108 8. The pcre_memmove() function, which is used only if the environment does not
109 have a standard memmove() function (and is therefore rarely compiled),
110 contained two bugs: (a) use of int instead of size_t, and (b) it was not
111 returning a result (though PCRE never actually uses the result).
112
113 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
114 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
115 returned instead of calling malloc() with an overflowing number that would
116 most likely cause subsequent chaos.
117
118 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
119
120 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
121 with this option is matched, the nmatch and pmatch options of regexec() are
122 ignored.
123
124 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
125 provided in case anyone wants to the the POSIX interface with UTF-8
126 strings.
127
128 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
129 C++ linking (needed for some HP-UX environments).
130
131 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
132 (unused parameter) and in the pcre_printint() function (omitted "default"
133 switch label when the default is to do nothing).
134
135 15. Added some code to make it possible, when PCRE is compiled as a C++
136 library, to replace subject pointers for pcre_exec() with a smart pointer
137 class, thus making it possible to process discontinuous strings.
138
139 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
140 much the same function. They were added by different people who were trying
141 to make PCRE easy to compile on non-Unix systems. It has been suggested
142 that PCRE_EXPORT be abolished now that there is more automatic apparatus
143 for compiling on Windows systems. I have therefore replaced it with
144 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
145 defaults to "extern" for C or "extern C" for C++, which works fine on
146 Unix-like systems. It is now possible to override the value of PCRE_DATA_
147 SCOPE with something explicit in config.h. In addition:
148
149 (a) pcreposix.h still had just "extern" instead of either of these macros;
150 I have replaced it with PCRE_DATA_SCOPE.
151
152 (b) Functions such as _pcre_xclass(), which are internal to the library,
153 but external in the C sense, all had PCRE_EXPORT in their definitions.
154 This is apparently wrong for the Windows case, so I have removed it.
155 (It makes no difference on Unix-like systems.)
156
157 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
158 of recursive calls to match(). This is different to MATCH_LIMIT because
159 that limits the total number of calls to match(), not all of which increase
160 the depth of recursion. Limiting the recursion depth limits the amount of
161 stack (or heap if NO_RECURSE is set) that is used. The default can be set
162 when PCRE is compiled, and changed at run time. A patch from Google adds
163 this functionality to the C++ interface.
164
165 18. Changes to the handling of Unicode character properties:
166
167 (a) Updated the table to Unicode 4.1.0.
168
169 (b) Recognize characters that are not in the table as "Cn" (undefined).
170
171 (c) I revised the way the table is implemented to a much improved format
172 which includes recognition of ranges. It now supports the ranges that
173 are defined in UnicodeData.txt, and it also amalgamates other
174 characters into ranges. This has reduced the number of entries in the
175 table from around 16,000 to around 3,000, thus reducing its size
176 considerably. I realized I did not need to use a tree structure after
177 all - a binary chop search is just as efficient. Having reduced the
178 number of entries, I extended their size from 6 bytes to 8 bytes to
179 allow for more data.
180
181 (d) Added support for Unicode script names via properties such as \p{Han}.
182
183 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
184 matching that character.
185
186 20. When matching a repeated Unicode property with a minimum greater than zero,
187 (for example \pL{2,}), PCRE could look past the end of the subject if it
188 reached it while seeking the minimum number of characters. This could
189 happen only if some of the characters were more than one byte long, because
190 there is a check for at least the minimum number of bytes.
191
192 21. Refactored the implementation of \p and \P so as to be more general, to
193 allow for more different types of property in future. This has changed the
194 compiled form incompatibly. Anybody with saved compiled patterns that use
195 \p or \P will have to recompile them.
196
197 22. Added "Any" and "L&" to the supported property types.
198
199 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
200 but give a compile time error if the value is greater than 0xff.
201
202 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
203 accidentally not being installed or uninstalled.
204
205 25. The pcre.h file was built from pcre.h.in, but the only changes that were
206 made were to insert the current release number. This seemed silly, because
207 it made things harder for people building PCRE on systems that don't run
208 "configure". I have turned pcre.h into a distributed file, no longer built
209 by "configure", with the version identification directly included. There is
210 no longer a pcre.h.in file.
211
212 However, this change necessitated a change to the pcre-config script as
213 well. It is built from pcre-config.in, and one of the substitutions was the
214 release number. I have updated configure.ac so that ./configure now finds
215 the release number by grepping pcre.h.
216
217 26. Added the ability to run the tests under valgrind.
218
219
220 Version 6.4 05-Sep-05
221 ---------------------
222
223 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
224 "--" to be printed when multiple files were scanned, even when none of the
225 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
226 consider it to be a bug, and have restored the previous behaviour.
227
228 2. A couple of code tidies to get rid of compiler warnings.
229
230 3. The pcretest program used to cheat by referring to symbols in the library
231 whose names begin with _pcre_. These are internal symbols that are not
232 really supposed to be visible externally, and in some environments it is
233 possible to suppress them. The cheating is now confined to including
234 certain files from the library's source, which is a bit cleaner.
235
236 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
237 file's purpose clearer.
238
239 5. Reorganized pcre_ucp_findchar().
240
241
242 Version 6.3 15-Aug-05
243 ---------------------
244
245 1. The file libpcre.pc.in did not have general read permission in the tarball.
246
247 2. There were some problems when building without C++ support:
248
249 (a) If C++ support was not built, "make install" and "make test" still
250 tried to test it.
251
252 (b) There were problems when the value of CXX was explicitly set. Some
253 changes have been made to try to fix these, and ...
254
255 (c) --disable-cpp can now be used to explicitly disable C++ support.
256
257 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
258 backslash in a target when C++ was disabled. This confuses some
259 versions of "make", apparently. Using an intermediate variable solves
260 this. (Same for CPP_LOBJ.)
261
262 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
263 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
264 necessary on certain architectures.
265
266 4. Added a setting of -export-symbols-regex to the link command to remove
267 those symbols that are exported in the C sense, but actually are local
268 within the library, and not documented. Their names all begin with
269 "_pcre_". This is not a perfect job, because (a) we have to except some
270 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
271 available (and never for static libraries). I have made a note to try to
272 find a way round (a) in the future.
273
274
275 Version 6.2 01-Aug-05
276 ---------------------
277
278 1. There was no test for integer overflow of quantifier values. A construction
279 such as {1111111111111111} would give undefined results. What is worse, if
280 a minimum quantifier for a parenthesized subpattern overflowed and became
281 negative, the calculation of the memory size went wrong. This could have
282 led to memory overwriting.
283
284 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
285
286 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
287 operating environments where this matters.
288
289 4. Applied Giuseppe Maxia's patch to add additional features for controlling
290 PCRE options from within the C++ wrapper.
291
292 5. Named capturing subpatterns were not being correctly counted when a pattern
293 was compiled. This caused two problems: (a) If there were more than 100
294 such subpatterns, the calculation of the memory needed for the whole
295 compiled pattern went wrong, leading to an overflow error. (b) Numerical
296 back references of the form \12, where the number was greater than 9, were
297 not recognized as back references, even though there were sufficient
298 previous subpatterns.
299
300 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
301 versions of gcc, e.g. 2.95.4.
302
303
304 Version 6.1 21-Jun-05
305 ---------------------
306
307 1. There was one reference to the variable "posix" in pcretest.c that was not
308 surrounded by "#if !defined NOPOSIX".
309
310 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
311 the cross-check on the old pcre_info() function, for the benefit of the
312 cut-down version of PCRE that is currently imported into Exim.
313
314 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
315 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
316 patterns that start (?i) but I don't think that matters. The use of (?i) is
317 just an example; this all applies to the other options as well.
318
319 4. Since libtool seems to echo the compile commands it is issuing, the output
320 from "make" can be reduced a bit by putting "@" in front of each libtool
321 compile command.
322
323 5. Patch from the folks at Google for configure.in to be a bit more thorough
324 in checking for a suitable C++ installation before trying to compile the
325 C++ stuff. This should fix a reported problem when a compiler was present,
326 but no suitable headers.
327
328 6. The man pages all had just "PCRE" as their title. I have changed them to
329 be the relevant file name. I have also arranged that these names are
330 retained in the file doc/pcre.txt, which is a concatenation in text format
331 of all the man pages except the little individual ones for each function.
332
333 7. The NON-UNIX-USE file had not been updated for the different set of source
334 files that come with release 6. I also added a few comments about the C++
335 wrapper.
336
337
338 Version 6.0 07-Jun-05
339 ---------------------
340
341 1. Some minor internal re-organization to help with my DFA experiments.
342
343 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
344 didn't matter for the library itself when fully configured, but did matter
345 when compiling without UCP support, or within Exim, where the ucp files are
346 not imported.
