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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 236 - (hide annotations) (download)
Tue Sep 11 12:57:06 2007 UTC (6 years, 10 months ago) by ph10
File size: 140336 byte(s)
<config.h> => "config.h" and also some cases of <pcre.h>.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 230 Version 7.4 10-Sep-07
5     ---------------------
6    
7     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
8 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
9     LF". That isn't really right - the whole point of the change was to try to
10     help when there was an actual mention of one of the two characters. So now
11     the change happens only if \r or \n (or a literal CR or LF) character is
12 ph10 230 encountered.
13 ph10 231
14     2. The 32-bit options word was also used for 6 internal flags, but the numbers
15     of both had grown to the point where there were only 3 bits left.
16     Fortunately, there was spare space in the data structure, and so I have
17     moved the internal flags into a new 16-bit field to free up more option
18 ph10 230 bits.
19 ph10 231
20     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
21     but did not set the internal JCHANGED flag - either of these is enough to
22     control the way the "get" function works - but the PCRE_INFO_JCHANGED
23     facility is supposed to tell if (?J) was ever used, so now (?J) at the
24 ph10 230 start sets both bits.
25    
26 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
27     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
28 ph10 232
29     5. doc/pcresyntax.html was missing from the distribution.
30 ph10 230
31 ph10 233 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
32     compatibility, even though it is no longer used.
33 ph10 234
34     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
35     strtoull to pcrecpp.cc to select the available functions in WIN32 (where
36     different names are used).
37 ph10 236
38     8. Changed all #include <config.h> to #include "config.h". There were also
39     some further <pcre.h> cases that I changed to "pcre.h".
40 ph10 231
41 ph10 233
42 ph10 228 Version 7.3 28-Aug-07
43 ph10 157 ---------------------
44    
45 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
46     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
47     brackets there is not right, since it causes compilers to look for an
48     installed pcre.h, not the version that is in the source that is being
49     compiled (which of course may be different). I have changed it back to:
50 ph10 197
51 ph10 189 #include "pcre.h"
52 ph10 197
53     I have a vague recollection that the change was concerned with compiling in
54     different directories, but in the new build system, that is taken care of
55     by the VPATH setting the Makefile.
56    
57 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
58     when the subject happened to end in the byte 0x85 (e.g. if the last
59     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
60     characters but of course it shouldn't be taken as a newline when it is part
61     of another character. The bug was that, for an unlimited repeat of . in
62     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
63     characters when looking for a newline.
64 ph10 197
65     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
66    
67     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
68     in debug output.
69    
70 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
71     long printing in the pcrecpp unittest when running under MinGW.
72 ph10 189
73 ph10 197 6. ESC_K was left out of the EBCDIC table.
74 ph10 189
75 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
76     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
77     limit also applies to "virtual nesting" when a pattern is recursive, and in
78     this case 1000 isn't so big. I have been able to remove this limit at the
79     expense of backing off one optimization in certain circumstances. Normally,
80     when pcre_exec() would call its internal match() function recursively and
81     immediately return the result unconditionally, it uses a "tail recursion"
82     feature to save stack. However, when a subpattern that can match an empty
83     string has an unlimited repetition quantifier, it no longer makes this
84     optimization. That gives it a stack frame in which to save the data for
85     checking that an empty string has been matched. Previously this was taken
86     from the 1000-entry workspace that had been reserved. So now there is no
87     explicit limit, but more stack is used.
88 ph10 189
89 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
90     syntax that is required for Windows, and which was going wrong for the
91     pcreposix and pcrecpp parts of the library. These were overlooked when this
92     problem was solved for the main library.
93 ph10 197
94 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
95     the size of patterns that contain repeated groups with explicit upper
96     limits. As the maximum quantifier is 65535, the maximum group length was
97     set at 30,000 so that the product of these two numbers did not overflow a
98     32-bit integer. However, it turns out that people want to use groups that
99     are longer than 30,000 bytes (though not repeat them that many times).
100     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
101     made it possible to implement the integer overflow checks in a much more
102     dynamic way, which I have now done. The artificial limitation on group
103     length has been removed - we now have only the limit on the total length of
104     the compiled pattern, which depends on the LINK_SIZE setting.
105 ph10 208
106     10. Fixed a bug in the documentation for get/copy named substring when
107     duplicate names are permitted. If none of the named substrings are set, the
108     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
109     empty string.
110    
111     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
112     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
113     because the ] is interpreted as the first data character and the
114     terminating ] is not found. PCRE has been made compatible with Perl in this
115     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
116     cause memory overwriting.
117    
118 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
119     string has been matched (to stop an infinite loop). It was not recognizing
120 ph10 208 a conditional subpattern that could match an empty string if that
121 ph10 206 subpattern was within another subpattern. For example, it looped when
122 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
123 ph10 206 condition was not nested. This bug has been fixed.
124 ph10 208
125 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
126     past the start of the subject in the presence of bytes with the top bit
127     set, for example "\x8aBCD".
128 ph10 211
129 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
130     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
131 ph10 200
132 ph10 211 14. Optimized (?!) to (*FAIL).
133 ph10 202
134 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
135     This restricts code points to be within the range 0 to 0x10FFFF, excluding
136     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
137     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
138     does: it's just the validity check that is more restrictive.
139 ph10 220
140     16. Inserted checks for integer overflows during escape sequence (backslash)
141     processing, and also fixed erroneous offset values for syntax errors during
142     backslash processing.
143    
144 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
145 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
146    
147 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
148     caused an overrun.
149 ph10 220
150     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
151     something other than just ASCII characters) inside a group that had an
152     unlimited repeat caused a loop at compile time (while checking to see
153     whether the group could match an empty string).
154    
155 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
156     [\P{Any}] did so. (Error in the code for printing property names.)
157 ph10 210
158 ph10 220 21. An orphan \E inside a character class could cause a crash.
159    
160     22. A repeated capturing bracket such as (A)? could cause a wild memory
161 ph10 218 reference during compilation.
162 ph10 220
163     23. There are several functions in pcre_compile() that scan along a compiled
164     expression for various reasons (e.g. to see if it's fixed length for look
165 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
166 ph10 220 present in the pattern. These operators have additional parameters compared
167     with \d, etc, and these were not being taken into account when moving along
168 ph10 218 the compiled data. Specifically:
169 ph10 220
170     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
171     length.
172    
173     (b) An item such as \pL+ within a repeated group could cause crashes or
174 ph10 218 loops.
175 ph10 220
176 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
177 ph10 220 "reference to non-existent subpattern" error.
178    
179 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
180    
181 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
182 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
183 ph10 211
184 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
185 ph10 218
186 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
187 ph10 223 character were causing crashes (broken optimization).
188 ph10 227
189 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
190 ph10 227 \p or \P) caused a compile-time loop.
191    
192 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
193     break. For example, the unstudied pattern [\r\n]A does not match the string
194     "\r\nA" because change 7.0/46 below moves the current point on by two
195     characters after failing to match at the start. However, the pattern \nA
196 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
197     the same is true. There doesn't seem any very clean way out of this, but
198     what I have chosen to do makes the common cases work: PCRE now takes note
199     of whether there can be an explicit match for \r or \n anywhere in the
200     pattern, and if so, 7.0/46 no longer applies. As part of this change,
201     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
202     pattern has explicit CR or LF references.
