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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 325 - (show annotations) (download)
Sat Mar 8 17:13:02 2008 UTC (6 years, 4 months ago) by ph10
File size: 152127 byte(s)
Add --include_dir and --exclude_dir to pcregrep.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.7 05-Mar-08
5 ---------------------
6
7 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
8 a string to a long long, pretend we don't even have a long long." This is
9 done by checking for the strtoq, strtoll, and _strtoi64 functions.
10
11 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
12 pre-7.6 versions, which defined a global no_arg variable instead of putting
13 it in the RE class.
14
15 3. Remove a line of dead code, identified by coverity and reported by Nuno
16 Lopes.
17
18 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
19
20 (1) The include/exclude patterns were being applied to the whole pathnames
21 of files, instead of just to the final components.
22
23 (2) If there was more than one level of directory, the subdirectories were
24 skipped unless they satisfied the include/exclude conditions. This is
25 inconsistent with GNU grep (and could even be seen as contrary to the
26 pcregrep specification - which I improved to make it absolutely clear).
27 The action now is always to scan all levels of directory, and just
28 apply the include/exclude patterns to regular files.
29
30 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
31 --exclude_dir in the tests to avoid scanning .svn directories.
32
33
34 Version 7.6 28-Jan-08
35 ---------------------
36
37 1. A character class containing a very large number of characters with
38 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
39 overflow.
40
41 2. Patch to cut out the "long long" test in pcrecpp_unittest when
42 HAVE_LONG_LONG is not defined.
43
44 3. Applied Christian Ehrlicher's patch to update the CMake build files to
45 bring them up to date and include new features. This patch includes:
46
47 - Fixed PH's badly added libz and libbz2 support.
48 - Fixed a problem with static linking.
49 - Added pcredemo. [But later removed - see 7 below.]
50 - Fixed dftables problem and added an option.
51 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
52 HAVE_LONG_LONG.
53 - Added readline support for pcretest.
54 - Added an listing of the option settings after cmake has run.
55
56 4. A user submitted a patch to Makefile that makes it easy to create
57 "pcre.dll" under mingw when using Configure/Make. I added stuff to
58 Makefile.am that cause it to include this special target, without
59 affecting anything else. Note that the same mingw target plus all
60 the other distribution libraries and programs are now supported
61 when configuring with CMake (see 6 below) instead of with
62 Configure/Make.
63
64 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
65 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
66 exported in the Windows port". It has not yet been confirmed that the patch
67 solves the problem, but it does no harm.
68
69 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
70 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
71 with CMake, and also correct the comment about stack recursion.
72
73 7. Remove the automatic building of pcredemo from the ./configure system and
74 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
75 of a program that users should build themselves after PCRE is installed, so
76 building it automatically is not really right. What is more, it gave
77 trouble in some build environments.
78
79 8. Further tidies to CMakeLists.txt from Sheri and Christian.
80
81
82 Version 7.5 10-Jan-08
83 ---------------------
84
85 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
86 values in parens when parsing an RE using the C++ wrapper."
87
88 2. Negative specials like \S did not work in character classes in UTF-8 mode.
89 Characters greater than 255 were excluded from the class instead of being
90 included.
91
92 3. The same bug as (2) above applied to negated POSIX classes such as
93 [:^space:].
94
95 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
96 defined or documented. It seems to have been a typo for PCRE_STATIC, so
97 I have changed it.
98
99 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
100 first named subpattern) and a construct such as (?&a) would reference the
101 first named subpattern whose name started with "a" (in other words, the
102 length check was missing). Both these problems are fixed. "Subpattern name
103 expected" is now given for (?&) (a zero-length name), and this patch also
104 makes it give the same error for \k'' (previously it complained that that
105 was a reference to a non-existent subpattern).
106
107 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
108 this is right because (?- can be followed by option settings as well as by
109 digits. I have, however, made the messages clearer.
110
111 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
112 than the number used in the conditional) now cause a compile-time error.
113 This is actually not compatible with Perl, which accepts such patterns, but
114 treats the conditional as always being FALSE (as PCRE used to), but it
115 seems to me that giving a diagnostic is better.
116
117 8. Change "alphameric" to the more common word "alphanumeric" in comments
118 and messages.
119
120 9. Fix two occurrences of "backslash" in comments that should have been
121 "backspace".
122
123 10. Remove two redundant lines of code that can never be obeyed (their function
124 was moved elsewhere).
125
126 11. The program that makes PCRE's Unicode character property table had a bug
127 which caused it to generate incorrect table entries for sequences of
128 characters that have the same character type, but are in different scripts.
129 It amalgamated them into a single range, with the script of the first of
130 them. In other words, some characters were in the wrong script. There were
131 thirteen such cases, affecting characters in the following ranges:
132
133 U+002b0 - U+002c1
134 U+0060c - U+0060d
135 U+0061e - U+00612
136 U+0064b - U+0065e
137 U+0074d - U+0076d
138 U+01800 - U+01805
139 U+01d00 - U+01d77
140 U+01d9b - U+01dbf
141 U+0200b - U+0200f
142 U+030fc - U+030fe
143 U+03260 - U+0327f
144 U+0fb46 - U+0fbb1
145 U+10450 - U+1049d
146
147 12. The -o option (show only the matching part of a line) for pcregrep was not
148 compatible with GNU grep in that, if there was more than one match in a
149 line, it showed only the first of them. It now behaves in the same way as
150 GNU grep.
151
152 13. If the -o and -v options were combined for pcregrep, it printed a blank
153 line for every non-matching line. GNU grep prints nothing, and pcregrep now
154 does the same. The return code can be used to tell if there were any
155 non-matching lines.
156
157 14. Added --file-offsets and --line-offsets to pcregrep.
158
159 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
160 infinitely looping recursion. The bug was that positive lookaheads were not
161 being skipped when checking for a possible empty match (negative lookaheads
162 and both kinds of lookbehind were skipped).
163
164 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
165 inclusion of <windows.h> to before rather than after the definition of
166 INVALID_FILE_ATTRIBUTES (patch from David Byron).
167
168 17. Specifying a possessive quantifier with a specific limit for a Unicode
169 character property caused pcre_compile() to compile bad code, which led at
170 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
171 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
172 caused the error; without that there was no problem.
173
174 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
175
176 19. Added --enable-pcretest-libreadline.
177
178 20. In pcrecpp.cc, the variable 'count' was incremented twice in
179 RE::GlobalReplace(). As a result, the number of replacements returned was
180 double what it should be. I removed one of the increments, but Craig sent a
181 later patch that removed the other one (the right fix) and added unit tests
182 that check the return values (which was not done before).
183
184 21. Several CMake things:
185
186 (1) Arranged that, when cmake is used on Unix, the libraries end up with
187 the names libpcre and libpcreposix, not just pcre and pcreposix.
188
189 (2) The above change means that pcretest and pcregrep are now correctly
190 linked with the newly-built libraries, not previously installed ones.
191
192 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
193
194 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
195 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
196 UTF-8 newline character). The key issue is that the pattern starts .*;
197 this means that the match must be either at the beginning, or after a
198 newline. The bug was in the code for advancing after a failed match and
199 checking that the new position followed a newline. It was not taking
200 account of UTF-8 characters correctly.
201
202 23. PCRE was behaving differently from Perl in the way it recognized POSIX
203 character classes. PCRE was not treating the sequence [:...:] as a
204 character class unless the ... were all letters. Perl, however, seems to
205 allow any characters between [: and :], though of course it rejects as
206 unknown any "names" that contain non-letters, because all the known class
207 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
208 for example, whereas PCRE did not - it did not recognize a POSIX character
209 class. This seemed a bit dangerous, so the code has been changed to be
210 closer to Perl. The behaviour is not identical to Perl, because PCRE will
211 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
212 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
213 Perl does, and where it didn't before.
214
215 24. Rewrite so as to remove the single use of %n from pcregrep because in some
216 Windows environments %n is disabled by default.
217
218
219 Version 7.4 21-Sep-07
220 ---------------------
221
222 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
223 means that a class such as [\s] counted as "explicit reference to CR or
224 LF". That isn't really right - the whole point of the change was to try to
225 help when there was an actual mention of one of the two characters. So now
226 the change happens only if \r or \n (or a literal CR or LF) character is
227 encountered.
228
229 2. The 32-bit options word was also used for 6 internal flags, but the numbers
230 of both had grown to the point where there were only 3 bits left.
231 Fortunately, there was spare space in the data structure, and so I have
232 moved the internal flags into a new 16-bit field to free up more option
233 bits.
234
235 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
236 but did not set the internal JCHANGED flag - either of these is enough to
237 control the way the "get" function works - but the PCRE_INFO_JCHANGED
238 facility is supposed to tell if (?J) was ever used, so now (?J) at the
239 start sets both bits.
240
241 4. Added options (at build time, compile time, exec time) to change \R from
242 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
243
244 5. doc/pcresyntax.html was missing from the distribution.
245
246 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
247 compatibility, even though it is no longer used.
248
249 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
250 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
251 windows.h file is present (where different names are used). [This was
252 reversed later after testing - see 16 below.]
253
254 8. Changed all #include <config.h> to #include "config.h". There were also
255 some further <pcre.h> cases that I changed to "pcre.h".
256
257 9. When pcregrep was used with the --colour option, it missed the line ending
258 sequence off the lines that it output.
259
260 10. It was pointed out to me that arrays of string pointers cause lots of
261 relocations when a shared library is dynamically loaded. A technique of
262 using a single long string with a table of offsets can drastically reduce
263 these. I have refactored PCRE in four places to do this. The result is
264 dramatic:
265
266 Originally: 290
267 After changing UCP table: 187
268 After changing error message table: 43
269 After changing table of "verbs" 36
270 After changing table of Posix names 22
271
272 Thanks to the folks working on Gregex for glib for this insight.
273
274 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
275 unicode-properties was also set.
276
277 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
278
279 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
280 checked only for CRLF.
281
282 14. Added casts to pcretest.c to avoid compiler warnings.
283
284 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
285
286 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
287 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
288 entirely. This removes changes made in 7 above.
289
290 17. The CMake files have been updated, and there is now more information about
291 building with CMake in the NON-UNIX-USE document.
292
293
294 Version 7.3 28-Aug-07
295 ---------------------
296
297 1. In the rejigging of the build system that eventually resulted in 7.1, the
298 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
299 brackets there is not right, since it causes compilers to look for an
300 installed pcre.h, not the version that is in the source that is being
301 compiled (which of course may be different). I have changed it back to:
302
303 #include "pcre.h"
304
305 I have a vague recollection that the change was concerned with compiling in
306 different directories, but in the new build system, that is taken care of
307 by the VPATH setting the Makefile.
308
309 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
310 when the subject happened to end in the byte 0x85 (e.g. if the last
311 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
312 characters but of course it shouldn't be taken as a newline when it is part
313 of another character. The bug was that, for an unlimited repeat of . in
314 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
315 characters when looking for a newline.
316
317 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
318
319 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
320 in debug output.
321
322 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
323 long printing in the pcrecpp unittest when running under MinGW.
324
325 6. ESC_K was left out of the EBCDIC table.
326
327 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
328 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
329 limit also applies to "virtual nesting" when a pattern is recursive, and in
330 this case 1000 isn't so big. I have been able to remove this limit at the
331 expense of backing off one optimization in certain circumstances. Normally,
332 when pcre_exec() would call its internal match() function recursively and
333 immediately return the result unconditionally, it uses a "tail recursion"
334 feature to save stack. However, when a subpattern that can match an empty
335 string has an unlimited repetition quantifier, it no longer makes this
336 optimization. That gives it a stack frame in which to save the data for
337 checking that an empty string has been matched. Previously this was taken
338 from the 1000-entry workspace that had been reserved. So now there is no
339 explicit limit, but more stack is used.
340
341 8. Applied Daniel's patches to solve problems with the import/export magic
342 syntax that is required for Windows, and which was going wrong for the
343 pcreposix and pcrecpp parts of the library. These were overlooked when this
344 problem was solved for the main library.
345
346 9. There were some crude static tests to avoid integer overflow when computing
347 the size of patterns that contain repeated groups with explicit upper
348 limits. As the maximum quantifier is 65535, the maximum group length was
349 set at 30,000 so that the product of these two numbers did not overflow a
350 32-bit integer. However, it turns out that people want to use groups that
351 are longer than 30,000 bytes (though not repeat them that many times).
352 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
353 made it possible to implement the integer overflow checks in a much more
354 dynamic way, which I have now done. The artificial limitation on group
355 length has been removed - we now have only the limit on the total length of
356 the compiled pattern, which depends on the LINK_SIZE setting.
357
358 10. Fixed a bug in the documentation for get/copy named substring when
359 duplicate names are permitted. If none of the named substrings are set, the
360 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
361 empty string.
362
363 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
364 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
365 because the ] is interpreted as the first data character and the
366 terminating ] is not found. PCRE has been made compatible with Perl in this
367 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
368 cause memory overwriting.
369
370 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
371 string has been matched (to stop an infinite loop). It was not recognizing
372 a conditional subpattern that could match an empty string if that
373 subpattern was within another subpattern. For example, it looped when
374 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
375 condition was not nested. This bug has been fixed.
376
377 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
378 past the start of the subject in the presence of bytes with the top bit
379 set, for example "\x8aBCD".
380
381 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
382 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
383
384 14. Optimized (?!) to (*FAIL).
385
386 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
387 This restricts code points to be within the range 0 to 0x10FFFF, excluding
388 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
389 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
390 does: it's just the validity check that is more restrictive.
391
392 16. Inserted checks for integer overflows during escape sequence (backslash)
393 processing, and also fixed erroneous offset values for syntax errors during
394 backslash processing.
395
396 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
397 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
398
399 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
400 caused an overrun.
401
402 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
403 something other than just ASCII characters) inside a group that had an
404 unlimited repeat caused a loop at compile time (while checking to see
405 whether the group could match an empty string).