347
348 3. Refactoring of the library code to split up the various functions into
349 different source modules. The addition of the new DFA matching code (see
350 below) to a single monolithic source would have made it really too
351 unwieldy, quite apart from causing all the code to be include in a
352 statically linked application, when only some functions are used. This is
353 relevant even without the DFA addition now that patterns can be compiled in
354 one application and matched in another.
355
356 The downside of splitting up is that there have to be some external
357 functions and data tables that are used internally in different modules of
358 the library but which are not part of the API. These have all had their
359 names changed to start with "_pcre_" so that they are unlikely to clash
360 with other external names.
361
362 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
363 a different (DFA) algorithm. Although it is slower than the original
364 function, it does have some advantages for certain types of matching
365 problem.
366
367 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
368 including restarting after a partial match.
369
370 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
371 defined when compiling for Windows was sent to me. I have put it into the
372 code, though I have no means of testing or verifying it.
373
374 7. Added the pcre_refcount() auxiliary function.
375
376 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
377 match before or at the first newline in the subject string. In pcretest,
378 the /f option on a pattern can be used to set this.
379
380 9. A repeated \w when used in UTF-8 mode with characters greater than 256
381 would behave wrongly. This has been present in PCRE since release 4.0.
382
383 10. A number of changes to the pcregrep command:
384
385 (a) Refactored how -x works; insert ^(...)$ instead of setting
386 PCRE_ANCHORED and checking the length, in preparation for adding
387 something similar for -w.
388
389 (b) Added the -w (match as a word) option.
390
391 (c) Refactored the way lines are read and buffered so as to have more
392 than one at a time available.
393
394 (d) Implemented a pcregrep test script.
395
396 (e) Added the -M (multiline match) option. This allows patterns to match
397 over several lines of the subject. The buffering ensures that at least
398 8K, or the rest of the document (whichever is the shorter) is available
399 for matching (and similarly the previous 8K for lookbehind assertions).
400
401 (f) Changed the --help output so that it now says
402
403 -w, --word-regex(p)
404
405 instead of two lines, one with "regex" and the other with "regexp"
406 because that confused at least one person since the short forms are the
407 same. (This required a bit of code, as the output is generated
408 automatically from a table. It wasn't just a text change.)
409
410 (g) -- can be used to terminate pcregrep options if the next thing isn't an
411 option but starts with a hyphen. Could be a pattern or a path name
412 starting with a hyphen, for instance.
413
414 (h) "-" can be given as a file name to represent stdin.
415
416 (i) When file names are being printed, "(standard input)" is used for
417 the standard input, for compatibility with GNU grep. Previously
418 "<stdin>" was used.
419
420 (j) The option --label=xxx can be used to supply a name to be used for
421 stdin when file names are being printed. There is no short form.
422
423 (k) Re-factored the options decoding logic because we are going to add
424 two more options that take data. Such options can now be given in four
425 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
426
427 (l) Added the -A, -B, and -C options for requesting that lines of context
428 around matches be printed.
429
430 (m) Added the -L option to print the names of files that do not contain
431 any matching lines, that is, the complement of -l.
432
433 (n) The return code is 2 if any file cannot be opened, but pcregrep does
434 continue to scan other files.
435
436 (o) The -s option was incorrectly implemented. For compatibility with other
437 greps, it now suppresses the error message for a non-existent or non-
438 accessible file (but not the return code). There is a new option called
439 -q that suppresses the output of matching lines, which was what -s was
440 previously doing.
441
442 (p) Added --include and --exclude options to specify files for inclusion
443 and exclusion when recursing.
444
445 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
446 Hopefully, it now does.
447
448 12. Missing cast in pcre_study().
449
450 13. Added an "uninstall" target to the makefile.
451
452 14. Replaced "extern" in the function prototypes in Makefile.in with
453 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
454 world, but is set differently for Windows.
455
456 15. Added a second compiling function called pcre_compile2(). The only
457 difference is that it has an extra argument, which is a pointer to an
458 integer error code. When there is a compile-time failure, this is set
459 non-zero, in addition to the error test pointer being set to point to an
460 error message. The new argument may be NULL if no error number is required
461 (but then you may as well call pcre_compile(), which is now just a
462 wrapper). This facility is provided because some applications need a
463 numeric error indication, but it has also enabled me to tidy up the way
464 compile-time errors are handled in the POSIX wrapper.
465
466 16. Added VPATH=.libs to the makefile; this should help when building with one
467 prefix path and installing with another. (Or so I'm told by someone who
468 knows more about this stuff than I do.)
469
470 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
471 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
472 match everything, including newlines. This is not POSIX-compatible, but
473 somebody wanted the feature. From pcretest it can be activated by using
474 both the P and the s flags.
475
476 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
477
478 19. libpcre.pc was being incorrectly installed as executable.
479
480 20. A couple of places in pcretest check for end-of-line by looking for '\n';
481 it now also looks for '\r' so that it will work unmodified on Windows.
482
483 21. Added Google's contributed C++ wrapper to the distribution.
484
485 22. Added some untidy missing memory free() calls in pcretest, to keep
486 Electric Fence happy when testing.
487
488
489
490 Version 5.0 13-Sep-04
491 ---------------------
492
493 1. Internal change: literal characters are no longer packed up into items
494 containing multiple characters in a single byte-string. Each character
495 is now matched using a separate opcode. However, there may be more than one
496 byte in the character in UTF-8 mode.
497
498 2. The pcre_callout_block structure has two new fields: pattern_position and
499 next_item_length. These contain the offset in the pattern to the next match
500 item, and its length, respectively.
501
502 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
503 insertion of callouts before each pattern item. Added the /C option to
504 pcretest to make use of this.
505
506 4. On the advice of a Windows user, the lines
507
508 #if defined(_WIN32) || defined(WIN32)
509 _setmode( _fileno( stdout ), 0x8000 );
510 #endif /* defined(_WIN32) || defined(WIN32) */
511
512 have been added to the source of pcretest. This apparently does useful
513 magic in relation to line terminators.
514
515 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
516 for the benefit of those environments where the "b" makes a difference.
517
518 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
519 to know about it. I have put a hack into configure.in that adds in code
520 to set GCC=yes if CC=icc. This seems to end up at a point in the
521 generated configure script that is early enough to affect the setting of
522 compiler options, which is what is needed, but I have no means of testing
523 whether it really works. (The user who reported this had patched the
524 generated configure script, which of course I cannot do.)
525
526 LATER: After change 22 below (new libtool files), the configure script
527 seems to know about icc (and also ecc). Therefore, I have commented out
528 this hack in configure.in.
529
530 7. Added support for pkg-config (2 patches were sent in).
531
532 8. Negated POSIX character classes that used a combination of internal tables
533 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
534 [[:^ascii]]. Typically, they would match almost any characters. The other
535 POSIX classes were not broken in this way.
536
537 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
538 to find the match, as PCRE was deluded into thinking that the match had to
539 start at the start point or following a newline. The same bug applied to
540 patterns with negative forward assertions or any backward assertions
541 preceding ".*" at the start, unless the pattern required a fixed first
542 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
543
544 10. In UTF-8 mode, when moving forwards in the subject after a failed match
545 starting at the last subject character, bytes beyond the end of the subject
546 string were read.
547
548 11. Renamed the variable "class" as "classbits" to make life easier for C++
549 users. (Previously there was a macro definition, but it apparently wasn't
550 enough.)
551
552 12. Added the new field "tables" to the extra data so that tables can be passed
553 in at exec time, or the internal tables can be re-selected. This allows
554 a compiled regex to be saved and re-used at a later time by a different
555 program that might have everything at different addresses.
556
557 13. Modified the pcre-config script so that, when run on Solaris, it shows a
558 -R library as well as a -L library.
559
560 14. The debugging options of pcretest (-d on the command line or D on a
561 pattern) showed incorrect output for anything following an extended class
562 that contained multibyte characters and which was followed by a quantifier.
563
564 15. Added optional support for general category Unicode character properties
565 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
566 support. It adds about 90K to the size of the library. The meanings of the
567 inbuilt class escapes such as \d and \s have NOT been changed.
568
569 16. Updated pcredemo.c to include calls to free() to release the memory for the
570 compiled pattern.
571
572 17. The generated file chartables.c was being created in the source directory
573 instead of in the building directory. This caused the build to fail if the
574 source directory was different from the building directory, and was
575 read-only.
576
577 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
578 file. No doubt somebody will tell me if they don't make sense... Also added
579 Dan Mooney's comments about building on OpenVMS.
580
581 19. Added support for partial matching via the PCRE_PARTIAL option for
582 pcre_exec() and the \P data escape in pcretest.
