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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 202 - (hide annotations) (download)
Fri Aug 3 09:44:26 2007 UTC (7 years ago) by ph10
File size: 133523 byte(s)
Refactor the integer overflow testing so as to avoid imposing an artificial 
limit on the size of subpatterns.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 189 Version 7.3 05-Jul-07
5 ph10 157 ---------------------
6    
7 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
8     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
9     brackets there is not right, since it causes compilers to look for an
10     installed pcre.h, not the version that is in the source that is being
11     compiled (which of course may be different). I have changed it back to:
12 ph10 197
13 ph10 189 #include "pcre.h"
14 ph10 197
15     I have a vague recollection that the change was concerned with compiling in
16     different directories, but in the new build system, that is taken care of
17     by the VPATH setting the Makefile.
18    
19 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
20     when the subject happened to end in the byte 0x85 (e.g. if the last
21     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
22     characters but of course it shouldn't be taken as a newline when it is part
23     of another character. The bug was that, for an unlimited repeat of . in
24     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
25     characters when looking for a newline.
26 ph10 197
27     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
28    
29     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
30     in debug output.
31    
32 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
33     long printing in the pcrecpp unittest when running under MinGW.
34 ph10 189
35 ph10 197 6. ESC_K was left out of the EBCDIC table.
36 ph10 189
37 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
38     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
39     limit also applies to "virtual nesting" when a pattern is recursive, and in
40     this case 1000 isn't so big. I have been able to remove this limit at the
41     expense of backing off one optimization in certain circumstances. Normally,
42     when pcre_exec() would call its internal match() function recursively and
43     immediately return the result unconditionally, it uses a "tail recursion"
44     feature to save stack. However, when a subpattern that can match an empty
45     string has an unlimited repetition quantifier, it no longer makes this
46     optimization. That gives it a stack frame in which to save the data for
47     checking that an empty string has been matched. Previously this was taken
48     from the 1000-entry workspace that had been reserved. So now there is no
49     explicit limit, but more stack is used.
50 ph10 189
51 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
52     syntax that is required for Windows, and which was going wrong for the
53     pcreposix and pcrecpp parts of the library. These were overlooked when this
54     problem was solved for the main library.
55 ph10 197
56 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
57     the size of patterns that contain repeated groups with explicit upper
58     limits. As the maximum quantifier is 65535, the maximum group length was
59     set at 30,000 so that the product of these two numbers did not overflow a
60     32-bit integer. However, it turns out that people want to use groups that
61     are longer than 30,000 bytes (though not repeat them that many times).
62     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
63     made it possible to implement the integer overflow checks in a much more
64     dynamic way, which I have now done. The artificial limitation on group
65     length has been removed - we now have only the limit on the total length of
66     the compiled pattern, which depends on the LINK_SIZE setting.
67 ph10 200
68 ph10 202
69 ph10 189 Version 7.2 19-Jun-07
70     ---------------------
71    
72 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
73     which is apparently normally available under Windows.
74 ph10 159
75     2. Re-jig the pcregrep tests with different newline settings in an attempt
76     to make them independent of the local environment's newline setting.
77 ph10 157
78 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
79 ph10 165
80 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
81     was not 2, because the output contained actual offsets. The recent new
82 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
83     usable with all link sizes.
84    
85 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
86     stack recursion. This gives a massive performance boost under BSD, but just
87 ph10 165 a small improvement under Linux. However, it saves one field in the frame
88 ph10 164 in all cases.
89 ph10 172
90 ph10 166 6. Added more features from the forthcoming Perl 5.10:
91 ph10 172
92 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
93     recursion call. It refers to the nth most recently opened parentheses.
94 ph10 172
95 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
96 ph10 172 to be opened parentheses.
97    
98     (c) Conditions that refer to capturing parentheses can be specified
99 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
100 ph10 172
101 ph10 168 (d) \K resets the start of the current match so that everything before
102 ph10 172 is not part of it.
103    
104 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
105 ph10 172
106 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
107 ph10 172 reference syntax.
108 ph10 182
109 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
110 ph10 182 alternative starts with the same number.
111 ph10 172
112 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
113    
114 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
115     PCRE_INFO_JCHANGED.
116    
117     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
118     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
119 ph10 170 for detecting groups that can match an empty string.
120 ph10 159
121 ph10 172 9. A pattern with a very large number of alternatives (more than several
122     hundred) was running out of internal workspace during the pre-compile
123     phase, where pcre_compile() figures out how much memory will be needed. A
124     bit of new cunning has reduced the workspace needed for groups with
125     alternatives. The 1000-alternative test pattern now uses 12 bytes of
126     workspace instead of running out of the 4096 that are available.
127 ph10 182
128 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
129 ph10 172
130 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
131     The report of the bug said:
132 ph10 182
133 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
134     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
135 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
136 ph10 185
137 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
138 ph10 185 it matched the wrong number of bytes.
139 ph10 172
140 ph10 179
141 ph10 155 Version 7.1 24-Apr-07
142 ph10 98 ---------------------
143    
144 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
145 ph10 122 that is more "standard", making use of automake and other Autotools. There
146 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
147     on this.
148 ph10 111
149     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
150     for recursive directory scanning broke on some systems because the files
151     are not scanned in any specific order and on different systems the order
152     was different. A call to "sort" has been inserted into RunGrepTest for the
153     approprate test as a short-term fix. In the longer term there may be an
154 ph10 100 alternative.
155 ph10 111
156 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
157 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
158     people do their own conversions for various reasons). The problems
159     concerned the use of low-level troff macros .br and .in. I have therefore
160     removed all such uses from the man pages (some were redundant, some could
161 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
162     HTML has been updated to handle .nf/.fi and to complain if it encounters
163     .br or .in.
164 ph10 111
165 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
166 ph10 123 arranged for config.h to be included in the distribution, with the name
167 ph10 111 config.h.generic, for the benefit of those who have to compile without
168     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
169    
170     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
171     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
172 ph10 127 makevp.bat for the current PCRE, using the additional files
173 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
174 ph10 111
175     6. A Windows user reported a minor discrepancy with test 2, which turned out
176     to be caused by a trailing space on an input line that had got lost in his
177 ph10 102 copy. The trailing space was an accident, so I've just removed it.
178 ph10 111
179 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
180 ph10 111 that is needed.
181    
182 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
183     as "const" (a) because they are and (b) because it helps the PHP
184     maintainers who have recently made a script to detect big data structures
185 ph10 111 in the php code that should be moved to the .rodata section. I remembered
186     to update Builducptable as well, so it won't revert if ucptable.h is ever
187 ph10 105 re-created.
188 ph10 111
189     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
190 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
191 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
192     support is not required. This saves 1.5-2K of code, which is important in
193 ph10 107 some applications.
194 ph10 111
195 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
196 ph10 111 so as not to refer to the tables, even though these functions will never be
197     called when UTF-8 support is disabled. Otherwise there are problems with a
198     shared library.
199    
200 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
201    
202     (a) It was defining its arguments as char * instead of void *.
203    
204 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
205     a long time ago when I wrote it, but is no longer the case.
206    
207 ph10 118 The emulated memove() is provided for those environments that have neither
208 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
209 ph10 118 is clearly not the case, as these two bugs were recently reported.
210 ph10 123
211 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
212 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
213     pages, and it removes trailing spaces from listed files. It also creates
214     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
215     case, it wraps all the #defines with #ifndefs. This script should be run
216 ph10 111 before "make dist".
217 ph10 123
218 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
219     with Unicode property support.
220 ph10 123
221     (a) For a maximizing quantifier, if the two different cases of the
222     character were of different lengths in their UTF-8 codings (there are
223     some cases like this - I found 11), and the matching function had to
224 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
225     were both the same length.
226 ph10 123
227     (b) When PCRE was configured to use the heap rather than the stack for
228     recursion during matching, it was not correctly preserving the data for
229     the other case of a UTF-8 character when checking ahead for a match
230     while processing a minimizing repeat. If the check also involved
231 ph10 115 matching a wide character, but failed, corruption could cause an
232     erroneous result when trying to check for a repeat of the original
233     character.
234 ph10 123
235 ph10 116 13. Some tidying changes to the testing mechanism:
236 ph10 98
237 ph10 116 (a) The RunTest script now detects the internal link size and whether there
238     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
239 ph10 123 values substituted by "configure". (The RunGrepTest script already did
240     this for UTF-8.) The configure.ac script no longer substitutes the
241     relevant variables.
242    
243 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
244     with length and offset values. This means that the output is different
245     for different internal link sizes. Test 2 is skipped for link sizes
246     other than 2 because of this, bypassing the problem. Unfortunately,
247     there was also a test in test 3 (the locale tests) that used /B and
248 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
249     I have added a new /Z option to pcretest that replaces the length and
250     offset values with spaces. This is now used to make test 3 independent
251 ph10 122 of link size. (Test 2 will be tidied up later.)
252 ph10 123
253     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
254 ph10 122 segmentation fault instead of returning the appropriate error message.
255 ph10 134
256 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
257 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
258     This doesn't seem right; it now treats the CRLF combination as the line
259     ending, and so does not match in that case. It's only a pattern such as ^$
260     that would hit this one: something like ^ABC$ would have failed after \r
261     and then tried again after \r\n.
262    
263 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
264 ph10 134 in an attempt to make files that differ only in their line terminators
265     compare equal. This works on Linux.
266 ph10 142
267 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
268     as it exited. This is now fixed, thanks to valgrind.
269 ph10 142
270 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
271 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
272     was because its rules for how to advance for /g after matching an empty
273 ph10 143 string at the end of a line did not allow for this case. They now check for
274     it specially.
275 ph10 150
276     20. pcretest is supposed to handle patterns and data of any length, by
277     extending its buffers when necessary. It was getting this wrong when the
278 ph10 147 buffer for a data line had to be extended.
279 ph10 150
280 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
281 ph10 150 CRLF as a newline sequence.
282 ph10 152
283 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
284 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
285     I have nevertheless tidied it up.
286 ph10 111
287 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
288 ph10 123
289 ph10 153 24. Added a man page for pcre-config.
290 ph10 152
291 ph10 153
292 nigel 93 Version 7.0 19-Dec-06
293     ---------------------
294    
295     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
296     moving to gcc 4.1.1.
297    
298     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
299     sys/time.h, which is documented as needed for this function. It doesn't
300     seem to matter on Linux, but it showed up on some releases of OS X.
301    
302     3. It seems that there are systems where bytes whose values are greater than
303     127 match isprint() in the "C" locale. The "C" locale should be the
304     default when a C program starts up. In most systems, only ASCII printing
305     characters match isprint(). This difference caused the output from pcretest
306     to vary, making some of the tests fail. I have changed pcretest so that:
307    
308     (a) When it is outputting text in the compiled version of a pattern, bytes
309     other than 32-126 are always shown as hex escapes.
310    
311     (b) When it is outputting text that is a matched part of a subject string,
312     it does the same, unless a different locale has been set for the match
313     (using the /L modifier). In this case, it uses isprint() to decide.
314    
315     4. Fixed a major bug that caused incorrect computation of the amount of memory
316     required for a compiled pattern when options that changed within the
317     pattern affected the logic of the preliminary scan that determines the
318     length. The relevant options are -x, and -i in UTF-8 mode. The result was
319     that the computed length was too small. The symptoms of this bug were
320     either the PCRE error "internal error: code overflow" from pcre_compile(),
321     or a glibc crash with a message such as "pcretest: free(): invalid next
322     size (fast)". Examples of patterns that provoked this bug (shown in
323     pcretest format) are:
324    
325     /(?-x: )/x
326     /(?x)(?-x: \s*#\s*)/
327     /((?i)[\x{c0}])/8
328     /(?i:[\x{c0}])/8
329    
330     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
331     is now done differently.
332    
333     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
334     wrapper classes; (b) implement a new function in the C++ scanner that is
335     more efficient than the old way of doing things because it avoids levels of
336     recursion in the regex matching; (c) add a paragraph to the documentation
337     for the FullMatch() function.
338    
339     6. The escape sequence \n was being treated as whatever was defined as
340     "newline". Not only was this contrary to the documentation, which states
341     that \n is character 10 (hex 0A), but it also went horribly wrong when
342     "newline" was defined as CRLF. This has been fixed.
343    
344     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
345     was being set to -1 for the "end of line" case (supposedly a value that no
346     character can have). Though this value is never used (the check for end of
347     line is "zero bytes in current character"), it caused compiler complaints.