406
407 20. Debugging a pattern containing \p or \P could cause a crash. For example,
408 [\P{Any}] did so. (Error in the code for printing property names.)
409
410 21. An orphan \E inside a character class could cause a crash.
411
412 22. A repeated capturing bracket such as (A)? could cause a wild memory
413 reference during compilation.
414
415 23. There are several functions in pcre_compile() that scan along a compiled
416 expression for various reasons (e.g. to see if it's fixed length for look
417 behind). There were bugs in these functions when a repeated \p or \P was
418 present in the pattern. These operators have additional parameters compared
419 with \d, etc, and these were not being taken into account when moving along
420 the compiled data. Specifically:
421
422 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
423 length.
424
425 (b) An item such as \pL+ within a repeated group could cause crashes or
426 loops.
427
428 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
429 "reference to non-existent subpattern" error.
430
431 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
432
433 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
434 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
435
436 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
437
438 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
439 character were causing crashes (broken optimization).
440
441 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
442 \p or \P) caused a compile-time loop.
443
444 28. More problems have arisen in unanchored patterns when CRLF is a valid line
445 break. For example, the unstudied pattern [\r\n]A does not match the string
446 "\r\nA" because change 7.0/46 below moves the current point on by two
447 characters after failing to match at the start. However, the pattern \nA
448 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
449 the same is true. There doesn't seem any very clean way out of this, but
450 what I have chosen to do makes the common cases work: PCRE now takes note
451 of whether there can be an explicit match for \r or \n anywhere in the
452 pattern, and if so, 7.0/46 no longer applies. As part of this change,
453 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
454 pattern has explicit CR or LF references.
455
456 29. Added (*CR) etc for changing newline setting at start of pattern.
457
458
459 Version 7.2 19-Jun-07
460 ---------------------
461
462 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
463 which is apparently normally available under Windows.
464
465 2. Re-jig the pcregrep tests with different newline settings in an attempt
466 to make them independent of the local environment's newline setting.
467
468 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
469
470 4. Some of the "internals" tests were previously cut out when the link size
471 was not 2, because the output contained actual offsets. The recent new
472 "Z" feature of pcretest means that these can be cut out, making the tests
473 usable with all link sizes.
474
475 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
476 stack recursion. This gives a massive performance boost under BSD, but just
477 a small improvement under Linux. However, it saves one field in the frame
478 in all cases.
479
480 6. Added more features from the forthcoming Perl 5.10:
481
482 (a) (?-n) (where n is a string of digits) is a relative subroutine or
483 recursion call. It refers to the nth most recently opened parentheses.
484
485 (b) (?+n) is also a relative subroutine call; it refers to the nth next
486 to be opened parentheses.
487
488 (c) Conditions that refer to capturing parentheses can be specified
489 relatively, for example, (?(-2)... or (?(+3)...
490
491 (d) \K resets the start of the current match so that everything before
492 is not part of it.
493
494 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
495
496 (f) \g{name} is another synonym - part of Perl 5.10's unification of
497 reference syntax.
498
499 (g) (?| introduces a group in which the numbering of parentheses in each
500 alternative starts with the same number.
501
502 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
503
504 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
505 PCRE_INFO_JCHANGED.
506
507 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
508 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
509 for detecting groups that can match an empty string.
510
511 9. A pattern with a very large number of alternatives (more than several
512 hundred) was running out of internal workspace during the pre-compile
513 phase, where pcre_compile() figures out how much memory will be needed. A
514 bit of new cunning has reduced the workspace needed for groups with
515 alternatives. The 1000-alternative test pattern now uses 12 bytes of
516 workspace instead of running out of the 4096 that are available.
517
518 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
519
520 11. Applied patch from Google to remove an optimization that didn't quite work.
521 The report of the bug said:
522
523 pcrecpp::RE("a*").FullMatch("aaa") matches, while
524 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
525 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
526
527 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
528 it matched the wrong number of bytes.
529
530
531 Version 7.1 24-Apr-07
532 ---------------------
533
534 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
535 that is more "standard", making use of automake and other Autotools. There
536 is some re-arrangement of the files and adjustment of comments consequent
537 on this.
538
539 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
540 for recursive directory scanning broke on some systems because the files
541 are not scanned in any specific order and on different systems the order
542 was different. A call to "sort" has been inserted into RunGrepTest for the
543 approprate test as a short-term fix. In the longer term there may be an
544 alternative.
545
546 3. I had an email from Eric Raymond about problems translating some of PCRE's
547 man pages to HTML (despite the fact that I distribute HTML pages, some
548 people do their own conversions for various reasons). The problems
549 concerned the use of low-level troff macros .br and .in. I have therefore
550 removed all such uses from the man pages (some were redundant, some could
551 be replaced by .nf/.fi pairs). The 132html script that I use to generate
552 HTML has been updated to handle .nf/.fi and to complain if it encounters
553 .br or .in.
554
555 4. Updated comments in configure.ac that get placed in config.h.in and also
556 arranged for config.h to be included in the distribution, with the name
557 config.h.generic, for the benefit of those who have to compile without
558 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
559
560 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
561 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
562 makevp.bat for the current PCRE, using the additional files
563 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
564
565 6. A Windows user reported a minor discrepancy with test 2, which turned out
566 to be caused by a trailing space on an input line that had got lost in his
567 copy. The trailing space was an accident, so I've just removed it.
568
569 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
570 that is needed.
571
572 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
573 as "const" (a) because they are and (b) because it helps the PHP
574 maintainers who have recently made a script to detect big data structures
575 in the php code that should be moved to the .rodata section. I remembered
576 to update Builducptable as well, so it won't revert if ucptable.h is ever
577 re-created.
578
579 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
580 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
581 order to be able to cut out the UTF-8 tables in the latter when UTF-8
582 support is not required. This saves 1.5-2K of code, which is important in
583 some applications.
584
585 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
586 so as not to refer to the tables, even though these functions will never be
587 called when UTF-8 support is disabled. Otherwise there are problems with a
588 shared library.
589
590 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
591
592 (a) It was defining its arguments as char * instead of void *.
593
594 (b) It was assuming that all moves were upwards in memory; this was true
595 a long time ago when I wrote it, but is no longer the case.
596
597 The emulated memove() is provided for those environments that have neither
598 memmove() nor bcopy(). I didn't think anyone used it these days, but that
599 is clearly not the case, as these two bugs were recently reported.
600
601 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
602 and Detrail to create the HTML documentation, the .txt form of the man
603 pages, and it removes trailing spaces from listed files. It also creates
604 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
605 case, it wraps all the #defines with #ifndefs. This script should be run
606 before "make dist".
607
608 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
609 with Unicode property support.
610
611 (a) For a maximizing quantifier, if the two different cases of the
612 character were of different lengths in their UTF-8 codings (there are
613 some cases like this - I found 11), and the matching function had to
614 back up over a mixture of the two cases, it incorrectly assumed they
615 were both the same length.
616
617 (b) When PCRE was configured to use the heap rather than the stack for
618 recursion during matching, it was not correctly preserving the data for
619 the other case of a UTF-8 character when checking ahead for a match
620 while processing a minimizing repeat. If the check also involved
621 matching a wide character, but failed, corruption could cause an
622 erroneous result when trying to check for a repeat of the original
623 character.
624
625 13. Some tidying changes to the testing mechanism:
626
627 (a) The RunTest script now detects the internal link size and whether there
628 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
629 values substituted by "configure". (The RunGrepTest script already did
630 this for UTF-8.) The configure.ac script no longer substitutes the
631 relevant variables.
632
633 (b) The debugging options /B and /D in pcretest show the compiled bytecode
634 with length and offset values. This means that the output is different
635 for different internal link sizes. Test 2 is skipped for link sizes
636 other than 2 because of this, bypassing the problem. Unfortunately,
637 there was also a test in test 3 (the locale tests) that used /B and
638 failed for link sizes other than 2. Rather than cut the whole test out,
639 I have added a new /Z option to pcretest that replaces the length and
640 offset values with spaces. This is now used to make test 3 independent
641 of link size. (Test 2 will be tidied up later.)
642
643 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
644 segmentation fault instead of returning the appropriate error message.
645
646 15. In multiline mode when the newline sequence was set to "any", the pattern
647 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
648 This doesn't seem right; it now treats the CRLF combination as the line
649 ending, and so does not match in that case. It's only a pattern such as ^$
650 that would hit this one: something like ^ABC$ would have failed after \r
651 and then tried again after \r\n.
652
653 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
654 in an attempt to make files that differ only in their line terminators
655 compare equal. This works on Linux.
656
657 17. Under certain error circumstances pcregrep might try to free random memory
658 as it exited. This is now fixed, thanks to valgrind.
659
660 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
661 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
662 was because its rules for how to advance for /g after matching an empty
663 string at the end of a line did not allow for this case. They now check for
664 it specially.
665
666 20. pcretest is supposed to handle patterns and data of any length, by
667 extending its buffers when necessary. It was getting this wrong when the
668 buffer for a data line had to be extended.
669
670 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
671 CRLF as a newline sequence.
672
673 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
674 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
675 I have nevertheless tidied it up.
676
677 23. Added some casts to kill warnings from HP-UX ia64 compiler.
678
679 24. Added a man page for pcre-config.
680
681
682 Version 7.0 19-Dec-06
683 ---------------------
684
685 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
686 moving to gcc 4.1.1.
687
688 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
689 sys/time.h, which is documented as needed for this function. It doesn't
690 seem to matter on Linux, but it showed up on some releases of OS X.
691
692 3. It seems that there are systems where bytes whose values are greater than
693 127 match isprint() in the "C" locale. The "C" locale should be the
694 default when a C program starts up. In most systems, only ASCII printing
695 characters match isprint(). This difference caused the output from pcretest
696 to vary, making some of the tests fail. I have changed pcretest so that:
697
698 (a) When it is outputting text in the compiled version of a pattern, bytes
699 other than 32-126 are always shown as hex escapes.
700
701 (b) When it is outputting text that is a matched part of a subject string,
702 it does the same, unless a different locale has been set for the match
703 (using the /L modifier). In this case, it uses isprint() to decide.
704
705 4. Fixed a major bug that caused incorrect computation of the amount of memory
706 required for a compiled pattern when options that changed within the
707 pattern affected the logic of the preliminary scan that determines the
708 length. The relevant options are -x, and -i in UTF-8 mode. The result was
709 that the computed length was too small. The symptoms of this bug were
710 either the PCRE error "internal error: code overflow" from pcre_compile(),
711 or a glibc crash with a message such as "pcretest: free(): invalid next
712 size (fast)". Examples of patterns that provoked this bug (shown in
713 pcretest format) are:
714
715 /(?-x: )/x
716 /(?x)(?-x: \s*#\s*)/
717 /((?i)[\x{c0}])/8
718 /(?i:[\x{c0}])/8
719
720 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
721 is now done differently.
722
723 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
724 wrapper classes; (b) implement a new function in the C++ scanner that is
725 more efficient than the old way of doing things because it avoids levels of
726 recursion in the regex matching; (c) add a paragraph to the documentation
727 for the FullMatch() function.
728
729 6. The escape sequence \n was being treated as whatever was defined as
730 "newline". Not only was this contrary to the documentation, which states
731 that \n is character 10 (hex 0A), but it also went horribly wrong when
732 "newline" was defined as CRLF. This has been fixed.
733
734 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
735 was being set to -1 for the "end of line" case (supposedly a value that no
736 character can have). Though this value is never used (the check for end of
737 line is "zero bytes in current character"), it caused compiler complaints.
738 I've changed it to 0xffffffff.
739
740 8. In pcre_version.c, the version string was being built by a sequence of
741 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
742 string (as it is for production releases) called a macro with an empty
743 argument. The C standard says the result of this is undefined. The gcc
744 compiler treats it as an empty string (which was what was wanted) but it is
745 reported that Visual C gives an error. The source has been hacked around to
746 avoid this problem.
747
748 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
749 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
750 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
751 of them did).
752
753 10. Originally, pcretest opened its input and output without "b"; then I was
754 told that "b" was needed in some environments, so it was added for release
755 5.0 to both the input and output. (It makes no difference on Unix-like
756 systems.) Later I was told that it is wrong for the input on Windows. I've
757 now abstracted the modes into two macros, to make it easier to fiddle with
758 them, and removed "b" from the input mode under Windows.
759
760 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
761
762 12. Added -help and --help to pcretest as an official way of being reminded
763 of the options.
764
765 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
766 and pcrecpp.cc because they annoy compilers at high warning levels.
767
768 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
769
770 15. Fixed an occurrence of == in configure.ac that should have been = (shell
771 scripts are not C programs :-) and which was not noticed because it works
772 on Linux.
773
774 16. pcretest is supposed to handle any length of pattern and data line (as one
775 line or as a continued sequence of lines) by extending its input buffer if
776 necessary. This feature was broken for very long pattern lines, leading to
777 a string of junk being passed to pcre_compile() if the pattern was longer
778 than about 50K.
779
780 17. I have done a major re-factoring of the way pcre_compile() computes the
781 amount of memory needed for a compiled pattern. Previously, there was code
782 that made a preliminary scan of the pattern in order to do this. That was
783 OK when PCRE was new, but as the facilities have expanded, it has become
784 harder and harder to keep it in step with the real compile phase, and there
785 have been a number of bugs (see for example, 4 above). I have now found a
786 cunning way of running the real compile function in a "fake" mode that
787 enables it to compute how much memory it would need, while actually only
788 ever using a few hundred bytes of working memory and without too many
789 tests of the mode. This should make future maintenance and development
790 easier. A side effect of this work is that the limit of 200 on the nesting
791 depth of parentheses has been removed (though this was never a serious
792 limitation, I suspect). However, there is a downside: pcre_compile() now
793 runs more slowly than before (30% or more, depending on the pattern). I
794 hope this isn't a big issue. There is no effect on runtime performance.