583
584 20. Extended pcretest with 3 new pattern features:
585
586 (i) A pattern option of the form ">rest-of-line" causes pcretest to
587 write the compiled pattern to the file whose name is "rest-of-line".
588 This is a straight binary dump of the data, with the saved pointer to
589 the character tables forced to be NULL. The study data, if any, is
590 written too. After writing, pcretest reads a new pattern.
591
592 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
593 compiled pattern from the given file. There must not be any
594 occurrences of "<" in the file name (pretty unlikely); if there are,
595 pcretest will instead treat the initial "<" as a pattern delimiter.
596 After reading in the pattern, pcretest goes on to read data lines as
597 usual.
598
599 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
600 and 16-bit fields in a compiled pattern, to simulate a pattern that
601 was compiled on a host of opposite endianness.
602
603 21. The pcre-exec() function can now cope with patterns that were compiled on
604 hosts of opposite endianness, with this restriction:
605
606 As for any compiled expression that is saved and used later, the tables
607 pointer field cannot be preserved; the extra_data field in the arguments
608 to pcre_exec() should be used to pass in a tables address if a value
609 other than the default internal tables were used at compile time.
610
611 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
612 now diagnosed as an error. Previously, most of the time, a negative number
613 would have been treated as zero, but if in addition "ovector" was passed as
614 NULL, a crash could occur.
615
616 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
617 new versions from the libtool 1.5 distribution (the last one is a copy of
618 a file called libtool.m4). This seems to have fixed the need to patch
619 "configure" to support Darwin 1.3 (which I used to do). However, I still
620 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
621 workstation).
622
623 24. Changed the PCRE licence to be the more standard "BSD" licence.
624
625
626 Version 4.5 01-Dec-03
627 ---------------------
628
629 1. There has been some re-arrangement of the code for the match() function so
630 that it can be compiled in a version that does not call itself recursively.
631 Instead, it keeps those local variables that need separate instances for
632 each "recursion" in a frame on the heap, and gets/frees frames whenever it
633 needs to "recurse". Keeping track of where control must go is done by means
634 of setjmp/longjmp. The whole thing is implemented by a set of macros that
635 hide most of the details from the main code, and operates only if
636 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
637 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
638 operating.
639
640 To make it easier for callers to provide specially tailored get/free
641 functions for this usage, two new functions, pcre_stack_malloc, and
642 pcre_stack_free, are used. They are always called in strict stacking order,
643 and the size of block requested is always the same.
644
645 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
646 PCRE has been compiled to use the stack or the heap for recursion. The
647 -C option of pcretest uses this to show which version is compiled.
648
649 A new data escape \S, is added to pcretest; it causes the amounts of store
650 obtained and freed by both kinds of malloc/free at match time to be added
651 to the output.
652
653 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
654 what's available on my current Linux desktop machine.
655
656 3. When matching a UTF-8 string, the test for a valid string at the start has
657 been extended. If start_offset is not zero, PCRE now checks that it points
658 to a byte that is the start of a UTF-8 character. If not, it returns
659 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
660 this is necessary because there may be backward assertions in the pattern.
661 When matching the same subject several times, it may save resources to use
662 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
663
664 4. The code for checking the validity of UTF-8 strings has been tightened so
665 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
666 containing "overlong sequences".
667
668 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
669 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
670 should have been "&", but it just so happened that all the cases this let
671 through by mistake were picked up later in the function.
672
673 6. I had used a variable called "isblank" - this is a C99 function, causing
674 some compilers to warn. To avoid this, I renamed it (as "blankclass").
675
676 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
677 prompting; (b) run "./pcretest /dev/null" at the start of the test script
678 so the version is shown; (c) stop "make test" echoing "./RunTest".
679
680 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
681
682 9. The prototype for memmove() for systems that don't have it was using
683 size_t, but the inclusion of the header that defines size_t was later. I've
684 moved the #includes for the C headers earlier to avoid this.
685
686 10. Added some adjustments to the code to make it easier to compiler on certain
687 special systems:
688
689 (a) Some "const" qualifiers were missing.
690 (b) Added the macro EXPORT before all exported functions; by default this
691 is defined to be empty.
692 (c) Changed the dftables auxiliary program (that builds chartables.c) so
693 that it reads its output file name as an argument instead of writing
694 to the standard output and assuming this can be redirected.
695
696 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
697 class containing characters with values greater than 255, PCRE compilation
698 went into a loop.
699
700 12. A recursive reference to a subpattern that was within another subpattern
701 that had a minimum quantifier of zero caused PCRE to crash. For example,
702 (x(y(?2))z)? provoked this bug with a subject that got as far as the
703 recursion. If the recursively-called subpattern itself had a zero repeat,
704 that was OK.
705
706 13. In pcretest, the buffer for reading a data line was set at 30K, but the
707 buffer into which it was copied (for escape processing) was still set at
708 1024, so long lines caused crashes.
709
710 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
711 "internal error: code overflow...". This applied to any character class
712 that was followed by a possessive quantifier.
713
714 15. Modified the Makefile to add libpcre.la as a prerequisite for
715 libpcreposix.la because I was told this is needed for a parallel build to
716 work.
717
718 16. If a pattern that contained .* following optional items at the start was
719 studied, the wrong optimizing data was generated, leading to matching
720 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
721 matching string must start with a or b or c. The correct conclusion for
722 this pattern is that a match can start with any character.
723
724
725 Version 4.4 13-Aug-03
726 ---------------------
727
728 1. In UTF-8 mode, a character class containing characters with values between
729 127 and 255 was not handled correctly if the compiled pattern was studied.
730 In fixing this, I have also improved the studying algorithm for such
731 classes (slightly).
732
733 2. Three internal functions had redundant arguments passed to them. Removal
734 might give a very teeny performance improvement.
735
736 3. Documentation bug: the value of the capture_top field in a callout is *one
737 more than* the number of the hightest numbered captured substring.
738
739 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
740 in incorrectly linking with a previously installed version. They now link
741 explicitly with libpcre.la.
742
743 5. configure.in no longer needs to recognize Cygwin specially.
744
745 6. A problem in pcre.in for Windows platforms is fixed.
746
747 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
748 pcretest, it used to include the size of the study block as part of its
749 output. Unfortunately, the structure contains a field that has a different
750 size on different hardware architectures. This meant that the tests that
751 showed this size failed. As the block is currently always of a fixed size,
752 this information isn't actually particularly useful in pcretest output, so
753 I have just removed it.
754
755 8. Three pre-processor statements accidentally did not start in column 1.
756 Sadly, there are *still* compilers around that complain, even though
757 standard C has not required this for well over a decade. Sigh.
758
759 9. In pcretest, the code for checking callouts passed small integers in the
760 callout_data field, which is a void * field. However, some picky compilers
761 complained about the casts involved for this on 64-bit systems. Now
762 pcretest passes the address of the small integer instead, which should get
763 rid of the warnings.
764
765 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
766 both compile and run time, and gives an error if an invalid UTF-8 sequence
767 is found. There is a option for disabling this check in cases where the
768 string is known to be correct and/or the maximum performance is wanted.
769
770 11. In response to a bug report, I changed one line in Makefile.in from
771
772 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
773 to
774 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
775
776 to look similar to other lines, but I have no way of telling whether this
777 is the right thing to do, as I do not use Windows. No doubt I'll get told
778 if it's wrong...
779
780
781 Version 4.3 21-May-03
782 ---------------------
783
784 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
785 Makefile.
786
787 2. Some refactoring to improve the quality of the code:
788
789 (i) The utf8_table... variables are now declared "const".
790
791 (ii) The code for \cx, which used the "case flipping" table to upper case
792 lower case letters, now just substracts 32. This is ASCII-specific,
793 but the whole concept of \cx is ASCII-specific, so it seems
794 reasonable.
795
796 (iii) PCRE was using its character types table to recognize decimal and
797 hexadecimal digits in the pattern. This is silly, because it handles
798 only 0-9, a-f, and A-F, but the character types table is locale-
799 specific, which means strange things might happen. A private
800 table is now used for this - though it costs 256 bytes, a table is
801 much faster than multiple explicit tests. Of course, the standard
802 character types table is still used for matching digits in subject
803 strings against \d.
804
805 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
806 ending in _t are). So I've renamed it as ESC_tee.
807
808 3. The first argument for regexec() in the POSIX wrapper should have been
809 defined as "const".