348     I've changed it to 0xffffffff.
349    
350     8. In pcre_version.c, the version string was being built by a sequence of
351     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
352     string (as it is for production releases) called a macro with an empty
353     argument. The C standard says the result of this is undefined. The gcc
354     compiler treats it as an empty string (which was what was wanted) but it is
355     reported that Visual C gives an error. The source has been hacked around to
356     avoid this problem.
357    
358     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
359     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
360     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
361     of them did).
362    
363     10. Originally, pcretest opened its input and output without "b"; then I was
364     told that "b" was needed in some environments, so it was added for release
365     5.0 to both the input and output. (It makes no difference on Unix-like
366     systems.) Later I was told that it is wrong for the input on Windows. I've
367     now abstracted the modes into two macros, to make it easier to fiddle with
368     them, and removed "b" from the input mode under Windows.
369    
370     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
371    
372     12. Added -help and --help to pcretest as an official way of being reminded
373     of the options.
374    
375     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
376     and pcrecpp.cc because they annoy compilers at high warning levels.
377    
378     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
379    
380     15. Fixed an occurrence of == in configure.ac that should have been = (shell
381     scripts are not C programs :-) and which was not noticed because it works
382     on Linux.
383    
384     16. pcretest is supposed to handle any length of pattern and data line (as one
385     line or as a continued sequence of lines) by extending its input buffer if
386     necessary. This feature was broken for very long pattern lines, leading to
387     a string of junk being passed to pcre_compile() if the pattern was longer
388     than about 50K.
389    
390     17. I have done a major re-factoring of the way pcre_compile() computes the
391     amount of memory needed for a compiled pattern. Previously, there was code
392     that made a preliminary scan of the pattern in order to do this. That was
393     OK when PCRE was new, but as the facilities have expanded, it has become
394     harder and harder to keep it in step with the real compile phase, and there
395     have been a number of bugs (see for example, 4 above). I have now found a
396     cunning way of running the real compile function in a "fake" mode that
397     enables it to compute how much memory it would need, while actually only
398     ever using a few hundred bytes of working memory and without too many
399     tests of the mode. This should make future maintenance and development
400     easier. A side effect of this work is that the limit of 200 on the nesting
401     depth of parentheses has been removed (though this was never a serious
402     limitation, I suspect). However, there is a downside: pcre_compile() now
403     runs more slowly than before (30% or more, depending on the pattern). I
404     hope this isn't a big issue. There is no effect on runtime performance.
405    
406     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
407     newline (only possible for the last line of a file) and it was a
408     pattern that set a locale (followed by /Lsomething), pcretest crashed.
409    
410     19. Added additional timing features to pcretest. (1) The -tm option now times
411     matching only, not compiling. (2) Both -t and -tm can be followed, as a
412     separate command line item, by a number that specifies the number of
413     repeats to use when timing. The default is 50000; this gives better
414     precision, but takes uncomfortably long for very large patterns.
415    
416     20. Extended pcre_study() to be more clever in cases where a branch of a
417     subpattern has no definite first character. For example, (a*|b*)[cd] would
418     previously give no result from pcre_study(). Now it recognizes that the
419     first character must be a, b, c, or d.
420    
421     21. There was an incorrect error "recursive call could loop indefinitely" if
422     a subpattern (or the entire pattern) that was being tested for matching an
423     empty string contained only one non-empty item after a nested subpattern.
424     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
425     incorrectly, because the \d was being skipped in the check.
426    
427     22. The pcretest program now has a new pattern option /B and a command line
428     option -b, which is equivalent to adding /B to every pattern. This causes
429     it to show the compiled bytecode, without the additional information that
430     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
431     is the same as /B/I).
432    
433     23. A new optimization is now able automatically to treat some sequences such
434     as a*b as a*+b. More specifically, if something simple (such as a character
435     or a simple class like \d) has an unlimited quantifier, and is followed by
436     something that cannot possibly match the quantified thing, the quantifier
437     is automatically "possessified".
438    
439     24. A recursive reference to a subpattern whose number was greater than 39
440     went wrong under certain circumstances in UTF-8 mode. This bug could also
441     have affected the operation of pcre_study().
442    
443     25. Realized that a little bit of performance could be had by replacing
444     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
445    
446     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
447    
448     27. Possessive quantifiers such as a++ were previously implemented by turning
449     them into atomic groups such as ($>a+). Now they have their own opcodes,
450     which improves performance. This includes the automatically created ones
451     from 23 above.
452    
453     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
454     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
455     the first matched character to be a colon. This applied both to named and
456     numbered groups.
457    
458     29. The ucpinternal.h header file was missing its idempotency #ifdef.
459    
460     30. I was sent a "project" file called libpcre.a.dev which I understand makes
461     building PCRE on Windows easier, so I have included it in the distribution.
462    
463     31. There is now a check in pcretest against a ridiculously large number being
464     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
465     loop, the loop is abandoned.
466    
467     32. Forward references to subpatterns in conditions such as (?(2)...) where
468     subpattern 2 is defined later cause pcre_compile() to search forwards in
469     the pattern for the relevant set of parentheses. This search went wrong
470     when there were unescaped parentheses in a character class, parentheses
471     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
472    
473     33. "Subroutine" calls and backreferences were previously restricted to
474     referencing subpatterns earlier in the regex. This restriction has now
475     been removed.
476    
477     34. Added a number of extra features that are going to be in Perl 5.10. On the
478     whole, these are just syntactic alternatives for features that PCRE had
479     previously implemented using the Python syntax or my own invention. The
480     other formats are all retained for compatibility.
481    
482     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
483     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
484     also .NET compatible.
485    
486     (b) A recursion or subroutine call to a named group can now be defined as
487     (?&name) as well as (?P>name).
488    
489     (c) A backreference to a named group can now be defined as \k<name> or
490     \k'name' as well as (?P=name). The new forms, as well as being in Perl
491     5.10, are also .NET compatible.
492    
493     (d) A conditional reference to a named group can now use the syntax
494     (?(<name>) or (?('name') as well as (?(name).
495    
496     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
497     groups (named and numbered) that are never evaluated inline, but can be
498     called as "subroutines" from elsewhere. In effect, the DEFINE condition
499     is always false. There may be only one alternative in such a group.
500    
501     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
502     as the simple (?(R). The condition is true only if the most recent
503     recursion is that of the given number or name. It does not search out
504     through the entire recursion stack.
505    
506     (g) The escape \gN or \g{N} has been added, where N is a positive or
507     negative number, specifying an absolute or relative reference.
508    
509     35. Tidied to get rid of some further signed/unsigned compiler warnings and
510     some "unreachable code" warnings.
511    
512     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
513     things, this adds five new scripts.
514    
515     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
516     There were also incompatibilities regarding the handling of \Q..\E inside
517     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
518     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
519    
520     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
521     matches an empty string, and forcibly breaks the loop. There were bugs in
522     this code in non-simple cases. For a pattern such as ^(a()*)* matched
523     against aaaa the result was just "a" rather than "aaaa", for example. Two
524     separate and independent bugs (that affected different cases) have been
525     fixed.
526    
527     39. Refactored the code to abolish the use of different opcodes for small
528     capturing bracket numbers. This is a tidy that I avoided doing when I
529     removed the limit on the number of capturing brackets for 3.5 back in 2001.
530     The new approach is not only tidier, it makes it possible to reduce the
531     memory needed to fix the previous bug (38).
532    
533     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
534     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
535     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
536     mode.
537    
538     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
539     report.
540    
541     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
542     copy construction and assignment in the C++ wrapper.
543    
544     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
545     couple of bugs that could have given wrong results in the "--newline=crlf"
546     case.
547    
548     44. Added a number of casts and did some reorganization of signed/unsigned int
549     variables following suggestions from Dair Grant. Also renamed the variable
550     "this" as "item" because it is a C++ keyword.
551    
552     45. Arranged for dftables to add
553    
554     #include "pcre_internal.h"
555    
556     to pcre_chartables.c because without it, gcc 4.x may remove the array
557     definition from the final binary if PCRE is built into a static library and
558     dead code stripping is activated.
559    
560     46. For an unanchored pattern, if a match attempt fails at the start of a
561     newline sequence, and the newline setting is CRLF or ANY, and the next two
562     characters are CRLF, advance by two characters instead of one.
563    
564    
565 nigel 91 Version 6.7 04-Jul-06
566     ---------------------
567    
568     1. In order to handle tests when input lines are enormously long, pcretest has
569     been re-factored so that it automatically extends its buffers when
570     necessary. The code is crude, but this _is_ just a test program. The
571     default size has been increased from 32K to 50K.
572    
573     2. The code in pcre_study() was using the value of the re argument before
574     testing it for NULL. (Of course, in any sensible call of the function, it
575     won't be NULL.)
576    
577     3. The memmove() emulation function in pcre_internal.h, which is used on
578     systems that lack both memmove() and bcopy() - that is, hardly ever -
579     was missing a "static" storage class specifier.
580    
581     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
582     containing an extended class (one that cannot be represented by a bitmap
583     because it contains high-valued characters or Unicode property items, e.g.
584     [\pZ]). Almost always one would set UTF-8 mode when processing such a
585     pattern, but PCRE should not loop if you do not (it no longer does).
586     [Detail: two cases were found: (a) a repeated subpattern containing an
587     extended class; (b) a recursive reference to a subpattern that followed a
588     previous extended class. It wasn't skipping over the extended class
589     correctly when UTF-8 mode was not set.]
590    
591     5. A negated single-character class was not being recognized as fixed-length
592     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
593     compile error "lookbehind assertion is not fixed length".
594    
595     6. The RunPerlTest auxiliary script was showing an unexpected difference
596     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
597     write a Perl script that can interpret lines of an input file either as
598     byte characters or as UTF-8, which is what "perltest" was being required to
599     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
600     can't do is switch easily at run time between having the "use utf8;" pragma
601     or not. In the end, I fudged it by using the RunPerlTest script to insert
602     "use utf8;" explicitly for the UTF-8 tests.
603    
604     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
605     the end of the subject string, contrary to the documentation and to what
606     Perl does. This was true of both matching functions. Now it matches only at
607     the start of the subject and immediately after *internal* newlines.
608    
609     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
610     a pointer to an int instead of a pointer to an unsigned long int. This
611     caused problems on 64-bit systems.
612    
613     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
614     instance of the 'standard' template library not being so standard".
615    
616     10. There was no check on the number of named subpatterns nor the maximum
617     length of a subpattern name. The product of these values is used to compute
618     the size of the memory block for a compiled pattern. By supplying a very
619     long subpattern name and a large number of named subpatterns, the size
620     computation could be caused to overflow. This is now prevented by limiting
621     the length of names to 32 characters, and the number of named subpatterns
622     to 10,000.
623    
624     11. Subpatterns that are repeated with specific counts have to be replicated in
625     the compiled pattern. The size of memory for this was computed from the
626     length of the subpattern and the repeat count. The latter is limited to
627     65535, but there was no limit on the former, meaning that integer overflow
628     could in principle occur. The compiled length of a repeated subpattern is
629     now limited to 30,000 bytes in order to prevent this.
630    
631     12. Added the optional facility to have named substrings with the same name.
632    
633     13. Added the ability to use a named substring as a condition, using the
634     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
635     are numbers (not recommended). Forward references are permitted.
636    
637     14. Added forward references in named backreferences (if you see what I mean).
638    
639     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
640     pattern could run off the end of the subject. For example, the pattern
641     "(?s)(.{1,5})"8 did this with the subject "ab".
642    
643     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
644     PCRE_CASELESS was set when matching characters that were quantified with ?
645     or *.
646    
647     17. A character class other than a single negated character that had a minimum
648     but no maximum quantifier - for example [ab]{6,} - was not handled
649     correctly by pce_dfa_exec(). It would match only one character.
650    
651     18. A valid (though odd) pattern that looked like a POSIX character
652     class but used an invalid character after [ (for example [[,abc,]]) caused
653     pcre_compile() to give the error "Failed: internal error: code overflow" or
654     in some cases to crash with a glibc free() error. This could even happen if
655     the pattern terminated after [[ but there just happened to be a sequence of
656     letters, a binary zero, and a closing ] in the memory that followed.
657    
658     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
659     over the years. Originally (before any Unicode support), just the bottom 8
660     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
661     output from "man perlunicode" includes this:
662    
663     The regular expression compiler produces polymorphic opcodes. That
664     is, the pattern adapts to the data and automatically switches to
665     the Unicode character scheme when presented with Unicode data--or
666     instead uses a traditional byte scheme when presented with byte
667     data.