795
796 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
797 newline (only possible for the last line of a file) and it was a
798 pattern that set a locale (followed by /Lsomething), pcretest crashed.
799
800 19. Added additional timing features to pcretest. (1) The -tm option now times
801 matching only, not compiling. (2) Both -t and -tm can be followed, as a
802 separate command line item, by a number that specifies the number of
803 repeats to use when timing. The default is 50000; this gives better
804 precision, but takes uncomfortably long for very large patterns.
805
806 20. Extended pcre_study() to be more clever in cases where a branch of a
807 subpattern has no definite first character. For example, (a*|b*)[cd] would
808 previously give no result from pcre_study(). Now it recognizes that the
809 first character must be a, b, c, or d.
810
811 21. There was an incorrect error "recursive call could loop indefinitely" if
812 a subpattern (or the entire pattern) that was being tested for matching an
813 empty string contained only one non-empty item after a nested subpattern.
814 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
815 incorrectly, because the \d was being skipped in the check.
816
817 22. The pcretest program now has a new pattern option /B and a command line
818 option -b, which is equivalent to adding /B to every pattern. This causes
819 it to show the compiled bytecode, without the additional information that
820 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
821 is the same as /B/I).
822
823 23. A new optimization is now able automatically to treat some sequences such
824 as a*b as a*+b. More specifically, if something simple (such as a character
825 or a simple class like \d) has an unlimited quantifier, and is followed by
826 something that cannot possibly match the quantified thing, the quantifier
827 is automatically "possessified".
828
829 24. A recursive reference to a subpattern whose number was greater than 39
830 went wrong under certain circumstances in UTF-8 mode. This bug could also
831 have affected the operation of pcre_study().
832
833 25. Realized that a little bit of performance could be had by replacing
834 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
835
836 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
837
838 27. Possessive quantifiers such as a++ were previously implemented by turning
839 them into atomic groups such as ($>a+). Now they have their own opcodes,
840 which improves performance. This includes the automatically created ones
841 from 23 above.
842
843 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
844 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
845 the first matched character to be a colon. This applied both to named and
846 numbered groups.
847
848 29. The ucpinternal.h header file was missing its idempotency #ifdef.
849
850 30. I was sent a "project" file called libpcre.a.dev which I understand makes
851 building PCRE on Windows easier, so I have included it in the distribution.
852
853 31. There is now a check in pcretest against a ridiculously large number being
854 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
855 loop, the loop is abandoned.
856
857 32. Forward references to subpatterns in conditions such as (?(2)...) where
858 subpattern 2 is defined later cause pcre_compile() to search forwards in
859 the pattern for the relevant set of parentheses. This search went wrong
860 when there were unescaped parentheses in a character class, parentheses
861 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
862
863 33. "Subroutine" calls and backreferences were previously restricted to
864 referencing subpatterns earlier in the regex. This restriction has now
865 been removed.
866
867 34. Added a number of extra features that are going to be in Perl 5.10. On the
868 whole, these are just syntactic alternatives for features that PCRE had
869 previously implemented using the Python syntax or my own invention. The
870 other formats are all retained for compatibility.
871
872 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
873 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
874 also .NET compatible.
875
876 (b) A recursion or subroutine call to a named group can now be defined as
877 (?&name) as well as (?P>name).
878
879 (c) A backreference to a named group can now be defined as \k<name> or
880 \k'name' as well as (?P=name). The new forms, as well as being in Perl
881 5.10, are also .NET compatible.
882
883 (d) A conditional reference to a named group can now use the syntax
884 (?(<name>) or (?('name') as well as (?(name).
885
886 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
887 groups (named and numbered) that are never evaluated inline, but can be
888 called as "subroutines" from elsewhere. In effect, the DEFINE condition
889 is always false. There may be only one alternative in such a group.
890
891 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
892 as the simple (?(R). The condition is true only if the most recent
893 recursion is that of the given number or name. It does not search out
894 through the entire recursion stack.
895
896 (g) The escape \gN or \g{N} has been added, where N is a positive or
897 negative number, specifying an absolute or relative reference.
898
899 35. Tidied to get rid of some further signed/unsigned compiler warnings and
900 some "unreachable code" warnings.
901
902 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
903 things, this adds five new scripts.
904
905 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
906 There were also incompatibilities regarding the handling of \Q..\E inside
907 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
908 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
909
910 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
911 matches an empty string, and forcibly breaks the loop. There were bugs in
912 this code in non-simple cases. For a pattern such as ^(a()*)* matched
913 against aaaa the result was just "a" rather than "aaaa", for example. Two
914 separate and independent bugs (that affected different cases) have been
915 fixed.
916
917 39. Refactored the code to abolish the use of different opcodes for small
918 capturing bracket numbers. This is a tidy that I avoided doing when I
919 removed the limit on the number of capturing brackets for 3.5 back in 2001.
920 The new approach is not only tidier, it makes it possible to reduce the
921 memory needed to fix the previous bug (38).
922
923 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
924 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
925 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
926 mode.
927
928 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
929 report.
930
931 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
932 copy construction and assignment in the C++ wrapper.
933
934 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
935 couple of bugs that could have given wrong results in the "--newline=crlf"
936 case.
937
938 44. Added a number of casts and did some reorganization of signed/unsigned int
939 variables following suggestions from Dair Grant. Also renamed the variable
940 "this" as "item" because it is a C++ keyword.
941
942 45. Arranged for dftables to add
943
944 #include "pcre_internal.h"
945
946 to pcre_chartables.c because without it, gcc 4.x may remove the array
947 definition from the final binary if PCRE is built into a static library and
948 dead code stripping is activated.
949
950 46. For an unanchored pattern, if a match attempt fails at the start of a
951 newline sequence, and the newline setting is CRLF or ANY, and the next two
952 characters are CRLF, advance by two characters instead of one.
953
954
955 Version 6.7 04-Jul-06
956 ---------------------
957
958 1. In order to handle tests when input lines are enormously long, pcretest has
959 been re-factored so that it automatically extends its buffers when
960 necessary. The code is crude, but this _is_ just a test program. The
961 default size has been increased from 32K to 50K.
962
963 2. The code in pcre_study() was using the value of the re argument before
964 testing it for NULL. (Of course, in any sensible call of the function, it
965 won't be NULL.)
966
967 3. The memmove() emulation function in pcre_internal.h, which is used on
968 systems that lack both memmove() and bcopy() - that is, hardly ever -
969 was missing a "static" storage class specifier.
970
971 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
972 containing an extended class (one that cannot be represented by a bitmap
973 because it contains high-valued characters or Unicode property items, e.g.
974 [\pZ]). Almost always one would set UTF-8 mode when processing such a
975 pattern, but PCRE should not loop if you do not (it no longer does).
976 [Detail: two cases were found: (a) a repeated subpattern containing an
977 extended class; (b) a recursive reference to a subpattern that followed a
978 previous extended class. It wasn't skipping over the extended class
979 correctly when UTF-8 mode was not set.]
980
981 5. A negated single-character class was not being recognized as fixed-length
982 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
983 compile error "lookbehind assertion is not fixed length".
984
985 6. The RunPerlTest auxiliary script was showing an unexpected difference
986 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
987 write a Perl script that can interpret lines of an input file either as
988 byte characters or as UTF-8, which is what "perltest" was being required to
989 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
990 can't do is switch easily at run time between having the "use utf8;" pragma
991 or not. In the end, I fudged it by using the RunPerlTest script to insert
992 "use utf8;" explicitly for the UTF-8 tests.
993
994 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
995 the end of the subject string, contrary to the documentation and to what
996 Perl does. This was true of both matching functions. Now it matches only at
997 the start of the subject and immediately after *internal* newlines.
998
999 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1000 a pointer to an int instead of a pointer to an unsigned long int. This
1001 caused problems on 64-bit systems.
1002
1003 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1004 instance of the 'standard' template library not being so standard".
1005
1006 10. There was no check on the number of named subpatterns nor the maximum
1007 length of a subpattern name. The product of these values is used to compute
1008 the size of the memory block for a compiled pattern. By supplying a very
1009 long subpattern name and a large number of named subpatterns, the size
1010 computation could be caused to overflow. This is now prevented by limiting
1011 the length of names to 32 characters, and the number of named subpatterns
1012 to 10,000.
1013
1014 11. Subpatterns that are repeated with specific counts have to be replicated in
1015 the compiled pattern. The size of memory for this was computed from the
1016 length of the subpattern and the repeat count. The latter is limited to
1017 65535, but there was no limit on the former, meaning that integer overflow
1018 could in principle occur. The compiled length of a repeated subpattern is
1019 now limited to 30,000 bytes in order to prevent this.
1020
1021 12. Added the optional facility to have named substrings with the same name.
1022
1023 13. Added the ability to use a named substring as a condition, using the
1024 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1025 are numbers (not recommended). Forward references are permitted.
1026
1027 14. Added forward references in named backreferences (if you see what I mean).
1028
1029 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1030 pattern could run off the end of the subject. For example, the pattern
1031 "(?s)(.{1,5})"8 did this with the subject "ab".
1032
1033 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1034 PCRE_CASELESS was set when matching characters that were quantified with ?
1035 or *.
1036
1037 17. A character class other than a single negated character that had a minimum
1038 but no maximum quantifier - for example [ab]{6,} - was not handled
1039 correctly by pce_dfa_exec(). It would match only one character.
1040
1041 18. A valid (though odd) pattern that looked like a POSIX character
1042 class but used an invalid character after [ (for example [[,abc,]]) caused
1043 pcre_compile() to give the error "Failed: internal error: code overflow" or
1044 in some cases to crash with a glibc free() error. This could even happen if
1045 the pattern terminated after [[ but there just happened to be a sequence of
1046 letters, a binary zero, and a closing ] in the memory that followed.
1047
1048 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1049 over the years. Originally (before any Unicode support), just the bottom 8
1050 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1051 output from "man perlunicode" includes this:
1052
1053 The regular expression compiler produces polymorphic opcodes. That
1054 is, the pattern adapts to the data and automatically switches to
1055 the Unicode character scheme when presented with Unicode data--or
1056 instead uses a traditional byte scheme when presented with byte
1057 data.
1058
1059 Sadly, a wide octal escape does not cause a switch, and in a string with
1060 no other multibyte characters, these octal escapes are treated as before.
1061 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1062 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1063 Unicode string.
1064
1065 I have not perpetrated such confusion in PCRE. Up till now, it took just
1066 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1067 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1068 translate to the appropriate multibyte character.
1069
1070 29. Applied some refactoring to reduce the number of warnings from Microsoft
1071 and Borland compilers. This has included removing the fudge introduced
1072 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1073 a warning about an unused variable.
1074
1075 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1076 characters since release 4.0, because Perl (from release 5.004) does not.
1077 [Or at least, is documented not to: some releases seem to be in conflict
1078 with the documentation.] However, when a pattern was studied with
1079 pcre_study() and all its branches started with \s, PCRE still included VT
1080 as a possible starting character. Of course, this did no harm; it just
1081 caused an unnecessary match attempt.
1082
1083 22. Removed a now-redundant internal flag bit that recorded the fact that case
1084 dependency changed within the pattern. This was once needed for "required
1085 byte" processing, but is no longer used. This recovers a now-scarce options
1086 bit. Also moved the least significant internal flag bit to the most-
1087 significant bit of the word, which was not previously used (hangover from
1088 the days when it was an int rather than a uint) to free up another bit for
1089 the future.
1090
1091 23. Added support for CRLF line endings as well as CR and LF. As well as the
1092 default being selectable at build time, it can now be changed at runtime
1093 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1094 specify that it is scanning data with non-default line endings.
1095
1096 24. Changed the definition of CXXLINK to make it agree with the definition of
1097 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1098
1099 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1100 recursions. This makes a big different to stack usage for some patterns.
1101
1102 26. If a subpattern containing a named recursion or subroutine reference such
1103 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1104 the space required for the compiled pattern went wrong and gave too small a
1105 value. Depending on the environment, this could lead to "Failed: internal
1106 error: code overflow at offset 49" or "glibc detected double free or
1107 corruption" errors.
1108
1109 27. Applied patches from Google (a) to support the new newline modes and (b) to
1110 advance over multibyte UTF-8 characters in GlobalReplace.
1111
1112 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1113 difference for some implementation of PCRE in some Windows version.
1114
1115 29. Added some extra testing facilities to pcretest:
1116
1117 \q<number> in a data line sets the "match limit" value
1118 \Q<number> in a data line sets the "match recursion limt" value
1119 -S <number> sets the stack size, where <number> is in megabytes
1120
1121 The -S option isn't available for Windows.
1122
1123
1124 Version 6.6 06-Feb-06
1125 ---------------------
1126
1127 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1128 in pcreposix.h. I have copied the definition from pcre.h.
1129
1130 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1131 because pcre.h is no longer a built file.
1132
1133 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1134 not normally included in the compiled code.
1135
1136
1137 Version 6.5 01-Feb-06
1138 ---------------------
1139
1140 1. When using the partial match feature with pcre_dfa_exec(), it was not
1141 anchoring the second and subsequent partial matches at the new starting
1142 point. This could lead to incorrect results. For example, with the pattern
1143 /1234/, partially matching against "123" and then "a4" gave a match.
1144
1145 2. Changes to pcregrep:
1146
1147 (a) All non-match returns from pcre_exec() were being treated as failures
1148 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1149 error message is output. Some extra information is given for the
1150 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1151 probably the only errors that are likely to be caused by users (by
1152 specifying a regex that has nested indefinite repeats, for instance).
1153 If there are more than 20 of these errors, pcregrep is abandoned.
1154
1155 (b) A binary zero was treated as data while matching, but terminated the
1156 output line if it was written out. This has been fixed: binary zeroes
1157 are now no different to any other data bytes.
