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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 207 - (hide annotations) (download)
Mon Aug 6 09:32:14 2007 UTC (7 years, 3 months ago) by ph10
File size: 134834 byte(s)
Fix backtrack past start of subject bugs caused by the use of \X, \p, or \P in 
non-UTF-8 mode and the presence of top-bit-set characters.

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 203
68     10. Fixed a bug in the documentation for get/copy named substring when
69     duplicate names are permitted. If none of the named substrings are set, the
70     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
71     empty string.
72 ph10 205
73     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
74     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
75     because the ] is interpreted as the first data character and the
76     terminating ] is not found. PCRE has been made compatible with Perl in this
77     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
78     cause memory overwriting.
79 ph10 206
80     10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
81     string has been matched (to stop an infinite loop). It was not recognizing
82     a conditional subpattern that could match an empty string if that
83     subpattern was within another subpattern. For example, it looped when
84     trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
85     condition was not nested. This bug has been fixed.
86 ph10 207
87     12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
88     past the start of the subject in the presence of bytes with the top bit
89     set, for example "\x8aBCD".
90 ph10 200
91 ph10 202
92 ph10 189 Version 7.2 19-Jun-07
93     ---------------------
94    
95 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
96     which is apparently normally available under Windows.
97 ph10 159
98     2. Re-jig the pcregrep tests with different newline settings in an attempt
99     to make them independent of the local environment's newline setting.
100 ph10 157
101 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
102 ph10 165
103 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
104     was not 2, because the output contained actual offsets. The recent new
105 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
106     usable with all link sizes.
107    
108 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
109     stack recursion. This gives a massive performance boost under BSD, but just
110 ph10 165 a small improvement under Linux. However, it saves one field in the frame
111 ph10 164 in all cases.
112 ph10 172
113 ph10 166 6. Added more features from the forthcoming Perl 5.10:
114 ph10 172
115 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
116     recursion call. It refers to the nth most recently opened parentheses.
117 ph10 172
118 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
119 ph10 172 to be opened parentheses.
120    
121     (c) Conditions that refer to capturing parentheses can be specified
122 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
123 ph10 172
124 ph10 168 (d) \K resets the start of the current match so that everything before
125 ph10 172 is not part of it.
126    
127 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
128 ph10 172
129 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
130 ph10 172 reference syntax.
131 ph10 182
132 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
133 ph10 182 alternative starts with the same number.
134 ph10 172
135 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
136    
137 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
138     PCRE_INFO_JCHANGED.
139    
140     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
141     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
142 ph10 170 for detecting groups that can match an empty string.
143 ph10 159
144 ph10 172 9. A pattern with a very large number of alternatives (more than several
145     hundred) was running out of internal workspace during the pre-compile
146     phase, where pcre_compile() figures out how much memory will be needed. A
147     bit of new cunning has reduced the workspace needed for groups with
148     alternatives. The 1000-alternative test pattern now uses 12 bytes of
149     workspace instead of running out of the 4096 that are available.
150 ph10 182
151 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
152 ph10 172
153 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
154     The report of the bug said:
155 ph10 182
156 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
157     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
158 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
159 ph10 185
160 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
161 ph10 185 it matched the wrong number of bytes.
162 ph10 172
163 ph10 179
164 ph10 155 Version 7.1 24-Apr-07
165 ph10 98 ---------------------
166    
167 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
168 ph10 122 that is more "standard", making use of automake and other Autotools. There
169 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
170     on this.
171 ph10 111
172     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
173     for recursive directory scanning broke on some systems because the files
174     are not scanned in any specific order and on different systems the order
175     was different. A call to "sort" has been inserted into RunGrepTest for the
176     approprate test as a short-term fix. In the longer term there may be an
177 ph10 100 alternative.
178 ph10 111
179 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
180 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
181     people do their own conversions for various reasons). The problems
182     concerned the use of low-level troff macros .br and .in. I have therefore
183     removed all such uses from the man pages (some were redundant, some could
184 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
185     HTML has been updated to handle .nf/.fi and to complain if it encounters
186     .br or .in.
187 ph10 111
188 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
189 ph10 123 arranged for config.h to be included in the distribution, with the name
190 ph10 111 config.h.generic, for the benefit of those who have to compile without
191     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
192    
193     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
194     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
195 ph10 127 makevp.bat for the current PCRE, using the additional files
196 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
197 ph10 111
198     6. A Windows user reported a minor discrepancy with test 2, which turned out
199     to be caused by a trailing space on an input line that had got lost in his
200 ph10 102 copy. The trailing space was an accident, so I've just removed it.
201 ph10 111
202 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
203 ph10 111 that is needed.
204    
205 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
206     as "const" (a) because they are and (b) because it helps the PHP
207     maintainers who have recently made a script to detect big data structures
208 ph10 111 in the php code that should be moved to the .rodata section. I remembered
209     to update Builducptable as well, so it won't revert if ucptable.h is ever
210 ph10 105 re-created.
211 ph10 111
212     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
213 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
214 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
215     support is not required. This saves 1.5-2K of code, which is important in
216 ph10 107 some applications.
217 ph10 111
218 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
219 ph10 111 so as not to refer to the tables, even though these functions will never be
220     called when UTF-8 support is disabled. Otherwise there are problems with a
221     shared library.
222    
223 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
224    
225     (a) It was defining its arguments as char * instead of void *.
226    
227 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
228     a long time ago when I wrote it, but is no longer the case.
229    
230 ph10 118 The emulated memove() is provided for those environments that have neither
231 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
232 ph10 118 is clearly not the case, as these two bugs were recently reported.
233 ph10 123
234 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
235 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
236     pages, and it removes trailing spaces from listed files. It also creates
237     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
238     case, it wraps all the #defines with #ifndefs. This script should be run
239 ph10 111 before "make dist".
240 ph10 123
241 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
242     with Unicode property support.
243 ph10 123
244     (a) For a maximizing quantifier, if the two different cases of the
245     character were of different lengths in their UTF-8 codings (there are
246     some cases like this - I found 11), and the matching function had to
247 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
248     were both the same length.
249 ph10 123
250     (b) When PCRE was configured to use the heap rather than the stack for
251     recursion during matching, it was not correctly preserving the data for
252     the other case of a UTF-8 character when checking ahead for a match
253     while processing a minimizing repeat. If the check also involved
254 ph10 115 matching a wide character, but failed, corruption could cause an
255     erroneous result when trying to check for a repeat of the original
256     character.
257 ph10 123
258 ph10 116 13. Some tidying changes to the testing mechanism:
259 ph10 98
260 ph10 116 (a) The RunTest script now detects the internal link size and whether there
261     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
262 ph10 123 values substituted by "configure". (The RunGrepTest script already did
263     this for UTF-8.) The configure.ac script no longer substitutes the
264     relevant variables.
265    
266 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
267     with length and offset values. This means that the output is different
268     for different internal link sizes. Test 2 is skipped for link sizes
269     other than 2 because of this, bypassing the problem. Unfortunately,
270     there was also a test in test 3 (the locale tests) that used /B and
271 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
272     I have added a new /Z option to pcretest that replaces the length and
273     offset values with spaces. This is now used to make test 3 independent
274 ph10 122 of link size. (Test 2 will be tidied up later.)
275 ph10 123
276     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
277 ph10 122 segmentation fault instead of returning the appropriate error message.
278 ph10 134
279 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
280 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
281     This doesn't seem right; it now treats the CRLF combination as the line
282     ending, and so does not match in that case. It's only a pattern such as ^$
283     that would hit this one: something like ^ABC$ would have failed after \r
284     and then tried again after \r\n.
285    
286 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
287 ph10 134 in an attempt to make files that differ only in their line terminators
288     compare equal. This works on Linux.
289 ph10 142
290 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
291     as it exited. This is now fixed, thanks to valgrind.
292 ph10 142
293 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
294 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
295     was because its rules for how to advance for /g after matching an empty
296 ph10 143 string at the end of a line did not allow for this case. They now check for
297     it specially.
298 ph10 150
299     20. pcretest is supposed to handle patterns and data of any length, by
300     extending its buffers when necessary. It was getting this wrong when the
301 ph10 147 buffer for a data line had to be extended.
302 ph10 150
303 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
304 ph10 150 CRLF as a newline sequence.
305 ph10 152
306 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
307 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
308     I have nevertheless tidied it up.
309 ph10 111
310 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
311 ph10 123
312 ph10 153 24. Added a man page for pcre-config.
313 ph10 152
314 ph10 153
315 nigel 93 Version 7.0 19-Dec-06
316     ---------------------
317    
318     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
319     moving to gcc 4.1.1.
320    
321     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
322     sys/time.h, which is documented as needed for this function. It doesn't
323     seem to matter on Linux, but it showed up on some releases of OS X.
324    
325     3. It seems that there are systems where bytes whose values are greater than
326     127 match isprint() in the "C" locale. The "C" locale should be the
327     default when a C program starts up. In most systems, only ASCII printing
328     characters match isprint(). This difference caused the output from pcretest
329     to vary, making some of the tests fail. I have changed pcretest so that:
330    
331     (a) When it is outputting text in the compiled version of a pattern, bytes
332     other than 32-126 are always shown as hex escapes.
333    
334     (b) When it is outputting text that is a matched part of a subject string,
335     it does the same, unless a different locale has been set for the match
336     (using the /L modifier). In this case, it uses isprint() to decide.
337    
338     4. Fixed a major bug that caused incorrect computation of the amount of memory
339     required for a compiled pattern when options that changed within the
340     pattern affected the logic of the preliminary scan that determines the
341     length. The relevant options are -x, and -i in UTF-8 mode. The result was
342     that the computed length was too small. The symptoms of this bug were
343     either the PCRE error "internal error: code overflow" from pcre_compile(),
344     or a glibc crash with a message such as "pcretest: free(): invalid next
345     size (fast)". Examples of patterns that provoked this bug (shown in
346     pcretest format) are:
347    
348     /(?-x: )/x
349     /(?x)(?-x: \s*#\s*)/
350     /((?i)[\x{c0}])/8
351     /(?i:[\x{c0}])/8
352    
353     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
354     is now done differently.
355    
356     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
357     wrapper classes; (b) implement a new function in the C++ scanner that is
358     more efficient than the old way of doing things because it avoids levels of
359     recursion in the regex matching; (c) add a paragraph to the documentation
360     for the FullMatch() function.
361    
362     6. The escape sequence \n was being treated as whatever was defined as
363     "newline". Not only was this contrary to the documentation, which states
364     that \n is character 10 (hex 0A), but it also went horribly wrong when
365     "newline" was defined as CRLF. This has been fixed.
366    
367     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
368     was being set to -1 for the "end of line" case (supposedly a value that no
369     character can have). Though this value is never used (the check for end of
370     line is "zero bytes in current character"), it caused compiler complaints.
371     I've changed it to 0xffffffff.
372    
373     8. In pcre_version.c, the version string was being built by a sequence of
374     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
375     string (as it is for production releases) called a macro with an empty
376     argument. The C standard says the result of this is undefined. The gcc
377     compiler treats it as an empty string (which was what was wanted) but it is
378     reported that Visual C gives an error. The source has been hacked around to
379     avoid this problem.
380    
381     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
382     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
383     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
384     of them did).
385    
386     10. Originally, pcretest opened its input and output without "b"; then I was
387     told that "b" was needed in some environments, so it was added for release
388     5.0 to both the input and output. (It makes no difference on Unix-like
389     systems.) Later I was told that it is wrong for the input on Windows. I've
390     now abstracted the modes into two macros, to make it easier to fiddle with
391     them, and removed "b" from the input mode under Windows.
392    
393     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
394    
395     12. Added -help and --help to pcretest as an official way of being reminded
396     of the options.
397    
398     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
399     and pcrecpp.cc because they annoy compilers at high warning levels.
400    
401     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
402    
403     15. Fixed an occurrence of == in configure.ac that should have been = (shell
404     scripts are not C programs :-) and which was not noticed because it works
405     on Linux.
406    
407     16. pcretest is supposed to handle any length of pattern and data line (as one
408     line or as a continued sequence of lines) by extending its input buffer if
409     necessary. This feature was broken for very long pattern lines, leading to
410     a string of junk being passed to pcre_compile() if the pattern was longer
411     than about 50K.