668    
669     Sadly, a wide octal escape does not cause a switch, and in a string with
670     no other multibyte characters, these octal escapes are treated as before.
671     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
672     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
673     Unicode string.
674    
675     I have not perpetrated such confusion in PCRE. Up till now, it took just
676     the bottom 8 bits, as in old Perl. I have now made octal escapes with
677     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
678     translate to the appropriate multibyte character.
679    
680     29. Applied some refactoring to reduce the number of warnings from Microsoft
681     and Borland compilers. This has included removing the fudge introduced
682     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
683     a warning about an unused variable.
684    
685     21. PCRE has not included VT (character 0x0b) in the set of whitespace
686     characters since release 4.0, because Perl (from release 5.004) does not.
687     [Or at least, is documented not to: some releases seem to be in conflict
688     with the documentation.] However, when a pattern was studied with
689     pcre_study() and all its branches started with \s, PCRE still included VT
690     as a possible starting character. Of course, this did no harm; it just
691     caused an unnecessary match attempt.
692    
693     22. Removed a now-redundant internal flag bit that recorded the fact that case
694     dependency changed within the pattern. This was once needed for "required
695     byte" processing, but is no longer used. This recovers a now-scarce options
696     bit. Also moved the least significant internal flag bit to the most-
697     significant bit of the word, which was not previously used (hangover from
698     the days when it was an int rather than a uint) to free up another bit for
699     the future.
700    
701     23. Added support for CRLF line endings as well as CR and LF. As well as the
702     default being selectable at build time, it can now be changed at runtime
703     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
704     specify that it is scanning data with non-default line endings.
705    
706     24. Changed the definition of CXXLINK to make it agree with the definition of
707     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
708    
709     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
710     recursions. This makes a big different to stack usage for some patterns.
711    
712     26. If a subpattern containing a named recursion or subroutine reference such
713     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
714     the space required for the compiled pattern went wrong and gave too small a
715     value. Depending on the environment, this could lead to "Failed: internal
716     error: code overflow at offset 49" or "glibc detected double free or
717     corruption" errors.
718    
719     27. Applied patches from Google (a) to support the new newline modes and (b) to
720     advance over multibyte UTF-8 characters in GlobalReplace.
721    
722     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
723     difference for some implementation of PCRE in some Windows version.
724    
725     29. Added some extra testing facilities to pcretest:
726    
727     \q<number> in a data line sets the "match limit" value
728     \Q<number> in a data line sets the "match recursion limt" value
729     -S <number> sets the stack size, where <number> is in megabytes
730    
731     The -S option isn't available for Windows.
732    
733    
734 nigel 89 Version 6.6 06-Feb-06
735     ---------------------
736    
737     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
738     in pcreposix.h. I have copied the definition from pcre.h.
739    
740     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
741     because pcre.h is no longer a built file.
742    
743     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
744     not normally included in the compiled code.
745    
746    
747 nigel 87 Version 6.5 01-Feb-06
748     ---------------------
749    
750     1. When using the partial match feature with pcre_dfa_exec(), it was not
751     anchoring the second and subsequent partial matches at the new starting
752     point. This could lead to incorrect results. For example, with the pattern
753     /1234/, partially matching against "123" and then "a4" gave a match.
754    
755     2. Changes to pcregrep:
756    
757     (a) All non-match returns from pcre_exec() were being treated as failures
758     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
759     error message is output. Some extra information is given for the
760     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
761     probably the only errors that are likely to be caused by users (by
762     specifying a regex that has nested indefinite repeats, for instance).
763     If there are more than 20 of these errors, pcregrep is abandoned.
764    
765     (b) A binary zero was treated as data while matching, but terminated the
766     output line if it was written out. This has been fixed: binary zeroes
767     are now no different to any other data bytes.
768    
769     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
770     used to set a locale for matching. The --locale=xxxx long option has
771     been added (no short equivalent) to specify a locale explicitly on the
772     pcregrep command, overriding the environment variables.
773    
774     (d) When -B was used with -n, some line numbers in the output were one less
775     than they should have been.
776    
777     (e) Added the -o (--only-matching) option.
778    
779     (f) If -A or -C was used with -c (count only), some lines of context were
780     accidentally printed for the final match.
781    
782     (g) Added the -H (--with-filename) option.
783    
784     (h) The combination of options -rh failed to suppress file names for files
785     that were found from directory arguments.
786    
787     (i) Added the -D (--devices) and -d (--directories) options.
788    
789     (j) Added the -F (--fixed-strings) option.
790    
791     (k) Allow "-" to be used as a file name for -f as well as for a data file.
792    
793     (l) Added the --colo(u)r option.
794    
795     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
796     is not present by default.
797    
798     3. A nasty bug was discovered in the handling of recursive patterns, that is,
799     items such as (?R) or (?1), when the recursion could match a number of
800     alternatives. If it matched one of the alternatives, but subsequently,
801     outside the recursion, there was a failure, the code tried to back up into
802     the recursion. However, because of the way PCRE is implemented, this is not
803     possible, and the result was an incorrect result from the match.
804    
805     In order to prevent this happening, the specification of recursion has
806     been changed so that all such subpatterns are automatically treated as
807     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
808    
809     4. I had overlooked the fact that, in some locales, there are characters for
810     which isalpha() is true but neither isupper() nor islower() are true. In
811     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
812     and ordfeminine) are like this. This affected the treatment of \w and \W
813     when they appeared in character classes, but not when they appeared outside
814     a character class. The bit map for "word" characters is now created
815     separately from the results of isalnum() instead of just taking it from the
816     upper, lower, and digit maps. (Plus the underscore character, of course.)
817    
818     5. The above bug also affected the handling of POSIX character classes such as
819     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
820     permanent tables. Instead, the bit maps for such a class were previously
821     created as the appropriate unions of the upper, lower, and digit bitmaps.
822     Now they are created by subtraction from the [[:word:]] class, which has
823     its own bitmap.
824    
825     6. The [[:blank:]] character class matches horizontal, but not vertical space.
826     It is created by subtracting the vertical space characters (\x09, \x0a,
827     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
828     subtraction was done in the overall bitmap for a character class, meaning
829     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
830     be recognized. This bug has been fixed.
831    
832     7. Patches from the folks at Google:
833    
834     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
835     real life, but is still worth protecting against".
836    
837     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
838     regular expressions".
839    
840     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
841     have it.
842    
843     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
844     "configure" and the latter not, in order to fix a problem somebody had
845     with compiling the Arg class on HP-UX.
846    
847     (e) Improve the error-handling of the C++ wrapper a little bit.
848    
849     (f) New tests for checking recursion limiting.
850    
851     8. The pcre_memmove() function, which is used only if the environment does not
852     have a standard memmove() function (and is therefore rarely compiled),
853     contained two bugs: (a) use of int instead of size_t, and (b) it was not
854     returning a result (though PCRE never actually uses the result).
855    
856     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
857     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
858     returned instead of calling malloc() with an overflowing number that would
859     most likely cause subsequent chaos.
860    
861     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
862    
863     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
864     with this option is matched, the nmatch and pmatch options of regexec() are
865     ignored.
866    
867     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
868     provided in case anyone wants to the the POSIX interface with UTF-8
869     strings.
870    
871     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
872     C++ linking (needed for some HP-UX environments).
873    
874     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
875     (unused parameter) and in the pcre_printint() function (omitted "default"
876     switch label when the default is to do nothing).
877    
878     15. Added some code to make it possible, when PCRE is compiled as a C++
879     library, to replace subject pointers for pcre_exec() with a smart pointer
880     class, thus making it possible to process discontinuous strings.
881    
882     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
883     much the same function. They were added by different people who were trying
884     to make PCRE easy to compile on non-Unix systems. It has been suggested
885     that PCRE_EXPORT be abolished now that there is more automatic apparatus
886     for compiling on Windows systems. I have therefore replaced it with
887     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
888     defaults to "extern" for C or "extern C" for C++, which works fine on
889     Unix-like systems. It is now possible to override the value of PCRE_DATA_
890     SCOPE with something explicit in config.h. In addition:
891    
892     (a) pcreposix.h still had just "extern" instead of either of these macros;
893     I have replaced it with PCRE_DATA_SCOPE.
894    
895     (b) Functions such as _pcre_xclass(), which are internal to the library,
896     but external in the C sense, all had PCRE_EXPORT in their definitions.
897     This is apparently wrong for the Windows case, so I have removed it.
898     (It makes no difference on Unix-like systems.)
899    
900     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
901     of recursive calls to match(). This is different to MATCH_LIMIT because
902     that limits the total number of calls to match(), not all of which increase
903     the depth of recursion. Limiting the recursion depth limits the amount of
904     stack (or heap if NO_RECURSE is set) that is used. The default can be set
905     when PCRE is compiled, and changed at run time. A patch from Google adds
906     this functionality to the C++ interface.
907    
908     18. Changes to the handling of Unicode character properties:
909    
910     (a) Updated the table to Unicode 4.1.0.
911    
912     (b) Recognize characters that are not in the table as "Cn" (undefined).
913    
914     (c) I revised the way the table is implemented to a much improved format
915     which includes recognition of ranges. It now supports the ranges that
916     are defined in UnicodeData.txt, and it also amalgamates other
917     characters into ranges. This has reduced the number of entries in the
918     table from around 16,000 to around 3,000, thus reducing its size
919     considerably. I realized I did not need to use a tree structure after
920     all - a binary chop search is just as efficient. Having reduced the
921     number of entries, I extended their size from 6 bytes to 8 bytes to
922     allow for more data.
923    
924     (d) Added support for Unicode script names via properties such as \p{Han}.
925    
926     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
927     matching that character.
928    
929     20. When matching a repeated Unicode property with a minimum greater than zero,
930     (for example \pL{2,}), PCRE could look past the end of the subject if it
931     reached it while seeking the minimum number of characters. This could
932     happen only if some of the characters were more than one byte long, because
933     there is a check for at least the minimum number of bytes.
934    
935     21. Refactored the implementation of \p and \P so as to be more general, to
936     allow for more different types of property in future. This has changed the
937     compiled form incompatibly. Anybody with saved compiled patterns that use
938     \p or \P will have to recompile them.
939    
940     22. Added "Any" and "L&" to the supported property types.
941    
942     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
943     but give a compile time error if the value is greater than 0xff.
944    
945     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
946     accidentally not being installed or uninstalled.
947    
948     25. The pcre.h file was built from pcre.h.in, but the only changes that were
949     made were to insert the current release number. This seemed silly, because
950     it made things harder for people building PCRE on systems that don't run
951     "configure". I have turned pcre.h into a distributed file, no longer built
952     by "configure", with the version identification directly included. There is
953     no longer a pcre.h.in file.
954    
955     However, this change necessitated a change to the pcre-config script as
956     well. It is built from pcre-config.in, and one of the substitutions was the
957     release number. I have updated configure.ac so that ./configure now finds
958     the release number by grepping pcre.h.
959    
960     26. Added the ability to run the tests under valgrind.
961    
962    
963 nigel 85 Version 6.4 05-Sep-05
964     ---------------------
965    
966     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
967     "--" to be printed when multiple files were scanned, even when none of the
968     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
969     consider it to be a bug, and have restored the previous behaviour.
970    
971     2. A couple of code tidies to get rid of compiler warnings.
972    
973     3. The pcretest program used to cheat by referring to symbols in the library
974     whose names begin with _pcre_. These are internal symbols that are not
975     really supposed to be visible externally, and in some environments it is
976     possible to suppress them. The cheating is now confined to including
977     certain files from the library's source, which is a bit cleaner.
978    
979     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
980     file's purpose clearer.
981    
982     5. Reorganized pcre_ucp_findchar().
983    
984    
985 nigel 83 Version 6.3 15-Aug-05
986     ---------------------
987    
988     1. The file libpcre.pc.in did not have general read permission in the tarball.
989    
990     2. There were some problems when building without C++ support:
991    
992     (a) If C++ support was not built, "make install" and "make test" still
993     tried to test it.
994    
995     (b) There were problems when the value of CXX was explicitly set. Some
996     changes have been made to try to fix these, and ...
997    
998     (c) --disable-cpp can now be used to explicitly disable C++ support.
999    
1000     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1001     backslash in a target when C++ was disabled. This confuses some
1002     versions of "make", apparently. Using an intermediate variable solves
1003     this. (Same for CPP_LOBJ.)
1004    
1005     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1006     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1007     necessary on certain architectures.