810
811 4. Changed pcretest to use malloc() for its buffers so that they can be
812 Electric Fenced for debugging.
813
814 5. There were several places in the code where, in UTF-8 mode, PCRE would try
815 to read one or more bytes before the start of the subject string. Often this
816 had no effect on PCRE's behaviour, but in some circumstances it could
817 provoke a segmentation fault.
818
819 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
820 to try to read one or more bytes before the start of the subject string.
821
822 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
823 UTF-8 support could misbehave in various ways if the subject string
824 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
825 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
826 back over UTF-8 characters.)
827
828
829 Version 4.2 14-Apr-03
830 ---------------------
831
832 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
833
834 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
835 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
836 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
837 [WIN_PREFIX]: new variable, "cyg" for Cygwin
838 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
839 and BUILD_EXEEXT
840 Note: automatic setting of the BUILD variables is not yet working
841 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
842 compile-time but not at link-time
843 [LINK]: use for linking executables only
844 make different versions for Windows and non-Windows
845 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
846 libraries
847 [LINK_FOR_BUILD]: new variable
848 [OBJEXT]: use throughout
849 [EXEEXT]: use throughout
850 <winshared>: new target
851 <wininstall>: new target
852 <dftables.o>: use native compiler
853 <dftables>: use native linker
854 <install>: handle Windows platform correctly
855 <clean>: ditto
856 <check>: ditto
857 copy DLL to top builddir before testing
858
859 As part of these changes, -no-undefined was removed again. This was reported
860 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
861 in any case.
862
863 3. Some tidies to get rid of compiler warnings:
864
865 . In the match_data structure, match_limit was an unsigned long int, whereas
866 match_call_count was an int. I've made them both unsigned long ints.
867
868 . In pcretest the fact that a const uschar * doesn't automatically cast to
869 a void * provoked a warning.
870
871 . Turning on some more compiler warnings threw up some "shadow" variables
872 and a few more missing casts.
873
874 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
875 option, a class that contained a single character with a value between 128
876 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
877
878 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
879 option, a class that contained several characters, but with at least one
880 whose value was between 128 and 255 caused PCRE to crash.
881
882
883 Version 4.1 12-Mar-03
884 ---------------------
885
886 1. Compiling with gcc -pedantic found a couple of places where casts were
887 needed, and a string in dftables.c that was longer than standard compilers are
888 required to support.
889
890 2. Compiling with Sun's compiler found a few more places where the code could
891 be tidied up in order to avoid warnings.
892
893 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
894 first of these names is deprecated in the latest Autoconf in favour of the name
895 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
896 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
897 analogy I have changed it to CFLAGS_FOR_BUILD.
898
899 4. Added -no-undefined to the linking command in the Makefile, because this is
900 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
901 linking step for the pcreposix library.
902
903 5. PCRE was failing to diagnose the case of two named groups with the same
904 name.
905
906 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
907 literal character that is needed in the subject for a match, and scans along to
908 ensure that it is present before embarking on the full matching process. This
909 saves time in cases of nested unlimited repeats that are never going to match.
910 Problem: the scan can take a lot of time if the subject is very long (e.g.
911 megabytes), thus penalizing straightforward matches. It is now done only if the
912 amount of subject to be scanned is less than 1000 bytes.
913
914 7. A lesser problem with the same optimization is that it was recording the
915 first character of an anchored pattern as "needed", thus provoking a search
916 right along the subject, even when the first match of the pattern was going to
917 fail. The "needed" character is now not set for anchored patterns, unless it
918 follows something in the pattern that is of non-fixed length. Thus, it still
919 fulfils its original purpose of finding quick non-matches in cases of nested
920 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
921
922
923 Version 4.0 17-Feb-03
924 ---------------------
925
926 1. If a comment in an extended regex that started immediately after a meta-item
927 extended to the end of string, PCRE compiled incorrect data. This could lead to
928 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
929
930 2. Moved to autoconf 2.53 and libtool 1.4.2.
931
932 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
933 the special perltest8 script is no longer needed - all the tests can be run
934 from a single perltest script.
935
936 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
937 by \s. It has now been removed in PCRE. This means it isn't recognized as
938 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
939 class [:space:] *does* include VT, thereby creating a mess.
940
941 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
942 space and tab.
943
944 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
945 its new features into the main test script, reducing the number of scripts.
946
947 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
948 were backward compatible, and made the (?i) apply to the whole pattern, as if
949 /i were given. Now it behaves more logically, and applies the option setting
950 only to what follows. PCRE has been changed to follow suit. However, if it
951 finds options settings right at the start of the pattern, it extracts them into
952 the global options, as before. Thus, they show up in the info data.
953
954 8. Added support for the \Q...\E escape sequence. Characters in between are
955 treated as literals. This is slightly different from Perl in that $ and @ are
956 also handled as literals inside the quotes. In Perl, they will cause variable
957 interpolation. Note the following examples:
958
959 Pattern PCRE matches Perl matches
960
961 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
962 \Qabc\$xyz\E abc\$xyz abc\$xyz
963 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
964
965 For compatibility with Perl, \Q...\E sequences are recognized inside character
966 classes as well as outside them.
967
968 9. Re-organized 3 code statements in pcretest to avoid "overflow in
969 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
970 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
971 signed/unsigned warnings.
972
973 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
974 option for pcretest, so I've replaced it by a simple function that does just
975 that job.
976
977 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
978 "pcregrep -".
979
980 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
981 Java package. This provides some syntactic sugar for simple cases of what my
982 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
983 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
984 item, you can use this simplified notation. Note that only makes sense with
985 greedy quantifiers. Consequently, the use of the possessive quantifier forces
986 greediness, whatever the setting of the PCRE_UNGREEDY option.
987
988 13. A change of greediness default within a pattern was not taking effect at
989 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
990 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
991 was abstracted outside.
992
993 14. PCRE now supports the \G assertion. It is true when the current matching
994 position is at the start point of the match. This differs from \A when the
995 starting offset is non-zero. Used with the /g option of pcretest (or similar
996 code), it works in the same way as it does for Perl's /g option. If all
997 alternatives of a regex begin with \G, the expression is anchored to the start
998 match position, and the "anchored" flag is set in the compiled expression.
999
1000 15. Some bugs concerning the handling of certain option changes within patterns
1001 have been fixed. These applied to options other than (?ims). For example,
1002 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1003 been the other way round. Some of this was related to change 7 above.
1004
1005 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1006 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1007 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1008 POSIX classes only within a class (e.g. /[[:alpha:]]/).
1009
1010 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1011 mode. Unlike ".", it always matches newline, whatever the setting of
1012 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1013 assertions. Perl allows it, but it doesn't (in general) work because it can't
1014 calculate the length of the lookbehind. At least, that's the case for Perl
1015 5.8.0 - I've been told they are going to document that it doesn't work in
1016 future.
1017
1018 18. Added an error diagnosis for escapes that PCRE does not support: these are
1019 \L, \l, \N, \P, \p, \U, \u, and \X.
1020
1021 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1022 reading past the end of the pattern in cases such as /[abcd/.
1023
1024 20. PCRE was getting more memory than necessary for patterns with classes that
1025 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1026
1027 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1028 compiling PCRE for use with Virtual Pascal.
1029
1030 22. Small fix to the Makefile to make it work properly if the build is done
1031 outside the source tree.
1032
1033 23. Added a new extension: a condition to go with recursion. If a conditional
1034 subpattern starts with (?(R) the "true" branch is used if recursion has
1035 happened, whereas the "false" branch is used only at the top level.
1036
1037 24. When there was a very long string of literal characters (over 255 bytes
1038 without UTF support, over 250 bytes with UTF support), the computation of how
1039 much memory was required could be incorrect, leading to segfaults or other
1040 strange effects.
1041
1042 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1043 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1044 there was a subsequent back reference to those brackets. This meant that, for
1045 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1046 possible to check for precisely this case. All we can do is abandon the
1047 optimization if .* occurs inside capturing brackets when there are any back
1048 references whatsoever. (See below for a better fix that came later.)
1049
1050 26. The handling of the optimization for finding the first character of a
1051 non-anchored pattern, and for finding a character that is required later in the
1052 match were failing in some cases. This didn't break the matching; it just
1053 failed to optimize when it could. The way this is done has been re-implemented.
1054
1055 27. Fixed typo in error message for invalid (?R item (it said "(?p").
1056
1057 28. Added a new feature that provides some of the functionality that Perl
1058 provides with (?{...}). The facility is termed a "callout". The way it is done
1059 in PCRE is for the caller to provide an optional function, by setting
1060 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1061 global variable. By default it is unset, which disables all calling out. To get
1062 the function called, the regex must include (?C) at appropriate points. This
1063 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1064 This provides a means of identifying different callout points. When PCRE
1065 reaches such a point in the regex, if pcre_callout has been set, the external
1066 function is called. It is provided with data in a structure called
1067 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1068 matching continues; if it returns a non-zero value, the match at the current
1069 point fails. However, backtracking will occur if possible. [This was changed
1070 later and other features added - see item 49 below.]