203 ph10 222
204 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
205 ph10 223
206 ph10 227
207 ph10 189 Version 7.2 19-Jun-07
208     ---------------------
209    
210 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
211     which is apparently normally available under Windows.
212 ph10 159
213     2. Re-jig the pcregrep tests with different newline settings in an attempt
214     to make them independent of the local environment's newline setting.
215 ph10 157
216 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
217 ph10 165
218 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
219     was not 2, because the output contained actual offsets. The recent new
220 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
221     usable with all link sizes.
222    
223 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
224     stack recursion. This gives a massive performance boost under BSD, but just
225 ph10 165 a small improvement under Linux. However, it saves one field in the frame
226 ph10 164 in all cases.
227 ph10 172
228 ph10 166 6. Added more features from the forthcoming Perl 5.10:
229 ph10 172
230 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
231     recursion call. It refers to the nth most recently opened parentheses.
232 ph10 172
233 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
234 ph10 172 to be opened parentheses.
235    
236     (c) Conditions that refer to capturing parentheses can be specified
237 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
238 ph10 172
239 ph10 168 (d) \K resets the start of the current match so that everything before
240 ph10 172 is not part of it.
241    
242 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
243 ph10 172
244 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
245 ph10 172 reference syntax.
246 ph10 182
247 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
248 ph10 182 alternative starts with the same number.
249 ph10 172
250 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
251    
252 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
253     PCRE_INFO_JCHANGED.
254    
255     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
256     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
257 ph10 170 for detecting groups that can match an empty string.
258 ph10 159
259 ph10 172 9. A pattern with a very large number of alternatives (more than several
260     hundred) was running out of internal workspace during the pre-compile
261     phase, where pcre_compile() figures out how much memory will be needed. A
262     bit of new cunning has reduced the workspace needed for groups with
263     alternatives. The 1000-alternative test pattern now uses 12 bytes of
264     workspace instead of running out of the 4096 that are available.
265 ph10 182
266 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
267 ph10 172
268 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
269     The report of the bug said:
270 ph10 182
271 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
272     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
273 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
274 ph10 185
275 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
276 ph10 185 it matched the wrong number of bytes.
277 ph10 172
278 ph10 179
279 ph10 155 Version 7.1 24-Apr-07
280 ph10 98 ---------------------
281    
282 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
283 ph10 122 that is more "standard", making use of automake and other Autotools. There
284 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
285     on this.
286 ph10 111
287     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
288     for recursive directory scanning broke on some systems because the files
289     are not scanned in any specific order and on different systems the order
290     was different. A call to "sort" has been inserted into RunGrepTest for the
291     approprate test as a short-term fix. In the longer term there may be an
292 ph10 100 alternative.
293 ph10 111
294 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
295 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
296     people do their own conversions for various reasons). The problems
297     concerned the use of low-level troff macros .br and .in. I have therefore
298     removed all such uses from the man pages (some were redundant, some could
299 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
300     HTML has been updated to handle .nf/.fi and to complain if it encounters
301     .br or .in.
302 ph10 111
303 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
304 ph10 123 arranged for config.h to be included in the distribution, with the name
305 ph10 111 config.h.generic, for the benefit of those who have to compile without
306     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
307    
308     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
309     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
310 ph10 127 makevp.bat for the current PCRE, using the additional files
311 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
312 ph10 111
313     6. A Windows user reported a minor discrepancy with test 2, which turned out
314     to be caused by a trailing space on an input line that had got lost in his
315 ph10 102 copy. The trailing space was an accident, so I've just removed it.
316 ph10 111
317 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
318 ph10 111 that is needed.
319    
320 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
321     as "const" (a) because they are and (b) because it helps the PHP
322     maintainers who have recently made a script to detect big data structures
323 ph10 111 in the php code that should be moved to the .rodata section. I remembered
324     to update Builducptable as well, so it won't revert if ucptable.h is ever
325 ph10 105 re-created.
326 ph10 111
327     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
328 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
329 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
330     support is not required. This saves 1.5-2K of code, which is important in
331 ph10 107 some applications.
332 ph10 111
333 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
334 ph10 111 so as not to refer to the tables, even though these functions will never be
335     called when UTF-8 support is disabled. Otherwise there are problems with a
336     shared library.
337    
338 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
339    
340     (a) It was defining its arguments as char * instead of void *.
341    
342 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
343     a long time ago when I wrote it, but is no longer the case.
344    
345 ph10 118 The emulated memove() is provided for those environments that have neither
346 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
347 ph10 118 is clearly not the case, as these two bugs were recently reported.
348 ph10 123
349 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
350 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
351     pages, and it removes trailing spaces from listed files. It also creates
352     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
353     case, it wraps all the #defines with #ifndefs. This script should be run
354 ph10 111 before "make dist".
355 ph10 123
356 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
357     with Unicode property support.
358 ph10 123
359     (a) For a maximizing quantifier, if the two different cases of the
360     character were of different lengths in their UTF-8 codings (there are
361     some cases like this - I found 11), and the matching function had to
362 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
363     were both the same length.
364 ph10 123
365     (b) When PCRE was configured to use the heap rather than the stack for
366     recursion during matching, it was not correctly preserving the data for
367     the other case of a UTF-8 character when checking ahead for a match
368     while processing a minimizing repeat. If the check also involved
369 ph10 115 matching a wide character, but failed, corruption could cause an
370     erroneous result when trying to check for a repeat of the original
371     character.
372 ph10 123
373 ph10 116 13. Some tidying changes to the testing mechanism:
374 ph10 98
375 ph10 116 (a) The RunTest script now detects the internal link size and whether there
376     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
377 ph10 123 values substituted by "configure". (The RunGrepTest script already did
378     this for UTF-8.) The configure.ac script no longer substitutes the
379     relevant variables.
380    
381 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
382     with length and offset values. This means that the output is different
383     for different internal link sizes. Test 2 is skipped for link sizes
384     other than 2 because of this, bypassing the problem. Unfortunately,
385     there was also a test in test 3 (the locale tests) that used /B and
386 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
387     I have added a new /Z option to pcretest that replaces the length and
388     offset values with spaces. This is now used to make test 3 independent
389 ph10 122 of link size. (Test 2 will be tidied up later.)
390 ph10 123
391     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
392 ph10 122 segmentation fault instead of returning the appropriate error message.
393 ph10 134
394 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
395 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
396     This doesn't seem right; it now treats the CRLF combination as the line
397     ending, and so does not match in that case. It's only a pattern such as ^$
398     that would hit this one: something like ^ABC$ would have failed after \r
399     and then tried again after \r\n.
400    
401 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
402 ph10 134 in an attempt to make files that differ only in their line terminators
403     compare equal. This works on Linux.
404 ph10 142
405 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
406     as it exited. This is now fixed, thanks to valgrind.
407 ph10 142
408 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
409 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
410     was because its rules for how to advance for /g after matching an empty
411 ph10 143 string at the end of a line did not allow for this case. They now check for
412     it specially.
413 ph10 150
414     20. pcretest is supposed to handle patterns and data of any length, by
415     extending its buffers when necessary. It was getting this wrong when the
416 ph10 147 buffer for a data line had to be extended.
417 ph10 150
418 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
419 ph10 150 CRLF as a newline sequence.
420 ph10 152
421 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
422 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
423     I have nevertheless tidied it up.
424 ph10 111
425 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
426 ph10 123
427 ph10 153 24. Added a man page for pcre-config.