412    
413     17. I have done a major re-factoring of the way pcre_compile() computes the
414     amount of memory needed for a compiled pattern. Previously, there was code
415     that made a preliminary scan of the pattern in order to do this. That was
416     OK when PCRE was new, but as the facilities have expanded, it has become
417     harder and harder to keep it in step with the real compile phase, and there
418     have been a number of bugs (see for example, 4 above). I have now found a
419     cunning way of running the real compile function in a "fake" mode that
420     enables it to compute how much memory it would need, while actually only
421     ever using a few hundred bytes of working memory and without too many
422     tests of the mode. This should make future maintenance and development
423     easier. A side effect of this work is that the limit of 200 on the nesting
424     depth of parentheses has been removed (though this was never a serious
425     limitation, I suspect). However, there is a downside: pcre_compile() now
426     runs more slowly than before (30% or more, depending on the pattern). I
427     hope this isn't a big issue. There is no effect on runtime performance.
428    
429     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
430     newline (only possible for the last line of a file) and it was a
431     pattern that set a locale (followed by /Lsomething), pcretest crashed.
432    
433     19. Added additional timing features to pcretest. (1) The -tm option now times
434     matching only, not compiling. (2) Both -t and -tm can be followed, as a
435     separate command line item, by a number that specifies the number of
436     repeats to use when timing. The default is 50000; this gives better
437     precision, but takes uncomfortably long for very large patterns.
438    
439     20. Extended pcre_study() to be more clever in cases where a branch of a
440     subpattern has no definite first character. For example, (a*|b*)[cd] would
441     previously give no result from pcre_study(). Now it recognizes that the
442     first character must be a, b, c, or d.
443    
444     21. There was an incorrect error "recursive call could loop indefinitely" if
445     a subpattern (or the entire pattern) that was being tested for matching an
446     empty string contained only one non-empty item after a nested subpattern.
447     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
448     incorrectly, because the \d was being skipped in the check.
449    
450     22. The pcretest program now has a new pattern option /B and a command line
451     option -b, which is equivalent to adding /B to every pattern. This causes
452     it to show the compiled bytecode, without the additional information that
453     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
454     is the same as /B/I).
455    
456     23. A new optimization is now able automatically to treat some sequences such
457     as a*b as a*+b. More specifically, if something simple (such as a character
458     or a simple class like \d) has an unlimited quantifier, and is followed by
459     something that cannot possibly match the quantified thing, the quantifier
460     is automatically "possessified".
461    
462     24. A recursive reference to a subpattern whose number was greater than 39
463     went wrong under certain circumstances in UTF-8 mode. This bug could also
464     have affected the operation of pcre_study().
465    
466     25. Realized that a little bit of performance could be had by replacing
467     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
468    
469     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
470    
471     27. Possessive quantifiers such as a++ were previously implemented by turning
472     them into atomic groups such as ($>a+). Now they have their own opcodes,
473     which improves performance. This includes the automatically created ones
474     from 23 above.
475    
476     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
477     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
478     the first matched character to be a colon. This applied both to named and
479     numbered groups.
480    
481     29. The ucpinternal.h header file was missing its idempotency #ifdef.
482    
483     30. I was sent a "project" file called libpcre.a.dev which I understand makes
484     building PCRE on Windows easier, so I have included it in the distribution.
485    
486     31. There is now a check in pcretest against a ridiculously large number being
487     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
488     loop, the loop is abandoned.
489    
490     32. Forward references to subpatterns in conditions such as (?(2)...) where
491     subpattern 2 is defined later cause pcre_compile() to search forwards in
492     the pattern for the relevant set of parentheses. This search went wrong
493     when there were unescaped parentheses in a character class, parentheses
494     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
495    
496     33. "Subroutine" calls and backreferences were previously restricted to
497     referencing subpatterns earlier in the regex. This restriction has now
498     been removed.
499    
500     34. Added a number of extra features that are going to be in Perl 5.10. On the
501     whole, these are just syntactic alternatives for features that PCRE had
502     previously implemented using the Python syntax or my own invention. The
503     other formats are all retained for compatibility.
504    
505     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
506     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
507     also .NET compatible.
508    
509     (b) A recursion or subroutine call to a named group can now be defined as
510     (?&name) as well as (?P>name).
511    
512     (c) A backreference to a named group can now be defined as \k<name> or
513     \k'name' as well as (?P=name). The new forms, as well as being in Perl
514     5.10, are also .NET compatible.
515    
516     (d) A conditional reference to a named group can now use the syntax
517     (?(<name>) or (?('name') as well as (?(name).
518    
519     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
520     groups (named and numbered) that are never evaluated inline, but can be
521     called as "subroutines" from elsewhere. In effect, the DEFINE condition
522     is always false. There may be only one alternative in such a group.
523    
524     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
525     as the simple (?(R). The condition is true only if the most recent
526     recursion is that of the given number or name. It does not search out
527     through the entire recursion stack.
528    
529     (g) The escape \gN or \g{N} has been added, where N is a positive or
530     negative number, specifying an absolute or relative reference.
531    
532     35. Tidied to get rid of some further signed/unsigned compiler warnings and
533     some "unreachable code" warnings.
534    
535     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
536     things, this adds five new scripts.
537    
538     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
539     There were also incompatibilities regarding the handling of \Q..\E inside
540     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
541     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
542    
543     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
544     matches an empty string, and forcibly breaks the loop. There were bugs in
545     this code in non-simple cases. For a pattern such as ^(a()*)* matched
546     against aaaa the result was just "a" rather than "aaaa", for example. Two
547     separate and independent bugs (that affected different cases) have been
548     fixed.
549    
550     39. Refactored the code to abolish the use of different opcodes for small
551     capturing bracket numbers. This is a tidy that I avoided doing when I
552     removed the limit on the number of capturing brackets for 3.5 back in 2001.
553     The new approach is not only tidier, it makes it possible to reduce the
554     memory needed to fix the previous bug (38).
555    
556     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
557     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
558     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
559     mode.
560    
561     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
562     report.
563    
564     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
565     copy construction and assignment in the C++ wrapper.
566    
567     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
568     couple of bugs that could have given wrong results in the "--newline=crlf"
569     case.
570    
571     44. Added a number of casts and did some reorganization of signed/unsigned int
572     variables following suggestions from Dair Grant. Also renamed the variable
573     "this" as "item" because it is a C++ keyword.
574    
575     45. Arranged for dftables to add
576    
577     #include "pcre_internal.h"
578    
579     to pcre_chartables.c because without it, gcc 4.x may remove the array
580     definition from the final binary if PCRE is built into a static library and
581     dead code stripping is activated.
582    
583     46. For an unanchored pattern, if a match attempt fails at the start of a
584     newline sequence, and the newline setting is CRLF or ANY, and the next two
585     characters are CRLF, advance by two characters instead of one.
586    
587    
588 nigel 91 Version 6.7 04-Jul-06
589     ---------------------
590    
591     1. In order to handle tests when input lines are enormously long, pcretest has
592     been re-factored so that it automatically extends its buffers when
593     necessary. The code is crude, but this _is_ just a test program. The
594     default size has been increased from 32K to 50K.
595    
596     2. The code in pcre_study() was using the value of the re argument before
597     testing it for NULL. (Of course, in any sensible call of the function, it
598     won't be NULL.)
599    
600     3. The memmove() emulation function in pcre_internal.h, which is used on
601     systems that lack both memmove() and bcopy() - that is, hardly ever -
602     was missing a "static" storage class specifier.
603    
604     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
605     containing an extended class (one that cannot be represented by a bitmap
606     because it contains high-valued characters or Unicode property items, e.g.
607     [\pZ]). Almost always one would set UTF-8 mode when processing such a
608     pattern, but PCRE should not loop if you do not (it no longer does).
609     [Detail: two cases were found: (a) a repeated subpattern containing an
610     extended class; (b) a recursive reference to a subpattern that followed a
611     previous extended class. It wasn't skipping over the extended class
612     correctly when UTF-8 mode was not set.]
613    
614     5. A negated single-character class was not being recognized as fixed-length
615     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
616     compile error "lookbehind assertion is not fixed length".
617    
618     6. The RunPerlTest auxiliary script was showing an unexpected difference
619     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
620     write a Perl script that can interpret lines of an input file either as
621     byte characters or as UTF-8, which is what "perltest" was being required to
622     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
623     can't do is switch easily at run time between having the "use utf8;" pragma
624     or not. In the end, I fudged it by using the RunPerlTest script to insert
625     "use utf8;" explicitly for the UTF-8 tests.
626    
627     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
628     the end of the subject string, contrary to the documentation and to what
629     Perl does. This was true of both matching functions. Now it matches only at
630     the start of the subject and immediately after *internal* newlines.
631    
632     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
633     a pointer to an int instead of a pointer to an unsigned long int. This
634     caused problems on 64-bit systems.
635    
636     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
637     instance of the 'standard' template library not being so standard".
638    
639     10. There was no check on the number of named subpatterns nor the maximum
640     length of a subpattern name. The product of these values is used to compute
641     the size of the memory block for a compiled pattern. By supplying a very
642     long subpattern name and a large number of named subpatterns, the size
643     computation could be caused to overflow. This is now prevented by limiting
644     the length of names to 32 characters, and the number of named subpatterns
645     to 10,000.
646    
647     11. Subpatterns that are repeated with specific counts have to be replicated in
648     the compiled pattern. The size of memory for this was computed from the
649     length of the subpattern and the repeat count. The latter is limited to
650     65535, but there was no limit on the former, meaning that integer overflow
651     could in principle occur. The compiled length of a repeated subpattern is
652     now limited to 30,000 bytes in order to prevent this.
653    
654     12. Added the optional facility to have named substrings with the same name.
655    
656     13. Added the ability to use a named substring as a condition, using the
657     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
658     are numbers (not recommended). Forward references are permitted.
659    
660     14. Added forward references in named backreferences (if you see what I mean).
661    
662     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
663     pattern could run off the end of the subject. For example, the pattern
664     "(?s)(.{1,5})"8 did this with the subject "ab".
665    
666     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
667     PCRE_CASELESS was set when matching characters that were quantified with ?
668     or *.
669    
670     17. A character class other than a single negated character that had a minimum
671     but no maximum quantifier - for example [ab]{6,} - was not handled
672     correctly by pce_dfa_exec(). It would match only one character.
673    
674     18. A valid (though odd) pattern that looked like a POSIX character
675     class but used an invalid character after [ (for example [[,abc,]]) caused
676     pcre_compile() to give the error "Failed: internal error: code overflow" or
677     in some cases to crash with a glibc free() error. This could even happen if
678     the pattern terminated after [[ but there just happened to be a sequence of
679     letters, a binary zero, and a closing ] in the memory that followed.
680    
681     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
682     over the years. Originally (before any Unicode support), just the bottom 8
683     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
684     output from "man perlunicode" includes this:
685    
686     The regular expression compiler produces polymorphic opcodes. That
687     is, the pattern adapts to the data and automatically switches to
688     the Unicode character scheme when presented with Unicode data--or
689     instead uses a traditional byte scheme when presented with byte
690     data.
691    
692     Sadly, a wide octal escape does not cause a switch, and in a string with
693     no other multibyte characters, these octal escapes are treated as before.
694     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
695     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
696     Unicode string.
697    
698     I have not perpetrated such confusion in PCRE. Up till now, it took just
699     the bottom 8 bits, as in old Perl. I have now made octal escapes with
700     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
701     translate to the appropriate multibyte character.
702    
703     29. Applied some refactoring to reduce the number of warnings from Microsoft
704     and Borland compilers. This has included removing the fudge introduced
705     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
706     a warning about an unused variable.
707    
708     21. PCRE has not included VT (character 0x0b) in the set of whitespace
709     characters since release 4.0, because Perl (from release 5.004) does not.
710     [Or at least, is documented not to: some releases seem to be in conflict
711     with the documentation.] However, when a pattern was studied with
712     pcre_study() and all its branches started with \s, PCRE still included VT
713     as a possible starting character. Of course, this did no harm; it just
714     caused an unnecessary match attempt.
715    
716     22. Removed a now-redundant internal flag bit that recorded the fact that case
717     dependency changed within the pattern. This was once needed for "required
718     byte" processing, but is no longer used. This recovers a now-scarce options
719     bit. Also moved the least significant internal flag bit to the most-
720     significant bit of the word, which was not previously used (hangover from
721     the days when it was an int rather than a uint) to free up another bit for
722     the future.