1008    
1009     4. Added a setting of -export-symbols-regex to the link command to remove
1010     those symbols that are exported in the C sense, but actually are local
1011     within the library, and not documented. Their names all begin with
1012     "_pcre_". This is not a perfect job, because (a) we have to except some
1013     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1014     available (and never for static libraries). I have made a note to try to
1015     find a way round (a) in the future.
1016    
1017    
1018 nigel 81 Version 6.2 01-Aug-05
1019     ---------------------
1020    
1021     1. There was no test for integer overflow of quantifier values. A construction
1022     such as {1111111111111111} would give undefined results. What is worse, if
1023     a minimum quantifier for a parenthesized subpattern overflowed and became
1024     negative, the calculation of the memory size went wrong. This could have
1025     led to memory overwriting.
1026    
1027     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1028    
1029     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1030     operating environments where this matters.
1031    
1032     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1033     PCRE options from within the C++ wrapper.
1034    
1035     5. Named capturing subpatterns were not being correctly counted when a pattern
1036     was compiled. This caused two problems: (a) If there were more than 100
1037     such subpatterns, the calculation of the memory needed for the whole
1038     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1039     back references of the form \12, where the number was greater than 9, were
1040     not recognized as back references, even though there were sufficient
1041     previous subpatterns.
1042    
1043     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1044     versions of gcc, e.g. 2.95.4.
1045    
1046    
1047 nigel 79 Version 6.1 21-Jun-05
1048     ---------------------
1049    
1050     1. There was one reference to the variable "posix" in pcretest.c that was not
1051     surrounded by "#if !defined NOPOSIX".
1052    
1053     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1054     the cross-check on the old pcre_info() function, for the benefit of the
1055     cut-down version of PCRE that is currently imported into Exim.
1056    
1057     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1058     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1059     patterns that start (?i) but I don't think that matters. The use of (?i) is
1060     just an example; this all applies to the other options as well.
1061    
1062     4. Since libtool seems to echo the compile commands it is issuing, the output
1063     from "make" can be reduced a bit by putting "@" in front of each libtool
1064     compile command.
1065    
1066     5. Patch from the folks at Google for configure.in to be a bit more thorough
1067     in checking for a suitable C++ installation before trying to compile the
1068     C++ stuff. This should fix a reported problem when a compiler was present,
1069     but no suitable headers.
1070    
1071     6. The man pages all had just "PCRE" as their title. I have changed them to
1072     be the relevant file name. I have also arranged that these names are
1073     retained in the file doc/pcre.txt, which is a concatenation in text format
1074     of all the man pages except the little individual ones for each function.
1075    
1076     7. The NON-UNIX-USE file had not been updated for the different set of source
1077     files that come with release 6. I also added a few comments about the C++
1078     wrapper.
1079    
1080    
1081 nigel 77 Version 6.0 07-Jun-05
1082     ---------------------
1083    
1084     1. Some minor internal re-organization to help with my DFA experiments.
1085    
1086     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1087     didn't matter for the library itself when fully configured, but did matter
1088     when compiling without UCP support, or within Exim, where the ucp files are
1089     not imported.
1090    
1091     3. Refactoring of the library code to split up the various functions into
1092     different source modules. The addition of the new DFA matching code (see
1093     below) to a single monolithic source would have made it really too
1094     unwieldy, quite apart from causing all the code to be include in a
1095     statically linked application, when only some functions are used. This is
1096     relevant even without the DFA addition now that patterns can be compiled in
1097     one application and matched in another.
1098    
1099     The downside of splitting up is that there have to be some external
1100     functions and data tables that are used internally in different modules of
1101     the library but which are not part of the API. These have all had their
1102     names changed to start with "_pcre_" so that they are unlikely to clash
1103     with other external names.
1104    
1105     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1106     a different (DFA) algorithm. Although it is slower than the original
1107     function, it does have some advantages for certain types of matching
1108     problem.
1109    
1110     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1111     including restarting after a partial match.
1112    
1113     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1114     defined when compiling for Windows was sent to me. I have put it into the
1115     code, though I have no means of testing or verifying it.
1116    
1117     7. Added the pcre_refcount() auxiliary function.
1118    
1119     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1120     match before or at the first newline in the subject string. In pcretest,
1121     the /f option on a pattern can be used to set this.
1122    
1123     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1124     would behave wrongly. This has been present in PCRE since release 4.0.
1125    
1126     10. A number of changes to the pcregrep command:
1127    
1128     (a) Refactored how -x works; insert ^(...)$ instead of setting
1129     PCRE_ANCHORED and checking the length, in preparation for adding
1130     something similar for -w.
1131    
1132     (b) Added the -w (match as a word) option.
1133    
1134     (c) Refactored the way lines are read and buffered so as to have more
1135     than one at a time available.
1136    
1137     (d) Implemented a pcregrep test script.
1138    
1139     (e) Added the -M (multiline match) option. This allows patterns to match
1140     over several lines of the subject. The buffering ensures that at least
1141     8K, or the rest of the document (whichever is the shorter) is available
1142     for matching (and similarly the previous 8K for lookbehind assertions).
1143    
1144     (f) Changed the --help output so that it now says
1145    
1146     -w, --word-regex(p)
1147    
1148     instead of two lines, one with "regex" and the other with "regexp"
1149     because that confused at least one person since the short forms are the
1150     same. (This required a bit of code, as the output is generated
1151     automatically from a table. It wasn't just a text change.)
1152    
1153     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1154     option but starts with a hyphen. Could be a pattern or a path name
1155     starting with a hyphen, for instance.
1156    
1157     (h) "-" can be given as a file name to represent stdin.
1158    
1159     (i) When file names are being printed, "(standard input)" is used for
1160     the standard input, for compatibility with GNU grep. Previously
1161     "<stdin>" was used.
1162    
1163     (j) The option --label=xxx can be used to supply a name to be used for
1164     stdin when file names are being printed. There is no short form.
1165    
1166     (k) Re-factored the options decoding logic because we are going to add
1167     two more options that take data. Such options can now be given in four
1168     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1169    
1170     (l) Added the -A, -B, and -C options for requesting that lines of context
1171     around matches be printed.
1172    
1173     (m) Added the -L option to print the names of files that do not contain
1174     any matching lines, that is, the complement of -l.
1175    
1176     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1177     continue to scan other files.
1178    
1179     (o) The -s option was incorrectly implemented. For compatibility with other
1180     greps, it now suppresses the error message for a non-existent or non-
1181     accessible file (but not the return code). There is a new option called
1182     -q that suppresses the output of matching lines, which was what -s was
1183     previously doing.
1184    
1185     (p) Added --include and --exclude options to specify files for inclusion
1186     and exclusion when recursing.
1187    
1188     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1189     Hopefully, it now does.
1190    
1191     12. Missing cast in pcre_study().
1192    
1193     13. Added an "uninstall" target to the makefile.
1194    
1195     14. Replaced "extern" in the function prototypes in Makefile.in with
1196     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1197     world, but is set differently for Windows.
1198    
1199     15. Added a second compiling function called pcre_compile2(). The only
1200     difference is that it has an extra argument, which is a pointer to an
1201     integer error code. When there is a compile-time failure, this is set
1202     non-zero, in addition to the error test pointer being set to point to an
1203     error message. The new argument may be NULL if no error number is required
1204     (but then you may as well call pcre_compile(), which is now just a
1205     wrapper). This facility is provided because some applications need a
1206     numeric error indication, but it has also enabled me to tidy up the way
1207     compile-time errors are handled in the POSIX wrapper.
1208    
1209     16. Added VPATH=.libs to the makefile; this should help when building with one
1210     prefix path and installing with another. (Or so I'm told by someone who
1211     knows more about this stuff than I do.)
1212    
1213     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1214     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1215     match everything, including newlines. This is not POSIX-compatible, but
1216     somebody wanted the feature. From pcretest it can be activated by using
1217     both the P and the s flags.
1218    
1219     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1220    
1221     19. libpcre.pc was being incorrectly installed as executable.
1222    
1223     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1224     it now also looks for '\r' so that it will work unmodified on Windows.
1225    
1226     21. Added Google's contributed C++ wrapper to the distribution.
1227    
1228     22. Added some untidy missing memory free() calls in pcretest, to keep
1229     Electric Fence happy when testing.
1230    
1231    
1232    
1233 nigel 75 Version 5.0 13-Sep-04
1234     ---------------------
1235    
1236     1. Internal change: literal characters are no longer packed up into items
1237     containing multiple characters in a single byte-string. Each character
1238     is now matched using a separate opcode. However, there may be more than one
1239     byte in the character in UTF-8 mode.
1240    
1241     2. The pcre_callout_block structure has two new fields: pattern_position and
1242     next_item_length. These contain the offset in the pattern to the next match
1243     item, and its length, respectively.
1244    
1245     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1246     insertion of callouts before each pattern item. Added the /C option to
1247     pcretest to make use of this.
1248    
1249     4. On the advice of a Windows user, the lines
1250    
1251     #if defined(_WIN32) || defined(WIN32)
1252     _setmode( _fileno( stdout ), 0x8000 );
1253     #endif /* defined(_WIN32) || defined(WIN32) */
1254    
1255     have been added to the source of pcretest. This apparently does useful
1256     magic in relation to line terminators.
1257    
1258     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1259     for the benefit of those environments where the "b" makes a difference.
1260    
1261     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1262     to know about it. I have put a hack into configure.in that adds in code
1263     to set GCC=yes if CC=icc. This seems to end up at a point in the
1264     generated configure script that is early enough to affect the setting of
1265     compiler options, which is what is needed, but I have no means of testing
1266     whether it really works. (The user who reported this had patched the
1267     generated configure script, which of course I cannot do.)
1268    
1269     LATER: After change 22 below (new libtool files), the configure script
1270     seems to know about icc (and also ecc). Therefore, I have commented out
1271     this hack in configure.in.
1272    
1273     7. Added support for pkg-config (2 patches were sent in).
1274    
1275     8. Negated POSIX character classes that used a combination of internal tables
1276     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1277     [[:^ascii]]. Typically, they would match almost any characters. The other
1278     POSIX classes were not broken in this way.
1279    
1280     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1281     to find the match, as PCRE was deluded into thinking that the match had to
1282     start at the start point or following a newline. The same bug applied to
1283     patterns with negative forward assertions or any backward assertions
1284     preceding ".*" at the start, unless the pattern required a fixed first
1285     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1286    
1287     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1288     starting at the last subject character, bytes beyond the end of the subject
1289     string were read.
1290    
1291     11. Renamed the variable "class" as "classbits" to make life easier for C++
1292     users. (Previously there was a macro definition, but it apparently wasn't
1293     enough.)
1294    
1295     12. Added the new field "tables" to the extra data so that tables can be passed
1296     in at exec time, or the internal tables can be re-selected. This allows
1297     a compiled regex to be saved and re-used at a later time by a different
1298     program that might have everything at different addresses.
1299    
1300     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1301     -R library as well as a -L library.
1302    
1303     14. The debugging options of pcretest (-d on the command line or D on a
1304     pattern) showed incorrect output for anything following an extended class
1305     that contained multibyte characters and which was followed by a quantifier.
1306    
1307     15. Added optional support for general category Unicode character properties
1308     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1309     support. It adds about 90K to the size of the library. The meanings of the
1310     inbuilt class escapes such as \d and \s have NOT been changed.
1311    
1312     16. Updated pcredemo.c to include calls to free() to release the memory for the
1313     compiled pattern.
1314    
1315     17. The generated file chartables.c was being created in the source directory
1316     instead of in the building directory. This caused the build to fail if the
1317     source directory was different from the building directory, and was
1318     read-only.
1319    
1320     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1321     file. No doubt somebody will tell me if they don't make sense... Also added
1322     Dan Mooney's comments about building on OpenVMS.
1323    
1324     19. Added support for partial matching via the PCRE_PARTIAL option for
1325     pcre_exec() and the \P data escape in pcretest.
1326    
1327     20. Extended pcretest with 3 new pattern features:
1328    
1329     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1330     write the compiled pattern to the file whose name is "rest-of-line".
1331     This is a straight binary dump of the data, with the saved pointer to
1332     the character tables forced to be NULL. The study data, if any, is
1333     written too. After writing, pcretest reads a new pattern.
1334    
1335     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1336     compiled pattern from the given file. There must not be any
1337     occurrences of "<" in the file name (pretty unlikely); if there are,
1338     pcretest will instead treat the initial "<" as a pattern delimiter.
1339     After reading in the pattern, pcretest goes on to read data lines as
1340     usual.
1341    
1342     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1343     and 16-bit fields in a compiled pattern, to simulate a pattern that
1344     was compiled on a host of opposite endianness.