1071
1072 29. pcretest is upgraded to test the callout functionality. It provides a
1073 callout function that displays information. By default, it shows the start of
1074 the match and the current position in the text. There are some new data escapes
1075 to vary what happens:
1076
1077 \C+ in addition, show current contents of captured substrings
1078 \C- do not supply a callout function
1079 \C!n return 1 when callout number n is reached
1080 \C!n!m return 1 when callout number n is reached for the mth time
1081
1082 30. If pcregrep was called with the -l option and just a single file name, it
1083 output "<stdin>" if a match was found, instead of the file name.
1084
1085 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1086 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1087 pcre_exec(). This saves a malloc/free per call. The default value of
1088 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1089 when configuring.
1090
1091 32. The default maximum size of a compiled pattern is 64K. There have been a
1092 few cases of people hitting this limit. The code now uses macros to handle the
1093 storing of links as offsets within the compiled pattern. It defaults to 2-byte
1094 links, but this can be changed to 3 or 4 bytes by --with-link-size when
1095 configuring. Tests 2 and 5 work only with 2-byte links because they output
1096 debugging information about compiled patterns.
1097
1098 33. Internal code re-arrangements:
1099
1100 (a) Moved the debugging function for printing out a compiled regex into
1101 its own source file (printint.c) and used #include to pull it into
1102 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1103 separate copies.
1104
1105 (b) Defined the list of op-code names for debugging as a macro in
1106 internal.h so that it is next to the definition of the opcodes.
1107
1108 (c) Defined a table of op-code lengths for simpler skipping along compiled
1109 code. This is again a macro in internal.h so that it is next to the
1110 definition of the opcodes.
1111
1112 34. Added support for recursive calls to individual subpatterns, along the
1113 lines of Robin Houston's patch (but implemented somewhat differently).
1114
1115 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1116 allow it to read and process whole directories in Win32. This code was
1117 contributed by Lionel Fourquaux; it has not been tested by me.
1118
1119 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1120 used to name a group. Names consist of alphanumerics and underscores, and must
1121 be unique. Back references use the syntax (?P=name) and recursive calls use
1122 (?P>name) which is a PCRE extension to the Python extension. Groups still have
1123 numbers. The function pcre_fullinfo() can be used after compilation to extract
1124 a name/number map. There are three relevant calls:
1125
1126 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1127 PCRE_INFO_NAMECOUNT yields the number of entries
1128 PCRE_INFO_NAMETABLE yields a pointer to the map.
1129
1130 The map is a vector of fixed-size entries. The size of each entry depends on
1131 the length of the longest name used. The first two bytes of each entry are the
1132 group number, most significant byte first. There follows the corresponding
1133 name, zero terminated. The names are in alphabetical order.
1134
1135 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1136 case instead of 255. Making it the same both with and without UTF-8 support
1137 means that the same test output works with both.
1138
1139 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1140 calling malloc() with a zero argument.
1141
1142 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1143 optimization. I've improved things by keeping a bitmap of backreferences with
1144 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1145 fact referenced, the optimization can be applied. It is unlikely that a
1146 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1147 the match to follow \n) will appear inside brackets with a number greater than
1148 31, but if it does, any back reference > 31 suppresses the optimization.
1149
1150 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1151 of disabling numbered capturing parentheses. Any opening parenthesis that is
1152 not followed by ? behaves as if it were followed by ?: but named parentheses
1153 can still be used for capturing (and they will acquire numbers in the usual
1154 way).
1155
1156 41. Redesigned the return codes from the match() function into yes/no/error so
1157 that errors can be passed back from deep inside the nested calls. A malloc
1158 failure while inside a recursive subpattern call now causes the
1159 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1160
1161 42. It is now possible to set a limit on the number of times the match()
1162 function is called in a call to pcre_exec(). This facility makes it possible to
1163 limit the amount of recursion and backtracking, though not in a directly
1164 obvious way, because the match() function is used in a number of different
1165 circumstances. The count starts from zero for each position in the subject
1166 string (for non-anchored patterns). The default limit is, for compatibility, a
1167 large number, namely 10 000 000. You can change this in two ways:
1168
1169 (a) When configuring PCRE before making, you can use --with-match-limit=n
1170 to set a default value for the compiled library.
1171
1172 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1173 a different value is set. See 45 below.
1174
1175 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1176
1177 43. Added a new function pcre_config(int, void *) to enable run-time extraction
1178 of things that can be changed at compile time. The first argument specifies
1179 what is wanted and the second points to where the information is to be placed.
1180 The current list of available information is:
1181
1182 PCRE_CONFIG_UTF8
1183
1184 The output is an integer that is set to one if UTF-8 support is available;
1185 otherwise it is set to zero.
1186
1187 PCRE_CONFIG_NEWLINE
1188
1189 The output is an integer that it set to the value of the code that is used for
1190 newline. It is either LF (10) or CR (13).
1191
1192 PCRE_CONFIG_LINK_SIZE
1193
1194 The output is an integer that contains the number of bytes used for internal
1195 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1196
1197 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1198
1199 The output is an integer that contains the threshold above which the POSIX
1200 interface uses malloc() for output vectors. See item 31 above.
1201
1202 PCRE_CONFIG_MATCH_LIMIT
1203
1204 The output is an unsigned integer that contains the default limit of the number
1205 of match() calls in a pcre_exec() execution. See 42 above.
1206
1207 44. pcretest has been upgraded by the addition of the -C option. This causes it
1208 to extract all the available output from the new pcre_config() function, and to
1209 output it. The program then exits immediately.
1210
1211 45. A need has arisen to pass over additional data with calls to pcre_exec() in
1212 order to support additional features. One way would have been to define
1213 pcre_exec2() (for example) with extra arguments, but this would not have been
1214 extensible, and would also have required all calls to the original function to
1215 be mapped to the new one. Instead, I have chosen to extend the mechanism that
1216 is used for passing in "extra" data from pcre_study().
1217
1218 The pcre_extra structure is now exposed and defined in pcre.h. It currently
1219 contains the following fields:
1220
1221 flags a bitmap indicating which of the following fields are set
1222 study_data opaque data from pcre_study()
1223 match_limit a way of specifying a limit on match() calls for a specific
1224 call to pcre_exec()
1225 callout_data data for callouts (see 49 below)
1226
1227 The flag bits are also defined in pcre.h, and are
1228
1229 PCRE_EXTRA_STUDY_DATA
1230 PCRE_EXTRA_MATCH_LIMIT
1231 PCRE_EXTRA_CALLOUT_DATA
1232
1233 The pcre_study() function now returns one of these new pcre_extra blocks, with
1234 the actual study data pointed to by the study_data field, and the
1235 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1236 before. That is, this change is entirely upwards-compatible and requires no
1237 change to existing code.
1238
1239 If you want to pass in additional data to pcre_exec(), you can either place it
1240 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1241 block.
1242
1243 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1244 data string contains the escape sequence \M, pcretest calls pcre_exec() several
1245 times with different match limits, until it finds the minimum value needed for
1246 pcre_exec() to complete. The value is then output. This can be instructive; for
1247 most simple matches the number is quite small, but for pathological cases it
1248 gets very large very quickly.
1249
1250 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1251 returns the size of the data block pointed to by the study_data field in a
1252 pcre_extra block, that is, the value that was passed as the argument to
1253 pcre_malloc() when PCRE was getting memory in which to place the information
1254 created by pcre_study(). The fourth argument should point to a size_t variable.
1255 pcretest has been extended so that this information is shown after a successful
1256 pcre_study() call when information about the compiled regex is being displayed.
1257
1258 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1259 because what follows is always an absolute path. (Later: it turns out that this
1260 is more than cosmetic for MinGW, because it doesn't like empty path
1261 components.)
1262
1263 49. Some changes have been made to the callout feature (see 28 above):
1264
1265 (i) A callout function now has three choices for what it returns:
1266
1267 0 => success, carry on matching
1268 > 0 => failure at this point, but backtrack if possible
1269 < 0 => serious error, return this value from pcre_exec()
1270
1271 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1272 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1273 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1274 use by callout functions. It will never be used by PCRE itself.