723    
724     23. Added support for CRLF line endings as well as CR and LF. As well as the
725     default being selectable at build time, it can now be changed at runtime
726     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
727     specify that it is scanning data with non-default line endings.
728    
729     24. Changed the definition of CXXLINK to make it agree with the definition of
730     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
731    
732     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
733     recursions. This makes a big different to stack usage for some patterns.
734    
735     26. If a subpattern containing a named recursion or subroutine reference such
736     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
737     the space required for the compiled pattern went wrong and gave too small a
738     value. Depending on the environment, this could lead to "Failed: internal
739     error: code overflow at offset 49" or "glibc detected double free or
740     corruption" errors.
741    
742     27. Applied patches from Google (a) to support the new newline modes and (b) to
743     advance over multibyte UTF-8 characters in GlobalReplace.
744    
745     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
746     difference for some implementation of PCRE in some Windows version.
747    
748     29. Added some extra testing facilities to pcretest:
749    
750     \q<number> in a data line sets the "match limit" value
751     \Q<number> in a data line sets the "match recursion limt" value
752     -S <number> sets the stack size, where <number> is in megabytes
753    
754     The -S option isn't available for Windows.
755    
756    
757 nigel 89 Version 6.6 06-Feb-06
758     ---------------------
759    
760     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
761     in pcreposix.h. I have copied the definition from pcre.h.
762    
763     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
764     because pcre.h is no longer a built file.
765    
766     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
767     not normally included in the compiled code.
768    
769    
770 nigel 87 Version 6.5 01-Feb-06
771     ---------------------
772    
773     1. When using the partial match feature with pcre_dfa_exec(), it was not
774     anchoring the second and subsequent partial matches at the new starting
775     point. This could lead to incorrect results. For example, with the pattern
776     /1234/, partially matching against "123" and then "a4" gave a match.
777    
778     2. Changes to pcregrep:
779    
780     (a) All non-match returns from pcre_exec() were being treated as failures
781     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
782     error message is output. Some extra information is given for the
783     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
784     probably the only errors that are likely to be caused by users (by
785     specifying a regex that has nested indefinite repeats, for instance).
786     If there are more than 20 of these errors, pcregrep is abandoned.
787    
788     (b) A binary zero was treated as data while matching, but terminated the
789     output line if it was written out. This has been fixed: binary zeroes
790     are now no different to any other data bytes.
791    
792     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
793     used to set a locale for matching. The --locale=xxxx long option has
794     been added (no short equivalent) to specify a locale explicitly on the
795     pcregrep command, overriding the environment variables.
796    
797     (d) When -B was used with -n, some line numbers in the output were one less
798     than they should have been.
799    
800     (e) Added the -o (--only-matching) option.
801    
802     (f) If -A or -C was used with -c (count only), some lines of context were
803     accidentally printed for the final match.
804    
805     (g) Added the -H (--with-filename) option.
806    
807     (h) The combination of options -rh failed to suppress file names for files
808     that were found from directory arguments.
809    
810     (i) Added the -D (--devices) and -d (--directories) options.
811    
812     (j) Added the -F (--fixed-strings) option.
813    
814     (k) Allow "-" to be used as a file name for -f as well as for a data file.
815    
816     (l) Added the --colo(u)r option.
817    
818     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
819     is not present by default.
820    
821     3. A nasty bug was discovered in the handling of recursive patterns, that is,
822     items such as (?R) or (?1), when the recursion could match a number of
823     alternatives. If it matched one of the alternatives, but subsequently,
824     outside the recursion, there was a failure, the code tried to back up into
825     the recursion. However, because of the way PCRE is implemented, this is not
826     possible, and the result was an incorrect result from the match.
827    
828     In order to prevent this happening, the specification of recursion has
829     been changed so that all such subpatterns are automatically treated as
830     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
831    
832     4. I had overlooked the fact that, in some locales, there are characters for
833     which isalpha() is true but neither isupper() nor islower() are true. In
834     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
835     and ordfeminine) are like this. This affected the treatment of \w and \W
836     when they appeared in character classes, but not when they appeared outside
837     a character class. The bit map for "word" characters is now created
838     separately from the results of isalnum() instead of just taking it from the
839     upper, lower, and digit maps. (Plus the underscore character, of course.)
840    
841     5. The above bug also affected the handling of POSIX character classes such as
842     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
843     permanent tables. Instead, the bit maps for such a class were previously
844     created as the appropriate unions of the upper, lower, and digit bitmaps.
845     Now they are created by subtraction from the [[:word:]] class, which has
846     its own bitmap.
847    
848     6. The [[:blank:]] character class matches horizontal, but not vertical space.
849     It is created by subtracting the vertical space characters (\x09, \x0a,
850     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
851     subtraction was done in the overall bitmap for a character class, meaning
852     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
853     be recognized. This bug has been fixed.
854    
855     7. Patches from the folks at Google:
856    
857     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
858     real life, but is still worth protecting against".
859    
860     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
861     regular expressions".
862    
863     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
864     have it.
865    
866     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
867     "configure" and the latter not, in order to fix a problem somebody had
868     with compiling the Arg class on HP-UX.
869    
870     (e) Improve the error-handling of the C++ wrapper a little bit.
871    
872     (f) New tests for checking recursion limiting.
873    
874     8. The pcre_memmove() function, which is used only if the environment does not
875     have a standard memmove() function (and is therefore rarely compiled),
876     contained two bugs: (a) use of int instead of size_t, and (b) it was not
877     returning a result (though PCRE never actually uses the result).
878    
879     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
880     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
881     returned instead of calling malloc() with an overflowing number that would
882     most likely cause subsequent chaos.
883    
884     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
885    
886     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
887     with this option is matched, the nmatch and pmatch options of regexec() are
888     ignored.
889    
890     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
891     provided in case anyone wants to the the POSIX interface with UTF-8
892     strings.
893    
894     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
895     C++ linking (needed for some HP-UX environments).
896    
897     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
898     (unused parameter) and in the pcre_printint() function (omitted "default"
899     switch label when the default is to do nothing).
900    
901     15. Added some code to make it possible, when PCRE is compiled as a C++
902     library, to replace subject pointers for pcre_exec() with a smart pointer
903     class, thus making it possible to process discontinuous strings.
904    
905     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
906     much the same function. They were added by different people who were trying
907     to make PCRE easy to compile on non-Unix systems. It has been suggested
908     that PCRE_EXPORT be abolished now that there is more automatic apparatus
909     for compiling on Windows systems. I have therefore replaced it with
910     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
911     defaults to "extern" for C or "extern C" for C++, which works fine on
912     Unix-like systems. It is now possible to override the value of PCRE_DATA_
913     SCOPE with something explicit in config.h. In addition:
914    
915     (a) pcreposix.h still had just "extern" instead of either of these macros;
916     I have replaced it with PCRE_DATA_SCOPE.
917    
918     (b) Functions such as _pcre_xclass(), which are internal to the library,
919     but external in the C sense, all had PCRE_EXPORT in their definitions.
920     This is apparently wrong for the Windows case, so I have removed it.
921     (It makes no difference on Unix-like systems.)
922    
923     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
924     of recursive calls to match(). This is different to MATCH_LIMIT because
925     that limits the total number of calls to match(), not all of which increase
926     the depth of recursion. Limiting the recursion depth limits the amount of
927     stack (or heap if NO_RECURSE is set) that is used. The default can be set
928     when PCRE is compiled, and changed at run time. A patch from Google adds
929     this functionality to the C++ interface.
930    
931     18. Changes to the handling of Unicode character properties:
932    
933     (a) Updated the table to Unicode 4.1.0.
934    
935     (b) Recognize characters that are not in the table as "Cn" (undefined).
936    
937     (c) I revised the way the table is implemented to a much improved format
938     which includes recognition of ranges. It now supports the ranges that
939     are defined in UnicodeData.txt, and it also amalgamates other
940     characters into ranges. This has reduced the number of entries in the
941     table from around 16,000 to around 3,000, thus reducing its size
942     considerably. I realized I did not need to use a tree structure after
943     all - a binary chop search is just as efficient. Having reduced the
944     number of entries, I extended their size from 6 bytes to 8 bytes to
945     allow for more data.
946    
947     (d) Added support for Unicode script names via properties such as \p{Han}.
948    
949     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
950     matching that character.
951    
952     20. When matching a repeated Unicode property with a minimum greater than zero,
953     (for example \pL{2,}), PCRE could look past the end of the subject if it
954     reached it while seeking the minimum number of characters. This could
955     happen only if some of the characters were more than one byte long, because
956     there is a check for at least the minimum number of bytes.
957    
958     21. Refactored the implementation of \p and \P so as to be more general, to
959     allow for more different types of property in future. This has changed the
960     compiled form incompatibly. Anybody with saved compiled patterns that use
961     \p or \P will have to recompile them.
962    
963     22. Added "Any" and "L&" to the supported property types.
964    
965     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
966     but give a compile time error if the value is greater than 0xff.
967    
968     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
969     accidentally not being installed or uninstalled.
970    
971     25. The pcre.h file was built from pcre.h.in, but the only changes that were
972     made were to insert the current release number. This seemed silly, because
973     it made things harder for people building PCRE on systems that don't run
974     "configure". I have turned pcre.h into a distributed file, no longer built
975     by "configure", with the version identification directly included. There is
976     no longer a pcre.h.in file.
977    
978     However, this change necessitated a change to the pcre-config script as
979     well. It is built from pcre-config.in, and one of the substitutions was the
980     release number. I have updated configure.ac so that ./configure now finds
981     the release number by grepping pcre.h.
982    
983     26. Added the ability to run the tests under valgrind.
984    
985    
986 nigel 85 Version 6.4 05-Sep-05
987     ---------------------
988    
989     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
990     "--" to be printed when multiple files were scanned, even when none of the
991     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
992     consider it to be a bug, and have restored the previous behaviour.
993    
994     2. A couple of code tidies to get rid of compiler warnings.
995    
996     3. The pcretest program used to cheat by referring to symbols in the library
997     whose names begin with _pcre_. These are internal symbols that are not
998     really supposed to be visible externally, and in some environments it is
999     possible to suppress them. The cheating is now confined to including
1000     certain files from the library's source, which is a bit cleaner.
1001    
1002     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1003     file's purpose clearer.
1004    
1005     5. Reorganized pcre_ucp_findchar().
1006    
1007    
1008 nigel 83 Version 6.3 15-Aug-05
1009     ---------------------
1010    
1011     1. The file libpcre.pc.in did not have general read permission in the tarball.
1012    
1013     2. There were some problems when building without C++ support:
1014    
1015     (a) If C++ support was not built, "make install" and "make test" still
1016     tried to test it.
1017    
1018     (b) There were problems when the value of CXX was explicitly set. Some
1019     changes have been made to try to fix these, and ...
1020    
1021     (c) --disable-cpp can now be used to explicitly disable C++ support.
1022    
1023     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1024     backslash in a target when C++ was disabled. This confuses some
1025     versions of "make", apparently. Using an intermediate variable solves
1026     this. (Same for CPP_LOBJ.)
1027    
1028     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1029     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1030     necessary on certain architectures.
1031    
1032     4. Added a setting of -export-symbols-regex to the link command to remove
1033     those symbols that are exported in the C sense, but actually are local
1034     within the library, and not documented. Their names all begin with
1035     "_pcre_". This is not a perfect job, because (a) we have to except some
1036     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1037     available (and never for static libraries). I have made a note to try to
1038     find a way round (a) in the future.
1039    
1040    
1041 nigel 81 Version 6.2 01-Aug-05
1042     ---------------------
1043    
1044     1. There was no test for integer overflow of quantifier values. A construction
1045     such as {1111111111111111} would give undefined results. What is worse, if
1046     a minimum quantifier for a parenthesized subpattern overflowed and became
1047     negative, the calculation of the memory size went wrong. This could have
1048     led to memory overwriting.
1049    
1050     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1051    
1052     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1053     operating environments where this matters.
1054    
1055     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1056     PCRE options from within the C++ wrapper.
1057    
1058     5. Named capturing subpatterns were not being correctly counted when a pattern
1059     was compiled. This caused two problems: (a) If there were more than 100
1060     such subpatterns, the calculation of the memory needed for the whole
1061     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1062     back references of the form \12, where the number was greater than 9, were
1063     not recognized as back references, even though there were sufficient
1064     previous subpatterns.