428 ph10 152
429 ph10 153
430 nigel 93 Version 7.0 19-Dec-06
431     ---------------------
432    
433     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
434     moving to gcc 4.1.1.
435    
436     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
437     sys/time.h, which is documented as needed for this function. It doesn't
438     seem to matter on Linux, but it showed up on some releases of OS X.
439    
440     3. It seems that there are systems where bytes whose values are greater than
441     127 match isprint() in the "C" locale. The "C" locale should be the
442     default when a C program starts up. In most systems, only ASCII printing
443     characters match isprint(). This difference caused the output from pcretest
444     to vary, making some of the tests fail. I have changed pcretest so that:
445    
446     (a) When it is outputting text in the compiled version of a pattern, bytes
447     other than 32-126 are always shown as hex escapes.
448    
449     (b) When it is outputting text that is a matched part of a subject string,
450     it does the same, unless a different locale has been set for the match
451     (using the /L modifier). In this case, it uses isprint() to decide.
452    
453     4. Fixed a major bug that caused incorrect computation of the amount of memory
454     required for a compiled pattern when options that changed within the
455     pattern affected the logic of the preliminary scan that determines the
456     length. The relevant options are -x, and -i in UTF-8 mode. The result was
457     that the computed length was too small. The symptoms of this bug were
458     either the PCRE error "internal error: code overflow" from pcre_compile(),
459     or a glibc crash with a message such as "pcretest: free(): invalid next
460     size (fast)". Examples of patterns that provoked this bug (shown in
461     pcretest format) are:
462    
463     /(?-x: )/x
464     /(?x)(?-x: \s*#\s*)/
465     /((?i)[\x{c0}])/8
466     /(?i:[\x{c0}])/8
467    
468     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
469     is now done differently.
470    
471     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
472     wrapper classes; (b) implement a new function in the C++ scanner that is
473     more efficient than the old way of doing things because it avoids levels of
474     recursion in the regex matching; (c) add a paragraph to the documentation
475     for the FullMatch() function.
476    
477     6. The escape sequence \n was being treated as whatever was defined as
478     "newline". Not only was this contrary to the documentation, which states
479     that \n is character 10 (hex 0A), but it also went horribly wrong when
480     "newline" was defined as CRLF. This has been fixed.
481    
482     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
483     was being set to -1 for the "end of line" case (supposedly a value that no
484     character can have). Though this value is never used (the check for end of
485     line is "zero bytes in current character"), it caused compiler complaints.
486     I've changed it to 0xffffffff.
487    
488     8. In pcre_version.c, the version string was being built by a sequence of
489     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
490     string (as it is for production releases) called a macro with an empty
491     argument. The C standard says the result of this is undefined. The gcc
492     compiler treats it as an empty string (which was what was wanted) but it is
493     reported that Visual C gives an error. The source has been hacked around to
494     avoid this problem.
495    
496     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
497     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
498     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
499     of them did).
500    
501     10. Originally, pcretest opened its input and output without "b"; then I was
502     told that "b" was needed in some environments, so it was added for release
503     5.0 to both the input and output. (It makes no difference on Unix-like
504     systems.) Later I was told that it is wrong for the input on Windows. I've
505     now abstracted the modes into two macros, to make it easier to fiddle with
506     them, and removed "b" from the input mode under Windows.
507    
508     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
509    
510     12. Added -help and --help to pcretest as an official way of being reminded
511     of the options.
512    
513     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
514     and pcrecpp.cc because they annoy compilers at high warning levels.
515    
516     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
517    
518     15. Fixed an occurrence of == in configure.ac that should have been = (shell
519     scripts are not C programs :-) and which was not noticed because it works
520     on Linux.
521    
522     16. pcretest is supposed to handle any length of pattern and data line (as one
523     line or as a continued sequence of lines) by extending its input buffer if
524     necessary. This feature was broken for very long pattern lines, leading to
525     a string of junk being passed to pcre_compile() if the pattern was longer
526     than about 50K.
527    
528     17. I have done a major re-factoring of the way pcre_compile() computes the
529     amount of memory needed for a compiled pattern. Previously, there was code
530     that made a preliminary scan of the pattern in order to do this. That was
531     OK when PCRE was new, but as the facilities have expanded, it has become
532     harder and harder to keep it in step with the real compile phase, and there
533     have been a number of bugs (see for example, 4 above). I have now found a
534     cunning way of running the real compile function in a "fake" mode that
535     enables it to compute how much memory it would need, while actually only
536     ever using a few hundred bytes of working memory and without too many
537     tests of the mode. This should make future maintenance and development
538     easier. A side effect of this work is that the limit of 200 on the nesting
539     depth of parentheses has been removed (though this was never a serious
540     limitation, I suspect). However, there is a downside: pcre_compile() now
541     runs more slowly than before (30% or more, depending on the pattern). I
542     hope this isn't a big issue. There is no effect on runtime performance.
543    
544     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
545     newline (only possible for the last line of a file) and it was a
546     pattern that set a locale (followed by /Lsomething), pcretest crashed.
547    
548     19. Added additional timing features to pcretest. (1) The -tm option now times
549     matching only, not compiling. (2) Both -t and -tm can be followed, as a
550     separate command line item, by a number that specifies the number of
551     repeats to use when timing. The default is 50000; this gives better
552     precision, but takes uncomfortably long for very large patterns.
553    
554     20. Extended pcre_study() to be more clever in cases where a branch of a
555     subpattern has no definite first character. For example, (a*|b*)[cd] would
556     previously give no result from pcre_study(). Now it recognizes that the
557     first character must be a, b, c, or d.
558    
559     21. There was an incorrect error "recursive call could loop indefinitely" if
560     a subpattern (or the entire pattern) that was being tested for matching an
561     empty string contained only one non-empty item after a nested subpattern.
562     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
563     incorrectly, because the \d was being skipped in the check.
564    
565     22. The pcretest program now has a new pattern option /B and a command line
566     option -b, which is equivalent to adding /B to every pattern. This causes
567     it to show the compiled bytecode, without the additional information that
568     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
569     is the same as /B/I).
570    
571     23. A new optimization is now able automatically to treat some sequences such
572     as a*b as a*+b. More specifically, if something simple (such as a character
573     or a simple class like \d) has an unlimited quantifier, and is followed by
574     something that cannot possibly match the quantified thing, the quantifier
575     is automatically "possessified".
576    
577     24. A recursive reference to a subpattern whose number was greater than 39
578     went wrong under certain circumstances in UTF-8 mode. This bug could also
579     have affected the operation of pcre_study().
580    
581     25. Realized that a little bit of performance could be had by replacing
582     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
583    
584     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
585    
586     27. Possessive quantifiers such as a++ were previously implemented by turning
587     them into atomic groups such as ($>a+). Now they have their own opcodes,
588     which improves performance. This includes the automatically created ones
589     from 23 above.
590    
591     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
592     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
593     the first matched character to be a colon. This applied both to named and
594     numbered groups.
595    
596     29. The ucpinternal.h header file was missing its idempotency #ifdef.
597    
598     30. I was sent a "project" file called libpcre.a.dev which I understand makes
599     building PCRE on Windows easier, so I have included it in the distribution.
600    
601     31. There is now a check in pcretest against a ridiculously large number being
602     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
603     loop, the loop is abandoned.