1345    
1346     21. The pcre-exec() function can now cope with patterns that were compiled on
1347     hosts of opposite endianness, with this restriction:
1348    
1349     As for any compiled expression that is saved and used later, the tables
1350     pointer field cannot be preserved; the extra_data field in the arguments
1351     to pcre_exec() should be used to pass in a tables address if a value
1352     other than the default internal tables were used at compile time.
1353    
1354     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1355     now diagnosed as an error. Previously, most of the time, a negative number
1356     would have been treated as zero, but if in addition "ovector" was passed as
1357     NULL, a crash could occur.
1358    
1359     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1360     new versions from the libtool 1.5 distribution (the last one is a copy of
1361     a file called libtool.m4). This seems to have fixed the need to patch
1362     "configure" to support Darwin 1.3 (which I used to do). However, I still
1363     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1364     workstation).
1365    
1366     24. Changed the PCRE licence to be the more standard "BSD" licence.
1367    
1368    
1369 nigel 73 Version 4.5 01-Dec-03
1370     ---------------------
1371    
1372     1. There has been some re-arrangement of the code for the match() function so
1373     that it can be compiled in a version that does not call itself recursively.
1374     Instead, it keeps those local variables that need separate instances for
1375     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1376     needs to "recurse". Keeping track of where control must go is done by means
1377     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1378     hide most of the details from the main code, and operates only if
1379     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1380     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1381     operating.
1382    
1383     To make it easier for callers to provide specially tailored get/free
1384     functions for this usage, two new functions, pcre_stack_malloc, and
1385     pcre_stack_free, are used. They are always called in strict stacking order,
1386     and the size of block requested is always the same.
1387    
1388     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1389     PCRE has been compiled to use the stack or the heap for recursion. The
1390     -C option of pcretest uses this to show which version is compiled.
1391    
1392     A new data escape \S, is added to pcretest; it causes the amounts of store
1393     obtained and freed by both kinds of malloc/free at match time to be added
1394     to the output.
1395    
1396     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1397     what's available on my current Linux desktop machine.
1398    
1399     3. When matching a UTF-8 string, the test for a valid string at the start has
1400     been extended. If start_offset is not zero, PCRE now checks that it points
1401     to a byte that is the start of a UTF-8 character. If not, it returns
1402     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1403     this is necessary because there may be backward assertions in the pattern.
1404     When matching the same subject several times, it may save resources to use
1405     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1406    
1407     4. The code for checking the validity of UTF-8 strings has been tightened so
1408     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1409     containing "overlong sequences".
1410    
1411     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1412     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1413     should have been "&", but it just so happened that all the cases this let
1414     through by mistake were picked up later in the function.
1415    
1416     6. I had used a variable called "isblank" - this is a C99 function, causing
1417     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1418    
1419     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1420     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1421     so the version is shown; (c) stop "make test" echoing "./RunTest".
1422    
1423     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1424    
1425     9. The prototype for memmove() for systems that don't have it was using
1426     size_t, but the inclusion of the header that defines size_t was later. I've
1427     moved the #includes for the C headers earlier to avoid this.
1428    
1429     10. Added some adjustments to the code to make it easier to compiler on certain
1430     special systems:
1431    
1432     (a) Some "const" qualifiers were missing.
1433     (b) Added the macro EXPORT before all exported functions; by default this
1434     is defined to be empty.
1435     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1436     that it reads its output file name as an argument instead of writing
1437     to the standard output and assuming this can be redirected.
1438    
1439     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1440     class containing characters with values greater than 255, PCRE compilation
1441     went into a loop.
1442    
1443     12. A recursive reference to a subpattern that was within another subpattern
1444     that had a minimum quantifier of zero caused PCRE to crash. For example,
1445     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1446     recursion. If the recursively-called subpattern itself had a zero repeat,
1447     that was OK.
1448    
1449     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1450     buffer into which it was copied (for escape processing) was still set at
1451     1024, so long lines caused crashes.
1452    
1453     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1454     "internal error: code overflow...". This applied to any character class
1455     that was followed by a possessive quantifier.
1456    
1457     15. Modified the Makefile to add libpcre.la as a prerequisite for
1458     libpcreposix.la because I was told this is needed for a parallel build to
1459     work.
1460    
1461     16. If a pattern that contained .* following optional items at the start was
1462     studied, the wrong optimizing data was generated, leading to matching
1463     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1464     matching string must start with a or b or c. The correct conclusion for
1465     this pattern is that a match can start with any character.
1466    
1467    
1468 nigel 71 Version 4.4 13-Aug-03
1469     ---------------------
1470    
1471     1. In UTF-8 mode, a character class containing characters with values between
1472     127 and 255 was not handled correctly if the compiled pattern was studied.
1473     In fixing this, I have also improved the studying algorithm for such
1474     classes (slightly).
1475    
1476     2. Three internal functions had redundant arguments passed to them. Removal
1477     might give a very teeny performance improvement.
1478    
1479     3. Documentation bug: the value of the capture_top field in a callout is *one
1480     more than* the number of the hightest numbered captured substring.
1481    
1482     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1483     in incorrectly linking with a previously installed version. They now link
1484     explicitly with libpcre.la.
1485    
1486     5. configure.in no longer needs to recognize Cygwin specially.
1487    
1488     6. A problem in pcre.in for Windows platforms is fixed.
1489    
1490     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1491     pcretest, it used to include the size of the study block as part of its
1492     output. Unfortunately, the structure contains a field that has a different
1493     size on different hardware architectures. This meant that the tests that
1494     showed this size failed. As the block is currently always of a fixed size,
1495     this information isn't actually particularly useful in pcretest output, so
1496     I have just removed it.
1497    
1498     8. Three pre-processor statements accidentally did not start in column 1.
1499     Sadly, there are *still* compilers around that complain, even though
1500     standard C has not required this for well over a decade. Sigh.
1501    
1502     9. In pcretest, the code for checking callouts passed small integers in the
1503     callout_data field, which is a void * field. However, some picky compilers
1504     complained about the casts involved for this on 64-bit systems. Now
1505     pcretest passes the address of the small integer instead, which should get
1506     rid of the warnings.
1507    
1508     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1509     both compile and run time, and gives an error if an invalid UTF-8 sequence
1510     is found. There is a option for disabling this check in cases where the
1511     string is known to be correct and/or the maximum performance is wanted.
1512    
1513     11. In response to a bug report, I changed one line in Makefile.in from
1514    
1515     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1516     to
1517     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1518    
1519     to look similar to other lines, but I have no way of telling whether this
1520     is the right thing to do, as I do not use Windows. No doubt I'll get told
1521     if it's wrong...
1522    
1523    
1524 nigel 69 Version 4.3 21-May-03
1525     ---------------------
1526    
1527     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1528     Makefile.
1529    
1530     2. Some refactoring to improve the quality of the code:
1531    
1532     (i) The utf8_table... variables are now declared "const".
1533    
1534     (ii) The code for \cx, which used the "case flipping" table to upper case
1535     lower case letters, now just substracts 32. This is ASCII-specific,
1536     but the whole concept of \cx is ASCII-specific, so it seems
1537     reasonable.
1538    
1539     (iii) PCRE was using its character types table to recognize decimal and
1540     hexadecimal digits in the pattern. This is silly, because it handles
1541     only 0-9, a-f, and A-F, but the character types table is locale-
1542     specific, which means strange things might happen. A private
1543     table is now used for this - though it costs 256 bytes, a table is
1544     much faster than multiple explicit tests. Of course, the standard
1545     character types table is still used for matching digits in subject
1546     strings against \d.
1547    
1548     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1549     ending in _t are). So I've renamed it as ESC_tee.
1550    
1551     3. The first argument for regexec() in the POSIX wrapper should have been
1552     defined as "const".
1553    
1554     4. Changed pcretest to use malloc() for its buffers so that they can be
1555     Electric Fenced for debugging.
1556    
1557     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1558     to read one or more bytes before the start of the subject string. Often this
1559     had no effect on PCRE's behaviour, but in some circumstances it could
1560     provoke a segmentation fault.
1561    
1562     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1563     to try to read one or more bytes before the start of the subject string.
1564    
1565     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1566     UTF-8 support could misbehave in various ways if the subject string
1567     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1568     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1569     back over UTF-8 characters.)
1570    
1571    
1572 nigel 67 Version 4.2 14-Apr-03
1573     ---------------------
1574    
1575     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1576    
1577     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1578     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1579     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1580     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1581     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1582     and BUILD_EXEEXT
1583     Note: automatic setting of the BUILD variables is not yet working
1584     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1585     compile-time but not at link-time
1586     [LINK]: use for linking executables only
1587     make different versions for Windows and non-Windows
1588     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1589     libraries
1590     [LINK_FOR_BUILD]: new variable
1591     [OBJEXT]: use throughout
1592     [EXEEXT]: use throughout
1593     <winshared>: new target
1594     <wininstall>: new target
1595     <dftables.o>: use native compiler
1596     <dftables>: use native linker
1597     <install>: handle Windows platform correctly
1598     <clean>: ditto
1599     <check>: ditto
1600     copy DLL to top builddir before testing
1601    
1602     As part of these changes, -no-undefined was removed again. This was reported
1603     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1604     in any case.
1605    
1606     3. Some tidies to get rid of compiler warnings:
1607    
1608     . In the match_data structure, match_limit was an unsigned long int, whereas
1609     match_call_count was an int. I've made them both unsigned long ints.
1610    
1611     . In pcretest the fact that a const uschar * doesn't automatically cast to
1612     a void * provoked a warning.
1613    
1614     . Turning on some more compiler warnings threw up some "shadow" variables
1615     and a few more missing casts.
1616    
1617     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1618     option, a class that contained a single character with a value between 128
1619     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1620    
1621     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1622     option, a class that contained several characters, but with at least one
1623     whose value was between 128 and 255 caused PCRE to crash.
1624    
1625    
1626 nigel 65 Version 4.1 12-Mar-03
1627     ---------------------
1628 nigel 63
1629 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1630     needed, and a string in dftables.c that was longer than standard compilers are
1631     required to support.
1632    
1633     2. Compiling with Sun's compiler found a few more places where the code could
1634     be tidied up in order to avoid warnings.
1635    
1636     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1637     first of these names is deprecated in the latest Autoconf in favour of the name
1638     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1639     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1640     analogy I have changed it to CFLAGS_FOR_BUILD.
1641    
1642     4. Added -no-undefined to the linking command in the Makefile, because this is
1643     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1644     linking step for the pcreposix library.
1645    
1646     5. PCRE was failing to diagnose the case of two named groups with the same
1647     name.
1648    
1649     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1650     literal character that is needed in the subject for a match, and scans along to
1651     ensure that it is present before embarking on the full matching process. This
1652     saves time in cases of nested unlimited repeats that are never going to match.
1653     Problem: the scan can take a lot of time if the subject is very long (e.g.
1654     megabytes), thus penalizing straightforward matches. It is now done only if the
1655     amount of subject to be scanned is less than 1000 bytes.
1656    
1657     7. A lesser problem with the same optimization is that it was recording the
1658     first character of an anchored pattern as "needed", thus provoking a search
1659     right along the subject, even when the first match of the pattern was going to
1660     fail. The "needed" character is now not set for anchored patterns, unless it
1661     follows something in the pattern that is of non-fixed length. Thus, it still
1662     fulfils its original purpose of finding quick non-matches in cases of nested
1663     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1664    
1665    
1666     Version 4.0 17-Feb-03
1667     ---------------------
1668    
1669 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1670     extended to the end of string, PCRE compiled incorrect data. This could lead to
1671     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1672    
1673     2. Moved to autoconf 2.53 and libtool 1.4.2.
1674    
1675     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1676     the special perltest8 script is no longer needed - all the tests can be run
1677     from a single perltest script.
1678    
1679     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1680     by \s. It has now been removed in PCRE. This means it isn't recognized as
1681     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1682     class [:space:] *does* include VT, thereby creating a mess.
1683    
1684     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1685     space and tab.
1686    
1687     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1688     its new features into the main test script, reducing the number of scripts.
1689    
1690     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1691     were backward compatible, and made the (?i) apply to the whole pattern, as if
1692     /i were given. Now it behaves more logically, and applies the option setting
1693     only to what follows. PCRE has been changed to follow suit. However, if it
1694     finds options settings right at the start of the pattern, it extracts them into
1695     the global options, as before. Thus, they show up in the info data.