1065    
1066     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1067     versions of gcc, e.g. 2.95.4.
1068    
1069    
1070 nigel 79 Version 6.1 21-Jun-05
1071     ---------------------
1072    
1073     1. There was one reference to the variable "posix" in pcretest.c that was not
1074     surrounded by "#if !defined NOPOSIX".
1075    
1076     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1077     the cross-check on the old pcre_info() function, for the benefit of the
1078     cut-down version of PCRE that is currently imported into Exim.
1079    
1080     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1081     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1082     patterns that start (?i) but I don't think that matters. The use of (?i) is
1083     just an example; this all applies to the other options as well.
1084    
1085     4. Since libtool seems to echo the compile commands it is issuing, the output
1086     from "make" can be reduced a bit by putting "@" in front of each libtool
1087     compile command.
1088    
1089     5. Patch from the folks at Google for configure.in to be a bit more thorough
1090     in checking for a suitable C++ installation before trying to compile the
1091     C++ stuff. This should fix a reported problem when a compiler was present,
1092     but no suitable headers.
1093    
1094     6. The man pages all had just "PCRE" as their title. I have changed them to
1095     be the relevant file name. I have also arranged that these names are
1096     retained in the file doc/pcre.txt, which is a concatenation in text format
1097     of all the man pages except the little individual ones for each function.
1098    
1099     7. The NON-UNIX-USE file had not been updated for the different set of source
1100     files that come with release 6. I also added a few comments about the C++
1101     wrapper.
1102    
1103    
1104 nigel 77 Version 6.0 07-Jun-05
1105     ---------------------
1106    
1107     1. Some minor internal re-organization to help with my DFA experiments.
1108    
1109     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1110     didn't matter for the library itself when fully configured, but did matter
1111     when compiling without UCP support, or within Exim, where the ucp files are
1112     not imported.
1113    
1114     3. Refactoring of the library code to split up the various functions into
1115     different source modules. The addition of the new DFA matching code (see
1116     below) to a single monolithic source would have made it really too
1117     unwieldy, quite apart from causing all the code to be include in a
1118     statically linked application, when only some functions are used. This is
1119     relevant even without the DFA addition now that patterns can be compiled in
1120     one application and matched in another.
1121    
1122     The downside of splitting up is that there have to be some external
1123     functions and data tables that are used internally in different modules of
1124     the library but which are not part of the API. These have all had their
1125     names changed to start with "_pcre_" so that they are unlikely to clash
1126     with other external names.
1127    
1128     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1129     a different (DFA) algorithm. Although it is slower than the original
1130     function, it does have some advantages for certain types of matching
1131     problem.
1132    
1133     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1134     including restarting after a partial match.
1135    
1136     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1137     defined when compiling for Windows was sent to me. I have put it into the
1138     code, though I have no means of testing or verifying it.
1139    
1140     7. Added the pcre_refcount() auxiliary function.
1141    
1142     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1143     match before or at the first newline in the subject string. In pcretest,
1144     the /f option on a pattern can be used to set this.
1145    
1146     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1147     would behave wrongly. This has been present in PCRE since release 4.0.
1148    
1149     10. A number of changes to the pcregrep command:
1150    
1151     (a) Refactored how -x works; insert ^(...)$ instead of setting
1152     PCRE_ANCHORED and checking the length, in preparation for adding
1153     something similar for -w.
1154    
1155     (b) Added the -w (match as a word) option.
1156    
1157     (c) Refactored the way lines are read and buffered so as to have more
1158     than one at a time available.
1159    
1160     (d) Implemented a pcregrep test script.
1161    
1162     (e) Added the -M (multiline match) option. This allows patterns to match
1163     over several lines of the subject. The buffering ensures that at least
1164     8K, or the rest of the document (whichever is the shorter) is available
1165     for matching (and similarly the previous 8K for lookbehind assertions).
1166    
1167     (f) Changed the --help output so that it now says
1168    
1169     -w, --word-regex(p)
1170    
1171     instead of two lines, one with "regex" and the other with "regexp"
1172     because that confused at least one person since the short forms are the
1173     same. (This required a bit of code, as the output is generated
1174     automatically from a table. It wasn't just a text change.)
1175    
1176     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1177     option but starts with a hyphen. Could be a pattern or a path name
1178     starting with a hyphen, for instance.
1179    
1180     (h) "-" can be given as a file name to represent stdin.
1181    
1182     (i) When file names are being printed, "(standard input)" is used for
1183     the standard input, for compatibility with GNU grep. Previously
1184     "<stdin>" was used.
1185    
1186     (j) The option --label=xxx can be used to supply a name to be used for
1187     stdin when file names are being printed. There is no short form.
1188    
1189     (k) Re-factored the options decoding logic because we are going to add
1190     two more options that take data. Such options can now be given in four
1191     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1192    
1193     (l) Added the -A, -B, and -C options for requesting that lines of context
1194     around matches be printed.
1195    
1196     (m) Added the -L option to print the names of files that do not contain
1197     any matching lines, that is, the complement of -l.
1198    
1199     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1200     continue to scan other files.
1201    
1202     (o) The -s option was incorrectly implemented. For compatibility with other
1203     greps, it now suppresses the error message for a non-existent or non-
1204     accessible file (but not the return code). There is a new option called
1205     -q that suppresses the output of matching lines, which was what -s was
1206     previously doing.
1207    
1208     (p) Added --include and --exclude options to specify files for inclusion
1209     and exclusion when recursing.
1210    
1211     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1212     Hopefully, it now does.
1213    
1214     12. Missing cast in pcre_study().
1215    
1216     13. Added an "uninstall" target to the makefile.
1217    
1218     14. Replaced "extern" in the function prototypes in Makefile.in with
1219     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1220     world, but is set differently for Windows.
1221    
1222     15. Added a second compiling function called pcre_compile2(). The only
1223     difference is that it has an extra argument, which is a pointer to an
1224     integer error code. When there is a compile-time failure, this is set
1225     non-zero, in addition to the error test pointer being set to point to an
1226     error message. The new argument may be NULL if no error number is required
1227     (but then you may as well call pcre_compile(), which is now just a
1228     wrapper). This facility is provided because some applications need a
1229     numeric error indication, but it has also enabled me to tidy up the way
1230     compile-time errors are handled in the POSIX wrapper.
1231    
1232     16. Added VPATH=.libs to the makefile; this should help when building with one
1233     prefix path and installing with another. (Or so I'm told by someone who
1234     knows more about this stuff than I do.)
1235    
1236     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1237     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1238     match everything, including newlines. This is not POSIX-compatible, but
1239     somebody wanted the feature. From pcretest it can be activated by using
1240     both the P and the s flags.
1241    
1242     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1243    
1244     19. libpcre.pc was being incorrectly installed as executable.
1245    
1246     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1247     it now also looks for '\r' so that it will work unmodified on Windows.
1248    
1249     21. Added Google's contributed C++ wrapper to the distribution.
1250    
1251     22. Added some untidy missing memory free() calls in pcretest, to keep
1252     Electric Fence happy when testing.
1253    
1254    
1255    
1256 nigel 75 Version 5.0 13-Sep-04
1257     ---------------------
1258    
1259     1. Internal change: literal characters are no longer packed up into items
1260     containing multiple characters in a single byte-string. Each character
1261     is now matched using a separate opcode. However, there may be more than one
1262     byte in the character in UTF-8 mode.
1263    
1264     2. The pcre_callout_block structure has two new fields: pattern_position and
1265     next_item_length. These contain the offset in the pattern to the next match
1266     item, and its length, respectively.
1267    
1268     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1269     insertion of callouts before each pattern item. Added the /C option to
1270     pcretest to make use of this.
1271    
1272     4. On the advice of a Windows user, the lines
1273    
1274     #if defined(_WIN32) || defined(WIN32)
1275     _setmode( _fileno( stdout ), 0x8000 );
1276     #endif /* defined(_WIN32) || defined(WIN32) */
1277    
1278     have been added to the source of pcretest. This apparently does useful
1279     magic in relation to line terminators.
1280    
1281     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1282     for the benefit of those environments where the "b" makes a difference.
1283    
1284     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1285     to know about it. I have put a hack into configure.in that adds in code
1286     to set GCC=yes if CC=icc. This seems to end up at a point in the
1287     generated configure script that is early enough to affect the setting of
1288     compiler options, which is what is needed, but I have no means of testing
1289     whether it really works. (The user who reported this had patched the
1290     generated configure script, which of course I cannot do.)
1291    
1292     LATER: After change 22 below (new libtool files), the configure script
1293     seems to know about icc (and also ecc). Therefore, I have commented out
1294     this hack in configure.in.
1295    
1296     7. Added support for pkg-config (2 patches were sent in).
1297    
1298     8. Negated POSIX character classes that used a combination of internal tables
1299     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1300     [[:^ascii]]. Typically, they would match almost any characters. The other
1301     POSIX classes were not broken in this way.
1302    
1303     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1304     to find the match, as PCRE was deluded into thinking that the match had to
1305     start at the start point or following a newline. The same bug applied to
1306     patterns with negative forward assertions or any backward assertions
1307     preceding ".*" at the start, unless the pattern required a fixed first
1308     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1309    
1310     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1311     starting at the last subject character, bytes beyond the end of the subject
1312     string were read.
1313    
1314     11. Renamed the variable "class" as "classbits" to make life easier for C++
1315     users. (Previously there was a macro definition, but it apparently wasn't
1316     enough.)
1317    
1318     12. Added the new field "tables" to the extra data so that tables can be passed
1319     in at exec time, or the internal tables can be re-selected. This allows
1320     a compiled regex to be saved and re-used at a later time by a different
1321     program that might have everything at different addresses.
1322    
1323     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1324     -R library as well as a -L library.
1325    
1326     14. The debugging options of pcretest (-d on the command line or D on a
1327     pattern) showed incorrect output for anything following an extended class
1328     that contained multibyte characters and which was followed by a quantifier.
1329    
1330     15. Added optional support for general category Unicode character properties
1331     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1332     support. It adds about 90K to the size of the library. The meanings of the
1333     inbuilt class escapes such as \d and \s have NOT been changed.
1334    
1335     16. Updated pcredemo.c to include calls to free() to release the memory for the
1336     compiled pattern.
1337    
1338     17. The generated file chartables.c was being created in the source directory
1339     instead of in the building directory. This caused the build to fail if the
1340     source directory was different from the building directory, and was
1341     read-only.
1342    
1343     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1344     file. No doubt somebody will tell me if they don't make sense... Also added
1345     Dan Mooney's comments about building on OpenVMS.
1346    
1347     19. Added support for partial matching via the PCRE_PARTIAL option for
1348     pcre_exec() and the \P data escape in pcretest.
1349    
1350     20. Extended pcretest with 3 new pattern features:
1351    
1352     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1353     write the compiled pattern to the file whose name is "rest-of-line".
1354     This is a straight binary dump of the data, with the saved pointer to
1355     the character tables forced to be NULL. The study data, if any, is
1356     written too. After writing, pcretest reads a new pattern.
1357    
1358     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1359     compiled pattern from the given file. There must not be any
1360     occurrences of "<" in the file name (pretty unlikely); if there are,
1361     pcretest will instead treat the initial "<" as a pattern delimiter.
1362     After reading in the pattern, pcretest goes on to read data lines as
1363     usual.
1364    
1365     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1366     and 16-bit fields in a compiled pattern, to simulate a pattern that
1367     was compiled on a host of opposite endianness.
1368    
1369     21. The pcre-exec() function can now cope with patterns that were compiled on
1370     hosts of opposite endianness, with this restriction:
1371    
1372     As for any compiled expression that is saved and used later, the tables
1373     pointer field cannot be preserved; the extra_data field in the arguments
1374     to pcre_exec() should be used to pass in a tables address if a value
1375     other than the default internal tables were used at compile time.
1376    
1377     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1378     now diagnosed as an error. Previously, most of the time, a negative number
1379     would have been treated as zero, but if in addition "ovector" was passed as
1380     NULL, a crash could occur.
1381    
1382     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1383     new versions from the libtool 1.5 distribution (the last one is a copy of
1384     a file called libtool.m4). This seems to have fixed the need to patch
1385     "configure" to support Darwin 1.3 (which I used to do). However, I still
1386     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1387     workstation).