604    
605     32. Forward references to subpatterns in conditions such as (?(2)...) where
606     subpattern 2 is defined later cause pcre_compile() to search forwards in
607     the pattern for the relevant set of parentheses. This search went wrong
608     when there were unescaped parentheses in a character class, parentheses
609     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
610    
611     33. "Subroutine" calls and backreferences were previously restricted to
612     referencing subpatterns earlier in the regex. This restriction has now
613     been removed.
614    
615     34. Added a number of extra features that are going to be in Perl 5.10. On the
616     whole, these are just syntactic alternatives for features that PCRE had
617     previously implemented using the Python syntax or my own invention. The
618     other formats are all retained for compatibility.
619    
620     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
621     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
622     also .NET compatible.
623    
624     (b) A recursion or subroutine call to a named group can now be defined as
625     (?&name) as well as (?P>name).
626    
627     (c) A backreference to a named group can now be defined as \k<name> or
628     \k'name' as well as (?P=name). The new forms, as well as being in Perl
629     5.10, are also .NET compatible.
630    
631     (d) A conditional reference to a named group can now use the syntax
632     (?(<name>) or (?('name') as well as (?(name).
633    
634     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
635     groups (named and numbered) that are never evaluated inline, but can be
636     called as "subroutines" from elsewhere. In effect, the DEFINE condition
637     is always false. There may be only one alternative in such a group.
638    
639     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
640     as the simple (?(R). The condition is true only if the most recent
641     recursion is that of the given number or name. It does not search out
642     through the entire recursion stack.
643    
644     (g) The escape \gN or \g{N} has been added, where N is a positive or
645     negative number, specifying an absolute or relative reference.
646    
647     35. Tidied to get rid of some further signed/unsigned compiler warnings and
648     some "unreachable code" warnings.
649    
650     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
651     things, this adds five new scripts.
652    
653     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
654     There were also incompatibilities regarding the handling of \Q..\E inside
655     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
656     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
657    
658     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
659     matches an empty string, and forcibly breaks the loop. There were bugs in
660     this code in non-simple cases. For a pattern such as ^(a()*)* matched
661     against aaaa the result was just "a" rather than "aaaa", for example. Two
662     separate and independent bugs (that affected different cases) have been
663     fixed.
664    
665     39. Refactored the code to abolish the use of different opcodes for small
666     capturing bracket numbers. This is a tidy that I avoided doing when I
667     removed the limit on the number of capturing brackets for 3.5 back in 2001.
668     The new approach is not only tidier, it makes it possible to reduce the
669     memory needed to fix the previous bug (38).
670    
671     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
672     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
673     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
674     mode.
675    
676     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
677     report.
678    
679     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
680     copy construction and assignment in the C++ wrapper.
681    
682     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
683     couple of bugs that could have given wrong results in the "--newline=crlf"
684     case.
685    
686     44. Added a number of casts and did some reorganization of signed/unsigned int
687     variables following suggestions from Dair Grant. Also renamed the variable
688     "this" as "item" because it is a C++ keyword.
689    
690     45. Arranged for dftables to add
691    
692     #include "pcre_internal.h"
693    
694     to pcre_chartables.c because without it, gcc 4.x may remove the array
695     definition from the final binary if PCRE is built into a static library and
696     dead code stripping is activated.
697    
698     46. For an unanchored pattern, if a match attempt fails at the start of a
699     newline sequence, and the newline setting is CRLF or ANY, and the next two
700     characters are CRLF, advance by two characters instead of one.
701    
702    
703 nigel 91 Version 6.7 04-Jul-06
704     ---------------------
705    
706     1. In order to handle tests when input lines are enormously long, pcretest has
707     been re-factored so that it automatically extends its buffers when
708     necessary. The code is crude, but this _is_ just a test program. The
709     default size has been increased from 32K to 50K.
710    
711     2. The code in pcre_study() was using the value of the re argument before
712     testing it for NULL. (Of course, in any sensible call of the function, it
713     won't be NULL.)
714    
715     3. The memmove() emulation function in pcre_internal.h, which is used on
716     systems that lack both memmove() and bcopy() - that is, hardly ever -
717     was missing a "static" storage class specifier.
718    
719     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
720     containing an extended class (one that cannot be represented by a bitmap
721     because it contains high-valued characters or Unicode property items, e.g.
722     [\pZ]). Almost always one would set UTF-8 mode when processing such a
723     pattern, but PCRE should not loop if you do not (it no longer does).
724     [Detail: two cases were found: (a) a repeated subpattern containing an
725     extended class; (b) a recursive reference to a subpattern that followed a
726     previous extended class. It wasn't skipping over the extended class
727     correctly when UTF-8 mode was not set.]
728    
729     5. A negated single-character class was not being recognized as fixed-length
730     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
731     compile error "lookbehind assertion is not fixed length".
732    
733     6. The RunPerlTest auxiliary script was showing an unexpected difference
734     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
735     write a Perl script that can interpret lines of an input file either as
736     byte characters or as UTF-8, which is what "perltest" was being required to
737     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
738     can't do is switch easily at run time between having the "use utf8;" pragma
739     or not. In the end, I fudged it by using the RunPerlTest script to insert
740     "use utf8;" explicitly for the UTF-8 tests.
741    
742     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
743     the end of the subject string, contrary to the documentation and to what
744     Perl does. This was true of both matching functions. Now it matches only at
745     the start of the subject and immediately after *internal* newlines.
746    
747     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
748     a pointer to an int instead of a pointer to an unsigned long int. This
749     caused problems on 64-bit systems.
750    
751     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
752     instance of the 'standard' template library not being so standard".
753    
754     10. There was no check on the number of named subpatterns nor the maximum
755     length of a subpattern name. The product of these values is used to compute
756     the size of the memory block for a compiled pattern. By supplying a very
757     long subpattern name and a large number of named subpatterns, the size
758     computation could be caused to overflow. This is now prevented by limiting
759     the length of names to 32 characters, and the number of named subpatterns
760     to 10,000.
761    
762     11. Subpatterns that are repeated with specific counts have to be replicated in
763     the compiled pattern. The size of memory for this was computed from the
764     length of the subpattern and the repeat count. The latter is limited to
765     65535, but there was no limit on the former, meaning that integer overflow
766     could in principle occur. The compiled length of a repeated subpattern is
767     now limited to 30,000 bytes in order to prevent this.
768    
769     12. Added the optional facility to have named substrings with the same name.
770    
771     13. Added the ability to use a named substring as a condition, using the
772     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
773     are numbers (not recommended). Forward references are permitted.
774    
775     14. Added forward references in named backreferences (if you see what I mean).
776    
777     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
778     pattern could run off the end of the subject. For example, the pattern
779     "(?s)(.{1,5})"8 did this with the subject "ab".
780    
781     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
782     PCRE_CASELESS was set when matching characters that were quantified with ?
783     or *.
784    
785     17. A character class other than a single negated character that had a minimum
786     but no maximum quantifier - for example [ab]{6,} - was not handled
787     correctly by pce_dfa_exec(). It would match only one character.
788    
789     18. A valid (though odd) pattern that looked like a POSIX character
790     class but used an invalid character after [ (for example [[,abc,]]) caused
791     pcre_compile() to give the error "Failed: internal error: code overflow" or
792     in some cases to crash with a glibc free() error. This could even happen if
793     the pattern terminated after [[ but there just happened to be a sequence of
794     letters, a binary zero, and a closing ] in the memory that followed.