1275
1276 (ii) The pcre_extra structure (see 45 above) has a void * field called
1277 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1278 pcre_callout_block structure has a field of the same name. The contents of
1279 the field passed in the pcre_extra structure are passed to the callout
1280 function in the corresponding field in the callout block. This makes it
1281 easier to use the same callout-containing regex from multiple threads. For
1282 testing, the pcretest program has a new data escape
1283
1284 \C*n pass the number n (may be negative) as callout_data
1285
1286 If the callout function in pcretest receives a non-zero value as
1287 callout_data, it returns that value.
1288
1289 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1290 there were some redundant $(CFLAGS) in commands that are now specified as
1291 $(LINK), which already includes $(CFLAGS).
1292
1293 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1294 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1295 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1296 one-byte characters throughout. Note that case-insensitive matching applies
1297 only to characters whose values are less than 256. PCRE doesn't support the
1298 notion of cases for higher-valued characters.
1299
1300 (i) A character class whose characters are all within 0-255 is handled as
1301 a bit map, and the map is inverted for negative classes. Previously, a
1302 character > 255 always failed to match such a class; however it should
1303 match if the class was a negative one (e.g. [^ab]). This has been fixed.
1304
1305 (ii) A negated character class with a single character < 255 is coded as
1306 "not this character" (OP_NOT). This wasn't working properly when the test
1307 character was multibyte, either singly or repeated.
1308
1309 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1310 mode, for example: \x{100}{2,3}.
1311
1312 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1313 singly or repeated) now correctly test multibyte characters. However,
1314 PCRE doesn't recognize any characters with values greater than 255 as
1315 digits, spaces, or word characters. Such characters always match \D, \S,
1316 and \W, and never match \d, \s, or \w.
1317
1318 (v) Classes may now contain characters and character ranges with values
1319 greater than 255. For example: [ab\x{100}-\x{400}].
1320
1321 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1322 PCRE in UTF-8 mode.
1323
1324 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1325 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1326 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1327 value.)
1328
1329 53. The single man page has become too large. I have therefore split it up into
1330 a number of separate man pages. These also give rise to individual HTML pages;
1331 these are now put in a separate directory, and there is an index.html page that
1332 lists them all. Some hyperlinking between the pages has been installed.
1333
1334 54. Added convenience functions for handling named capturing parentheses.
1335
1336 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1337 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1338 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1339 are faulted.
1340
1341 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1342 calling configure. These values are used when compiling the dftables.c program
1343 which is run to generate the source of the default character tables. They
1344 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1345 you will need to set these values.
1346
1347 57. Updated the building process for Windows DLL, as provided by Fred Cox.
1348
1349
1350 Version 3.9 02-Jan-02
1351 ---------------------
1352
1353 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1354
1355 2. If --disable-static was given, the building process failed when trying to
1356 build pcretest and pcregrep. (For some reason it was using libtool to compile
1357 them, which is not right, as they aren't part of the library.)
1358
1359
1360 Version 3.8 18-Dec-01
1361 ---------------------
1362
1363 1. The experimental UTF-8 code was completely screwed up. It was packing the
1364 bytes in the wrong order. How dumb can you get?
1365
1366
1367 Version 3.7 29-Oct-01
1368 ---------------------
1369
1370 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1371 This caused pcretest, when used on the test data, to segfault. Unfortunately,
1372 this didn't happen under Solaris 8, where I normally test things.
1373
1374 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1375 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1376 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1377 this fix an hour or so after the initial 3.7 release.)
1378
1379
1380 Version 3.6 23-Oct-01
1381 ---------------------
1382
1383 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1384 offsets passed as NULL with zero offset count.
1385
1386 2. The config.guess and config.sub files had not been updated when I moved to
1387 the latest autoconf.
1388
1389
1390 Version 3.5 15-Aug-01
1391 ---------------------
1392
1393 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1394 had been forgotten.
1395
1396 2. By using declared but undefined structures, we can avoid using "void"
1397 definitions in pcre.h while keeping the internal definitions of the structures
1398 private.
1399
1400 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1401 user point of view, this means that both static and shared libraries are built
1402 by default, but this can be individually controlled. More of the work of
1403 handling this static/shared cases is now inside libtool instead of PCRE's make
1404 file.
1405
1406 4. The pcretest utility is now installed along with pcregrep because it is
1407 useful for users (to test regexs) and by doing this, it automatically gets
1408 relinked by libtool. The documentation has been turned into a man page, so
1409 there are now .1, .txt, and .html versions in /doc.
1410
1411 5. Upgrades to pcregrep:
1412 (i) Added long-form option names like gnu grep.
1413 (ii) Added --help to list all options with an explanatory phrase.
1414 (iii) Added -r, --recursive to recurse into sub-directories.
1415 (iv) Added -f, --file to read patterns from a file.
1416
1417 6. pcre_exec() was referring to its "code" argument before testing that
1418 argument for NULL (and giving an error if it was NULL).
1419
1420 7. Upgraded Makefile.in to allow for compiling in a different directory from
1421 the source directory.
1422
1423 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1424 options bits, the pointer it was passed was to an int instead of to an unsigned
1425 long int. This mattered only on 64-bit systems.
1426
1427 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1428 generated) instead of pcre.in, which it its source. Also made the same change
1429 in several of the .c files.
1430
1431 10. A new release of gcc defines printf() as a macro, which broke pcretest
1432 because it had an ifdef in the middle of a string argument for printf(). Fixed
1433 by using separate calls to printf().
1434
1435 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1436 script, to force use of CR or LF instead of \n in the source. On non-Unix
1437 systems, the value can be set in config.h.
1438
1439 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1440 absolute limit. Changed the text of the error message to make this clear, and
1441 likewise updated the man page.
1442
1443 13. The limit of 99 on the number of capturing subpatterns has been removed.
1444 The new limit is 65535, which I hope will not be a "real" limit.
1445
1446
1447 Version 3.4 22-Aug-00
1448 ---------------------
1449
1450 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1451
1452 2. Diagnose condition (?(0) as an error instead of crashing on matching.
1453
1454
1455 Version 3.3 01-Aug-00
1456 ---------------------
1457
1458 1. If an octal character was given, but the value was greater than \377, it
1459 was not getting masked to the least significant bits, as documented. This could
1460 lead to crashes in some systems.
1461
1462 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1463 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1464
1465 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1466 These just pass their arguments on to (pcre_free)(), but they are provided
1467 because some uses of PCRE bind it to non-C systems that can call its functions,
1468 but cannot call free() or pcre_free() directly.
1469
1470 4. Add "make test" as a synonym for "make check". Corrected some comments in
1471 the Makefile.
1472
1473 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1474 Makefile.
1475
1476 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1477 command called pgrep for grepping around the active processes.
1478
1479 7. Added the beginnings of support for UTF-8 character strings.
1480
1481 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1482 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1483 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1484 out for the ar command.)
1485
1486
1487 Version 3.2 12-May-00
1488 ---------------------
1489
1490 This is purely a bug fixing release.
1491
1492 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1493 of ZA. This was just one example of several cases that could provoke this bug,
1494 which was introduced by change 9 of version 2.00. The code for breaking
1495 infinite loops after an iteration that matches an empty string was't working
1496 correctly.
1497
1498 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1499 when matched against abbab (for example). After matching an empty string, it
1500 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1501 caused it to match further down the string than it should.
1502
1503 3. The code contained an inclusion of sys/types.h. It isn't clear why this
1504 was there because it doesn't seem to be needed, and it causes trouble on some
1505 systems, as it is not a Standard C header. It has been removed.
1506
1507 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1508 were reported on the Macintosh. The changes were from
1509
1510 while ((c = *(++ptr)) != 0 && c != '\n');
1511 to
1512 while ((c = *(++ptr)) != 0 && c != '\n') ;
1513
1514 Totally extraordinary, but if that's what it takes...
1515
1516 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1517 available. Added HAVE_BCOPY and an autoconf test for it; if neither
1518 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1519 assumes the way PCRE uses memmove() (always moving upwards).
1520
1521 6. PCRE is being used in one environment where strchr() is not available. There
1522 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1523 faster code anyway.
1524
1525
1526 Version 3.1 09-Feb-00
1527 ---------------------
1528
1529 The only change in this release is the fixing of some bugs in Makefile.in for
1530 the "install" target:
1531
1532 (1) It was failing to install pcreposix.h.
1533
1534 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
1535
1536
1537 Version 3.0 01-Feb-00
1538 ---------------------
1539
1540 1. Add support for the /+ modifier to perltest (to output $` like it does in
1541 pcretest).
1542
1543 2. Add support for the /g modifier to perltest.
1544
1545 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
1546 matches null strings.
1547
1548 4. Fix perltest so that it doesn't do unwanted things when fed an empty
1549 pattern. Perl treats empty patterns specially - it reuses the most recent
1550 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
1551 effect.