1158
1159 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1160 used to set a locale for matching. The --locale=xxxx long option has
1161 been added (no short equivalent) to specify a locale explicitly on the
1162 pcregrep command, overriding the environment variables.
1163
1164 (d) When -B was used with -n, some line numbers in the output were one less
1165 than they should have been.
1166
1167 (e) Added the -o (--only-matching) option.
1168
1169 (f) If -A or -C was used with -c (count only), some lines of context were
1170 accidentally printed for the final match.
1171
1172 (g) Added the -H (--with-filename) option.
1173
1174 (h) The combination of options -rh failed to suppress file names for files
1175 that were found from directory arguments.
1176
1177 (i) Added the -D (--devices) and -d (--directories) options.
1178
1179 (j) Added the -F (--fixed-strings) option.
1180
1181 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1182
1183 (l) Added the --colo(u)r option.
1184
1185 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1186 is not present by default.
1187
1188 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1189 items such as (?R) or (?1), when the recursion could match a number of
1190 alternatives. If it matched one of the alternatives, but subsequently,
1191 outside the recursion, there was a failure, the code tried to back up into
1192 the recursion. However, because of the way PCRE is implemented, this is not
1193 possible, and the result was an incorrect result from the match.
1194
1195 In order to prevent this happening, the specification of recursion has
1196 been changed so that all such subpatterns are automatically treated as
1197 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1198
1199 4. I had overlooked the fact that, in some locales, there are characters for
1200 which isalpha() is true but neither isupper() nor islower() are true. In
1201 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1202 and ordfeminine) are like this. This affected the treatment of \w and \W
1203 when they appeared in character classes, but not when they appeared outside
1204 a character class. The bit map for "word" characters is now created
1205 separately from the results of isalnum() instead of just taking it from the
1206 upper, lower, and digit maps. (Plus the underscore character, of course.)
1207
1208 5. The above bug also affected the handling of POSIX character classes such as
1209 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1210 permanent tables. Instead, the bit maps for such a class were previously
1211 created as the appropriate unions of the upper, lower, and digit bitmaps.
1212 Now they are created by subtraction from the [[:word:]] class, which has
1213 its own bitmap.
1214
1215 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1216 It is created by subtracting the vertical space characters (\x09, \x0a,
1217 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1218 subtraction was done in the overall bitmap for a character class, meaning
1219 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1220 be recognized. This bug has been fixed.
1221
1222 7. Patches from the folks at Google:
1223
1224 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1225 real life, but is still worth protecting against".
1226
1227 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1228 regular expressions".
1229
1230 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1231 have it.
1232
1233 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1234 "configure" and the latter not, in order to fix a problem somebody had
1235 with compiling the Arg class on HP-UX.
1236
1237 (e) Improve the error-handling of the C++ wrapper a little bit.
1238
1239 (f) New tests for checking recursion limiting.
1240
1241 8. The pcre_memmove() function, which is used only if the environment does not
1242 have a standard memmove() function (and is therefore rarely compiled),
1243 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1244 returning a result (though PCRE never actually uses the result).
1245
1246 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1247 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1248 returned instead of calling malloc() with an overflowing number that would
1249 most likely cause subsequent chaos.
1250
1251 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1252
1253 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1254 with this option is matched, the nmatch and pmatch options of regexec() are
1255 ignored.
1256
1257 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1258 provided in case anyone wants to the the POSIX interface with UTF-8
1259 strings.
1260
1261 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1262 C++ linking (needed for some HP-UX environments).
1263
1264 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1265 (unused parameter) and in the pcre_printint() function (omitted "default"
1266 switch label when the default is to do nothing).
1267
1268 15. Added some code to make it possible, when PCRE is compiled as a C++
1269 library, to replace subject pointers for pcre_exec() with a smart pointer
1270 class, thus making it possible to process discontinuous strings.
1271
1272 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1273 much the same function. They were added by different people who were trying
1274 to make PCRE easy to compile on non-Unix systems. It has been suggested
1275 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1276 for compiling on Windows systems. I have therefore replaced it with
1277 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1278 defaults to "extern" for C or "extern C" for C++, which works fine on
1279 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1280 SCOPE with something explicit in config.h. In addition:
1281
1282 (a) pcreposix.h still had just "extern" instead of either of these macros;
1283 I have replaced it with PCRE_DATA_SCOPE.
1284
1285 (b) Functions such as _pcre_xclass(), which are internal to the library,
1286 but external in the C sense, all had PCRE_EXPORT in their definitions.
1287 This is apparently wrong for the Windows case, so I have removed it.
1288 (It makes no difference on Unix-like systems.)
1289
1290 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1291 of recursive calls to match(). This is different to MATCH_LIMIT because
1292 that limits the total number of calls to match(), not all of which increase
1293 the depth of recursion. Limiting the recursion depth limits the amount of
1294 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1295 when PCRE is compiled, and changed at run time. A patch from Google adds
1296 this functionality to the C++ interface.
1297
1298 18. Changes to the handling of Unicode character properties:
1299
1300 (a) Updated the table to Unicode 4.1.0.
1301
1302 (b) Recognize characters that are not in the table as "Cn" (undefined).
1303
1304 (c) I revised the way the table is implemented to a much improved format
1305 which includes recognition of ranges. It now supports the ranges that
1306 are defined in UnicodeData.txt, and it also amalgamates other
1307 characters into ranges. This has reduced the number of entries in the
1308 table from around 16,000 to around 3,000, thus reducing its size
1309 considerably. I realized I did not need to use a tree structure after
1310 all - a binary chop search is just as efficient. Having reduced the
1311 number of entries, I extended their size from 6 bytes to 8 bytes to
1312 allow for more data.
1313
1314 (d) Added support for Unicode script names via properties such as \p{Han}.
1315
1316 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1317 matching that character.
1318
1319 20. When matching a repeated Unicode property with a minimum greater than zero,
1320 (for example \pL{2,}), PCRE could look past the end of the subject if it
1321 reached it while seeking the minimum number of characters. This could
1322 happen only if some of the characters were more than one byte long, because
1323 there is a check for at least the minimum number of bytes.
1324
1325 21. Refactored the implementation of \p and \P so as to be more general, to
1326 allow for more different types of property in future. This has changed the
1327 compiled form incompatibly. Anybody with saved compiled patterns that use
1328 \p or \P will have to recompile them.
1329
1330 22. Added "Any" and "L&" to the supported property types.
1331
1332 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1333 but give a compile time error if the value is greater than 0xff.
1334
1335 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1336 accidentally not being installed or uninstalled.
1337
1338 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1339 made were to insert the current release number. This seemed silly, because
1340 it made things harder for people building PCRE on systems that don't run
1341 "configure". I have turned pcre.h into a distributed file, no longer built
1342 by "configure", with the version identification directly included. There is
1343 no longer a pcre.h.in file.
1344
1345 However, this change necessitated a change to the pcre-config script as
1346 well. It is built from pcre-config.in, and one of the substitutions was the
1347 release number. I have updated configure.ac so that ./configure now finds
1348 the release number by grepping pcre.h.
1349
1350 26. Added the ability to run the tests under valgrind.
1351
1352
1353 Version 6.4 05-Sep-05
1354 ---------------------
1355
1356 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1357 "--" to be printed when multiple files were scanned, even when none of the
1358 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1359 consider it to be a bug, and have restored the previous behaviour.
1360
1361 2. A couple of code tidies to get rid of compiler warnings.
1362
1363 3. The pcretest program used to cheat by referring to symbols in the library
1364 whose names begin with _pcre_. These are internal symbols that are not
1365 really supposed to be visible externally, and in some environments it is
1366 possible to suppress them. The cheating is now confined to including
1367 certain files from the library's source, which is a bit cleaner.
1368
1369 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1370 file's purpose clearer.
1371
1372 5. Reorganized pcre_ucp_findchar().
1373
1374
1375 Version 6.3 15-Aug-05
1376 ---------------------
1377
1378 1. The file libpcre.pc.in did not have general read permission in the tarball.
1379
1380 2. There were some problems when building without C++ support:
1381
1382 (a) If C++ support was not built, "make install" and "make test" still
1383 tried to test it.
1384
1385 (b) There were problems when the value of CXX was explicitly set. Some
1386 changes have been made to try to fix these, and ...
1387
1388 (c) --disable-cpp can now be used to explicitly disable C++ support.
1389
1390 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1391 backslash in a target when C++ was disabled. This confuses some
1392 versions of "make", apparently. Using an intermediate variable solves
1393 this. (Same for CPP_LOBJ.)
1394
1395 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1396 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1397 necessary on certain architectures.
1398
1399 4. Added a setting of -export-symbols-regex to the link command to remove
1400 those symbols that are exported in the C sense, but actually are local
1401 within the library, and not documented. Their names all begin with
1402 "_pcre_". This is not a perfect job, because (a) we have to except some
1403 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1404 available (and never for static libraries). I have made a note to try to
1405 find a way round (a) in the future.
1406
1407
1408 Version 6.2 01-Aug-05
1409 ---------------------
1410
1411 1. There was no test for integer overflow of quantifier values. A construction
1412 such as {1111111111111111} would give undefined results. What is worse, if
1413 a minimum quantifier for a parenthesized subpattern overflowed and became
1414 negative, the calculation of the memory size went wrong. This could have
1415 led to memory overwriting.
1416
1417 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1418
1419 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1420 operating environments where this matters.
1421
1422 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1423 PCRE options from within the C++ wrapper.
1424
1425 5. Named capturing subpatterns were not being correctly counted when a pattern
1426 was compiled. This caused two problems: (a) If there were more than 100
1427 such subpatterns, the calculation of the memory needed for the whole
1428 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1429 back references of the form \12, where the number was greater than 9, were
1430 not recognized as back references, even though there were sufficient
1431 previous subpatterns.
1432
1433 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1434 versions of gcc, e.g. 2.95.4.
1435
1436
1437 Version 6.1 21-Jun-05
1438 ---------------------
1439
1440 1. There was one reference to the variable "posix" in pcretest.c that was not
1441 surrounded by "#if !defined NOPOSIX".
1442
1443 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1444 the cross-check on the old pcre_info() function, for the benefit of the
1445 cut-down version of PCRE that is currently imported into Exim.
1446
1447 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1448 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1449 patterns that start (?i) but I don't think that matters. The use of (?i) is
1450 just an example; this all applies to the other options as well.
1451
1452 4. Since libtool seems to echo the compile commands it is issuing, the output
1453 from "make" can be reduced a bit by putting "@" in front of each libtool
1454 compile command.
1455
1456 5. Patch from the folks at Google for configure.in to be a bit more thorough
1457 in checking for a suitable C++ installation before trying to compile the
1458 C++ stuff. This should fix a reported problem when a compiler was present,
1459 but no suitable headers.
1460
1461 6. The man pages all had just "PCRE" as their title. I have changed them to
1462 be the relevant file name. I have also arranged that these names are
1463 retained in the file doc/pcre.txt, which is a concatenation in text format
1464 of all the man pages except the little individual ones for each function.
1465
1466 7. The NON-UNIX-USE file had not been updated for the different set of source
1467 files that come with release 6. I also added a few comments about the C++
1468 wrapper.
1469
1470
1471 Version 6.0 07-Jun-05
1472 ---------------------
1473
1474 1. Some minor internal re-organization to help with my DFA experiments.
1475
1476 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1477 didn't matter for the library itself when fully configured, but did matter
1478 when compiling without UCP support, or within Exim, where the ucp files are
1479 not imported.
1480
1481 3. Refactoring of the library code to split up the various functions into
1482 different source modules. The addition of the new DFA matching code (see
1483 below) to a single monolithic source would have made it really too
1484 unwieldy, quite apart from causing all the code to be include in a
1485 statically linked application, when only some functions are used. This is
1486 relevant even without the DFA addition now that patterns can be compiled in
1487 one application and matched in another.
1488
1489 The downside of splitting up is that there have to be some external
1490 functions and data tables that are used internally in different modules of
1491 the library but which are not part of the API. These have all had their
1492 names changed to start with "_pcre_" so that they are unlikely to clash
1493 with other external names.
1494
1495 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1496 a different (DFA) algorithm. Although it is slower than the original
1497 function, it does have some advantages for certain types of matching
1498 problem.
1499
1500 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1501 including restarting after a partial match.
1502
1503 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1504 defined when compiling for Windows was sent to me. I have put it into the
1505 code, though I have no means of testing or verifying it.
1506
1507 7. Added the pcre_refcount() auxiliary function.
1508
1509 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1510 match before or at the first newline in the subject string. In pcretest,
1511 the /f option on a pattern can be used to set this.
1512
1513 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1514 would behave wrongly. This has been present in PCRE since release 4.0.
1515
1516 10. A number of changes to the pcregrep command:
1517
1518 (a) Refactored how -x works; insert ^(...)$ instead of setting
1519 PCRE_ANCHORED and checking the length, in preparation for adding
1520 something similar for -w.
1521
1522 (b) Added the -w (match as a word) option.
1523
1524 (c) Refactored the way lines are read and buffered so as to have more
1525 than one at a time available.
1526
1527 (d) Implemented a pcregrep test script.
1528
1529 (e) Added the -M (multiline match) option. This allows patterns to match
1530 over several lines of the subject. The buffering ensures that at least
1531 8K, or the rest of the document (whichever is the shorter) is available
1532 for matching (and similarly the previous 8K for lookbehind assertions).
1533
1534 (f) Changed the --help output so that it now says
1535
1536 -w, --word-regex(p)
1537
1538 instead of two lines, one with "regex" and the other with "regexp"
1539 because that confused at least one person since the short forms are the
1540 same. (This required a bit of code, as the output is generated
1541 automatically from a table. It wasn't just a text change.)
1542
1543 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1544 option but starts with a hyphen. Could be a pattern or a path name
1545 starting with a hyphen, for instance.
1546
1547 (h) "-" can be given as a file name to represent stdin.