795    
796     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
797     over the years. Originally (before any Unicode support), just the bottom 8
798     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
799     output from "man perlunicode" includes this:
800    
801     The regular expression compiler produces polymorphic opcodes. That
802     is, the pattern adapts to the data and automatically switches to
803     the Unicode character scheme when presented with Unicode data--or
804     instead uses a traditional byte scheme when presented with byte
805     data.
806    
807     Sadly, a wide octal escape does not cause a switch, and in a string with
808     no other multibyte characters, these octal escapes are treated as before.
809     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
810     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
811     Unicode string.
812    
813     I have not perpetrated such confusion in PCRE. Up till now, it took just
814     the bottom 8 bits, as in old Perl. I have now made octal escapes with
815     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
816     translate to the appropriate multibyte character.
817    
818     29. Applied some refactoring to reduce the number of warnings from Microsoft
819     and Borland compilers. This has included removing the fudge introduced
820     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
821     a warning about an unused variable.
822    
823     21. PCRE has not included VT (character 0x0b) in the set of whitespace
824     characters since release 4.0, because Perl (from release 5.004) does not.
825     [Or at least, is documented not to: some releases seem to be in conflict
826     with the documentation.] However, when a pattern was studied with
827     pcre_study() and all its branches started with \s, PCRE still included VT
828     as a possible starting character. Of course, this did no harm; it just
829     caused an unnecessary match attempt.
830    
831     22. Removed a now-redundant internal flag bit that recorded the fact that case
832     dependency changed within the pattern. This was once needed for "required
833     byte" processing, but is no longer used. This recovers a now-scarce options
834     bit. Also moved the least significant internal flag bit to the most-
835     significant bit of the word, which was not previously used (hangover from
836     the days when it was an int rather than a uint) to free up another bit for
837     the future.
838    
839     23. Added support for CRLF line endings as well as CR and LF. As well as the
840     default being selectable at build time, it can now be changed at runtime
841     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
842     specify that it is scanning data with non-default line endings.
843    
844     24. Changed the definition of CXXLINK to make it agree with the definition of
845     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
846    
847     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
848     recursions. This makes a big different to stack usage for some patterns.
849    
850     26. If a subpattern containing a named recursion or subroutine reference such
851     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
852     the space required for the compiled pattern went wrong and gave too small a
853     value. Depending on the environment, this could lead to "Failed: internal
854     error: code overflow at offset 49" or "glibc detected double free or
855     corruption" errors.
856    
857     27. Applied patches from Google (a) to support the new newline modes and (b) to
858     advance over multibyte UTF-8 characters in GlobalReplace.
859    
860     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
861     difference for some implementation of PCRE in some Windows version.
862    
863     29. Added some extra testing facilities to pcretest:
864    
865     \q<number> in a data line sets the "match limit" value
866     \Q<number> in a data line sets the "match recursion limt" value
867     -S <number> sets the stack size, where <number> is in megabytes
868    
869     The -S option isn't available for Windows.
870    
871    
872 nigel 89 Version 6.6 06-Feb-06
873     ---------------------
874    
875     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
876     in pcreposix.h. I have copied the definition from pcre.h.
877    
878     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
879     because pcre.h is no longer a built file.
880    
881     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
882     not normally included in the compiled code.
883    
884    
885 nigel 87 Version 6.5 01-Feb-06
886     ---------------------
887    
888     1. When using the partial match feature with pcre_dfa_exec(), it was not
889     anchoring the second and subsequent partial matches at the new starting
890     point. This could lead to incorrect results. For example, with the pattern
891     /1234/, partially matching against "123" and then "a4" gave a match.
892    
893     2. Changes to pcregrep:
894    
895     (a) All non-match returns from pcre_exec() were being treated as failures
896     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
897     error message is output. Some extra information is given for the
898     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
899     probably the only errors that are likely to be caused by users (by
900     specifying a regex that has nested indefinite repeats, for instance).
901     If there are more than 20 of these errors, pcregrep is abandoned.
902    
903     (b) A binary zero was treated as data while matching, but terminated the
904     output line if it was written out. This has been fixed: binary zeroes
905     are now no different to any other data bytes.
906    
907     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
908     used to set a locale for matching. The --locale=xxxx long option has
909     been added (no short equivalent) to specify a locale explicitly on the
910     pcregrep command, overriding the environment variables.
911    
912     (d) When -B was used with -n, some line numbers in the output were one less
913     than they should have been.
914    
915     (e) Added the -o (--only-matching) option.
916    
917     (f) If -A or -C was used with -c (count only), some lines of context were
918     accidentally printed for the final match.
919    
920     (g) Added the -H (--with-filename) option.
921    
922     (h) The combination of options -rh failed to suppress file names for files
923     that were found from directory arguments.
924    
925     (i) Added the -D (--devices) and -d (--directories) options.
926    
927     (j) Added the -F (--fixed-strings) option.
928    
929     (k) Allow "-" to be used as a file name for -f as well as for a data file.
930    
931     (l) Added the --colo(u)r option.
932    
933     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
934     is not present by default.
935    
936     3. A nasty bug was discovered in the handling of recursive patterns, that is,
937     items such as (?R) or (?1), when the recursion could match a number of
938     alternatives. If it matched one of the alternatives, but subsequently,
939     outside the recursion, there was a failure, the code tried to back up into
940     the recursion. However, because of the way PCRE is implemented, this is not
941     possible, and the result was an incorrect result from the match.
942    
943     In order to prevent this happening, the specification of recursion has
944     been changed so that all such subpatterns are automatically treated as
945     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
946    
947     4. I had overlooked the fact that, in some locales, there are characters for
948     which isalpha() is true but neither isupper() nor islower() are true. In
949     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
950     and ordfeminine) are like this. This affected the treatment of \w and \W
951     when they appeared in character classes, but not when they appeared outside
952     a character class. The bit map for "word" characters is now created
953     separately from the results of isalnum() instead of just taking it from the
954     upper, lower, and digit maps. (Plus the underscore character, of course.)
955    
956     5. The above bug also affected the handling of POSIX character classes such as
957     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
958     permanent tables. Instead, the bit maps for such a class were previously
959     created as the appropriate unions of the upper, lower, and digit bitmaps.
960     Now they are created by subtraction from the [[:word:]] class, which has
961     its own bitmap.
962    
963     6. The [[:blank:]] character class matches horizontal, but not vertical space.
964     It is created by subtracting the vertical space characters (\x09, \x0a,
965     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
966     subtraction was done in the overall bitmap for a character class, meaning
967     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
968     be recognized. This bug has been fixed.
969    
970     7. Patches from the folks at Google:
971    
972     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
973     real life, but is still worth protecting against".
974    
975     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
976     regular expressions".
977    
978     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
979     have it.
980    
981     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
982     "configure" and the latter not, in order to fix a problem somebody had
983     with compiling the Arg class on HP-UX.
984    
985     (e) Improve the error-handling of the C++ wrapper a little bit.
986    
987     (f) New tests for checking recursion limiting.
988    
989     8. The pcre_memmove() function, which is used only if the environment does not
990     have a standard memmove() function (and is therefore rarely compiled),
991     contained two bugs: (a) use of int instead of size_t, and (b) it was not
992     returning a result (though PCRE never actually uses the result).