1552
1553 5. The POSIX interface was broken in that it was just handing over the POSIX
1554 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
1555 required a bigger vector, with some working space on the end. This means that
1556 the POSIX wrapper now has to get and free some memory, and copy the results.
1557
1558 6. Added some simple autoconf support, placing the test data and the
1559 documentation in separate directories, re-organizing some of the
1560 information files, and making it build pcre-config (a GNU standard). Also added
1561 libtool support for building PCRE as a shared library, which is now the
1562 default.
1563
1564 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
1565 09 are not valid octal constants. Single digits will be used for minor values
1566 less than 10.
1567
1568 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
1569 existing programs that set these in the POSIX interface can use PCRE without
1570 modification.
1571
1572 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
1573 return all that pcre_info() returns, plus additional data. The pcre_info()
1574 function is retained for compatibility, but is considered to be obsolete.
1575
1576 10. Added experimental recursion feature (?R) to handle one common case that
1577 Perl 5.6 will be able to do with (?p{...}).
1578
1579 11. Added support for POSIX character classes like [:alpha:], which Perl is
1580 adopting.
1581
1582
1583 Version 2.08 31-Aug-99
1584 ----------------------
1585
1586 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
1587 trying to match at the startoffset position, but instead was moving forward to
1588 the next newline as if a previous match had failed.
1589
1590 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
1591 and could get into a loop if a null string was matched other than at the start
1592 of the subject.
1593
1594 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
1595 be distinguished at compile time, and for completeness also added PCRE_DATE.
1596
1597 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
1598 in GnuWin32 environments.
1599
1600
1601 Version 2.07 29-Jul-99
1602 ----------------------
1603
1604 1. The documentation is now supplied in plain text form and HTML as well as in
1605 the form of man page sources.
1606
1607 2. C++ compilers don't like assigning (void *) values to other pointer types.
1608 In particular this affects malloc(). Although there is no problem in Standard
1609 C, I've put in casts to keep C++ compilers happy.
1610
1611 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
1612 should be (const char *).
1613
1614 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
1615 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
1616 However, I haven't made this a standard facility. The documentation doesn't
1617 mention it, and the Makefile doesn't support it.
1618
1619 5. The Makefile now contains an "install" target, with editable destinations at
1620 the top of the file. The pcretest program is not installed.
1621
1622 6. pgrep -V now gives the PCRE version number and date.
1623
1624 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1625 causing the entire string to be ignored, instead of just the last character.
1626
1627 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1628 non-matching string, it can take a very, very long time, even for strings of
1629 quite modest length, because of the nested recursion. PCRE now does better in
1630 some of these cases. It does this by remembering the last required literal
1631 character in the pattern, and pre-searching the subject to ensure it is present
1632 before running the real match. In other words, it applies a heuristic to detect
1633 some types of certain failure quickly, and in the above example, if presented
1634 with a string that has no trailing " it gives "no match" very quickly.
1635
1636 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1637 other alternatives are tried instead.
1638
1639
1640 Version 2.06 09-Jun-99
1641 ----------------------
1642
1643 1. Change pcretest's output for amount of store used to show just the code
1644 space, because the remainder (the data block) varies in size between 32-bit and
1645 64-bit systems.
1646
1647 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1648 start matching at. This allows lookbehinds to work when searching for multiple
1649 occurrences in a string.
1650
1651 3. Added additional options to pcretest for testing multiple occurrences:
1652
1653 /+ outputs the rest of the string that follows a match
1654 /g loops for multiple occurrences, using the new startoffset argument
1655 /G loops for multiple occurrences by passing an incremented pointer
1656
1657 4. PCRE wasn't doing the "first character" optimization for patterns starting
1658 with \b or \B, though it was doing it for other lookbehind assertions. That is,
1659 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1660 the letter 'x'. On long subject strings, this gives a significant speed-up.
1661
1662
1663 Version 2.05 21-Apr-99
1664 ----------------------
1665
1666 1. Changed the type of magic_number from int to long int so that it works
1667 properly on 16-bit systems.
1668
1669 2. Fixed a bug which caused patterns starting with .* not to work correctly
1670 when the subject string contained newline characters. PCRE was assuming
1671 anchoring for such patterns in all cases, which is not correct because .* will
1672 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1673 DOTALL is set at top level; otherwise it knows that patterns starting with .*
1674 must be retried after every newline in the subject.
1675
1676
1677 Version 2.04 18-Feb-99
1678 ----------------------
1679
1680 1. For parenthesized subpatterns with repeats whose minimum was zero, the
1681 computation of the store needed to hold the pattern was incorrect (too large).
1682 If such patterns were nested a few deep, this could multiply and become a real
1683 problem.
1684
1685 2. Added /M option to pcretest to show the memory requirement of a specific
1686 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1687
1688 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1689 compiled in such a way that the backtracking after subsequent failure was
1690 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1691 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1692
1693
1694 Version 2.03 02-Feb-99
1695 ----------------------
1696
1697 1. Fixed typo and small mistake in man page.
1698
1699 2. Added 4th condition (GPL supersedes if conflict) and created separate
1700 LICENCE file containing the conditions.
1701
1702 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1703 Perl, that is the internal \ allows the delimiter to be included in the
1704 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1705 the final delimiter, add \ to the end of the pattern (to test the error).
1706
1707 4. Added the convenience functions for extracting substrings after a successful
1708 match. Updated pcretest to make it able to test these functions.
1709
1710
1711 Version 2.02 14-Jan-99
1712 ----------------------
1713
1714 1. Initialized the working variables associated with each extraction so that
1715 their saving and restoring doesn't refer to uninitialized store.
1716
1717 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1718 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1719 fix the problem.
1720
1721 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1722 calls, and wasn't printing the correct value for compiling calls. Increased the
1723 default value of LOOPREPEAT, and the number of significant figures in the
1724 times.
1725
1726 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1727
1728 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1729 a building problem on Windows NT with a FAT file system.
1730
1731
1732 Version 2.01 21-Oct-98
1733 ----------------------
1734
1735 1. Changed the API for pcre_compile() to allow for the provision of a pointer
1736 to character tables built by pcre_maketables() in the current locale. If NULL
1737 is passed, the default tables are used.
1738
1739
1740 Version 2.00 24-Sep-98
1741 ----------------------
1742
1743 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1744 it any more.
1745
1746 2. Allow quantification of (?>) groups, and make it work correctly.
1747
1748 3. The first character computation wasn't working for (?>) groups.
1749
1750 4. Correct the implementation of \Z (it is permitted to match on the \n at the
1751 end of the subject) and add 5.005's \z, which really does match only at the
1752 very end of the subject.
1753
1754 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1755
1756 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1757 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1758 localized options. All options to pcre_study() were also removed.
1759
1760 7. Add other new features from 5.005:
1761
1762 $(?<= positive lookbehind
1763 $(?<! negative lookbehind
1764 (?imsx-imsx) added the unsetting capability
1765 such a setting is global if at outer level; local otherwise
1766 (?imsx-imsx:) non-capturing groups with option setting
1767 (?(cond)re|re) conditional pattern matching
1768
1769 A backreference to itself in a repeated group matches the previous
1770 captured string.
1771
1772 8. General tidying up of studying (both automatic and via "study")
1773 consequential on the addition of new assertions.
1774
1775 9. As in 5.005, unlimited repeated groups that could match an empty substring
1776 are no longer faulted at compile time. Instead, the loop is forcibly broken at
1777 runtime if any iteration does actually match an empty substring.
1778
1779 10. Include the RunTest script in the distribution.
1780
1781 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1782 discrepancies, some of which were old and were also with respect to 5.004. They
1783 have now been fixed.
1784
1785
1786 Version 1.09 28-Apr-98
1787 ----------------------
1788
1789 1. A negated single character class followed by a quantifier with a minimum
1790 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
1791 program crashes, or just wrong answers. This did not apply to negated classes
1792 containing more than one character, or to minima other than one.
1793
1794
1795 Version 1.08 27-Mar-98
1796 ----------------------
1797
1798 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
1799
1800 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
1801 latter must appear before anything that relies on it in the pattern.
1802
1803
1804 Version 1.07 16-Feb-98
1805 ----------------------
1806
1807 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
1808 repeat of a potentially empty string).
1809
1810
1811 Version 1.06 23-Jan-98
1812 ----------------------
1813
1814 1. Added Markus Oberhumer's little patches for C++.
1815
1816 2. Literal strings longer than 255 characters were broken.