1548
1549 (i) When file names are being printed, "(standard input)" is used for
1550 the standard input, for compatibility with GNU grep. Previously
1551 "<stdin>" was used.
1552
1553 (j) The option --label=xxx can be used to supply a name to be used for
1554 stdin when file names are being printed. There is no short form.
1555
1556 (k) Re-factored the options decoding logic because we are going to add
1557 two more options that take data. Such options can now be given in four
1558 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1559
1560 (l) Added the -A, -B, and -C options for requesting that lines of context
1561 around matches be printed.
1562
1563 (m) Added the -L option to print the names of files that do not contain
1564 any matching lines, that is, the complement of -l.
1565
1566 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1567 continue to scan other files.
1568
1569 (o) The -s option was incorrectly implemented. For compatibility with other
1570 greps, it now suppresses the error message for a non-existent or non-
1571 accessible file (but not the return code). There is a new option called
1572 -q that suppresses the output of matching lines, which was what -s was
1573 previously doing.
1574
1575 (p) Added --include and --exclude options to specify files for inclusion
1576 and exclusion when recursing.
1577
1578 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1579 Hopefully, it now does.
1580
1581 12. Missing cast in pcre_study().
1582
1583 13. Added an "uninstall" target to the makefile.
1584
1585 14. Replaced "extern" in the function prototypes in Makefile.in with
1586 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1587 world, but is set differently for Windows.
1588
1589 15. Added a second compiling function called pcre_compile2(). The only
1590 difference is that it has an extra argument, which is a pointer to an
1591 integer error code. When there is a compile-time failure, this is set
1592 non-zero, in addition to the error test pointer being set to point to an
1593 error message. The new argument may be NULL if no error number is required
1594 (but then you may as well call pcre_compile(), which is now just a
1595 wrapper). This facility is provided because some applications need a
1596 numeric error indication, but it has also enabled me to tidy up the way
1597 compile-time errors are handled in the POSIX wrapper.
1598
1599 16. Added VPATH=.libs to the makefile; this should help when building with one
1600 prefix path and installing with another. (Or so I'm told by someone who
1601 knows more about this stuff than I do.)
1602
1603 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1604 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1605 match everything, including newlines. This is not POSIX-compatible, but
1606 somebody wanted the feature. From pcretest it can be activated by using
1607 both the P and the s flags.
1608
1609 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1610
1611 19. libpcre.pc was being incorrectly installed as executable.
1612
1613 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1614 it now also looks for '\r' so that it will work unmodified on Windows.
1615
1616 21. Added Google's contributed C++ wrapper to the distribution.
1617
1618 22. Added some untidy missing memory free() calls in pcretest, to keep
1619 Electric Fence happy when testing.
1620
1621
1622
1623 Version 5.0 13-Sep-04
1624 ---------------------
1625
1626 1. Internal change: literal characters are no longer packed up into items
1627 containing multiple characters in a single byte-string. Each character
1628 is now matched using a separate opcode. However, there may be more than one
1629 byte in the character in UTF-8 mode.
1630
1631 2. The pcre_callout_block structure has two new fields: pattern_position and
1632 next_item_length. These contain the offset in the pattern to the next match
1633 item, and its length, respectively.
1634
1635 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1636 insertion of callouts before each pattern item. Added the /C option to
1637 pcretest to make use of this.
1638
1639 4. On the advice of a Windows user, the lines
1640
1641 #if defined(_WIN32) || defined(WIN32)
1642 _setmode( _fileno( stdout ), 0x8000 );
1643 #endif /* defined(_WIN32) || defined(WIN32) */
1644
1645 have been added to the source of pcretest. This apparently does useful
1646 magic in relation to line terminators.
1647
1648 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1649 for the benefit of those environments where the "b" makes a difference.
1650
1651 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1652 to know about it. I have put a hack into configure.in that adds in code
1653 to set GCC=yes if CC=icc. This seems to end up at a point in the
1654 generated configure script that is early enough to affect the setting of
1655 compiler options, which is what is needed, but I have no means of testing
1656 whether it really works. (The user who reported this had patched the
1657 generated configure script, which of course I cannot do.)
1658
1659 LATER: After change 22 below (new libtool files), the configure script
1660 seems to know about icc (and also ecc). Therefore, I have commented out
1661 this hack in configure.in.
1662
1663 7. Added support for pkg-config (2 patches were sent in).
1664
1665 8. Negated POSIX character classes that used a combination of internal tables
1666 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1667 [[:^ascii]]. Typically, they would match almost any characters. The other
1668 POSIX classes were not broken in this way.
1669
1670 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1671 to find the match, as PCRE was deluded into thinking that the match had to
1672 start at the start point or following a newline. The same bug applied to
1673 patterns with negative forward assertions or any backward assertions
1674 preceding ".*" at the start, unless the pattern required a fixed first
1675 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1676
1677 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1678 starting at the last subject character, bytes beyond the end of the subject
1679 string were read.
1680
1681 11. Renamed the variable "class" as "classbits" to make life easier for C++
1682 users. (Previously there was a macro definition, but it apparently wasn't
1683 enough.)
1684
1685 12. Added the new field "tables" to the extra data so that tables can be passed
1686 in at exec time, or the internal tables can be re-selected. This allows
1687 a compiled regex to be saved and re-used at a later time by a different
1688 program that might have everything at different addresses.
1689
1690 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1691 -R library as well as a -L library.
1692
1693 14. The debugging options of pcretest (-d on the command line or D on a
1694 pattern) showed incorrect output for anything following an extended class
1695 that contained multibyte characters and which was followed by a quantifier.
1696
1697 15. Added optional support for general category Unicode character properties
1698 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1699 support. It adds about 90K to the size of the library. The meanings of the
1700 inbuilt class escapes such as \d and \s have NOT been changed.
1701
1702 16. Updated pcredemo.c to include calls to free() to release the memory for the
1703 compiled pattern.
1704
1705 17. The generated file chartables.c was being created in the source directory
1706 instead of in the building directory. This caused the build to fail if the
1707 source directory was different from the building directory, and was
1708 read-only.
1709
1710 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1711 file. No doubt somebody will tell me if they don't make sense... Also added
1712 Dan Mooney's comments about building on OpenVMS.
1713
1714 19. Added support for partial matching via the PCRE_PARTIAL option for
1715 pcre_exec() and the \P data escape in pcretest.
1716
1717 20. Extended pcretest with 3 new pattern features:
1718
1719 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1720 write the compiled pattern to the file whose name is "rest-of-line".
1721 This is a straight binary dump of the data, with the saved pointer to
1722 the character tables forced to be NULL. The study data, if any, is
1723 written too. After writing, pcretest reads a new pattern.
1724
1725 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1726 compiled pattern from the given file. There must not be any
1727 occurrences of "<" in the file name (pretty unlikely); if there are,
1728 pcretest will instead treat the initial "<" as a pattern delimiter.
1729 After reading in the pattern, pcretest goes on to read data lines as
1730 usual.
1731
1732 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1733 and 16-bit fields in a compiled pattern, to simulate a pattern that
1734 was compiled on a host of opposite endianness.
1735
1736 21. The pcre-exec() function can now cope with patterns that were compiled on
1737 hosts of opposite endianness, with this restriction:
1738
1739 As for any compiled expression that is saved and used later, the tables
1740 pointer field cannot be preserved; the extra_data field in the arguments
1741 to pcre_exec() should be used to pass in a tables address if a value
1742 other than the default internal tables were used at compile time.
1743
1744 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1745 now diagnosed as an error. Previously, most of the time, a negative number
1746 would have been treated as zero, but if in addition "ovector" was passed as
1747 NULL, a crash could occur.
1748
1749 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1750 new versions from the libtool 1.5 distribution (the last one is a copy of
1751 a file called libtool.m4). This seems to have fixed the need to patch
1752 "configure" to support Darwin 1.3 (which I used to do). However, I still
1753 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1754 workstation).
1755
1756 24. Changed the PCRE licence to be the more standard "BSD" licence.
1757
1758
1759 Version 4.5 01-Dec-03
1760 ---------------------
1761
1762 1. There has been some re-arrangement of the code for the match() function so
1763 that it can be compiled in a version that does not call itself recursively.
1764 Instead, it keeps those local variables that need separate instances for
1765 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1766 needs to "recurse". Keeping track of where control must go is done by means
1767 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1768 hide most of the details from the main code, and operates only if
1769 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1770 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1771 operating.
1772
1773 To make it easier for callers to provide specially tailored get/free
1774 functions for this usage, two new functions, pcre_stack_malloc, and
1775 pcre_stack_free, are used. They are always called in strict stacking order,
1776 and the size of block requested is always the same.
1777
1778 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1779 PCRE has been compiled to use the stack or the heap for recursion. The
1780 -C option of pcretest uses this to show which version is compiled.
1781
1782 A new data escape \S, is added to pcretest; it causes the amounts of store
1783 obtained and freed by both kinds of malloc/free at match time to be added
1784 to the output.
1785
1786 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1787 what's available on my current Linux desktop machine.
1788
1789 3. When matching a UTF-8 string, the test for a valid string at the start has
1790 been extended. If start_offset is not zero, PCRE now checks that it points
1791 to a byte that is the start of a UTF-8 character. If not, it returns
1792 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1793 this is necessary because there may be backward assertions in the pattern.
1794 When matching the same subject several times, it may save resources to use
1795 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1796
1797 4. The code for checking the validity of UTF-8 strings has been tightened so
1798 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1799 containing "overlong sequences".
1800
1801 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1802 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1803 should have been "&", but it just so happened that all the cases this let
1804 through by mistake were picked up later in the function.
1805
1806 6. I had used a variable called "isblank" - this is a C99 function, causing
1807 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1808
1809 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1810 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1811 so the version is shown; (c) stop "make test" echoing "./RunTest".
1812
1813 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1814
1815 9. The prototype for memmove() for systems that don't have it was using
1816 size_t, but the inclusion of the header that defines size_t was later. I've
1817 moved the #includes for the C headers earlier to avoid this.
1818
1819 10. Added some adjustments to the code to make it easier to compiler on certain
1820 special systems:
1821
1822 (a) Some "const" qualifiers were missing.
1823 (b) Added the macro EXPORT before all exported functions; by default this
1824 is defined to be empty.
1825 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1826 that it reads its output file name as an argument instead of writing
1827 to the standard output and assuming this can be redirected.
1828
1829 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1830 class containing characters with values greater than 255, PCRE compilation
1831 went into a loop.
1832
1833 12. A recursive reference to a subpattern that was within another subpattern
1834 that had a minimum quantifier of zero caused PCRE to crash. For example,
1835 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1836 recursion. If the recursively-called subpattern itself had a zero repeat,
1837 that was OK.
1838
1839 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1840 buffer into which it was copied (for escape processing) was still set at
1841 1024, so long lines caused crashes.
1842
1843 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1844 "internal error: code overflow...". This applied to any character class
1845 that was followed by a possessive quantifier.
1846
1847 15. Modified the Makefile to add libpcre.la as a prerequisite for
1848 libpcreposix.la because I was told this is needed for a parallel build to
1849 work.
1850
1851 16. If a pattern that contained .* following optional items at the start was
1852 studied, the wrong optimizing data was generated, leading to matching
1853 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1854 matching string must start with a or b or c. The correct conclusion for
1855 this pattern is that a match can start with any character.
1856
1857
1858 Version 4.4 13-Aug-03
1859 ---------------------
1860
1861 1. In UTF-8 mode, a character class containing characters with values between
1862 127 and 255 was not handled correctly if the compiled pattern was studied.
1863 In fixing this, I have also improved the studying algorithm for such
1864 classes (slightly).
1865
1866 2. Three internal functions had redundant arguments passed to them. Removal
1867 might give a very teeny performance improvement.
1868
1869 3. Documentation bug: the value of the capture_top field in a callout is *one
1870 more than* the number of the hightest numbered captured substring.
1871
1872 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1873 in incorrectly linking with a previously installed version. They now link
1874 explicitly with libpcre.la.
1875
1876 5. configure.in no longer needs to recognize Cygwin specially.
1877
1878 6. A problem in pcre.in for Windows platforms is fixed.
1879
1880 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1881 pcretest, it used to include the size of the study block as part of its
1882 output. Unfortunately, the structure contains a field that has a different
1883 size on different hardware architectures. This meant that the tests that
1884 showed this size failed. As the block is currently always of a fixed size,
1885 this information isn't actually particularly useful in pcretest output, so
1886 I have just removed it.
1887
1888 8. Three pre-processor statements accidentally did not start in column 1.
1889 Sadly, there are *still* compilers around that complain, even though
1890 standard C has not required this for well over a decade. Sigh.
1891
1892 9. In pcretest, the code for checking callouts passed small integers in the
1893 callout_data field, which is a void * field. However, some picky compilers
1894 complained about the casts involved for this on 64-bit systems. Now
1895 pcretest passes the address of the small integer instead, which should get
1896 rid of the warnings.
1897
1898 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1899 both compile and run time, and gives an error if an invalid UTF-8 sequence
1900 is found. There is a option for disabling this check in cases where the
1901 string is known to be correct and/or the maximum performance is wanted.
1902
1903 11. In response to a bug report, I changed one line in Makefile.in from
1904
1905 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1906 to
1907 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1908
1909 to look similar to other lines, but I have no way of telling whether this
1910 is the right thing to do, as I do not use Windows. No doubt I'll get told
1911 if it's wrong...
1912
1913
1914 Version 4.3 21-May-03
1915 ---------------------
1916
1917 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1918 Makefile.
1919
1920 2. Some refactoring to improve the quality of the code:
1921
1922 (i) The utf8_table... variables are now declared "const".
1923
1924 (ii) The code for \cx, which used the "case flipping" table to upper case
1925 lower case letters, now just substracts 32. This is ASCII-specific,
1926 but the whole concept of \cx is ASCII-specific, so it seems
1927 reasonable.