993    
994     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
995     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
996     returned instead of calling malloc() with an overflowing number that would
997     most likely cause subsequent chaos.
998    
999     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1000    
1001     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1002     with this option is matched, the nmatch and pmatch options of regexec() are
1003     ignored.
1004    
1005     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1006     provided in case anyone wants to the the POSIX interface with UTF-8
1007     strings.
1008    
1009     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1010     C++ linking (needed for some HP-UX environments).
1011    
1012     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1013     (unused parameter) and in the pcre_printint() function (omitted "default"
1014     switch label when the default is to do nothing).
1015    
1016     15. Added some code to make it possible, when PCRE is compiled as a C++
1017     library, to replace subject pointers for pcre_exec() with a smart pointer
1018     class, thus making it possible to process discontinuous strings.
1019    
1020     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1021     much the same function. They were added by different people who were trying
1022     to make PCRE easy to compile on non-Unix systems. It has been suggested
1023     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1024     for compiling on Windows systems. I have therefore replaced it with
1025     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1026     defaults to "extern" for C or "extern C" for C++, which works fine on
1027     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1028     SCOPE with something explicit in config.h. In addition:
1029    
1030     (a) pcreposix.h still had just "extern" instead of either of these macros;
1031     I have replaced it with PCRE_DATA_SCOPE.
1032    
1033     (b) Functions such as _pcre_xclass(), which are internal to the library,
1034     but external in the C sense, all had PCRE_EXPORT in their definitions.
1035     This is apparently wrong for the Windows case, so I have removed it.
1036     (It makes no difference on Unix-like systems.)
1037    
1038     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1039     of recursive calls to match(). This is different to MATCH_LIMIT because
1040     that limits the total number of calls to match(), not all of which increase
1041     the depth of recursion. Limiting the recursion depth limits the amount of
1042     stack (or heap if NO_RECURSE is set) that is used. The default can be set
1043     when PCRE is compiled, and changed at run time. A patch from Google adds
1044     this functionality to the C++ interface.
1045    
1046     18. Changes to the handling of Unicode character properties:
1047    
1048     (a) Updated the table to Unicode 4.1.0.
1049    
1050     (b) Recognize characters that are not in the table as "Cn" (undefined).
1051    
1052     (c) I revised the way the table is implemented to a much improved format
1053     which includes recognition of ranges. It now supports the ranges that
1054     are defined in UnicodeData.txt, and it also amalgamates other
1055     characters into ranges. This has reduced the number of entries in the
1056     table from around 16,000 to around 3,000, thus reducing its size
1057     considerably. I realized I did not need to use a tree structure after
1058     all - a binary chop search is just as efficient. Having reduced the
1059     number of entries, I extended their size from 6 bytes to 8 bytes to
1060     allow for more data.
1061    
1062     (d) Added support for Unicode script names via properties such as \p{Han}.
1063    
1064     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1065     matching that character.
1066    
1067     20. When matching a repeated Unicode property with a minimum greater than zero,
1068     (for example \pL{2,}), PCRE could look past the end of the subject if it
1069     reached it while seeking the minimum number of characters. This could
1070     happen only if some of the characters were more than one byte long, because
1071     there is a check for at least the minimum number of bytes.
1072    
1073     21. Refactored the implementation of \p and \P so as to be more general, to
1074     allow for more different types of property in future. This has changed the
1075     compiled form incompatibly. Anybody with saved compiled patterns that use
1076     \p or \P will have to recompile them.
1077    
1078     22. Added "Any" and "L&" to the supported property types.
1079    
1080     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1081     but give a compile time error if the value is greater than 0xff.
1082    
1083     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1084     accidentally not being installed or uninstalled.
1085    
1086     25. The pcre.h file was built from pcre.h.in, but the only changes that were
1087     made were to insert the current release number. This seemed silly, because
1088     it made things harder for people building PCRE on systems that don't run
1089     "configure". I have turned pcre.h into a distributed file, no longer built
1090     by "configure", with the version identification directly included. There is
1091     no longer a pcre.h.in file.
1092    
1093     However, this change necessitated a change to the pcre-config script as
1094     well. It is built from pcre-config.in, and one of the substitutions was the
1095     release number. I have updated configure.ac so that ./configure now finds
1096     the release number by grepping pcre.h.
1097    
1098     26. Added the ability to run the tests under valgrind.
1099    
1100    
1101 nigel 85 Version 6.4 05-Sep-05
1102     ---------------------
1103    
1104     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1105     "--" to be printed when multiple files were scanned, even when none of the
1106     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1107     consider it to be a bug, and have restored the previous behaviour.
1108    
1109     2. A couple of code tidies to get rid of compiler warnings.
1110    
1111     3. The pcretest program used to cheat by referring to symbols in the library
1112     whose names begin with _pcre_. These are internal symbols that are not
1113     really supposed to be visible externally, and in some environments it is
1114     possible to suppress them. The cheating is now confined to including
1115     certain files from the library's source, which is a bit cleaner.
1116    
1117     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1118     file's purpose clearer.
1119    
1120     5. Reorganized pcre_ucp_findchar().
1121    
1122    
1123 nigel 83 Version 6.3 15-Aug-05
1124     ---------------------
1125    
1126     1. The file libpcre.pc.in did not have general read permission in the tarball.
1127    
1128     2. There were some problems when building without C++ support:
1129    
1130     (a) If C++ support was not built, "make install" and "make test" still
1131     tried to test it.
1132    
1133     (b) There were problems when the value of CXX was explicitly set. Some
1134     changes have been made to try to fix these, and ...
1135    
1136     (c) --disable-cpp can now be used to explicitly disable C++ support.
1137    
1138     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1139     backslash in a target when C++ was disabled. This confuses some
1140     versions of "make", apparently. Using an intermediate variable solves
1141     this. (Same for CPP_LOBJ.)
1142    
1143     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1144     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1145     necessary on certain architectures.
1146    
1147     4. Added a setting of -export-symbols-regex to the link command to remove
1148     those symbols that are exported in the C sense, but actually are local
1149     within the library, and not documented. Their names all begin with
1150     "_pcre_". This is not a perfect job, because (a) we have to except some
1151     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1152     available (and never for static libraries). I have made a note to try to
1153     find a way round (a) in the future.
1154    
1155    
1156 nigel 81 Version 6.2 01-Aug-05
1157     ---------------------
1158    
1159     1. There was no test for integer overflow of quantifier values. A construction
1160     such as {1111111111111111} would give undefined results. What is worse, if
1161     a minimum quantifier for a parenthesized subpattern overflowed and became
1162     negative, the calculation of the memory size went wrong. This could have
1163     led to memory overwriting.
1164    
1165     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1166    
1167     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1168     operating environments where this matters.
1169    
1170     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1171     PCRE options from within the C++ wrapper.
1172    
1173     5. Named capturing subpatterns were not being correctly counted when a pattern
1174     was compiled. This caused two problems: (a) If there were more than 100
1175     such subpatterns, the calculation of the memory needed for the whole
1176     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1177     back references of the form \12, where the number was greater than 9, were
1178     not recognized as back references, even though there were sufficient
1179     previous subpatterns.
1180    
1181     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1182     versions of gcc, e.g. 2.95.4.