1817
1818
1819 Version 1.05 23-Dec-97
1820 ----------------------
1821
1822 1. Negated character classes containing more than one character were failing if
1823 PCRE_CASELESS was set at run time.
1824
1825
1826 Version 1.04 19-Dec-97
1827 ----------------------
1828
1829 1. Corrected the man page, where some "const" qualifiers had been omitted.
1830
1831 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
1832 input syntax.
1833
1834 3. Fixed memory leak which occurred when a regex with back references was
1835 matched with an offsets vector that wasn't big enough. The temporary memory
1836 that is used in this case wasn't being freed if the match failed.
1837
1838 4. Tidied pcretest to ensure it frees memory that it gets.
1839
1840 5. Temporary memory was being obtained in the case where the passed offsets
1841 vector was exactly big enough.
1842
1843 6. Corrected definition of offsetof() from change 5 below.
1844
1845 7. I had screwed up change 6 below and broken the rules for the use of
1846 setjmp(). Now fixed.
1847
1848
1849 Version 1.03 18-Dec-97
1850 ----------------------
1851
1852 1. A erroneous regex with a missing opening parenthesis was correctly
1853 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
1854 on some systems.
1855
1856 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
1857 it was reported that one broken compiler failed on the former because "code" is
1858 also an independent variable.
1859
1860 3. The erroneous regex a[]b caused an array overrun reference.
1861
1862 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
1863 fail on data ending with that character. (It was going on too far, and checking
1864 the next character, typically a binary zero.) This was specific to the
1865 optimized code for single-character negative classes.
1866
1867 5. Added a contributed patch from the TIN world which does the following:
1868
1869 + Add an undef for memmove, in case the the system defines a macro for it.
1870
1871 + Add a definition of offsetof(), in case there isn't one. (I don't know
1872 the reason behind this - offsetof() is part of the ANSI standard - but
1873 it does no harm).
1874
1875 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
1876 most of the places where whitespace preceded '#'. I have given up and
1877 allowed the remaining 2 cases to be at the margin.
1878
1879 + Rename some variables in pcre to eliminate shadowing. This seems very
1880 pedantic, but does no harm, of course.
1881
1882 6. Moved the call to setjmp() into its own function, to get rid of warnings
1883 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
1884
1885 7. Constructs such as \d{8,} were compiling into the equivalent of
1886 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
1887 outcome, but in this particular case used more store than had been allocated,
1888 which caused the bug to be discovered because it threw up an internal error.
1889
1890 8. The debugging code in both pcre and pcretest for outputting the compiled
1891 form of a regex was going wrong in the case of back references followed by
1892 curly-bracketed repeats.
1893
1894
1895 Version 1.02 12-Dec-97
1896 ----------------------
1897
1898 1. Typos in pcre.3 and comments in the source fixed.
1899
1900 2. Applied a contributed patch to get rid of places where it used to remove
1901 'const' from variables, and fixed some signed/unsigned and uninitialized
1902 variable warnings.
1903
1904 3. Added the "runtest" target to Makefile.
1905
1906 4. Set default compiler flag to -O2 rather than just -O.
1907
1908
1909 Version 1.01 19-Nov-97
1910 ----------------------
1911
1912 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
1913 like /([ab]*)*/, that is, for classes with more than one character in them.
1914
1915 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
1916 as /((?>a*))*/ (a PCRE_EXTRA facility).
1917
1918
1919 Version 1.00 18-Nov-97
1920 ----------------------
1921
1922 1. Added compile-time macros to support systems such as SunOS4 which don't have
1923 memmove() or strerror() but have other things that can be used instead.
1924
1925 2. Arranged that "make clean" removes the executables.
1926
1927
1928 Version 0.99 27-Oct-97
1929 ----------------------
1930
1931 1. Fixed bug in code for optimizing classes with only one character. It was
1932 initializing a 32-byte map regardless, which could cause it to run off the end
1933 of the memory it had got.
1934
1935 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1936
1937
1938 Version 0.98 22-Oct-97
1939 ----------------------
1940
1941 1. Fixed bug in code for handling temporary memory usage when there are more
1942 back references than supplied space in the ovector. This could cause segfaults.
1943
1944
1945 Version 0.97 21-Oct-97
1946 ----------------------
1947
1948 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1949
1950 2. Optimized negated single characters not to use a bit map.
1951
1952 3. Brought error texts together as macro definitions; clarified some of them;
1953 fixed one that was wrong - it said "range out of order" when it meant "invalid
1954 escape sequence".
1955
1956 4. Changed some char * arguments to const char *.
1957
1958 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1959
1960 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1961 pcretest.
1962
1963
1964 Version 0.96 16-Oct-97
1965 ----------------------
1966
1967 1. Added a simple "pgrep" utility to the distribution.
1968
1969 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1970 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1971 where "ddd" means "one or more decimal digits".
1972
1973 3. Fixed serious bug. If a pattern had a back reference, but the call to
1974 pcre_exec() didn't supply a large enough ovector to record the related
1975 identifying subpattern, the match always failed. PCRE now remembers the number
1976 of the largest back reference, and gets some temporary memory in which to save
1977 the offsets during matching if necessary, in order to ensure that
1978 backreferences always work.
1979
1980 4. Increased the compatibility with Perl in a number of ways:
1981
1982 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1983 to request this handling. The option can be set at compile or exec time.
1984
1985 (b) $ matches before a terminating newline by default; an option
1986 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
1987 mode). The option can be set at compile or exec time.
1988
1989 (c) The handling of \ followed by a digit other than 0 is now supposed to be
1990 the same as Perl's. If the decimal number it represents is less than 10
1991 or there aren't that many previous left capturing parentheses, an octal
1992 escape is read. Inside a character class, it's always an octal escape,
1993 even if it is a single digit.
1994
1995 (d) An escaped but undefined alphabetic character is taken as a literal,
1996 unless PCRE_EXTRA is set. Currently this just reserves the remaining
1997 escapes.
1998
1999 (e) {0} is now permitted. (The previous item is removed from the compiled
2000 pattern).
2001
2002 5. Changed all the names of code files so that the basic parts are no longer
2003 than 10 characters, and abolished the teeny "globals.c" file.
2004
2005 6. Changed the handling of character classes; they are now done with a 32-byte
2006 bit map always.
2007
2008 7. Added the -d and /D options to pcretest to make it possible to look at the
2009 internals of compilation without having to recompile pcre.
2010
2011
2012 Version 0.95 23-Sep-97
2013 ----------------------
2014
2015 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2016 \x20 at the start of a run of normal characters. These were being treated as
2017 real characters, instead of the source characters being re-checked.
2018
2019
2020 Version 0.94 18-Sep-97
2021 ----------------------
2022
2023 1. The functions are now thread-safe, with the caveat that the global variables
2024 containing pointers to malloc() and free() or alternative functions are the
2025 same for all threads.
2026
2027 2. Get pcre_study() to generate a bitmap of initial characters for non-
2028 anchored patterns when this is possible, and use it if passed to pcre_exec().
2029
2030
2031 Version 0.93 15-Sep-97
2032 ----------------------
2033
2034 1. /(b)|(:+)/ was computing an incorrect first character.
2035
2036 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2037 but not actually doing anything yet.
2038
2039 3. Treat "-" characters in classes that cannot be part of ranges as literals,
2040 as Perl does (e.g. [-az] or [az-]).
2041
2042 4. Set the anchored flag if a branch starts with .* or .*? because that tests
2043 all possible positions.
2044
2045 5. Split up into different modules to avoid including unneeded functions in a
2046 compiled binary. However, compile and exec are still in one module. The "study"
2047 function is split off.
2048
2049 6. The character tables are now in a separate module whose source is generated
2050 by an auxiliary program - but can then be edited by hand if required. There are
2051 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2052 toupper() in the code.
2053
2054 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2055 make them global. Abolish the function for setting them, as the caller can now
2056 set them directly.
2057
2058
2059 Version 0.92 11-Sep-97
2060 ----------------------
2061
2062 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2063 (e.g. /a{1,3}/) was broken (I mis-optimized it).
2064
2065 2. Caseless matching was not working in character classes if the characters in
2066 the pattern were in upper case.
2067
2068 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2069
2070 4. Make PCRE_ANCHORED public and accept as a compile option.
2071
2072 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2073 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2074 pass them.
2075
2076 6. Give an error if bad option bits passed at compile or run time.
2077
2078 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2079 pcretest to cause it to pass that flag.
2080
2081 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2082 options, and the first character, if set.
2083
2084 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2085
2086
2087 Version 0.91 10-Sep-97
2088 ----------------------
2089
2090 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2091 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2092
2093 2. PCRE was looping on encountering an indefinitely repeated back reference to
2094 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2095 Perl does - treats the match as successful.
2096
2097 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12