1388    
1389     24. Changed the PCRE licence to be the more standard "BSD" licence.
1390    
1391    
1392 nigel 73 Version 4.5 01-Dec-03
1393     ---------------------
1394    
1395     1. There has been some re-arrangement of the code for the match() function so
1396     that it can be compiled in a version that does not call itself recursively.
1397     Instead, it keeps those local variables that need separate instances for
1398     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1399     needs to "recurse". Keeping track of where control must go is done by means
1400     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1401     hide most of the details from the main code, and operates only if
1402     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1403     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1404     operating.
1405    
1406     To make it easier for callers to provide specially tailored get/free
1407     functions for this usage, two new functions, pcre_stack_malloc, and
1408     pcre_stack_free, are used. They are always called in strict stacking order,
1409     and the size of block requested is always the same.
1410    
1411     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1412     PCRE has been compiled to use the stack or the heap for recursion. The
1413     -C option of pcretest uses this to show which version is compiled.
1414    
1415     A new data escape \S, is added to pcretest; it causes the amounts of store
1416     obtained and freed by both kinds of malloc/free at match time to be added
1417     to the output.
1418    
1419     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1420     what's available on my current Linux desktop machine.
1421    
1422     3. When matching a UTF-8 string, the test for a valid string at the start has
1423     been extended. If start_offset is not zero, PCRE now checks that it points
1424     to a byte that is the start of a UTF-8 character. If not, it returns
1425     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1426     this is necessary because there may be backward assertions in the pattern.
1427     When matching the same subject several times, it may save resources to use
1428     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1429    
1430     4. The code for checking the validity of UTF-8 strings has been tightened so
1431     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1432     containing "overlong sequences".
1433    
1434     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1435     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1436     should have been "&", but it just so happened that all the cases this let
1437     through by mistake were picked up later in the function.
1438    
1439     6. I had used a variable called "isblank" - this is a C99 function, causing
1440     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1441    
1442     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1443     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1444     so the version is shown; (c) stop "make test" echoing "./RunTest".
1445    
1446     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1447    
1448     9. The prototype for memmove() for systems that don't have it was using
1449     size_t, but the inclusion of the header that defines size_t was later. I've
1450     moved the #includes for the C headers earlier to avoid this.
1451    
1452     10. Added some adjustments to the code to make it easier to compiler on certain
1453     special systems:
1454    
1455     (a) Some "const" qualifiers were missing.
1456     (b) Added the macro EXPORT before all exported functions; by default this
1457     is defined to be empty.
1458     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1459     that it reads its output file name as an argument instead of writing
1460     to the standard output and assuming this can be redirected.
1461    
1462     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1463     class containing characters with values greater than 255, PCRE compilation
1464     went into a loop.
1465    
1466     12. A recursive reference to a subpattern that was within another subpattern
1467     that had a minimum quantifier of zero caused PCRE to crash. For example,
1468     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1469     recursion. If the recursively-called subpattern itself had a zero repeat,
1470     that was OK.
1471    
1472     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1473     buffer into which it was copied (for escape processing) was still set at
1474     1024, so long lines caused crashes.
1475    
1476     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1477     "internal error: code overflow...". This applied to any character class
1478     that was followed by a possessive quantifier.
1479    
1480     15. Modified the Makefile to add libpcre.la as a prerequisite for
1481     libpcreposix.la because I was told this is needed for a parallel build to
1482     work.
1483    
1484     16. If a pattern that contained .* following optional items at the start was
1485     studied, the wrong optimizing data was generated, leading to matching
1486     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1487     matching string must start with a or b or c. The correct conclusion for
1488     this pattern is that a match can start with any character.
1489    
1490    
1491 nigel 71 Version 4.4 13-Aug-03
1492     ---------------------
1493    
1494     1. In UTF-8 mode, a character class containing characters with values between
1495     127 and 255 was not handled correctly if the compiled pattern was studied.
1496     In fixing this, I have also improved the studying algorithm for such
1497     classes (slightly).
1498    
1499     2. Three internal functions had redundant arguments passed to them. Removal
1500     might give a very teeny performance improvement.
1501    
1502     3. Documentation bug: the value of the capture_top field in a callout is *one
1503     more than* the number of the hightest numbered captured substring.
1504    
1505     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1506     in incorrectly linking with a previously installed version. They now link
1507     explicitly with libpcre.la.
1508    
1509     5. configure.in no longer needs to recognize Cygwin specially.
1510    
1511     6. A problem in pcre.in for Windows platforms is fixed.
1512    
1513     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1514     pcretest, it used to include the size of the study block as part of its
1515     output. Unfortunately, the structure contains a field that has a different
1516     size on different hardware architectures. This meant that the tests that
1517     showed this size failed. As the block is currently always of a fixed size,
1518     this information isn't actually particularly useful in pcretest output, so
1519     I have just removed it.
1520    
1521     8. Three pre-processor statements accidentally did not start in column 1.
1522     Sadly, there are *still* compilers around that complain, even though
1523     standard C has not required this for well over a decade. Sigh.
1524    
1525     9. In pcretest, the code for checking callouts passed small integers in the
1526     callout_data field, which is a void * field. However, some picky compilers
1527     complained about the casts involved for this on 64-bit systems. Now
1528     pcretest passes the address of the small integer instead, which should get
1529     rid of the warnings.
1530    
1531     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1532     both compile and run time, and gives an error if an invalid UTF-8 sequence
1533     is found. There is a option for disabling this check in cases where the
1534     string is known to be correct and/or the maximum performance is wanted.
1535    
1536     11. In response to a bug report, I changed one line in Makefile.in from
1537    
1538     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1539     to
1540     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1541    
1542     to look similar to other lines, but I have no way of telling whether this
1543     is the right thing to do, as I do not use Windows. No doubt I'll get told
1544     if it's wrong...
1545    
1546    
1547 nigel 69 Version 4.3 21-May-03
1548     ---------------------
1549    
1550     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1551     Makefile.
1552    
1553     2. Some refactoring to improve the quality of the code:
1554    
1555     (i) The utf8_table... variables are now declared "const".
1556    
1557     (ii) The code for \cx, which used the "case flipping" table to upper case
1558     lower case letters, now just substracts 32. This is ASCII-specific,
1559     but the whole concept of \cx is ASCII-specific, so it seems
1560     reasonable.
1561    
1562     (iii) PCRE was using its character types table to recognize decimal and
1563     hexadecimal digits in the pattern. This is silly, because it handles
1564     only 0-9, a-f, and A-F, but the character types table is locale-
1565     specific, which means strange things might happen. A private
1566     table is now used for this - though it costs 256 bytes, a table is
1567     much faster than multiple explicit tests. Of course, the standard
1568     character types table is still used for matching digits in subject
1569     strings against \d.
1570    
1571     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1572     ending in _t are). So I've renamed it as ESC_tee.
1573    
1574     3. The first argument for regexec() in the POSIX wrapper should have been
1575     defined as "const".
1576    
1577     4. Changed pcretest to use malloc() for its buffers so that they can be
1578     Electric Fenced for debugging.
1579    
1580     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1581     to read one or more bytes before the start of the subject string. Often this
1582     had no effect on PCRE's behaviour, but in some circumstances it could
1583     provoke a segmentation fault.
1584    
1585     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1586     to try to read one or more bytes before the start of the subject string.
1587    
1588     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1589     UTF-8 support could misbehave in various ways if the subject string
1590     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1591     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1592     back over UTF-8 characters.)
1593    
1594    
1595 nigel 67 Version 4.2 14-Apr-03
1596     ---------------------
1597    
1598     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1599    
1600     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1601     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1602     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1603     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1604     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1605     and BUILD_EXEEXT
1606     Note: automatic setting of the BUILD variables is not yet working
1607     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1608     compile-time but not at link-time
1609     [LINK]: use for linking executables only
1610     make different versions for Windows and non-Windows
1611     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1612     libraries
1613     [LINK_FOR_BUILD]: new variable
1614     [OBJEXT]: use throughout
1615     [EXEEXT]: use throughout
1616     <winshared>: new target
1617     <wininstall>: new target
1618     <dftables.o>: use native compiler
1619     <dftables>: use native linker
1620     <install>: handle Windows platform correctly
1621     <clean>: ditto
1622     <check>: ditto
1623     copy DLL to top builddir before testing
1624    
1625     As part of these changes, -no-undefined was removed again. This was reported
1626     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1627     in any case.
1628    
1629     3. Some tidies to get rid of compiler warnings:
1630    
1631     . In the match_data structure, match_limit was an unsigned long int, whereas
1632     match_call_count was an int. I've made them both unsigned long ints.
1633    
1634     . In pcretest the fact that a const uschar * doesn't automatically cast to
1635     a void * provoked a warning.
1636    
1637     . Turning on some more compiler warnings threw up some "shadow" variables
1638     and a few more missing casts.
1639    
1640     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1641     option, a class that contained a single character with a value between 128
1642     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1643    
1644     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1645     option, a class that contained several characters, but with at least one
1646     whose value was between 128 and 255 caused PCRE to crash.
1647    
1648    
1649 nigel 65 Version 4.1 12-Mar-03
1650     ---------------------
1651 nigel 63
1652 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1653     needed, and a string in dftables.c that was longer than standard compilers are
1654     required to support.
1655    
1656     2. Compiling with Sun's compiler found a few more places where the code could
1657     be tidied up in order to avoid warnings.
1658    
1659     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1660     first of these names is deprecated in the latest Autoconf in favour of the name
1661     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1662     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1663     analogy I have changed it to CFLAGS_FOR_BUILD.
1664    
1665     4. Added -no-undefined to the linking command in the Makefile, because this is
1666     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1667     linking step for the pcreposix library.
1668    
1669     5. PCRE was failing to diagnose the case of two named groups with the same
1670     name.
1671    
1672     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1673     literal character that is needed in the subject for a match, and scans along to
1674     ensure that it is present before embarking on the full matching process. This
1675     saves time in cases of nested unlimited repeats that are never going to match.
1676     Problem: the scan can take a lot of time if the subject is very long (e.g.
1677     megabytes), thus penalizing straightforward matches. It is now done only if the
1678     amount of subject to be scanned is less than 1000 bytes.
1679    
1680     7. A lesser problem with the same optimization is that it was recording the
1681     first character of an anchored pattern as "needed", thus provoking a search
1682     right along the subject, even when the first match of the pattern was going to
1683     fail. The "needed" character is now not set for anchored patterns, unless it
1684     follows something in the pattern that is of non-fixed length. Thus, it still
1685     fulfils its original purpose of finding quick non-matches in cases of nested
1686     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1687    
1688    
1689     Version 4.0 17-Feb-03
1690     ---------------------
1691    
1692 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1693     extended to the end of string, PCRE compiled incorrect data. This could lead to
1694     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1695    
1696     2. Moved to autoconf 2.53 and libtool 1.4.2.
1697    
1698     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1699     the special perltest8 script is no longer needed - all the tests can be run
1700     from a single perltest script.
1701    
1702     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1703     by \s. It has now been removed in PCRE. This means it isn't recognized as
1704     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1705     class [:space:] *does* include VT, thereby creating a mess.
1706    
1707     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1708     space and tab.
1709    
1710     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1711     its new features into the main test script, reducing the number of scripts.
1712    
1713     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1714     were backward compatible, and made the (?i) apply to the whole pattern, as if
1715     /i were given. Now it behaves more logically, and applies the option setting
1716     only to what follows. PCRE has been changed to follow suit. However, if it
1717     finds options settings right at the start of the pattern, it extracts them into
1718     the global options, as before. Thus, they show up in the info data.
1719    
1720     8. Added support for the \Q...\E escape sequence. Characters in between are
1721     treated as literals. This is slightly different from Perl in that $ and @ are
1722     also handled as literals inside the quotes. In Perl, they will cause variable
1723     interpolation. Note the following examples:
1724    
1725     Pattern PCRE matches Perl matches
1726    
1727     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1728     \Qabc\$xyz\E abc\$xyz abc\$xyz
1729     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1730    
1731     For compatibility with Perl, \Q...\E sequences are recognized inside character
1732     classes as well as outside them.
1733    
1734     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1735     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1736     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1737     signed/unsigned warnings.
1738    
1739     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1740     option for pcretest, so I've replaced it by a simple function that does just
1741     that job.