1183    
1184    
1185 nigel 79 Version 6.1 21-Jun-05
1186     ---------------------
1187    
1188     1. There was one reference to the variable "posix" in pcretest.c that was not
1189     surrounded by "#if !defined NOPOSIX".
1190    
1191     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1192     the cross-check on the old pcre_info() function, for the benefit of the
1193     cut-down version of PCRE that is currently imported into Exim.
1194    
1195     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1196     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1197     patterns that start (?i) but I don't think that matters. The use of (?i) is
1198     just an example; this all applies to the other options as well.
1199    
1200     4. Since libtool seems to echo the compile commands it is issuing, the output
1201     from "make" can be reduced a bit by putting "@" in front of each libtool
1202     compile command.
1203    
1204     5. Patch from the folks at Google for configure.in to be a bit more thorough
1205     in checking for a suitable C++ installation before trying to compile the
1206     C++ stuff. This should fix a reported problem when a compiler was present,
1207     but no suitable headers.
1208    
1209     6. The man pages all had just "PCRE" as their title. I have changed them to
1210     be the relevant file name. I have also arranged that these names are
1211     retained in the file doc/pcre.txt, which is a concatenation in text format
1212     of all the man pages except the little individual ones for each function.
1213    
1214     7. The NON-UNIX-USE file had not been updated for the different set of source
1215     files that come with release 6. I also added a few comments about the C++
1216     wrapper.
1217    
1218    
1219 nigel 77 Version 6.0 07-Jun-05
1220     ---------------------
1221    
1222     1. Some minor internal re-organization to help with my DFA experiments.
1223    
1224     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1225     didn't matter for the library itself when fully configured, but did matter
1226     when compiling without UCP support, or within Exim, where the ucp files are
1227     not imported.
1228    
1229     3. Refactoring of the library code to split up the various functions into
1230     different source modules. The addition of the new DFA matching code (see
1231     below) to a single monolithic source would have made it really too
1232     unwieldy, quite apart from causing all the code to be include in a
1233     statically linked application, when only some functions are used. This is
1234     relevant even without the DFA addition now that patterns can be compiled in
1235     one application and matched in another.
1236    
1237     The downside of splitting up is that there have to be some external
1238     functions and data tables that are used internally in different modules of
1239     the library but which are not part of the API. These have all had their
1240     names changed to start with "_pcre_" so that they are unlikely to clash
1241     with other external names.
1242    
1243     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1244     a different (DFA) algorithm. Although it is slower than the original
1245     function, it does have some advantages for certain types of matching
1246     problem.
1247    
1248     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1249     including restarting after a partial match.
1250    
1251     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1252     defined when compiling for Windows was sent to me. I have put it into the
1253     code, though I have no means of testing or verifying it.
1254    
1255     7. Added the pcre_refcount() auxiliary function.
1256    
1257     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1258     match before or at the first newline in the subject string. In pcretest,
1259     the /f option on a pattern can be used to set this.
1260    
1261     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1262     would behave wrongly. This has been present in PCRE since release 4.0.
1263    
1264     10. A number of changes to the pcregrep command:
1265    
1266     (a) Refactored how -x works; insert ^(...)$ instead of setting
1267     PCRE_ANCHORED and checking the length, in preparation for adding
1268     something similar for -w.
1269    
1270     (b) Added the -w (match as a word) option.
1271    
1272     (c) Refactored the way lines are read and buffered so as to have more
1273     than one at a time available.
1274    
1275     (d) Implemented a pcregrep test script.
1276    
1277     (e) Added the -M (multiline match) option. This allows patterns to match
1278     over several lines of the subject. The buffering ensures that at least
1279     8K, or the rest of the document (whichever is the shorter) is available
1280     for matching (and similarly the previous 8K for lookbehind assertions).
1281    
1282     (f) Changed the --help output so that it now says
1283    
1284     -w, --word-regex(p)
1285    
1286     instead of two lines, one with "regex" and the other with "regexp"
1287     because that confused at least one person since the short forms are the
1288     same. (This required a bit of code, as the output is generated
1289     automatically from a table. It wasn't just a text change.)
1290    
1291     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1292     option but starts with a hyphen. Could be a pattern or a path name
1293     starting with a hyphen, for instance.
1294    
1295     (h) "-" can be given as a file name to represent stdin.
1296    
1297     (i) When file names are being printed, "(standard input)" is used for
1298     the standard input, for compatibility with GNU grep. Previously
1299     "<stdin>" was used.
1300    
1301     (j) The option --label=xxx can be used to supply a name to be used for
1302     stdin when file names are being printed. There is no short form.
1303    
1304     (k) Re-factored the options decoding logic because we are going to add
1305     two more options that take data. Such options can now be given in four
1306     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1307    
1308     (l) Added the -A, -B, and -C options for requesting that lines of context
1309     around matches be printed.
1310    
1311     (m) Added the -L option to print the names of files that do not contain
1312     any matching lines, that is, the complement of -l.
1313    
1314     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1315     continue to scan other files.
1316    
1317     (o) The -s option was incorrectly implemented. For compatibility with other
1318     greps, it now suppresses the error message for a non-existent or non-
1319     accessible file (but not the return code). There is a new option called
1320     -q that suppresses the output of matching lines, which was what -s was
1321     previously doing.
1322    
1323     (p) Added --include and --exclude options to specify files for inclusion
1324     and exclusion when recursing.
1325    
1326     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1327     Hopefully, it now does.
1328    
1329     12. Missing cast in pcre_study().
1330    
1331     13. Added an "uninstall" target to the makefile.
1332    
1333     14. Replaced "extern" in the function prototypes in Makefile.in with
1334     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1335     world, but is set differently for Windows.
1336    
1337     15. Added a second compiling function called pcre_compile2(). The only
1338     difference is that it has an extra argument, which is a pointer to an
1339     integer error code. When there is a compile-time failure, this is set
1340     non-zero, in addition to the error test pointer being set to point to an
1341     error message. The new argument may be NULL if no error number is required
1342     (but then you may as well call pcre_compile(), which is now just a
1343     wrapper). This facility is provided because some applications need a
1344     numeric error indication, but it has also enabled me to tidy up the way
1345     compile-time errors are handled in the POSIX wrapper.
1346    
1347     16. Added VPATH=.libs to the makefile; this should help when building with one
1348     prefix path and installing with another. (Or so I'm told by someone who
1349     knows more about this stuff than I do.)
1350    
1351     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1352     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1353     match everything, including newlines. This is not POSIX-compatible, but
1354     somebody wanted the feature. From pcretest it can be activated by using
1355     both the P and the s flags.
1356    
1357     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1358    
1359     19. libpcre.pc was being incorrectly installed as executable.
1360    
1361     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1362     it now also looks for '\r' so that it will work unmodified on Windows.
1363    
1364     21. Added Google's contributed C++ wrapper to the distribution.
1365    
1366     22. Added some untidy missing memory free() calls in pcretest, to keep
1367     Electric Fence happy when testing.
1368    
1369    
1370    
1371 nigel 75 Version 5.0 13-Sep-04
1372     ---------------------
1373    
1374     1. Internal change: literal characters are no longer packed up into items
1375     containing multiple characters in a single byte-string. Each character
1376     is now matched using a separate opcode. However, there may be more than one
1377     byte in the character in UTF-8 mode.
1378    
1379     2. The pcre_callout_block structure has two new fields: pattern_position and
1380     next_item_length. These contain the offset in the pattern to the next match
1381     item, and its length, respectively.