1928
1929 (iii) PCRE was using its character types table to recognize decimal and
1930 hexadecimal digits in the pattern. This is silly, because it handles
1931 only 0-9, a-f, and A-F, but the character types table is locale-
1932 specific, which means strange things might happen. A private
1933 table is now used for this - though it costs 256 bytes, a table is
1934 much faster than multiple explicit tests. Of course, the standard
1935 character types table is still used for matching digits in subject
1936 strings against \d.
1937
1938 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1939 ending in _t are). So I've renamed it as ESC_tee.
1940
1941 3. The first argument for regexec() in the POSIX wrapper should have been
1942 defined as "const".
1943
1944 4. Changed pcretest to use malloc() for its buffers so that they can be
1945 Electric Fenced for debugging.
1946
1947 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1948 to read one or more bytes before the start of the subject string. Often this
1949 had no effect on PCRE's behaviour, but in some circumstances it could
1950 provoke a segmentation fault.
1951
1952 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1953 to try to read one or more bytes before the start of the subject string.
1954
1955 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1956 UTF-8 support could misbehave in various ways if the subject string
1957 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1958 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1959 back over UTF-8 characters.)
1960
1961
1962 Version 4.2 14-Apr-03
1963 ---------------------
1964
1965 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1966
1967 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1968 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1969 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1970 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1971 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1972 and BUILD_EXEEXT
1973 Note: automatic setting of the BUILD variables is not yet working
1974 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1975 compile-time but not at link-time
1976 [LINK]: use for linking executables only
1977 make different versions for Windows and non-Windows
1978 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1979 libraries
1980 [LINK_FOR_BUILD]: new variable
1981 [OBJEXT]: use throughout
1982 [EXEEXT]: use throughout
1983 <winshared>: new target
1984 <wininstall>: new target
1985 <dftables.o>: use native compiler
1986 <dftables>: use native linker
1987 <install>: handle Windows platform correctly
1988 <clean>: ditto
1989 <check>: ditto
1990 copy DLL to top builddir before testing
1991
1992 As part of these changes, -no-undefined was removed again. This was reported
1993 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1994 in any case.
1995
1996 3. Some tidies to get rid of compiler warnings:
1997
1998 . In the match_data structure, match_limit was an unsigned long int, whereas
1999 match_call_count was an int. I've made them both unsigned long ints.
2000
2001 . In pcretest the fact that a const uschar * doesn't automatically cast to
2002 a void * provoked a warning.
2003
2004 . Turning on some more compiler warnings threw up some "shadow" variables
2005 and a few more missing casts.
2006
2007 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2008 option, a class that contained a single character with a value between 128
2009 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2010
2011 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2012 option, a class that contained several characters, but with at least one
2013 whose value was between 128 and 255 caused PCRE to crash.
2014
2015
2016 Version 4.1 12-Mar-03
2017 ---------------------
2018
2019 1. Compiling with gcc -pedantic found a couple of places where casts were
2020 needed, and a string in dftables.c that was longer than standard compilers are
2021 required to support.
2022
2023 2. Compiling with Sun's compiler found a few more places where the code could
2024 be tidied up in order to avoid warnings.
2025
2026 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2027 first of these names is deprecated in the latest Autoconf in favour of the name
2028 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2029 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2030 analogy I have changed it to CFLAGS_FOR_BUILD.
2031
2032 4. Added -no-undefined to the linking command in the Makefile, because this is
2033 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2034 linking step for the pcreposix library.
2035
2036 5. PCRE was failing to diagnose the case of two named groups with the same
2037 name.
2038
2039 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2040 literal character that is needed in the subject for a match, and scans along to
2041 ensure that it is present before embarking on the full matching process. This
2042 saves time in cases of nested unlimited repeats that are never going to match.
2043 Problem: the scan can take a lot of time if the subject is very long (e.g.
2044 megabytes), thus penalizing straightforward matches. It is now done only if the
2045 amount of subject to be scanned is less than 1000 bytes.
2046
2047 7. A lesser problem with the same optimization is that it was recording the
2048 first character of an anchored pattern as "needed", thus provoking a search
2049 right along the subject, even when the first match of the pattern was going to
2050 fail. The "needed" character is now not set for anchored patterns, unless it
2051 follows something in the pattern that is of non-fixed length. Thus, it still
2052 fulfils its original purpose of finding quick non-matches in cases of nested
2053 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2054
2055
2056 Version 4.0 17-Feb-03
2057 ---------------------
2058
2059 1. If a comment in an extended regex that started immediately after a meta-item
2060 extended to the end of string, PCRE compiled incorrect data. This could lead to
2061 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2062
2063 2. Moved to autoconf 2.53 and libtool 1.4.2.
2064
2065 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2066 the special perltest8 script is no longer needed - all the tests can be run
2067 from a single perltest script.
2068
2069 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2070 by \s. It has now been removed in PCRE. This means it isn't recognized as
2071 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2072 class [:space:] *does* include VT, thereby creating a mess.
2073
2074 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2075 space and tab.
2076
2077 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2078 its new features into the main test script, reducing the number of scripts.
2079
2080 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2081 were backward compatible, and made the (?i) apply to the whole pattern, as if
2082 /i were given. Now it behaves more logically, and applies the option setting
2083 only to what follows. PCRE has been changed to follow suit. However, if it
2084 finds options settings right at the start of the pattern, it extracts them into
2085 the global options, as before. Thus, they show up in the info data.
2086
2087 8. Added support for the \Q...\E escape sequence. Characters in between are
2088 treated as literals. This is slightly different from Perl in that $ and @ are
2089 also handled as literals inside the quotes. In Perl, they will cause variable
2090 interpolation. Note the following examples:
2091
2092 Pattern PCRE matches Perl matches
2093
2094 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2095 \Qabc\$xyz\E abc\$xyz abc\$xyz
2096 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2097
2098 For compatibility with Perl, \Q...\E sequences are recognized inside character
2099 classes as well as outside them.
2100
2101 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2102 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2103 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2104 signed/unsigned warnings.
2105
2106 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2107 option for pcretest, so I've replaced it by a simple function that does just
2108 that job.
2109
2110 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2111 "pcregrep -".
2112
2113 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2114 Java package. This provides some syntactic sugar for simple cases of what my
2115 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2116 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2117 item, you can use this simplified notation. Note that only makes sense with
2118 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2119 greediness, whatever the setting of the PCRE_UNGREEDY option.
2120
2121 13. A change of greediness default within a pattern was not taking effect at
2122 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2123 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2124 was abstracted outside.
2125
2126 14. PCRE now supports the \G assertion. It is true when the current matching
2127 position is at the start point of the match. This differs from \A when the
2128 starting offset is non-zero. Used with the /g option of pcretest (or similar
2129 code), it works in the same way as it does for Perl's /g option. If all
2130 alternatives of a regex begin with \G, the expression is anchored to the start
2131 match position, and the "anchored" flag is set in the compiled expression.
2132
2133 15. Some bugs concerning the handling of certain option changes within patterns
2134 have been fixed. These applied to options other than (?ims). For example,
2135 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2136 been the other way round. Some of this was related to change 7 above.
2137
2138 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2139 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2140 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2141 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2142
2143 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2144 mode. Unlike ".", it always matches newline, whatever the setting of
2145 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2146 assertions. Perl allows it, but it doesn't (in general) work because it can't
2147 calculate the length of the lookbehind. At least, that's the case for Perl
2148 5.8.0 - I've been told they are going to document that it doesn't work in
2149 future.
2150
2151 18. Added an error diagnosis for escapes that PCRE does not support: these are
2152 \L, \l, \N, \P, \p, \U, \u, and \X.
2153
2154 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2155 reading past the end of the pattern in cases such as /[abcd/.
2156
2157 20. PCRE was getting more memory than necessary for patterns with classes that
2158 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2159
2160 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2161 compiling PCRE for use with Virtual Pascal.
2162
2163 22. Small fix to the Makefile to make it work properly if the build is done
2164 outside the source tree.
2165
2166 23. Added a new extension: a condition to go with recursion. If a conditional
2167 subpattern starts with (?(R) the "true" branch is used if recursion has
2168 happened, whereas the "false" branch is used only at the top level.
2169
2170 24. When there was a very long string of literal characters (over 255 bytes
2171 without UTF support, over 250 bytes with UTF support), the computation of how
2172 much memory was required could be incorrect, leading to segfaults or other
2173 strange effects.
2174
2175 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2176 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2177 there was a subsequent back reference to those brackets. This meant that, for
2178 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2179 possible to check for precisely this case. All we can do is abandon the
2180 optimization if .* occurs inside capturing brackets when there are any back
2181 references whatsoever. (See below for a better fix that came later.)
2182
2183 26. The handling of the optimization for finding the first character of a
2184 non-anchored pattern, and for finding a character that is required later in the
2185 match were failing in some cases. This didn't break the matching; it just
2186 failed to optimize when it could. The way this is done has been re-implemented.
2187
2188 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2189
2190 28. Added a new feature that provides some of the functionality that Perl
2191 provides with (?{...}). The facility is termed a "callout". The way it is done
2192 in PCRE is for the caller to provide an optional function, by setting
2193 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2194 global variable. By default it is unset, which disables all calling out. To get
2195 the function called, the regex must include (?C) at appropriate points. This
2196 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2197 This provides a means of identifying different callout points. When PCRE
2198 reaches such a point in the regex, if pcre_callout has been set, the external
2199 function is called. It is provided with data in a structure called
2200 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2201 matching continues; if it returns a non-zero value, the match at the current
2202 point fails. However, backtracking will occur if possible. [This was changed
2203 later and other features added - see item 49 below.]
2204
2205 29. pcretest is upgraded to test the callout functionality. It provides a
2206 callout function that displays information. By default, it shows the start of
2207 the match and the current position in the text. There are some new data escapes
2208 to vary what happens:
2209
2210 \C+ in addition, show current contents of captured substrings
2211 \C- do not supply a callout function
2212 \C!n return 1 when callout number n is reached
2213 \C!n!m return 1 when callout number n is reached for the mth time
2214
2215 30. If pcregrep was called with the -l option and just a single file name, it
2216 output "<stdin>" if a match was found, instead of the file name.
2217
2218 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2219 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2220 pcre_exec(). This saves a malloc/free per call. The default value of
2221 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2222 when configuring.
2223
2224 32. The default maximum size of a compiled pattern is 64K. There have been a
2225 few cases of people hitting this limit. The code now uses macros to handle the
2226 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2227 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2228 configuring. Tests 2 and 5 work only with 2-byte links because they output
2229 debugging information about compiled patterns.
2230
2231 33. Internal code re-arrangements:
2232
2233 (a) Moved the debugging function for printing out a compiled regex into
2234 its own source file (printint.c) and used #include to pull it into
2235 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2236 separate copies.
2237
2238 (b) Defined the list of op-code names for debugging as a macro in
2239 internal.h so that it is next to the definition of the opcodes.
2240
2241 (c) Defined a table of op-code lengths for simpler skipping along compiled
2242 code. This is again a macro in internal.h so that it is next to the
2243 definition of the opcodes.
2244
2245 34. Added support for recursive calls to individual subpatterns, along the
2246 lines of Robin Houston's patch (but implemented somewhat differently).
2247
2248 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2249 allow it to read and process whole directories in Win32. This code was
2250 contributed by Lionel Fourquaux; it has not been tested by me.
2251
2252 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2253 used to name a group. Names consist of alphanumerics and underscores, and must
2254 be unique. Back references use the syntax (?P=name) and recursive calls use
2255 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2256 numbers. The function pcre_fullinfo() can be used after compilation to extract
2257 a name/number map. There are three relevant calls:
2258
2259 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2260 PCRE_INFO_NAMECOUNT yields the number of entries
2261 PCRE_INFO_NAMETABLE yields a pointer to the map.
2262
2263 The map is a vector of fixed-size entries. The size of each entry depends on
2264 the length of the longest name used. The first two bytes of each entry are the
2265 group number, most significant byte first. There follows the corresponding
2266 name, zero terminated. The names are in alphabetical order.
2267
2268 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2269 case instead of 255. Making it the same both with and without UTF-8 support
2270 means that the same test output works with both.
2271
2272 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2273 calling malloc() with a zero argument.
2274
2275 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2276 optimization. I've improved things by keeping a bitmap of backreferences with
2277 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2278 fact referenced, the optimization can be applied. It is unlikely that a
2279 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2280 the match to follow \n) will appear inside brackets with a number greater than
2281 31, but if it does, any back reference > 31 suppresses the optimization.
2282
2283 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2284 of disabling numbered capturing parentheses. Any opening parenthesis that is
2285 not followed by ? behaves as if it were followed by ?: but named parentheses
2286 can still be used for capturing (and they will acquire numbers in the usual
2287 way).
2288
2289 41. Redesigned the return codes from the match() function into yes/no/error so
2290 that errors can be passed back from deep inside the nested calls. A malloc
2291 failure while inside a recursive subpattern call now causes the
2292 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2293
2294 42. It is now possible to set a limit on the number of times the match()
2295 function is called in a call to pcre_exec(). This facility makes it possible to
2296 limit the amount of recursion and backtracking, though not in a directly
2297 obvious way, because the match() function is used in a number of different
2298 circumstances. The count starts from zero for each position in the subject
2299 string (for non-anchored patterns). The default limit is, for compatibility, a
2300 large number, namely 10 000 000. You can change this in two ways:
2301
2302 (a) When configuring PCRE before making, you can use --with-match-limit=n
2303 to set a default value for the compiled library.
2304
2305 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2306 a different value is set. See 45 below.
2307
2308 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2309
2310 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2311 of things that can be changed at compile time. The first argument specifies
2312 what is wanted and the second points to where the information is to be placed.
2313 The current list of available information is:
2314
2315 PCRE_CONFIG_UTF8
2316
2317 The output is an integer that is set to one if UTF-8 support is available;
2318 otherwise it is set to zero.