1742    
1743     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1744     "pcregrep -".
1745    
1746     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1747     Java package. This provides some syntactic sugar for simple cases of what my
1748     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1749     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1750     item, you can use this simplified notation. Note that only makes sense with
1751     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1752     greediness, whatever the setting of the PCRE_UNGREEDY option.
1753    
1754     13. A change of greediness default within a pattern was not taking effect at
1755     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1756     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1757     was abstracted outside.
1758    
1759     14. PCRE now supports the \G assertion. It is true when the current matching
1760     position is at the start point of the match. This differs from \A when the
1761     starting offset is non-zero. Used with the /g option of pcretest (or similar
1762     code), it works in the same way as it does for Perl's /g option. If all
1763     alternatives of a regex begin with \G, the expression is anchored to the start
1764     match position, and the "anchored" flag is set in the compiled expression.
1765    
1766     15. Some bugs concerning the handling of certain option changes within patterns
1767     have been fixed. These applied to options other than (?ims). For example,
1768     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1769     been the other way round. Some of this was related to change 7 above.
1770    
1771     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1772     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1773     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1774     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1775    
1776     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1777     mode. Unlike ".", it always matches newline, whatever the setting of
1778     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1779     assertions. Perl allows it, but it doesn't (in general) work because it can't
1780     calculate the length of the lookbehind. At least, that's the case for Perl
1781     5.8.0 - I've been told they are going to document that it doesn't work in
1782     future.
1783    
1784     18. Added an error diagnosis for escapes that PCRE does not support: these are
1785     \L, \l, \N, \P, \p, \U, \u, and \X.
1786    
1787     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1788     reading past the end of the pattern in cases such as /[abcd/.
1789    
1790     20. PCRE was getting more memory than necessary for patterns with classes that
1791     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1792    
1793     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1794     compiling PCRE for use with Virtual Pascal.
1795    
1796     22. Small fix to the Makefile to make it work properly if the build is done
1797     outside the source tree.
1798    
1799     23. Added a new extension: a condition to go with recursion. If a conditional
1800     subpattern starts with (?(R) the "true" branch is used if recursion has
1801     happened, whereas the "false" branch is used only at the top level.
1802    
1803     24. When there was a very long string of literal characters (over 255 bytes
1804     without UTF support, over 250 bytes with UTF support), the computation of how
1805     much memory was required could be incorrect, leading to segfaults or other
1806     strange effects.
1807    
1808     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1809     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1810     there was a subsequent back reference to those brackets. This meant that, for
1811     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1812     possible to check for precisely this case. All we can do is abandon the
1813     optimization if .* occurs inside capturing brackets when there are any back
1814     references whatsoever. (See below for a better fix that came later.)
1815    
1816     26. The handling of the optimization for finding the first character of a
1817     non-anchored pattern, and for finding a character that is required later in the
1818     match were failing in some cases. This didn't break the matching; it just
1819     failed to optimize when it could. The way this is done has been re-implemented.
1820    
1821     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1822    
1823     28. Added a new feature that provides some of the functionality that Perl
1824     provides with (?{...}). The facility is termed a "callout". The way it is done
1825     in PCRE is for the caller to provide an optional function, by setting
1826     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1827     global variable. By default it is unset, which disables all calling out. To get
1828     the function called, the regex must include (?C) at appropriate points. This
1829     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1830     This provides a means of identifying different callout points. When PCRE
1831     reaches such a point in the regex, if pcre_callout has been set, the external
1832     function is called. It is provided with data in a structure called
1833     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1834     matching continues; if it returns a non-zero value, the match at the current
1835     point fails. However, backtracking will occur if possible. [This was changed
1836     later and other features added - see item 49 below.]
1837    
1838     29. pcretest is upgraded to test the callout functionality. It provides a
1839     callout function that displays information. By default, it shows the start of
1840     the match and the current position in the text. There are some new data escapes
1841     to vary what happens:
1842    
1843     \C+ in addition, show current contents of captured substrings
1844     \C- do not supply a callout function
1845     \C!n return 1 when callout number n is reached
1846     \C!n!m return 1 when callout number n is reached for the mth time
1847    
1848     30. If pcregrep was called with the -l option and just a single file name, it
1849     output "<stdin>" if a match was found, instead of the file name.
1850    
1851     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1852     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1853     pcre_exec(). This saves a malloc/free per call. The default value of
1854     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1855     when configuring.
1856    
1857     32. The default maximum size of a compiled pattern is 64K. There have been a
1858     few cases of people hitting this limit. The code now uses macros to handle the
1859     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1860     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1861     configuring. Tests 2 and 5 work only with 2-byte links because they output
1862     debugging information about compiled patterns.
1863    
1864     33. Internal code re-arrangements:
1865    
1866     (a) Moved the debugging function for printing out a compiled regex into
1867     its own source file (printint.c) and used #include to pull it into
1868     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1869     separate copies.
1870    
1871     (b) Defined the list of op-code names for debugging as a macro in
1872     internal.h so that it is next to the definition of the opcodes.
1873    
1874     (c) Defined a table of op-code lengths for simpler skipping along compiled
1875     code. This is again a macro in internal.h so that it is next to the
1876     definition of the opcodes.
1877    
1878     34. Added support for recursive calls to individual subpatterns, along the
1879     lines of Robin Houston's patch (but implemented somewhat differently).
1880    
1881     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1882     allow it to read and process whole directories in Win32. This code was
1883     contributed by Lionel Fourquaux; it has not been tested by me.
1884    
1885     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1886     used to name a group. Names consist of alphanumerics and underscores, and must
1887     be unique. Back references use the syntax (?P=name) and recursive calls use
1888     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1889     numbers. The function pcre_fullinfo() can be used after compilation to extract
1890     a name/number map. There are three relevant calls:
1891    
1892     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1893     PCRE_INFO_NAMECOUNT yields the number of entries
1894     PCRE_INFO_NAMETABLE yields a pointer to the map.
1895    
1896     The map is a vector of fixed-size entries. The size of each entry depends on
1897     the length of the longest name used. The first two bytes of each entry are the
1898     group number, most significant byte first. There follows the corresponding
1899     name, zero terminated. The names are in alphabetical order.
1900    
1901     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1902     case instead of 255. Making it the same both with and without UTF-8 support
1903     means that the same test output works with both.
1904    
1905     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1906     calling malloc() with a zero argument.
1907    
1908     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1909     optimization. I've improved things by keeping a bitmap of backreferences with
1910     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1911     fact referenced, the optimization can be applied. It is unlikely that a
1912     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1913     the match to follow \n) will appear inside brackets with a number greater than
1914     31, but if it does, any back reference > 31 suppresses the optimization.
1915    
1916     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1917     of disabling numbered capturing parentheses. Any opening parenthesis that is
1918     not followed by ? behaves as if it were followed by ?: but named parentheses
1919     can still be used for capturing (and they will acquire numbers in the usual
1920     way).
1921    
1922     41. Redesigned the return codes from the match() function into yes/no/error so
1923     that errors can be passed back from deep inside the nested calls. A malloc
1924     failure while inside a recursive subpattern call now causes the
1925     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1926    
1927     42. It is now possible to set a limit on the number of times the match()
1928     function is called in a call to pcre_exec(). This facility makes it possible to
1929     limit the amount of recursion and backtracking, though not in a directly
1930     obvious way, because the match() function is used in a number of different
1931     circumstances. The count starts from zero for each position in the subject
1932     string (for non-anchored patterns). The default limit is, for compatibility, a
1933     large number, namely 10 000 000. You can change this in two ways:
1934    
1935     (a) When configuring PCRE before making, you can use --with-match-limit=n
1936     to set a default value for the compiled library.
1937    
1938     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1939     a different value is set. See 45 below.
1940    
1941     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1942    
1943     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1944     of things that can be changed at compile time. The first argument specifies
1945     what is wanted and the second points to where the information is to be placed.
1946     The current list of available information is:
1947    
1948     PCRE_CONFIG_UTF8
1949    
1950     The output is an integer that is set to one if UTF-8 support is available;
1951     otherwise it is set to zero.
1952    
1953     PCRE_CONFIG_NEWLINE
1954    
1955     The output is an integer that it set to the value of the code that is used for
1956     newline. It is either LF (10) or CR (13).
1957    
1958     PCRE_CONFIG_LINK_SIZE
1959    
1960     The output is an integer that contains the number of bytes used for internal
1961     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1962    
1963     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1964    
1965     The output is an integer that contains the threshold above which the POSIX
1966     interface uses malloc() for output vectors. See item 31 above.
1967    
1968     PCRE_CONFIG_MATCH_LIMIT
1969    
1970     The output is an unsigned integer that contains the default limit of the number
1971     of match() calls in a pcre_exec() execution. See 42 above.
1972    
1973     44. pcretest has been upgraded by the addition of the -C option. This causes it
1974     to extract all the available output from the new pcre_config() function, and to
1975     output it. The program then exits immediately.
1976    
1977     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1978     order to support additional features. One way would have been to define
1979     pcre_exec2() (for example) with extra arguments, but this would not have been
1980     extensible, and would also have required all calls to the original function to
1981     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1982     is used for passing in "extra" data from pcre_study().
1983    
1984     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1985     contains the following fields:
1986    
1987     flags a bitmap indicating which of the following fields are set
1988     study_data opaque data from pcre_study()
1989     match_limit a way of specifying a limit on match() calls for a specific
1990     call to pcre_exec()
1991     callout_data data for callouts (see 49 below)
1992    
1993     The flag bits are also defined in pcre.h, and are
1994    
1995     PCRE_EXTRA_STUDY_DATA
1996     PCRE_EXTRA_MATCH_LIMIT
1997     PCRE_EXTRA_CALLOUT_DATA
1998    
1999     The pcre_study() function now returns one of these new pcre_extra blocks, with
2000     the actual study data pointed to by the study_data field, and the
2001     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2002     before. That is, this change is entirely upwards-compatible and requires no
2003     change to existing code.
2004    
2005     If you want to pass in additional data to pcre_exec(), you can either place it
2006     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2007     block.
2008    
2009     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2010     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2011     times with different match limits, until it finds the minimum value needed for
2012     pcre_exec() to complete. The value is then output. This can be instructive; for
2013     most simple matches the number is quite small, but for pathological cases it
2014     gets very large very quickly.
2015    
2016     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2017     returns the size of the data block pointed to by the study_data field in a
2018     pcre_extra block, that is, the value that was passed as the argument to
2019     pcre_malloc() when PCRE was getting memory in which to place the information
2020     created by pcre_study(). The fourth argument should point to a size_t variable.
2021     pcretest has been extended so that this information is shown after a successful
2022     pcre_study() call when information about the compiled regex is being displayed.
2023    
2024     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2025     because what follows is always an absolute path. (Later: it turns out that this
2026     is more than cosmetic for MinGW, because it doesn't like empty path
2027     components.)
2028    
2029     49. Some changes have been made to the callout feature (see 28 above):
2030    
2031     (i) A callout function now has three choices for what it returns:
2032    
2033     0 => success, carry on matching
2034     > 0 => failure at this point, but backtrack if possible
2035     < 0 => serious error, return this value from pcre_exec()
2036    
2037     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2038     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2039     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2040     use by callout functions. It will never be used by PCRE itself.
2041    
2042     (ii) The pcre_extra structure (see 45 above) has a void * field called
2043     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2044     pcre_callout_block structure has a field of the same name. The contents of
2045     the field passed in the pcre_extra structure are passed to the callout
2046     function in the corresponding field in the callout block. This makes it
2047     easier to use the same callout-containing regex from multiple threads. For
2048     testing, the pcretest program has a new data escape
2049    
2050     \C*n pass the number n (may be negative) as callout_data
2051    
2052     If the callout function in pcretest receives a non-zero value as
2053     callout_data, it returns that value.
2054    
2055     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2056     there were some redundant $(CFLAGS) in commands that are now specified as
2057     $(LINK), which already includes $(CFLAGS).
2058    
2059     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2060     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2061     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2062     one-byte characters throughout. Note that case-insensitive matching applies
2063     only to characters whose values are less than 256. PCRE doesn't support the
2064     notion of cases for higher-valued characters.