1696    
1697     8. Added support for the \Q...\E escape sequence. Characters in between are
1698     treated as literals. This is slightly different from Perl in that $ and @ are
1699     also handled as literals inside the quotes. In Perl, they will cause variable
1700     interpolation. Note the following examples:
1701    
1702     Pattern PCRE matches Perl matches
1703    
1704     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1705     \Qabc\$xyz\E abc\$xyz abc\$xyz
1706     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1707    
1708     For compatibility with Perl, \Q...\E sequences are recognized inside character
1709     classes as well as outside them.
1710    
1711     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1712     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1713     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1714     signed/unsigned warnings.
1715    
1716     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1717     option for pcretest, so I've replaced it by a simple function that does just
1718     that job.
1719    
1720     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1721     "pcregrep -".
1722    
1723     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1724     Java package. This provides some syntactic sugar for simple cases of what my
1725     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1726     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1727     item, you can use this simplified notation. Note that only makes sense with
1728     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1729     greediness, whatever the setting of the PCRE_UNGREEDY option.
1730    
1731     13. A change of greediness default within a pattern was not taking effect at
1732     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1733     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1734     was abstracted outside.
1735    
1736     14. PCRE now supports the \G assertion. It is true when the current matching
1737     position is at the start point of the match. This differs from \A when the
1738     starting offset is non-zero. Used with the /g option of pcretest (or similar
1739     code), it works in the same way as it does for Perl's /g option. If all
1740     alternatives of a regex begin with \G, the expression is anchored to the start
1741     match position, and the "anchored" flag is set in the compiled expression.
1742    
1743     15. Some bugs concerning the handling of certain option changes within patterns
1744     have been fixed. These applied to options other than (?ims). For example,
1745     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1746     been the other way round. Some of this was related to change 7 above.
1747    
1748     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1749     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1750     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1751     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1752    
1753     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1754     mode. Unlike ".", it always matches newline, whatever the setting of
1755     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1756     assertions. Perl allows it, but it doesn't (in general) work because it can't
1757     calculate the length of the lookbehind. At least, that's the case for Perl
1758     5.8.0 - I've been told they are going to document that it doesn't work in
1759     future.
1760    
1761     18. Added an error diagnosis for escapes that PCRE does not support: these are
1762     \L, \l, \N, \P, \p, \U, \u, and \X.
1763    
1764     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1765     reading past the end of the pattern in cases such as /[abcd/.
1766    
1767     20. PCRE was getting more memory than necessary for patterns with classes that
1768     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1769    
1770     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1771     compiling PCRE for use with Virtual Pascal.
1772    
1773     22. Small fix to the Makefile to make it work properly if the build is done
1774     outside the source tree.
1775    
1776     23. Added a new extension: a condition to go with recursion. If a conditional
1777     subpattern starts with (?(R) the "true" branch is used if recursion has
1778     happened, whereas the "false" branch is used only at the top level.
1779    
1780     24. When there was a very long string of literal characters (over 255 bytes
1781     without UTF support, over 250 bytes with UTF support), the computation of how
1782     much memory was required could be incorrect, leading to segfaults or other
1783     strange effects.
1784    
1785     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1786     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1787     there was a subsequent back reference to those brackets. This meant that, for
1788     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1789     possible to check for precisely this case. All we can do is abandon the
1790     optimization if .* occurs inside capturing brackets when there are any back
1791     references whatsoever. (See below for a better fix that came later.)
1792    
1793     26. The handling of the optimization for finding the first character of a
1794     non-anchored pattern, and for finding a character that is required later in the
1795     match were failing in some cases. This didn't break the matching; it just
1796     failed to optimize when it could. The way this is done has been re-implemented.
1797    
1798     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1799    
1800     28. Added a new feature that provides some of the functionality that Perl
1801     provides with (?{...}). The facility is termed a "callout". The way it is done
1802     in PCRE is for the caller to provide an optional function, by setting
1803     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1804     global variable. By default it is unset, which disables all calling out. To get
1805     the function called, the regex must include (?C) at appropriate points. This
1806     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1807     This provides a means of identifying different callout points. When PCRE
1808     reaches such a point in the regex, if pcre_callout has been set, the external
1809     function is called. It is provided with data in a structure called
1810     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1811     matching continues; if it returns a non-zero value, the match at the current
1812     point fails. However, backtracking will occur if possible. [This was changed
1813     later and other features added - see item 49 below.]
1814    
1815     29. pcretest is upgraded to test the callout functionality. It provides a
1816     callout function that displays information. By default, it shows the start of
1817     the match and the current position in the text. There are some new data escapes
1818     to vary what happens:
1819    
1820     \C+ in addition, show current contents of captured substrings
1821     \C- do not supply a callout function
1822     \C!n return 1 when callout number n is reached
1823     \C!n!m return 1 when callout number n is reached for the mth time
1824    
1825     30. If pcregrep was called with the -l option and just a single file name, it
1826     output "<stdin>" if a match was found, instead of the file name.
1827    
1828     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1829     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1830     pcre_exec(). This saves a malloc/free per call. The default value of
1831     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1832     when configuring.
1833    
1834     32. The default maximum size of a compiled pattern is 64K. There have been a
1835     few cases of people hitting this limit. The code now uses macros to handle the
1836     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1837     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1838     configuring. Tests 2 and 5 work only with 2-byte links because they output
1839     debugging information about compiled patterns.
1840    
1841     33. Internal code re-arrangements:
1842    
1843     (a) Moved the debugging function for printing out a compiled regex into
1844     its own source file (printint.c) and used #include to pull it into
1845     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1846     separate copies.
1847    
1848     (b) Defined the list of op-code names for debugging as a macro in
1849     internal.h so that it is next to the definition of the opcodes.
1850    
1851     (c) Defined a table of op-code lengths for simpler skipping along compiled
1852     code. This is again a macro in internal.h so that it is next to the
1853     definition of the opcodes.
1854    
1855     34. Added support for recursive calls to individual subpatterns, along the
1856     lines of Robin Houston's patch (but implemented somewhat differently).
1857    
1858     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1859     allow it to read and process whole directories in Win32. This code was
1860     contributed by Lionel Fourquaux; it has not been tested by me.
1861    
1862     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1863     used to name a group. Names consist of alphanumerics and underscores, and must
1864     be unique. Back references use the syntax (?P=name) and recursive calls use
1865     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1866     numbers. The function pcre_fullinfo() can be used after compilation to extract
1867     a name/number map. There are three relevant calls:
1868    
1869     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1870     PCRE_INFO_NAMECOUNT yields the number of entries
1871     PCRE_INFO_NAMETABLE yields a pointer to the map.
1872    
1873     The map is a vector of fixed-size entries. The size of each entry depends on
1874     the length of the longest name used. The first two bytes of each entry are the
1875     group number, most significant byte first. There follows the corresponding
1876     name, zero terminated. The names are in alphabetical order.
1877    
1878     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1879     case instead of 255. Making it the same both with and without UTF-8 support
1880     means that the same test output works with both.
1881    
1882     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1883     calling malloc() with a zero argument.
1884    
1885     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1886     optimization. I've improved things by keeping a bitmap of backreferences with
1887     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1888     fact referenced, the optimization can be applied. It is unlikely that a
1889     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1890     the match to follow \n) will appear inside brackets with a number greater than
1891     31, but if it does, any back reference > 31 suppresses the optimization.
1892    
1893     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1894     of disabling numbered capturing parentheses. Any opening parenthesis that is
1895     not followed by ? behaves as if it were followed by ?: but named parentheses
1896     can still be used for capturing (and they will acquire numbers in the usual
1897     way).
1898    
1899     41. Redesigned the return codes from the match() function into yes/no/error so
1900     that errors can be passed back from deep inside the nested calls. A malloc
1901     failure while inside a recursive subpattern call now causes the
1902     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1903    
1904     42. It is now possible to set a limit on the number of times the match()
1905     function is called in a call to pcre_exec(). This facility makes it possible to
1906     limit the amount of recursion and backtracking, though not in a directly
1907     obvious way, because the match() function is used in a number of different
1908     circumstances. The count starts from zero for each position in the subject
1909     string (for non-anchored patterns). The default limit is, for compatibility, a
1910     large number, namely 10 000 000. You can change this in two ways:
1911    
1912     (a) When configuring PCRE before making, you can use --with-match-limit=n
1913     to set a default value for the compiled library.
1914    
1915     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1916     a different value is set. See 45 below.
1917    
1918     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1919    
1920     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1921     of things that can be changed at compile time. The first argument specifies
1922     what is wanted and the second points to where the information is to be placed.
1923     The current list of available information is:
1924    
1925     PCRE_CONFIG_UTF8
1926    
1927     The output is an integer that is set to one if UTF-8 support is available;
1928     otherwise it is set to zero.
1929    
1930     PCRE_CONFIG_NEWLINE
1931    
1932     The output is an integer that it set to the value of the code that is used for
1933     newline. It is either LF (10) or CR (13).
1934    
1935     PCRE_CONFIG_LINK_SIZE
1936    
1937     The output is an integer that contains the number of bytes used for internal
1938     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1939    
1940     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1941    
1942     The output is an integer that contains the threshold above which the POSIX
1943     interface uses malloc() for output vectors. See item 31 above.
1944    
1945     PCRE_CONFIG_MATCH_LIMIT
1946    
1947     The output is an unsigned integer that contains the default limit of the number
1948     of match() calls in a pcre_exec() execution. See 42 above.
1949    
1950     44. pcretest has been upgraded by the addition of the -C option. This causes it
1951     to extract all the available output from the new pcre_config() function, and to
1952     output it. The program then exits immediately.
1953    
1954     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1955     order to support additional features. One way would have been to define
1956     pcre_exec2() (for example) with extra arguments, but this would not have been
1957     extensible, and would also have required all calls to the original function to
1958     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1959     is used for passing in "extra" data from pcre_study().
1960    
1961     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1962     contains the following fields:
1963    
1964     flags a bitmap indicating which of the following fields are set
1965     study_data opaque data from pcre_study()
1966     match_limit a way of specifying a limit on match() calls for a specific
1967     call to pcre_exec()
1968     callout_data data for callouts (see 49 below)
1969    
1970     The flag bits are also defined in pcre.h, and are
1971    
1972     PCRE_EXTRA_STUDY_DATA
1973     PCRE_EXTRA_MATCH_LIMIT
1974     PCRE_EXTRA_CALLOUT_DATA
1975    
1976     The pcre_study() function now returns one of these new pcre_extra blocks, with
1977     the actual study data pointed to by the study_data field, and the
1978     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1979     before. That is, this change is entirely upwards-compatible and requires no
1980     change to existing code.
1981    
1982     If you want to pass in additional data to pcre_exec(), you can either place it
1983     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1984     block.
1985    
1986     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1987     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1988     times with different match limits, until it finds the minimum value needed for
1989     pcre_exec() to complete. The value is then output. This can be instructive; for
1990     most simple matches the number is quite small, but for pathological cases it
1991     gets very large very quickly.
1992    
1993     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1994     returns the size of the data block pointed to by the study_data field in a
1995     pcre_extra block, that is, the value that was passed as the argument to
1996     pcre_malloc() when PCRE was getting memory in which to place the information
1997     created by pcre_study(). The fourth argument should point to a size_t variable.
1998     pcretest has been extended so that this information is shown after a successful
1999     pcre_study() call when information about the compiled regex is being displayed.
2000    
2001     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2002     because what follows is always an absolute path. (Later: it turns out that this
2003     is more than cosmetic for MinGW, because it doesn't like empty path
2004     components.)
2005    
2006     49. Some changes have been made to the callout feature (see 28 above):
2007    
2008     (i) A callout function now has three choices for what it returns:
2009    
2010     0 => success, carry on matching
2011     > 0 => failure at this point, but backtrack if possible
2012     < 0 => serious error, return this value from pcre_exec()
2013    
2014     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2015     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2016     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2017     use by callout functions. It will never be used by PCRE itself.
2018    
2019     (ii) The pcre_extra structure (see 45 above) has a void * field called
2020     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2021     pcre_callout_block structure has a field of the same name. The contents of
2022     the field passed in the pcre_extra structure are passed to the callout
2023     function in the corresponding field in the callout block. This makes it
2024     easier to use the same callout-containing regex from multiple threads. For
2025     testing, the pcretest program has a new data escape
2026    
2027     \C*n pass the number n (may be negative) as callout_data
2028    
2029     If the callout function in pcretest receives a non-zero value as
2030     callout_data, it returns that value.
2031    
2032     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2033     there were some redundant $(CFLAGS) in commands that are now specified as
2034     $(LINK), which already includes $(CFLAGS).