2319
2320 PCRE_CONFIG_NEWLINE
2321
2322 The output is an integer that it set to the value of the code that is used for
2323 newline. It is either LF (10) or CR (13).
2324
2325 PCRE_CONFIG_LINK_SIZE
2326
2327 The output is an integer that contains the number of bytes used for internal
2328 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2329
2330 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2331
2332 The output is an integer that contains the threshold above which the POSIX
2333 interface uses malloc() for output vectors. See item 31 above.
2334
2335 PCRE_CONFIG_MATCH_LIMIT
2336
2337 The output is an unsigned integer that contains the default limit of the number
2338 of match() calls in a pcre_exec() execution. See 42 above.
2339
2340 44. pcretest has been upgraded by the addition of the -C option. This causes it
2341 to extract all the available output from the new pcre_config() function, and to
2342 output it. The program then exits immediately.
2343
2344 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2345 order to support additional features. One way would have been to define
2346 pcre_exec2() (for example) with extra arguments, but this would not have been
2347 extensible, and would also have required all calls to the original function to
2348 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2349 is used for passing in "extra" data from pcre_study().
2350
2351 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2352 contains the following fields:
2353
2354 flags a bitmap indicating which of the following fields are set
2355 study_data opaque data from pcre_study()
2356 match_limit a way of specifying a limit on match() calls for a specific
2357 call to pcre_exec()
2358 callout_data data for callouts (see 49 below)
2359
2360 The flag bits are also defined in pcre.h, and are
2361
2362 PCRE_EXTRA_STUDY_DATA
2363 PCRE_EXTRA_MATCH_LIMIT
2364 PCRE_EXTRA_CALLOUT_DATA
2365
2366 The pcre_study() function now returns one of these new pcre_extra blocks, with
2367 the actual study data pointed to by the study_data field, and the
2368 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2369 before. That is, this change is entirely upwards-compatible and requires no
2370 change to existing code.
2371
2372 If you want to pass in additional data to pcre_exec(), you can either place it
2373 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2374 block.
2375
2376 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2377 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2378 times with different match limits, until it finds the minimum value needed for
2379 pcre_exec() to complete. The value is then output. This can be instructive; for
2380 most simple matches the number is quite small, but for pathological cases it
2381 gets very large very quickly.
2382
2383 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2384 returns the size of the data block pointed to by the study_data field in a
2385 pcre_extra block, that is, the value that was passed as the argument to
2386 pcre_malloc() when PCRE was getting memory in which to place the information
2387 created by pcre_study(). The fourth argument should point to a size_t variable.
2388 pcretest has been extended so that this information is shown after a successful
2389 pcre_study() call when information about the compiled regex is being displayed.
2390
2391 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2392 because what follows is always an absolute path. (Later: it turns out that this
2393 is more than cosmetic for MinGW, because it doesn't like empty path
2394 components.)
2395
2396 49. Some changes have been made to the callout feature (see 28 above):
2397
2398 (i) A callout function now has three choices for what it returns:
2399
2400 0 => success, carry on matching
2401 > 0 => failure at this point, but backtrack if possible
2402 < 0 => serious error, return this value from pcre_exec()
2403
2404 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2405 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2406 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2407 use by callout functions. It will never be used by PCRE itself.
2408
2409 (ii) The pcre_extra structure (see 45 above) has a void * field called
2410 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2411 pcre_callout_block structure has a field of the same name. The contents of
2412 the field passed in the pcre_extra structure are passed to the callout
2413 function in the corresponding field in the callout block. This makes it
2414 easier to use the same callout-containing regex from multiple threads. For
2415 testing, the pcretest program has a new data escape
2416
2417 \C*n pass the number n (may be negative) as callout_data
2418
2419 If the callout function in pcretest receives a non-zero value as
2420 callout_data, it returns that value.
2421
2422 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2423 there were some redundant $(CFLAGS) in commands that are now specified as
2424 $(LINK), which already includes $(CFLAGS).
2425
2426 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2427 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2428 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2429 one-byte characters throughout. Note that case-insensitive matching applies
2430 only to characters whose values are less than 256. PCRE doesn't support the
2431 notion of cases for higher-valued characters.
2432
2433 (i) A character class whose characters are all within 0-255 is handled as
2434 a bit map, and the map is inverted for negative classes. Previously, a
2435 character > 255 always failed to match such a class; however it should
2436 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2437
2438 (ii) A negated character class with a single character < 255 is coded as
2439 "not this character" (OP_NOT). This wasn't working properly when the test
2440 character was multibyte, either singly or repeated.
2441
2442 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2443 mode, for example: \x{100}{2,3}.
2444
2445 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2446 singly or repeated) now correctly test multibyte characters. However,
2447 PCRE doesn't recognize any characters with values greater than 255 as
2448 digits, spaces, or word characters. Such characters always match \D, \S,
2449 and \W, and never match \d, \s, or \w.
2450
2451 (v) Classes may now contain characters and character ranges with values
2452 greater than 255. For example: [ab\x{100}-\x{400}].
2453
2454 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2455 PCRE in UTF-8 mode.
2456
2457 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2458 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2459 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2460 value.)
2461
2462 53. The single man page has become too large. I have therefore split it up into
2463 a number of separate man pages. These also give rise to individual HTML pages;
2464 these are now put in a separate directory, and there is an index.html page that
2465 lists them all. Some hyperlinking between the pages has been installed.
2466
2467 54. Added convenience functions for handling named capturing parentheses.
2468
2469 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2470 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2471 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2472 are faulted.
2473
2474 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2475 calling configure. These values are used when compiling the dftables.c program
2476 which is run to generate the source of the default character tables. They
2477 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2478 you will need to set these values.
2479
2480 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2481
2482
2483 Version 3.9 02-Jan-02
2484 ---------------------
2485
2486 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2487
2488 2. If --disable-static was given, the building process failed when trying to
2489 build pcretest and pcregrep. (For some reason it was using libtool to compile
2490 them, which is not right, as they aren't part of the library.)
2491
2492
2493 Version 3.8 18-Dec-01
2494 ---------------------
2495
2496 1. The experimental UTF-8 code was completely screwed up. It was packing the
2497 bytes in the wrong order. How dumb can you get?
2498
2499
2500 Version 3.7 29-Oct-01
2501 ---------------------
2502
2503 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2504 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2505 this didn't happen under Solaris 8, where I normally test things.
2506
2507 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2508 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2509 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2510 this fix an hour or so after the initial 3.7 release.)
2511
2512
2513 Version 3.6 23-Oct-01
2514 ---------------------
2515
2516 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2517 offsets passed as NULL with zero offset count.
2518
2519 2. The config.guess and config.sub files had not been updated when I moved to
2520 the latest autoconf.
2521
2522
2523 Version 3.5 15-Aug-01
2524 ---------------------
2525
2526 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2527 had been forgotten.
2528
2529 2. By using declared but undefined structures, we can avoid using "void"
2530 definitions in pcre.h while keeping the internal definitions of the structures
2531 private.
2532
2533 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2534 user point of view, this means that both static and shared libraries are built
2535 by default, but this can be individually controlled. More of the work of
2536 handling this static/shared cases is now inside libtool instead of PCRE's make
2537 file.
2538
2539 4. The pcretest utility is now installed along with pcregrep because it is
2540 useful for users (to test regexs) and by doing this, it automatically gets
2541 relinked by libtool. The documentation has been turned into a man page, so
2542 there are now .1, .txt, and .html versions in /doc.
2543
2544 5. Upgrades to pcregrep:
2545 (i) Added long-form option names like gnu grep.
2546 (ii) Added --help to list all options with an explanatory phrase.
2547 (iii) Added -r, --recursive to recurse into sub-directories.
2548 (iv) Added -f, --file to read patterns from a file.
2549
2550 6. pcre_exec() was referring to its "code" argument before testing that
2551 argument for NULL (and giving an error if it was NULL).
2552
2553 7. Upgraded Makefile.in to allow for compiling in a different directory from
2554 the source directory.
2555
2556 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2557 options bits, the pointer it was passed was to an int instead of to an unsigned
2558 long int. This mattered only on 64-bit systems.
2559
2560 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2561 generated) instead of pcre.in, which it its source. Also made the same change
2562 in several of the .c files.
2563
2564 10. A new release of gcc defines printf() as a macro, which broke pcretest
2565 because it had an ifdef in the middle of a string argument for printf(). Fixed
2566 by using separate calls to printf().
2567
2568 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2569 script, to force use of CR or LF instead of \n in the source. On non-Unix
2570 systems, the value can be set in config.h.
2571
2572 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2573 absolute limit. Changed the text of the error message to make this clear, and
2574 likewise updated the man page.
2575
2576 13. The limit of 99 on the number of capturing subpatterns has been removed.
2577 The new limit is 65535, which I hope will not be a "real" limit.
2578
2579
2580 Version 3.4 22-Aug-00
2581 ---------------------
2582
2583 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2584
2585 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2586
2587
2588 Version 3.3 01-Aug-00
2589 ---------------------
2590
2591 1. If an octal character was given, but the value was greater than \377, it
2592 was not getting masked to the least significant bits, as documented. This could
2593 lead to crashes in some systems.
2594
2595 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2596 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2597
2598 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2599 These just pass their arguments on to (pcre_free)(), but they are provided
2600 because some uses of PCRE bind it to non-C systems that can call its functions,
2601 but cannot call free() or pcre_free() directly.
2602
2603 4. Add "make test" as a synonym for "make check". Corrected some comments in
2604 the Makefile.
2605
2606 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2607 Makefile.
2608
2609 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2610 command called pgrep for grepping around the active processes.
2611
2612 7. Added the beginnings of support for UTF-8 character strings.
2613
2614 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2615 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2616 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2617 out for the ar command.)
2618
2619
2620 Version 3.2 12-May-00
2621 ---------------------
2622
2623 This is purely a bug fixing release.
2624
2625 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2626 of ZA. This was just one example of several cases that could provoke this bug,
2627 which was introduced by change 9 of version 2.00. The code for breaking
2628 infinite loops after an iteration that matches an empty string was't working
2629 correctly.
2630
2631 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2632 when matched against abbab (for example). After matching an empty string, it
2633 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2634 caused it to match further down the string than it should.
2635
2636 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2637 was there because it doesn't seem to be needed, and it causes trouble on some
2638 systems, as it is not a Standard C header. It has been removed.
2639
2640 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2641 were reported on the Macintosh. The changes were from
2642
2643 while ((c = *(++ptr)) != 0 && c != '\n');
2644 to
2645 while ((c = *(++ptr)) != 0 && c != '\n') ;
2646
2647 Totally extraordinary, but if that's what it takes...
2648
2649 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2650 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2651 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2652 assumes the way PCRE uses memmove() (always moving upwards).
2653
2654 6. PCRE is being used in one environment where strchr() is not available. There
2655 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2656 faster code anyway.
2657
2658
2659 Version 3.1 09-Feb-00
2660 ---------------------
2661
2662 The only change in this release is the fixing of some bugs in Makefile.in for
2663 the "install" target:
2664
2665 (1) It was failing to install pcreposix.h.
2666
2667 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2668
2669
2670 Version 3.0 01-Feb-00
2671 ---------------------
2672
2673 1. Add support for the /+ modifier to perltest (to output $` like it does in
2674 pcretest).
2675
2676 2. Add support for the /g modifier to perltest.
2677
2678 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2679 matches null strings.
2680
2681 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2682 pattern. Perl treats empty patterns specially - it reuses the most recent
2683 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2684 effect.
2685
2686 5. The POSIX interface was broken in that it was just handing over the POSIX
2687 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2688 required a bigger vector, with some working space on the end. This means that
2689 the POSIX wrapper now has to get and free some memory, and copy the results.
2690
2691 6. Added some simple autoconf support, placing the test data and the
2692 documentation in separate directories, re-organizing some of the
2693 information files, and making it build pcre-config (a GNU standard). Also added
2694 libtool support for building PCRE as a shared library, which is now the
2695 default.
2696
2697 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2698 09 are not valid octal constants. Single digits will be used for minor values
2699 less than 10.
2700
2701 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2702 existing programs that set these in the POSIX interface can use PCRE without
2703 modification.
2704
2705 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2706 return all that pcre_info() returns, plus additional data. The pcre_info()
2707 function is retained for compatibility, but is considered to be obsolete.
2708
2709 10. Added experimental recursion feature (?R) to handle one common case that
2710 Perl 5.6 will be able to do with (?p{...}).
2711
2712 11. Added support for POSIX character classes like [:alpha:], which Perl is
2713 adopting.
2714
2715
2716 Version 2.08 31-Aug-99
2717 ----------------------
2718
2719 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2720 trying to match at the startoffset position, but instead was moving forward to
2721 the next newline as if a previous match had failed.
2722
2723 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2724 and could get into a loop if a null string was matched other than at the start
2725 of the subject.
2726
2727 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2728 be distinguished at compile time, and for completeness also added PCRE_DATE.
2729
2730 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2731 in GnuWin32 environments.
2732
2733
2734 Version 2.07 29-Jul-99
2735 ----------------------
2736
2737 1. The documentation is now supplied in plain text form and HTML as well as in
2738 the form of man page sources.
2739
2740 2. C++ compilers don't like assigning (void *) values to other pointer types.
2741 In particular this affects malloc(). Although there is no problem in Standard
2742 C, I've put in casts to keep C++ compilers happy.
2743
2744 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2745 should be (const char *).
2746
2747 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2748 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2749 However, I haven't made this a standard facility. The documentation doesn't
2750 mention it, and the Makefile doesn't support it.
2751
2752 5. The Makefile now contains an "install" target, with editable destinations at
2753 the top of the file. The pcretest program is not installed.
2754
2755 6. pgrep -V now gives the PCRE version number and date.
2756
2757 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2758 causing the entire string to be ignored, instead of just the last character.