2065    
2066     (i) A character class whose characters are all within 0-255 is handled as
2067     a bit map, and the map is inverted for negative classes. Previously, a
2068     character > 255 always failed to match such a class; however it should
2069     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2070    
2071     (ii) A negated character class with a single character < 255 is coded as
2072     "not this character" (OP_NOT). This wasn't working properly when the test
2073     character was multibyte, either singly or repeated.
2074    
2075     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2076     mode, for example: \x{100}{2,3}.
2077    
2078     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2079     singly or repeated) now correctly test multibyte characters. However,
2080     PCRE doesn't recognize any characters with values greater than 255 as
2081     digits, spaces, or word characters. Such characters always match \D, \S,
2082     and \W, and never match \d, \s, or \w.
2083    
2084     (v) Classes may now contain characters and character ranges with values
2085     greater than 255. For example: [ab\x{100}-\x{400}].
2086    
2087     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2088     PCRE in UTF-8 mode.
2089    
2090     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2091     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2092     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2093     value.)
2094    
2095     53. The single man page has become too large. I have therefore split it up into
2096     a number of separate man pages. These also give rise to individual HTML pages;
2097     these are now put in a separate directory, and there is an index.html page that
2098     lists them all. Some hyperlinking between the pages has been installed.
2099    
2100     54. Added convenience functions for handling named capturing parentheses.
2101    
2102     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2103     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2104     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2105     are faulted.
2106    
2107     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2108     calling configure. These values are used when compiling the dftables.c program
2109     which is run to generate the source of the default character tables. They
2110     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2111     you will need to set these values.
2112    
2113     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2114    
2115    
2116     Version 3.9 02-Jan-02
2117 nigel 61 ---------------------
2118    
2119     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2120    
2121     2. If --disable-static was given, the building process failed when trying to
2122     build pcretest and pcregrep. (For some reason it was using libtool to compile
2123     them, which is not right, as they aren't part of the library.)
2124    
2125    
2126 nigel 59 Version 3.8 18-Dec-01
2127     ---------------------
2128    
2129     1. The experimental UTF-8 code was completely screwed up. It was packing the
2130     bytes in the wrong order. How dumb can you get?
2131    
2132    
2133 nigel 57 Version 3.7 29-Oct-01
2134     ---------------------
2135    
2136     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2137     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2138     this didn't happen under Solaris 8, where I normally test things.
2139    
2140 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2141     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2142     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2143     this fix an hour or so after the initial 3.7 release.)
2144 nigel 57
2145 nigel 59
2146 nigel 55 Version 3.6 23-Oct-01
2147     ---------------------
2148    
2149     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2150     offsets passed as NULL with zero offset count.
2151    
2152     2. The config.guess and config.sub files had not been updated when I moved to
2153     the latest autoconf.
2154    
2155    
2156 nigel 53 Version 3.5 15-Aug-01
2157     ---------------------
2158 nigel 5
2159 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2160     had been forgotten.
2161    
2162     2. By using declared but undefined structures, we can avoid using "void"
2163     definitions in pcre.h while keeping the internal definitions of the structures
2164     private.
2165    
2166     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2167     user point of view, this means that both static and shared libraries are built
2168     by default, but this can be individually controlled. More of the work of
2169     handling this static/shared cases is now inside libtool instead of PCRE's make
2170     file.
2171    
2172     4. The pcretest utility is now installed along with pcregrep because it is
2173     useful for users (to test regexs) and by doing this, it automatically gets
2174     relinked by libtool. The documentation has been turned into a man page, so
2175     there are now .1, .txt, and .html versions in /doc.
2176    
2177     5. Upgrades to pcregrep:
2178     (i) Added long-form option names like gnu grep.
2179     (ii) Added --help to list all options with an explanatory phrase.
2180     (iii) Added -r, --recursive to recurse into sub-directories.
2181     (iv) Added -f, --file to read patterns from a file.
2182    
2183     6. pcre_exec() was referring to its "code" argument before testing that
2184     argument for NULL (and giving an error if it was NULL).
2185    
2186     7. Upgraded Makefile.in to allow for compiling in a different directory from
2187     the source directory.
2188    
2189     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2190     options bits, the pointer it was passed was to an int instead of to an unsigned
2191     long int. This mattered only on 64-bit systems.
2192    
2193     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2194     generated) instead of pcre.in, which it its source. Also made the same change
2195     in several of the .c files.
2196    
2197     10. A new release of gcc defines printf() as a macro, which broke pcretest
2198     because it had an ifdef in the middle of a string argument for printf(). Fixed
2199     by using separate calls to printf().
2200    
2201     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2202     script, to force use of CR or LF instead of \n in the source. On non-Unix
2203     systems, the value can be set in config.h.
2204    
2205     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2206     absolute limit. Changed the text of the error message to make this clear, and
2207     likewise updated the man page.
2208    
2209     13. The limit of 99 on the number of capturing subpatterns has been removed.
2210     The new limit is 65535, which I hope will not be a "real" limit.
2211    
2212    
2213 nigel 51 Version 3.4 22-Aug-00
2214     ---------------------
2215    
2216     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2217    
2218     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2219    
2220    
2221 nigel 49 Version 3.3 01-Aug-00
2222     ---------------------
2223    
2224     1. If an octal character was given, but the value was greater than \377, it
2225     was not getting masked to the least significant bits, as documented. This could
2226     lead to crashes in some systems.
2227    
2228     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2229     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2230    
2231     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2232     These just pass their arguments on to (pcre_free)(), but they are provided
2233     because some uses of PCRE bind it to non-C systems that can call its functions,
2234     but cannot call free() or pcre_free() directly.
2235    
2236     4. Add "make test" as a synonym for "make check". Corrected some comments in
2237     the Makefile.
2238    
2239     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2240     Makefile.
2241    
2242     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2243     command called pgrep for grepping around the active processes.
2244    
2245     7. Added the beginnings of support for UTF-8 character strings.
2246    
2247     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2248     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2249     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2250     out for the ar command.)
2251    
2252    
2253 nigel 47 Version 3.2 12-May-00
2254     ---------------------
2255    
2256     This is purely a bug fixing release.
2257    
2258     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2259     of ZA. This was just one example of several cases that could provoke this bug,
2260     which was introduced by change 9 of version 2.00. The code for breaking
2261     infinite loops after an iteration that matches an empty string was't working
2262     correctly.
2263    
2264     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2265     when matched against abbab (for example). After matching an empty string, it
2266     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2267     caused it to match further down the string than it should.
2268    
2269     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2270     was there because it doesn't seem to be needed, and it causes trouble on some
2271     systems, as it is not a Standard C header. It has been removed.
2272    
2273     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2274     were reported on the Macintosh. The changes were from
2275    
2276     while ((c = *(++ptr)) != 0 && c != '\n');
2277     to
2278     while ((c = *(++ptr)) != 0 && c != '\n') ;
2279    
2280     Totally extraordinary, but if that's what it takes...
2281    
2282     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2283     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2284     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2285     assumes the way PCRE uses memmove() (always moving upwards).
2286    
2287     6. PCRE is being used in one environment where strchr() is not available. There
2288     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2289     faster code anyway.
2290    
2291    
2292 nigel 45 Version 3.1 09-Feb-00
2293     ---------------------
2294    
2295     The only change in this release is the fixing of some bugs in Makefile.in for
2296     the "install" target:
2297    
2298     (1) It was failing to install pcreposix.h.
2299    
2300     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2301    
2302    
2303 nigel 43 Version 3.0 01-Feb-00
2304     ---------------------
2305 nigel 41
2306     1. Add support for the /+ modifier to perltest (to output $` like it does in
2307     pcretest).
2308    
2309     2. Add support for the /g modifier to perltest.
2310    
2311     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2312     matches null strings.
2313    
2314     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2315     pattern. Perl treats empty patterns specially - it reuses the most recent
2316     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2317     effect.
2318    
2319     5. The POSIX interface was broken in that it was just handing over the POSIX
2320     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2321     required a bigger vector, with some working space on the end. This means that
2322     the POSIX wrapper now has to get and free some memory, and copy the results.
2323    
2324 nigel 43 6. Added some simple autoconf support, placing the test data and the
2325     documentation in separate directories, re-organizing some of the
2326     information files, and making it build pcre-config (a GNU standard). Also added
2327     libtool support for building PCRE as a shared library, which is now the
2328     default.
2329 nigel 41
2330 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2331     09 are not valid octal constants. Single digits will be used for minor values
2332     less than 10.
2333    
2334     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2335     existing programs that set these in the POSIX interface can use PCRE without
2336     modification.
2337    
2338     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2339     return all that pcre_info() returns, plus additional data. The pcre_info()
2340     function is retained for compatibility, but is considered to be obsolete.
2341    
2342     10. Added experimental recursion feature (?R) to handle one common case that
2343     Perl 5.6 will be able to do with (?p{...}).
2344    
2345     11. Added support for POSIX character classes like [:alpha:], which Perl is
2346     adopting.
2347    
2348    
2349 nigel 39 Version 2.08 31-Aug-99
2350     ----------------------
2351    
2352     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2353     trying to match at the startoffset position, but instead was moving forward to
2354     the next newline as if a previous match had failed.
2355    
2356     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2357     and could get into a loop if a null string was matched other than at the start
2358     of the subject.
2359    
2360     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2361     be distinguished at compile time, and for completeness also added PCRE_DATE.
2362    
2363     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2364     in GnuWin32 environments.
2365    
2366    
2367 nigel 37 Version 2.07 29-Jul-99
2368     ----------------------
2369    
2370     1. The documentation is now supplied in plain text form and HTML as well as in
2371     the form of man page sources.
2372    
2373     2. C++ compilers don't like assigning (void *) values to other pointer types.
2374     In particular this affects malloc(). Although there is no problem in Standard
2375     C, I've put in casts to keep C++ compilers happy.
2376    
2377     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2378     should be (const char *).
2379    
2380     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2381     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2382     However, I haven't made this a standard facility. The documentation doesn't
2383     mention it, and the Makefile doesn't support it.
2384    
2385     5. The Makefile now contains an "install" target, with editable destinations at
2386     the top of the file. The pcretest program is not installed.
2387    
2388     6. pgrep -V now gives the PCRE version number and date.
2389    
2390     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2391     causing the entire string to be ignored, instead of just the last character.
2392    
2393     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2394     non-matching string, it can take a very, very long time, even for strings of
2395     quite modest length, because of the nested recursion. PCRE now does better in
2396     some of these cases. It does this by remembering the last required literal
2397     character in the pattern, and pre-searching the subject to ensure it is present
2398     before running the real match. In other words, it applies a heuristic to detect
2399     some types of certain failure quickly, and in the above example, if presented
2400     with a string that has no trailing " it gives "no match" very quickly.
2401    
2402     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2403     other alternatives are tried instead.
2404    
2405    
2406 nigel 35 Version 2.06 09-Jun-99
2407     ----------------------
2408    
2409     1. Change pcretest's output for amount of store used to show just the code
2410     space, because the remainder (the data block) varies in size between 32-bit and
2411     64-bit systems.
2412    
2413     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2414     start matching at. This allows lookbehinds to work when searching for multiple
2415     occurrences in a string.
2416    
2417     3. Added additional options to pcretest for testing multiple occurrences:
2418    
2419     /+ outputs the rest of the string that follows a match
2420     /g loops for multiple occurrences, using the new startoffset argument
2421     /G loops for multiple occurrences by passing an incremented pointer
2422    
2423     4. PCRE wasn't doing the "first character" optimization for patterns starting
2424     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2425     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2426     the letter 'x'. On long subject strings, this gives a significant speed-up.
2427    
2428    
2429 nigel 33 Version 2.05 21-Apr-99
2430     ----------------------
2431    
2432     1. Changed the type of magic_number from int to long int so that it works
2433     properly on 16-bit systems.
2434    
2435     2. Fixed a bug which caused patterns starting with .* not to work correctly
2436     when the subject string contained newline characters. PCRE was assuming
2437     anchoring for such patterns in all cases, which is not correct because .* will
2438     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2439     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2440     must be retried after every newline in the subject.
2441    
2442    
2443 nigel 31 Version 2.04 18-Feb-99
2444     ----------------------
2445    
2446     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2447     computation of the store needed to hold the pattern was incorrect (too large).
2448     If such patterns were nested a few deep, this could multiply and become a real
2449     problem.