2035    
2036     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2037     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2038     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2039     one-byte characters throughout. Note that case-insensitive matching applies
2040     only to characters whose values are less than 256. PCRE doesn't support the
2041     notion of cases for higher-valued characters.
2042    
2043     (i) A character class whose characters are all within 0-255 is handled as
2044     a bit map, and the map is inverted for negative classes. Previously, a
2045     character > 255 always failed to match such a class; however it should
2046     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2047    
2048     (ii) A negated character class with a single character < 255 is coded as
2049     "not this character" (OP_NOT). This wasn't working properly when the test
2050     character was multibyte, either singly or repeated.
2051    
2052     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2053     mode, for example: \x{100}{2,3}.
2054    
2055     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2056     singly or repeated) now correctly test multibyte characters. However,
2057     PCRE doesn't recognize any characters with values greater than 255 as
2058     digits, spaces, or word characters. Such characters always match \D, \S,
2059     and \W, and never match \d, \s, or \w.
2060    
2061     (v) Classes may now contain characters and character ranges with values
2062     greater than 255. For example: [ab\x{100}-\x{400}].
2063    
2064     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2065     PCRE in UTF-8 mode.
2066    
2067     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2068     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2069     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2070     value.)
2071    
2072     53. The single man page has become too large. I have therefore split it up into
2073     a number of separate man pages. These also give rise to individual HTML pages;
2074     these are now put in a separate directory, and there is an index.html page that
2075     lists them all. Some hyperlinking between the pages has been installed.
2076    
2077     54. Added convenience functions for handling named capturing parentheses.
2078    
2079     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2080     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2081     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2082     are faulted.
2083    
2084     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2085     calling configure. These values are used when compiling the dftables.c program
2086     which is run to generate the source of the default character tables. They
2087     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2088     you will need to set these values.
2089    
2090     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2091    
2092    
2093     Version 3.9 02-Jan-02
2094 nigel 61 ---------------------
2095    
2096     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2097    
2098     2. If --disable-static was given, the building process failed when trying to
2099     build pcretest and pcregrep. (For some reason it was using libtool to compile
2100     them, which is not right, as they aren't part of the library.)
2101    
2102    
2103 nigel 59 Version 3.8 18-Dec-01
2104     ---------------------
2105    
2106     1. The experimental UTF-8 code was completely screwed up. It was packing the
2107     bytes in the wrong order. How dumb can you get?
2108    
2109    
2110 nigel 57 Version 3.7 29-Oct-01
2111     ---------------------
2112    
2113     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2114     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2115     this didn't happen under Solaris 8, where I normally test things.
2116    
2117 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2118     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2119     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2120     this fix an hour or so after the initial 3.7 release.)
2121 nigel 57
2122 nigel 59
2123 nigel 55 Version 3.6 23-Oct-01
2124     ---------------------
2125    
2126     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2127     offsets passed as NULL with zero offset count.
2128    
2129     2. The config.guess and config.sub files had not been updated when I moved to
2130     the latest autoconf.
2131    
2132    
2133 nigel 53 Version 3.5 15-Aug-01
2134     ---------------------
2135 nigel 5
2136 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2137     had been forgotten.
2138    
2139     2. By using declared but undefined structures, we can avoid using "void"
2140     definitions in pcre.h while keeping the internal definitions of the structures
2141     private.
2142    
2143     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2144     user point of view, this means that both static and shared libraries are built
2145     by default, but this can be individually controlled. More of the work of
2146     handling this static/shared cases is now inside libtool instead of PCRE's make
2147     file.
2148    
2149     4. The pcretest utility is now installed along with pcregrep because it is
2150     useful for users (to test regexs) and by doing this, it automatically gets
2151     relinked by libtool. The documentation has been turned into a man page, so
2152     there are now .1, .txt, and .html versions in /doc.
2153    
2154     5. Upgrades to pcregrep:
2155     (i) Added long-form option names like gnu grep.
2156     (ii) Added --help to list all options with an explanatory phrase.
2157     (iii) Added -r, --recursive to recurse into sub-directories.
2158     (iv) Added -f, --file to read patterns from a file.
2159    
2160     6. pcre_exec() was referring to its "code" argument before testing that
2161     argument for NULL (and giving an error if it was NULL).
2162    
2163     7. Upgraded Makefile.in to allow for compiling in a different directory from
2164     the source directory.
2165    
2166     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2167     options bits, the pointer it was passed was to an int instead of to an unsigned
2168     long int. This mattered only on 64-bit systems.
2169    
2170     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2171     generated) instead of pcre.in, which it its source. Also made the same change
2172     in several of the .c files.
2173    
2174     10. A new release of gcc defines printf() as a macro, which broke pcretest
2175     because it had an ifdef in the middle of a string argument for printf(). Fixed
2176     by using separate calls to printf().
2177    
2178     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2179     script, to force use of CR or LF instead of \n in the source. On non-Unix
2180     systems, the value can be set in config.h.
2181    
2182     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2183     absolute limit. Changed the text of the error message to make this clear, and
2184     likewise updated the man page.
2185    
2186     13. The limit of 99 on the number of capturing subpatterns has been removed.
2187     The new limit is 65535, which I hope will not be a "real" limit.
2188    
2189    
2190 nigel 51 Version 3.4 22-Aug-00
2191     ---------------------
2192    
2193     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2194    
2195     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2196    
2197    
2198 nigel 49 Version 3.3 01-Aug-00
2199     ---------------------
2200    
2201     1. If an octal character was given, but the value was greater than \377, it
2202     was not getting masked to the least significant bits, as documented. This could
2203     lead to crashes in some systems.
2204    
2205     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2206     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2207    
2208     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2209     These just pass their arguments on to (pcre_free)(), but they are provided
2210     because some uses of PCRE bind it to non-C systems that can call its functions,
2211     but cannot call free() or pcre_free() directly.
2212    
2213     4. Add "make test" as a synonym for "make check". Corrected some comments in
2214     the Makefile.
2215    
2216     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2217     Makefile.
2218    
2219     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2220     command called pgrep for grepping around the active processes.
2221    
2222     7. Added the beginnings of support for UTF-8 character strings.
2223    
2224     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2225     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2226     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2227     out for the ar command.)
2228    
2229    
2230 nigel 47 Version 3.2 12-May-00
2231     ---------------------
2232    
2233     This is purely a bug fixing release.
2234    
2235     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2236     of ZA. This was just one example of several cases that could provoke this bug,
2237     which was introduced by change 9 of version 2.00. The code for breaking
2238     infinite loops after an iteration that matches an empty string was't working
2239     correctly.
2240    
2241     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2242     when matched against abbab (for example). After matching an empty string, it
2243     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2244     caused it to match further down the string than it should.
2245    
2246     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2247     was there because it doesn't seem to be needed, and it causes trouble on some
2248     systems, as it is not a Standard C header. It has been removed.
2249    
2250     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2251     were reported on the Macintosh. The changes were from
2252    
2253     while ((c = *(++ptr)) != 0 && c != '\n');
2254     to
2255     while ((c = *(++ptr)) != 0 && c != '\n') ;
2256    
2257     Totally extraordinary, but if that's what it takes...
2258    
2259     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2260     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2261     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2262     assumes the way PCRE uses memmove() (always moving upwards).
2263    
2264     6. PCRE is being used in one environment where strchr() is not available. There
2265     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2266     faster code anyway.
2267    
2268    
2269 nigel 45 Version 3.1 09-Feb-00
2270     ---------------------
2271    
2272     The only change in this release is the fixing of some bugs in Makefile.in for
2273     the "install" target:
2274    
2275     (1) It was failing to install pcreposix.h.
2276    
2277     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2278    
2279    
2280 nigel 43 Version 3.0 01-Feb-00
2281     ---------------------
2282 nigel 41
2283     1. Add support for the /+ modifier to perltest (to output $` like it does in
2284     pcretest).
2285    
2286     2. Add support for the /g modifier to perltest.
2287    
2288     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2289     matches null strings.
2290    
2291     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2292     pattern. Perl treats empty patterns specially - it reuses the most recent
2293     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2294     effect.
2295    
2296     5. The POSIX interface was broken in that it was just handing over the POSIX
2297     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2298     required a bigger vector, with some working space on the end. This means that
2299     the POSIX wrapper now has to get and free some memory, and copy the results.
2300    
2301 nigel 43 6. Added some simple autoconf support, placing the test data and the
2302     documentation in separate directories, re-organizing some of the
2303     information files, and making it build pcre-config (a GNU standard). Also added
2304     libtool support for building PCRE as a shared library, which is now the
2305     default.
2306 nigel 41
2307 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2308     09 are not valid octal constants. Single digits will be used for minor values
2309     less than 10.
2310    
2311     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2312     existing programs that set these in the POSIX interface can use PCRE without
2313     modification.
2314    
2315     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2316     return all that pcre_info() returns, plus additional data. The pcre_info()
2317     function is retained for compatibility, but is considered to be obsolete.
2318    
2319     10. Added experimental recursion feature (?R) to handle one common case that
2320     Perl 5.6 will be able to do with (?p{...}).
2321    
2322     11. Added support for POSIX character classes like [:alpha:], which Perl is
2323     adopting.
2324    
2325    
2326 nigel 39 Version 2.08 31-Aug-99
2327     ----------------------
2328    
2329     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2330     trying to match at the startoffset position, but instead was moving forward to
2331     the next newline as if a previous match had failed.
2332    
2333     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2334     and could get into a loop if a null string was matched other than at the start
2335     of the subject.
2336    
2337     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2338     be distinguished at compile time, and for completeness also added PCRE_DATE.
2339    
2340     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2341     in GnuWin32 environments.
2342    
2343    
2344 nigel 37 Version 2.07 29-Jul-99
2345     ----------------------
2346    
2347     1. The documentation is now supplied in plain text form and HTML as well as in
2348     the form of man page sources.
2349    
2350     2. C++ compilers don't like assigning (void *) values to other pointer types.
2351     In particular this affects malloc(). Although there is no problem in Standard
2352     C, I've put in casts to keep C++ compilers happy.
2353    
2354     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2355     should be (const char *).
2356    
2357     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2358     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2359     However, I haven't made this a standard facility. The documentation doesn't
2360     mention it, and the Makefile doesn't support it.
2361    
2362     5. The Makefile now contains an "install" target, with editable destinations at
2363     the top of the file. The pcretest program is not installed.
2364    
2365     6. pgrep -V now gives the PCRE version number and date.
2366    
2367     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2368     causing the entire string to be ignored, instead of just the last character.
2369    
2370     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2371     non-matching string, it can take a very, very long time, even for strings of
2372     quite modest length, because of the nested recursion. PCRE now does better in
2373     some of these cases. It does this by remembering the last required literal
2374     character in the pattern, and pre-searching the subject to ensure it is present
2375     before running the real match. In other words, it applies a heuristic to detect
2376     some types of certain failure quickly, and in the above example, if presented
2377     with a string that has no trailing " it gives "no match" very quickly.
2378    
2379     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2380     other alternatives are tried instead.
2381    
2382    
2383 nigel 35 Version 2.06 09-Jun-99
2384     ----------------------
2385    
2386     1. Change pcretest's output for amount of store used to show just the code
2387     space, because the remainder (the data block) varies in size between 32-bit and
2388     64-bit systems.
2389    
2390     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2391     start matching at. This allows lookbehinds to work when searching for multiple
2392     occurrences in a string.
2393    
2394     3. Added additional options to pcretest for testing multiple occurrences:
2395    
2396     /+ outputs the rest of the string that follows a match
2397     /g loops for multiple occurrences, using the new startoffset argument
2398     /G loops for multiple occurrences by passing an incremented pointer
2399    
2400     4. PCRE wasn't doing the "first character" optimization for patterns starting
2401     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2402     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2403     the letter 'x'. On long subject strings, this gives a significant speed-up.
2404    
2405    
2406 nigel 33 Version 2.05 21-Apr-99
2407     ----------------------
2408    
2409     1. Changed the type of magic_number from int to long int so that it works
2410     properly on 16-bit systems.
2411    
2412     2. Fixed a bug which caused patterns starting with .* not to work correctly
2413     when the subject string contained newline characters. PCRE was assuming
2414     anchoring for such patterns in all cases, which is not correct because .* will
2415     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2416     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2417     must be retried after every newline in the subject.
2418    
2419    
2420 nigel 31 Version 2.04 18-Feb-99
2421     ----------------------
2422    
2423     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2424     computation of the store needed to hold the pattern was incorrect (too large).