2759
2760 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2761 non-matching string, it can take a very, very long time, even for strings of
2762 quite modest length, because of the nested recursion. PCRE now does better in
2763 some of these cases. It does this by remembering the last required literal
2764 character in the pattern, and pre-searching the subject to ensure it is present
2765 before running the real match. In other words, it applies a heuristic to detect
2766 some types of certain failure quickly, and in the above example, if presented
2767 with a string that has no trailing " it gives "no match" very quickly.
2768
2769 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2770 other alternatives are tried instead.
2771
2772
2773 Version 2.06 09-Jun-99
2774 ----------------------
2775
2776 1. Change pcretest's output for amount of store used to show just the code
2777 space, because the remainder (the data block) varies in size between 32-bit and
2778 64-bit systems.
2779
2780 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2781 start matching at. This allows lookbehinds to work when searching for multiple
2782 occurrences in a string.
2783
2784 3. Added additional options to pcretest for testing multiple occurrences:
2785
2786 /+ outputs the rest of the string that follows a match
2787 /g loops for multiple occurrences, using the new startoffset argument
2788 /G loops for multiple occurrences by passing an incremented pointer
2789
2790 4. PCRE wasn't doing the "first character" optimization for patterns starting
2791 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2792 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2793 the letter 'x'. On long subject strings, this gives a significant speed-up.
2794
2795
2796 Version 2.05 21-Apr-99
2797 ----------------------
2798
2799 1. Changed the type of magic_number from int to long int so that it works
2800 properly on 16-bit systems.
2801
2802 2. Fixed a bug which caused patterns starting with .* not to work correctly
2803 when the subject string contained newline characters. PCRE was assuming
2804 anchoring for such patterns in all cases, which is not correct because .* will
2805 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2806 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2807 must be retried after every newline in the subject.
2808
2809
2810 Version 2.04 18-Feb-99
2811 ----------------------
2812
2813 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2814 computation of the store needed to hold the pattern was incorrect (too large).
2815 If such patterns were nested a few deep, this could multiply and become a real
2816 problem.
2817
2818 2. Added /M option to pcretest to show the memory requirement of a specific
2819 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2820
2821 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2822 compiled in such a way that the backtracking after subsequent failure was
2823 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2824 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2825
2826
2827 Version 2.03 02-Feb-99
2828 ----------------------
2829
2830 1. Fixed typo and small mistake in man page.
2831
2832 2. Added 4th condition (GPL supersedes if conflict) and created separate
2833 LICENCE file containing the conditions.
2834
2835 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2836 Perl, that is the internal \ allows the delimiter to be included in the
2837 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2838 the final delimiter, add \ to the end of the pattern (to test the error).
2839
2840 4. Added the convenience functions for extracting substrings after a successful
2841 match. Updated pcretest to make it able to test these functions.
2842
2843
2844 Version 2.02 14-Jan-99
2845 ----------------------
2846
2847 1. Initialized the working variables associated with each extraction so that
2848 their saving and restoring doesn't refer to uninitialized store.
2849
2850 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2851 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2852 fix the problem.
2853
2854 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2855 calls, and wasn't printing the correct value for compiling calls. Increased the
2856 default value of LOOPREPEAT, and the number of significant figures in the
2857 times.
2858
2859 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2860
2861 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2862 a building problem on Windows NT with a FAT file system.
2863
2864
2865 Version 2.01 21-Oct-98
2866 ----------------------
2867
2868 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2869 to character tables built by pcre_maketables() in the current locale. If NULL
2870 is passed, the default tables are used.
2871
2872
2873 Version 2.00 24-Sep-98
2874 ----------------------
2875
2876 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2877 it any more.
2878
2879 2. Allow quantification of (?>) groups, and make it work correctly.
2880
2881 3. The first character computation wasn't working for (?>) groups.
2882
2883 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2884 end of the subject) and add 5.005's \z, which really does match only at the
2885 very end of the subject.
2886
2887 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2888
2889 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2890 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2891 localized options. All options to pcre_study() were also removed.
2892
2893 7. Add other new features from 5.005:
2894
2895 $(?<= positive lookbehind
2896 $(?<! negative lookbehind
2897 (?imsx-imsx) added the unsetting capability
2898 such a setting is global if at outer level; local otherwise
2899 (?imsx-imsx:) non-capturing groups with option setting
2900 (?(cond)re|re) conditional pattern matching
2901
2902 A backreference to itself in a repeated group matches the previous
2903 captured string.
2904
2905 8. General tidying up of studying (both automatic and via "study")
2906 consequential on the addition of new assertions.
2907
2908 9. As in 5.005, unlimited repeated groups that could match an empty substring
2909 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2910 runtime if any iteration does actually match an empty substring.
2911
2912 10. Include the RunTest script in the distribution.
2913
2914 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2915 discrepancies, some of which were old and were also with respect to 5.004. They
2916 have now been fixed.
2917
2918
2919 Version 1.09 28-Apr-98
2920 ----------------------
2921
2922 1. A negated single character class followed by a quantifier with a minimum
2923 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2924 program crashes, or just wrong answers. This did not apply to negated classes
2925 containing more than one character, or to minima other than one.
2926
2927
2928 Version 1.08 27-Mar-98
2929 ----------------------
2930
2931 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2932
2933 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2934 latter must appear before anything that relies on it in the pattern.
2935
2936
2937 Version 1.07 16-Feb-98
2938 ----------------------
2939
2940 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2941 repeat of a potentially empty string).
2942
2943
2944 Version 1.06 23-Jan-98
2945 ----------------------
2946
2947 1. Added Markus Oberhumer's little patches for C++.
2948
2949 2. Literal strings longer than 255 characters were broken.
2950
2951
2952 Version 1.05 23-Dec-97
2953 ----------------------
2954
2955 1. Negated character classes containing more than one character were failing if
2956 PCRE_CASELESS was set at run time.
2957
2958
2959 Version 1.04 19-Dec-97
2960 ----------------------
2961
2962 1. Corrected the man page, where some "const" qualifiers had been omitted.
2963
2964 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2965 input syntax.
2966
2967 3. Fixed memory leak which occurred when a regex with back references was
2968 matched with an offsets vector that wasn't big enough. The temporary memory
2969 that is used in this case wasn't being freed if the match failed.
2970
2971 4. Tidied pcretest to ensure it frees memory that it gets.
2972
2973 5. Temporary memory was being obtained in the case where the passed offsets
2974 vector was exactly big enough.
2975
2976 6. Corrected definition of offsetof() from change 5 below.
2977
2978 7. I had screwed up change 6 below and broken the rules for the use of
2979 setjmp(). Now fixed.
2980
2981
2982 Version 1.03 18-Dec-97
2983 ----------------------
2984
2985 1. A erroneous regex with a missing opening parenthesis was correctly
2986 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2987 on some systems.
2988
2989 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2990 it was reported that one broken compiler failed on the former because "code" is
2991 also an independent variable.
2992
2993 3. The erroneous regex a[]b caused an array overrun reference.
2994
2995 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2996 fail on data ending with that character. (It was going on too far, and checking
2997 the next character, typically a binary zero.) This was specific to the
2998 optimized code for single-character negative classes.
2999
3000 5. Added a contributed patch from the TIN world which does the following:
3001
3002 + Add an undef for memmove, in case the the system defines a macro for it.
3003
3004 + Add a definition of offsetof(), in case there isn't one. (I don't know
3005 the reason behind this - offsetof() is part of the ANSI standard - but
3006 it does no harm).
3007
3008 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3009 most of the places where whitespace preceded '#'. I have given up and
3010 allowed the remaining 2 cases to be at the margin.
3011
3012 + Rename some variables in pcre to eliminate shadowing. This seems very
3013 pedantic, but does no harm, of course.
3014
3015 6. Moved the call to setjmp() into its own function, to get rid of warnings
3016 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3017
3018 7. Constructs such as \d{8,} were compiling into the equivalent of
3019 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3020 outcome, but in this particular case used more store than had been allocated,
3021 which caused the bug to be discovered because it threw up an internal error.
3022
3023 8. The debugging code in both pcre and pcretest for outputting the compiled
3024 form of a regex was going wrong in the case of back references followed by
3025 curly-bracketed repeats.
3026
3027
3028 Version 1.02 12-Dec-97
3029 ----------------------
3030
3031 1. Typos in pcre.3 and comments in the source fixed.
3032
3033 2. Applied a contributed patch to get rid of places where it used to remove
3034 'const' from variables, and fixed some signed/unsigned and uninitialized
3035 variable warnings.
3036
3037 3. Added the "runtest" target to Makefile.
3038
3039 4. Set default compiler flag to -O2 rather than just -O.
3040
3041
3042 Version 1.01 19-Nov-97
3043 ----------------------
3044
3045 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3046 like /([ab]*)*/, that is, for classes with more than one character in them.
3047
3048 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3049 as /((?>a*))*/ (a PCRE_EXTRA facility).
3050
3051
3052 Version 1.00 18-Nov-97
3053 ----------------------
3054
3055 1. Added compile-time macros to support systems such as SunOS4 which don't have
3056 memmove() or strerror() but have other things that can be used instead.
3057
3058 2. Arranged that "make clean" removes the executables.
3059
3060
3061 Version 0.99 27-Oct-97
3062 ----------------------
3063
3064 1. Fixed bug in code for optimizing classes with only one character. It was
3065 initializing a 32-byte map regardless, which could cause it to run off the end
3066 of the memory it had got.
3067
3068 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3069
3070
3071 Version 0.98 22-Oct-97
3072 ----------------------
3073
3074 1. Fixed bug in code for handling temporary memory usage when there are more
3075 back references than supplied space in the ovector. This could cause segfaults.
3076
3077
3078 Version 0.97 21-Oct-97
3079 ----------------------
3080
3081 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3082
3083 2. Optimized negated single characters not to use a bit map.
3084
3085 3. Brought error texts together as macro definitions; clarified some of them;
3086 fixed one that was wrong - it said "range out of order" when it meant "invalid
3087 escape sequence".
3088
3089 4. Changed some char * arguments to const char *.
3090
3091 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3092
3093 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3094 pcretest.
3095
3096
3097 Version 0.96 16-Oct-97
3098 ----------------------
3099
3100 1. Added a simple "pgrep" utility to the distribution.
3101
3102 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3103 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3104 where "ddd" means "one or more decimal digits".
3105
3106 3. Fixed serious bug. If a pattern had a back reference, but the call to
3107 pcre_exec() didn't supply a large enough ovector to record the related
3108 identifying subpattern, the match always failed. PCRE now remembers the number
3109 of the largest back reference, and gets some temporary memory in which to save
3110 the offsets during matching if necessary, in order to ensure that
3111 backreferences always work.
3112
3113 4. Increased the compatibility with Perl in a number of ways:
3114
3115 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3116 to request this handling. The option can be set at compile or exec time.
3117
3118 (b) $ matches before a terminating newline by default; an option
3119 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3120 mode). The option can be set at compile or exec time.
3121
3122 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3123 the same as Perl's. If the decimal number it represents is less than 10
3124 or there aren't that many previous left capturing parentheses, an octal
3125 escape is read. Inside a character class, it's always an octal escape,
3126 even if it is a single digit.
3127
3128 (d) An escaped but undefined alphabetic character is taken as a literal,
3129 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3130 escapes.
3131
3132 (e) {0} is now permitted. (The previous item is removed from the compiled
3133 pattern).
3134
3135 5. Changed all the names of code files so that the basic parts are no longer
3136 than 10 characters, and abolished the teeny "globals.c" file.
3137
3138 6. Changed the handling of character classes; they are now done with a 32-byte
3139 bit map always.
3140
3141 7. Added the -d and /D options to pcretest to make it possible to look at the
3142 internals of compilation without having to recompile pcre.
3143
3144
3145 Version 0.95 23-Sep-97
3146 ----------------------
3147
3148 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3149 \x20 at the start of a run of normal characters. These were being treated as
3150 real characters, instead of the source characters being re-checked.
3151
3152
3153 Version 0.94 18-Sep-97
3154 ----------------------
3155
3156 1. The functions are now thread-safe, with the caveat that the global variables
3157 containing pointers to malloc() and free() or alternative functions are the
3158 same for all threads.
3159
3160 2. Get pcre_study() to generate a bitmap of initial characters for non-
3161 anchored patterns when this is possible, and use it if passed to pcre_exec().
3162
3163
3164 Version 0.93 15-Sep-97
3165 ----------------------
3166
3167 1. /(b)|(:+)/ was computing an incorrect first character.
3168
3169 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3170 but not actually doing anything yet.
3171
3172 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3173 as Perl does (e.g. [-az] or [az-]).
3174
3175 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3176 all possible positions.
3177
3178 5. Split up into different modules to avoid including unneeded functions in a
3179 compiled binary. However, compile and exec are still in one module. The "study"
3180 function is split off.
3181
3182 6. The character tables are now in a separate module whose source is generated
3183 by an auxiliary program - but can then be edited by hand if required. There are
3184 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3185 toupper() in the code.
3186
3187 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3188 make them global. Abolish the function for setting them, as the caller can now
3189 set them directly.
3190
3191
3192 Version 0.92 11-Sep-97
3193 ----------------------
3194
3195 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3196 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3197
3198 2. Caseless matching was not working in character classes if the characters in
3199 the pattern were in upper case.
3200
3201 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3202
3203 4. Make PCRE_ANCHORED public and accept as a compile option.
3204
3205 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3206 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3207 pass them.
3208
3209 6. Give an error if bad option bits passed at compile or run time.
3210
3211 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3212 pcretest to cause it to pass that flag.
3213
3214 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3215 options, and the first character, if set.
3216
3217 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3218
3219
3220 Version 0.91 10-Sep-97
3221 ----------------------
3222
3223 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3224 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3225
3226 2. PCRE was looping on encountering an indefinitely repeated back reference to
3227 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3228 Perl does - treats the match as successful.
3229
3230 ****

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12