2450    
2451     2. Added /M option to pcretest to show the memory requirement of a specific
2452     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2453    
2454     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2455     compiled in such a way that the backtracking after subsequent failure was
2456     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2457     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2458    
2459    
2460 nigel 29 Version 2.03 02-Feb-99
2461     ----------------------
2462    
2463     1. Fixed typo and small mistake in man page.
2464    
2465 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2466     LICENCE file containing the conditions.
2467 nigel 29
2468     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2469     Perl, that is the internal \ allows the delimiter to be included in the
2470     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2471     the final delimiter, add \ to the end of the pattern (to test the error).
2472    
2473     4. Added the convenience functions for extracting substrings after a successful
2474     match. Updated pcretest to make it able to test these functions.
2475    
2476    
2477 nigel 27 Version 2.02 14-Jan-99
2478     ----------------------
2479    
2480     1. Initialized the working variables associated with each extraction so that
2481     their saving and restoring doesn't refer to uninitialized store.
2482    
2483     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2484     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2485     fix the problem.
2486    
2487     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2488     calls, and wasn't printing the correct value for compiling calls. Increased the
2489     default value of LOOPREPEAT, and the number of significant figures in the
2490     times.
2491    
2492     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2493    
2494     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2495     a building problem on Windows NT with a FAT file system.
2496    
2497    
2498 nigel 25 Version 2.01 21-Oct-98
2499     ----------------------
2500    
2501     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2502     to character tables built by pcre_maketables() in the current locale. If NULL
2503     is passed, the default tables are used.
2504    
2505    
2506 nigel 23 Version 2.00 24-Sep-98
2507 nigel 21 ----------------------
2508    
2509 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2510     it any more.
2511    
2512     2. Allow quantification of (?>) groups, and make it work correctly.
2513    
2514     3. The first character computation wasn't working for (?>) groups.
2515    
2516     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2517     end of the subject) and add 5.005's \z, which really does match only at the
2518     very end of the subject.
2519    
2520     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2521    
2522     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2523     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2524     localized options. All options to pcre_study() were also removed.
2525    
2526     7. Add other new features from 5.005:
2527    
2528     $(?<= positive lookbehind
2529     $(?<! negative lookbehind
2530     (?imsx-imsx) added the unsetting capability
2531     such a setting is global if at outer level; local otherwise
2532     (?imsx-imsx:) non-capturing groups with option setting
2533     (?(cond)re|re) conditional pattern matching
2534    
2535     A backreference to itself in a repeated group matches the previous
2536     captured string.
2537    
2538     8. General tidying up of studying (both automatic and via "study")
2539     consequential on the addition of new assertions.
2540    
2541     9. As in 5.005, unlimited repeated groups that could match an empty substring
2542     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2543     runtime if any iteration does actually match an empty substring.
2544    
2545     10. Include the RunTest script in the distribution.
2546    
2547     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2548     discrepancies, some of which were old and were also with respect to 5.004. They
2549     have now been fixed.
2550    
2551    
2552     Version 1.09 28-Apr-98
2553     ----------------------
2554    
2555 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2556     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2557     program crashes, or just wrong answers. This did not apply to negated classes
2558     containing more than one character, or to minima other than one.
2559    
2560    
2561 nigel 19 Version 1.08 27-Mar-98
2562     ----------------------
2563    
2564     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2565    
2566     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2567     latter must appear before anything that relies on it in the pattern.
2568    
2569    
2570 nigel 17 Version 1.07 16-Feb-98
2571     ----------------------
2572    
2573     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2574     repeat of a potentially empty string).
2575    
2576    
2577 nigel 15 Version 1.06 23-Jan-98
2578     ----------------------
2579    
2580     1. Added Markus Oberhumer's little patches for C++.
2581    
2582     2. Literal strings longer than 255 characters were broken.
2583    
2584    
2585 nigel 13 Version 1.05 23-Dec-97
2586     ----------------------
2587    
2588     1. Negated character classes containing more than one character were failing if
2589     PCRE_CASELESS was set at run time.
2590    
2591    
2592 nigel 11 Version 1.04 19-Dec-97
2593     ----------------------
2594    
2595     1. Corrected the man page, where some "const" qualifiers had been omitted.
2596    
2597     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2598     input syntax.
2599    
2600     3. Fixed memory leak which occurred when a regex with back references was
2601     matched with an offsets vector that wasn't big enough. The temporary memory
2602     that is used in this case wasn't being freed if the match failed.
2603    
2604     4. Tidied pcretest to ensure it frees memory that it gets.
2605    
2606     5. Temporary memory was being obtained in the case where the passed offsets
2607     vector was exactly big enough.
2608    
2609     6. Corrected definition of offsetof() from change 5 below.
2610    
2611     7. I had screwed up change 6 below and broken the rules for the use of
2612     setjmp(). Now fixed.
2613    
2614    
2615 nigel 9 Version 1.03 18-Dec-97
2616     ----------------------
2617    
2618     1. A erroneous regex with a missing opening parenthesis was correctly
2619     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2620     on some systems.
2621    
2622     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2623     it was reported that one broken compiler failed on the former because "code" is
2624     also an independent variable.
2625    
2626     3. The erroneous regex a[]b caused an array overrun reference.
2627    
2628     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2629     fail on data ending with that character. (It was going on too far, and checking
2630     the next character, typically a binary zero.) This was specific to the
2631     optimized code for single-character negative classes.
2632    
2633     5. Added a contributed patch from the TIN world which does the following:
2634    
2635     + Add an undef for memmove, in case the the system defines a macro for it.
2636    
2637     + Add a definition of offsetof(), in case there isn't one. (I don't know
2638     the reason behind this - offsetof() is part of the ANSI standard - but
2639     it does no harm).
2640    
2641     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2642     most of the places where whitespace preceded '#'. I have given up and
2643     allowed the remaining 2 cases to be at the margin.
2644    
2645     + Rename some variables in pcre to eliminate shadowing. This seems very
2646     pedantic, but does no harm, of course.
2647    
2648     6. Moved the call to setjmp() into its own function, to get rid of warnings
2649     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2650    
2651     7. Constructs such as \d{8,} were compiling into the equivalent of
2652 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2653 nigel 9 outcome, but in this particular case used more store than had been allocated,
2654     which caused the bug to be discovered because it threw up an internal error.
2655    
2656     8. The debugging code in both pcre and pcretest for outputting the compiled
2657     form of a regex was going wrong in the case of back references followed by
2658     curly-bracketed repeats.
2659    
2660    
2661 nigel 7 Version 1.02 12-Dec-97
2662     ----------------------
2663    
2664     1. Typos in pcre.3 and comments in the source fixed.
2665    
2666     2. Applied a contributed patch to get rid of places where it used to remove
2667     'const' from variables, and fixed some signed/unsigned and uninitialized
2668     variable warnings.
2669    
2670     3. Added the "runtest" target to Makefile.
2671    
2672     4. Set default compiler flag to -O2 rather than just -O.
2673    
2674    
2675 nigel 5 Version 1.01 19-Nov-97
2676     ----------------------
2677    
2678     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2679     like /([ab]*)*/, that is, for classes with more than one character in them.
2680    
2681     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2682     as /((?>a*))*/ (a PCRE_EXTRA facility).
2683    
2684    
2685     Version 1.00 18-Nov-97
2686     ----------------------
2687    
2688     1. Added compile-time macros to support systems such as SunOS4 which don't have
2689     memmove() or strerror() but have other things that can be used instead.
2690    
2691     2. Arranged that "make clean" removes the executables.
2692    
2693    
2694 nigel 3 Version 0.99 27-Oct-97
2695     ----------------------
2696    
2697     1. Fixed bug in code for optimizing classes with only one character. It was
2698     initializing a 32-byte map regardless, which could cause it to run off the end
2699     of the memory it had got.
2700    
2701     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2702    
2703    
2704     Version 0.98 22-Oct-97
2705     ----------------------
2706    
2707     1. Fixed bug in code for handling temporary memory usage when there are more
2708     back references than supplied space in the ovector. This could cause segfaults.
2709    
2710    
2711     Version 0.97 21-Oct-97
2712     ----------------------
2713    
2714     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2715    
2716     2. Optimized negated single characters not to use a bit map.
2717    
2718     3. Brought error texts together as macro definitions; clarified some of them;
2719     fixed one that was wrong - it said "range out of order" when it meant "invalid
2720     escape sequence".
2721    
2722     4. Changed some char * arguments to const char *.
2723    
2724     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2725    
2726     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2727     pcretest.
2728    
2729    
2730     Version 0.96 16-Oct-97
2731     ----------------------
2732    
2733     1. Added a simple "pgrep" utility to the distribution.
2734    
2735     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2736     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2737     where "ddd" means "one or more decimal digits".
2738    
2739     3. Fixed serious bug. If a pattern had a back reference, but the call to
2740     pcre_exec() didn't supply a large enough ovector to record the related
2741     identifying subpattern, the match always failed. PCRE now remembers the number
2742     of the largest back reference, and gets some temporary memory in which to save
2743     the offsets during matching if necessary, in order to ensure that
2744     backreferences always work.
2745    
2746     4. Increased the compatibility with Perl in a number of ways:
2747    
2748     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2749     to request this handling. The option can be set at compile or exec time.
2750    
2751     (b) $ matches before a terminating newline by default; an option
2752     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2753     mode). The option can be set at compile or exec time.
2754    
2755     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2756     the same as Perl's. If the decimal number it represents is less than 10
2757     or there aren't that many previous left capturing parentheses, an octal
2758     escape is read. Inside a character class, it's always an octal escape,
2759     even if it is a single digit.
2760    
2761     (d) An escaped but undefined alphabetic character is taken as a literal,
2762     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2763     escapes.
2764    
2765     (e) {0} is now permitted. (The previous item is removed from the compiled
2766     pattern).
2767    
2768     5. Changed all the names of code files so that the basic parts are no longer
2769     than 10 characters, and abolished the teeny "globals.c" file.
2770    
2771     6. Changed the handling of character classes; they are now done with a 32-byte
2772     bit map always.
2773    
2774     7. Added the -d and /D options to pcretest to make it possible to look at the
2775     internals of compilation without having to recompile pcre.
2776    
2777    
2778     Version 0.95 23-Sep-97
2779     ----------------------
2780    
2781     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2782     \x20 at the start of a run of normal characters. These were being treated as
2783     real characters, instead of the source characters being re-checked.
2784    
2785    
2786     Version 0.94 18-Sep-97
2787     ----------------------
2788    
2789     1. The functions are now thread-safe, with the caveat that the global variables
2790     containing pointers to malloc() and free() or alternative functions are the
2791     same for all threads.
2792    
2793     2. Get pcre_study() to generate a bitmap of initial characters for non-
2794     anchored patterns when this is possible, and use it if passed to pcre_exec().
2795    
2796    
2797     Version 0.93 15-Sep-97
2798     ----------------------
2799    
2800     1. /(b)|(:+)/ was computing an incorrect first character.
2801    
2802     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2803     but not actually doing anything yet.
2804    
2805     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2806     as Perl does (e.g. [-az] or [az-]).
2807    
2808     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2809     all possible positions.
2810    
2811     5. Split up into different modules to avoid including unneeded functions in a
2812     compiled binary. However, compile and exec are still in one module. The "study"
2813     function is split off.
2814    
2815     6. The character tables are now in a separate module whose source is generated
2816     by an auxiliary program - but can then be edited by hand if required. There are
2817     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2818     toupper() in the code.
2819    
2820     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2821     make them global. Abolish the function for setting them, as the caller can now
2822     set them directly.
2823    
2824    
2825     Version 0.92 11-Sep-97
2826     ----------------------
2827    
2828     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2829     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2830    
2831     2. Caseless matching was not working in character classes if the characters in
2832     the pattern were in upper case.
2833    
2834     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2835    
2836     4. Make PCRE_ANCHORED public and accept as a compile option.
2837    
2838     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2839     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2840     pass them.
2841    
2842     6. Give an error if bad option bits passed at compile or run time.
2843    
2844     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2845     pcretest to cause it to pass that flag.
2846    
2847     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2848     options, and the first character, if set.
2849    
2850     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2851    
2852    
2853     Version 0.91 10-Sep-97
2854     ----------------------
2855    
2856     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2857     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2858    
2859     2. PCRE was looping on encountering an indefinitely repeated back reference to
2860     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2861     Perl does - treats the match as successful.
2862    
2863     ****

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