2425     If such patterns were nested a few deep, this could multiply and become a real
2426     problem.
2427    
2428     2. Added /M option to pcretest to show the memory requirement of a specific
2429     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2430    
2431     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2432     compiled in such a way that the backtracking after subsequent failure was
2433     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2434     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2435    
2436    
2437 nigel 29 Version 2.03 02-Feb-99
2438     ----------------------
2439    
2440     1. Fixed typo and small mistake in man page.
2441    
2442 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2443     LICENCE file containing the conditions.
2444 nigel 29
2445     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2446     Perl, that is the internal \ allows the delimiter to be included in the
2447     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2448     the final delimiter, add \ to the end of the pattern (to test the error).
2449    
2450     4. Added the convenience functions for extracting substrings after a successful
2451     match. Updated pcretest to make it able to test these functions.
2452    
2453    
2454 nigel 27 Version 2.02 14-Jan-99
2455     ----------------------
2456    
2457     1. Initialized the working variables associated with each extraction so that
2458     their saving and restoring doesn't refer to uninitialized store.
2459    
2460     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2461     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2462     fix the problem.
2463    
2464     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2465     calls, and wasn't printing the correct value for compiling calls. Increased the
2466     default value of LOOPREPEAT, and the number of significant figures in the
2467     times.
2468    
2469     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2470    
2471     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2472     a building problem on Windows NT with a FAT file system.
2473    
2474    
2475 nigel 25 Version 2.01 21-Oct-98
2476     ----------------------
2477    
2478     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2479     to character tables built by pcre_maketables() in the current locale. If NULL
2480     is passed, the default tables are used.
2481    
2482    
2483 nigel 23 Version 2.00 24-Sep-98
2484 nigel 21 ----------------------
2485    
2486 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2487     it any more.
2488    
2489     2. Allow quantification of (?>) groups, and make it work correctly.
2490    
2491     3. The first character computation wasn't working for (?>) groups.
2492    
2493     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2494     end of the subject) and add 5.005's \z, which really does match only at the
2495     very end of the subject.
2496    
2497     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2498    
2499     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2500     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2501     localized options. All options to pcre_study() were also removed.
2502    
2503     7. Add other new features from 5.005:
2504    
2505     $(?<= positive lookbehind
2506     $(?<! negative lookbehind
2507     (?imsx-imsx) added the unsetting capability
2508     such a setting is global if at outer level; local otherwise
2509     (?imsx-imsx:) non-capturing groups with option setting
2510     (?(cond)re|re) conditional pattern matching
2511    
2512     A backreference to itself in a repeated group matches the previous
2513     captured string.
2514    
2515     8. General tidying up of studying (both automatic and via "study")
2516     consequential on the addition of new assertions.
2517    
2518     9. As in 5.005, unlimited repeated groups that could match an empty substring
2519     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2520     runtime if any iteration does actually match an empty substring.
2521    
2522     10. Include the RunTest script in the distribution.
2523    
2524     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2525     discrepancies, some of which were old and were also with respect to 5.004. They
2526     have now been fixed.
2527    
2528    
2529     Version 1.09 28-Apr-98
2530     ----------------------
2531    
2532 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2533     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2534     program crashes, or just wrong answers. This did not apply to negated classes
2535     containing more than one character, or to minima other than one.
2536    
2537    
2538 nigel 19 Version 1.08 27-Mar-98
2539     ----------------------
2540    
2541     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2542    
2543     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2544     latter must appear before anything that relies on it in the pattern.
2545    
2546    
2547 nigel 17 Version 1.07 16-Feb-98
2548     ----------------------
2549    
2550     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2551     repeat of a potentially empty string).
2552    
2553    
2554 nigel 15 Version 1.06 23-Jan-98
2555     ----------------------
2556    
2557     1. Added Markus Oberhumer's little patches for C++.
2558    
2559     2. Literal strings longer than 255 characters were broken.
2560    
2561    
2562 nigel 13 Version 1.05 23-Dec-97
2563     ----------------------
2564    
2565     1. Negated character classes containing more than one character were failing if
2566     PCRE_CASELESS was set at run time.
2567    
2568    
2569 nigel 11 Version 1.04 19-Dec-97
2570     ----------------------
2571    
2572     1. Corrected the man page, where some "const" qualifiers had been omitted.
2573    
2574     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2575     input syntax.
2576    
2577     3. Fixed memory leak which occurred when a regex with back references was
2578     matched with an offsets vector that wasn't big enough. The temporary memory
2579     that is used in this case wasn't being freed if the match failed.
2580    
2581     4. Tidied pcretest to ensure it frees memory that it gets.
2582    
2583     5. Temporary memory was being obtained in the case where the passed offsets
2584     vector was exactly big enough.
2585    
2586     6. Corrected definition of offsetof() from change 5 below.
2587    
2588     7. I had screwed up change 6 below and broken the rules for the use of
2589     setjmp(). Now fixed.
2590    
2591    
2592 nigel 9 Version 1.03 18-Dec-97
2593     ----------------------
2594    
2595     1. A erroneous regex with a missing opening parenthesis was correctly
2596     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2597     on some systems.
2598    
2599     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2600     it was reported that one broken compiler failed on the former because "code" is
2601     also an independent variable.
2602    
2603     3. The erroneous regex a[]b caused an array overrun reference.
2604    
2605     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2606     fail on data ending with that character. (It was going on too far, and checking
2607     the next character, typically a binary zero.) This was specific to the
2608     optimized code for single-character negative classes.
2609    
2610     5. Added a contributed patch from the TIN world which does the following:
2611    
2612     + Add an undef for memmove, in case the the system defines a macro for it.
2613    
2614     + Add a definition of offsetof(), in case there isn't one. (I don't know
2615     the reason behind this - offsetof() is part of the ANSI standard - but
2616     it does no harm).
2617    
2618     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2619     most of the places where whitespace preceded '#'. I have given up and
2620     allowed the remaining 2 cases to be at the margin.
2621    
2622     + Rename some variables in pcre to eliminate shadowing. This seems very
2623     pedantic, but does no harm, of course.
2624    
2625     6. Moved the call to setjmp() into its own function, to get rid of warnings
2626     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2627    
2628     7. Constructs such as \d{8,} were compiling into the equivalent of
2629 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2630 nigel 9 outcome, but in this particular case used more store than had been allocated,
2631     which caused the bug to be discovered because it threw up an internal error.
2632    
2633     8. The debugging code in both pcre and pcretest for outputting the compiled
2634     form of a regex was going wrong in the case of back references followed by
2635     curly-bracketed repeats.
2636    
2637    
2638 nigel 7 Version 1.02 12-Dec-97
2639     ----------------------
2640    
2641     1. Typos in pcre.3 and comments in the source fixed.
2642    
2643     2. Applied a contributed patch to get rid of places where it used to remove
2644     'const' from variables, and fixed some signed/unsigned and uninitialized
2645     variable warnings.
2646    
2647     3. Added the "runtest" target to Makefile.
2648    
2649     4. Set default compiler flag to -O2 rather than just -O.
2650    
2651    
2652 nigel 5 Version 1.01 19-Nov-97
2653     ----------------------
2654    
2655     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2656     like /([ab]*)*/, that is, for classes with more than one character in them.
2657    
2658     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2659     as /((?>a*))*/ (a PCRE_EXTRA facility).
2660    
2661    
2662     Version 1.00 18-Nov-97
2663     ----------------------
2664    
2665     1. Added compile-time macros to support systems such as SunOS4 which don't have
2666     memmove() or strerror() but have other things that can be used instead.
2667    
2668     2. Arranged that "make clean" removes the executables.
2669    
2670    
2671 nigel 3 Version 0.99 27-Oct-97
2672     ----------------------
2673    
2674     1. Fixed bug in code for optimizing classes with only one character. It was
2675     initializing a 32-byte map regardless, which could cause it to run off the end
2676     of the memory it had got.
2677    
2678     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2679    
2680    
2681     Version 0.98 22-Oct-97
2682     ----------------------
2683    
2684     1. Fixed bug in code for handling temporary memory usage when there are more
2685     back references than supplied space in the ovector. This could cause segfaults.
2686    
2687    
2688     Version 0.97 21-Oct-97
2689     ----------------------
2690    
2691     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2692    
2693     2. Optimized negated single characters not to use a bit map.
2694    
2695     3. Brought error texts together as macro definitions; clarified some of them;
2696     fixed one that was wrong - it said "range out of order" when it meant "invalid
2697     escape sequence".
2698    
2699     4. Changed some char * arguments to const char *.
2700    
2701     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2702    
2703     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2704     pcretest.
2705    
2706    
2707     Version 0.96 16-Oct-97
2708     ----------------------
2709    
2710     1. Added a simple "pgrep" utility to the distribution.
2711    
2712     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2713     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2714     where "ddd" means "one or more decimal digits".
2715    
2716     3. Fixed serious bug. If a pattern had a back reference, but the call to
2717     pcre_exec() didn't supply a large enough ovector to record the related
2718     identifying subpattern, the match always failed. PCRE now remembers the number
2719     of the largest back reference, and gets some temporary memory in which to save
2720     the offsets during matching if necessary, in order to ensure that
2721     backreferences always work.
2722    
2723     4. Increased the compatibility with Perl in a number of ways:
2724    
2725     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2726     to request this handling. The option can be set at compile or exec time.
2727    
2728     (b) $ matches before a terminating newline by default; an option
2729     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2730     mode). The option can be set at compile or exec time.
2731    
2732     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2733     the same as Perl's. If the decimal number it represents is less than 10
2734     or there aren't that many previous left capturing parentheses, an octal
2735     escape is read. Inside a character class, it's always an octal escape,
2736     even if it is a single digit.
2737    
2738     (d) An escaped but undefined alphabetic character is taken as a literal,
2739     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2740     escapes.
2741    
2742     (e) {0} is now permitted. (The previous item is removed from the compiled
2743     pattern).
2744    
2745     5. Changed all the names of code files so that the basic parts are no longer
2746     than 10 characters, and abolished the teeny "globals.c" file.
2747    
2748     6. Changed the handling of character classes; they are now done with a 32-byte
2749     bit map always.
2750    
2751     7. Added the -d and /D options to pcretest to make it possible to look at the
2752     internals of compilation without having to recompile pcre.
2753    
2754    
2755     Version 0.95 23-Sep-97
2756     ----------------------
2757    
2758     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2759     \x20 at the start of a run of normal characters. These were being treated as
2760     real characters, instead of the source characters being re-checked.
2761    
2762    
2763     Version 0.94 18-Sep-97
2764     ----------------------
2765    
2766     1. The functions are now thread-safe, with the caveat that the global variables
2767     containing pointers to malloc() and free() or alternative functions are the
2768     same for all threads.
2769    
2770     2. Get pcre_study() to generate a bitmap of initial characters for non-
2771     anchored patterns when this is possible, and use it if passed to pcre_exec().
2772    
2773    
2774     Version 0.93 15-Sep-97
2775     ----------------------
2776    
2777     1. /(b)|(:+)/ was computing an incorrect first character.
2778    
2779     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2780     but not actually doing anything yet.
2781    
2782     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2783     as Perl does (e.g. [-az] or [az-]).
2784    
2785     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2786     all possible positions.
2787    
2788     5. Split up into different modules to avoid including unneeded functions in a
2789     compiled binary. However, compile and exec are still in one module. The "study"
2790     function is split off.
2791    
2792     6. The character tables are now in a separate module whose source is generated
2793     by an auxiliary program - but can then be edited by hand if required. There are
2794     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2795     toupper() in the code.
2796    
2797     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2798     make them global. Abolish the function for setting them, as the caller can now
2799     set them directly.
2800    
2801    
2802     Version 0.92 11-Sep-97
2803     ----------------------
2804    
2805     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2806     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2807    
2808     2. Caseless matching was not working in character classes if the characters in
2809     the pattern were in upper case.
2810    
2811     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2812    
2813     4. Make PCRE_ANCHORED public and accept as a compile option.
2814    
2815     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2816     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2817     pass them.
2818    
2819     6. Give an error if bad option bits passed at compile or run time.
2820    
2821     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2822     pcretest to cause it to pass that flag.
2823    
2824     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2825     options, and the first character, if set.
2826    
2827     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2828    
2829    
2830     Version 0.91 10-Sep-97
2831     ----------------------
2832    
2833     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2834     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2835    
2836     2. PCRE was looping on encountering an indefinitely repeated back reference to
2837     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2838     Perl does - treats the match as successful.
2839    
2840     ****

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