1382    
1383     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1384     insertion of callouts before each pattern item. Added the /C option to
1385     pcretest to make use of this.
1386    
1387     4. On the advice of a Windows user, the lines
1388    
1389     #if defined(_WIN32) || defined(WIN32)
1390     _setmode( _fileno( stdout ), 0x8000 );
1391     #endif /* defined(_WIN32) || defined(WIN32) */
1392    
1393     have been added to the source of pcretest. This apparently does useful
1394     magic in relation to line terminators.
1395    
1396     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1397     for the benefit of those environments where the "b" makes a difference.
1398    
1399     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1400     to know about it. I have put a hack into configure.in that adds in code
1401     to set GCC=yes if CC=icc. This seems to end up at a point in the
1402     generated configure script that is early enough to affect the setting of
1403     compiler options, which is what is needed, but I have no means of testing
1404     whether it really works. (The user who reported this had patched the
1405     generated configure script, which of course I cannot do.)
1406    
1407     LATER: After change 22 below (new libtool files), the configure script
1408     seems to know about icc (and also ecc). Therefore, I have commented out
1409     this hack in configure.in.
1410    
1411     7. Added support for pkg-config (2 patches were sent in).
1412    
1413     8. Negated POSIX character classes that used a combination of internal tables
1414     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1415     [[:^ascii]]. Typically, they would match almost any characters. The other
1416     POSIX classes were not broken in this way.
1417    
1418     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1419     to find the match, as PCRE was deluded into thinking that the match had to
1420     start at the start point or following a newline. The same bug applied to
1421     patterns with negative forward assertions or any backward assertions
1422     preceding ".*" at the start, unless the pattern required a fixed first
1423     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1424    
1425     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1426     starting at the last subject character, bytes beyond the end of the subject
1427     string were read.
1428    
1429     11. Renamed the variable "class" as "classbits" to make life easier for C++
1430     users. (Previously there was a macro definition, but it apparently wasn't
1431     enough.)
1432    
1433     12. Added the new field "tables" to the extra data so that tables can be passed
1434     in at exec time, or the internal tables can be re-selected. This allows
1435     a compiled regex to be saved and re-used at a later time by a different
1436     program that might have everything at different addresses.
1437    
1438     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1439     -R library as well as a -L library.
1440    
1441     14. The debugging options of pcretest (-d on the command line or D on a
1442     pattern) showed incorrect output for anything following an extended class
1443     that contained multibyte characters and which was followed by a quantifier.
1444    
1445     15. Added optional support for general category Unicode character properties
1446     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1447     support. It adds about 90K to the size of the library. The meanings of the
1448     inbuilt class escapes such as \d and \s have NOT been changed.
1449    
1450     16. Updated pcredemo.c to include calls to free() to release the memory for the
1451     compiled pattern.
1452    
1453     17. The generated file chartables.c was being created in the source directory
1454     instead of in the building directory. This caused the build to fail if the
1455     source directory was different from the building directory, and was
1456     read-only.
1457    
1458     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1459     file. No doubt somebody will tell me if they don't make sense... Also added
1460     Dan Mooney's comments about building on OpenVMS.
1461    
1462     19. Added support for partial matching via the PCRE_PARTIAL option for
1463     pcre_exec() and the \P data escape in pcretest.
1464    
1465     20. Extended pcretest with 3 new pattern features:
1466    
1467     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1468     write the compiled pattern to the file whose name is "rest-of-line".
1469     This is a straight binary dump of the data, with the saved pointer to
1470     the character tables forced to be NULL. The study data, if any, is
1471     written too. After writing, pcretest reads a new pattern.
1472    
1473     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1474     compiled pattern from the given file. There must not be any
1475     occurrences of "<" in the file name (pretty unlikely); if there are,
1476     pcretest will instead treat the initial "<" as a pattern delimiter.
1477     After reading in the pattern, pcretest goes on to read data lines as
1478     usual.
1479    
1480     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1481     and 16-bit fields in a compiled pattern, to simulate a pattern that
1482     was compiled on a host of opposite endianness.
1483    
1484     21. The pcre-exec() function can now cope with patterns that were compiled on
1485     hosts of opposite endianness, with this restriction:
1486    
1487     As for any compiled expression that is saved and used later, the tables
1488     pointer field cannot be preserved; the extra_data field in the arguments
1489     to pcre_exec() should be used to pass in a tables address if a value
1490     other than the default internal tables were used at compile time.
1491    
1492     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1493     now diagnosed as an error. Previously, most of the time, a negative number
1494     would have been treated as zero, but if in addition "ovector" was passed as
1495     NULL, a crash could occur.
1496    
1497     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1498     new versions from the libtool 1.5 distribution (the last one is a copy of
1499     a file called libtool.m4). This seems to have fixed the need to patch
1500     "configure" to support Darwin 1.3 (which I used to do). However, I still
1501     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1502     workstation).
1503    
1504     24. Changed the PCRE licence to be the more standard "BSD" licence.
1505    
1506    
1507 nigel 73 Version 4.5 01-Dec-03
1508     ---------------------
1509    
1510     1. There has been some re-arrangement of the code for the match() function so
1511     that it can be compiled in a version that does not call itself recursively.
1512     Instead, it keeps those local variables that need separate instances for
1513     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1514     needs to "recurse". Keeping track of where control must go is done by means
1515     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1516     hide most of the details from the main code, and operates only if
1517     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1518     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1519     operating.
1520    
1521     To make it easier for callers to provide specially tailored get/free
1522     functions for this usage, two new functions, pcre_stack_malloc, and
1523     pcre_stack_free, are used. They are always called in strict stacking order,
1524     and the size of block requested is always the same.
1525    
1526     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1527     PCRE has been compiled to use the stack or the heap for recursion. The
1528     -C option of pcretest uses this to show which version is compiled.
1529    
1530     A new data escape \S, is added to pcretest; it causes the amounts of store
1531     obtained and freed by both kinds of malloc/free at match time to be added
1532     to the output.
1533    
1534     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1535     what's available on my current Linux desktop machine.
1536    
1537     3. When matching a UTF-8 string, the test for a valid string at the start has
1538     been extended. If start_offset is not zero, PCRE now checks that it points
1539     to a byte that is the start of a UTF-8 character. If not, it returns
1540     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1541     this is necessary because there may be backward assertions in the pattern.
1542     When matching the same subject several times, it may save resources to use
1543     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1544    
1545     4. The code for checking the validity of UTF-8 strings has been tightened so
1546     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1547     containing "overlong sequences".
1548    
1549     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1550     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1551     should have been "&", but it just so happened that all the cases this let
1552     through by mistake were picked up later in the function.
1553    
1554     6. I had used a variable called "isblank" - this is a C99 function, causing
1555     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1556    
1557     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1558     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1559     so the version is shown; (c) stop "make test" echoing "./RunTest".
1560    
1561     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1562    
1563     9. The prototype for memmove() for systems that don't have it was using
1564     size_t, but the inclusion of the header that defines size_t was later. I've
1565     moved the #includes for the C headers earlier to avoid this.
1566    
1567     10. Added some adjustments to the code to make it easier to compiler on certain
1568     special systems:
1569    
1570     (a) Some "const" qualifiers were missing.
1571     (b) Added the macro EXPORT before all exported functions; by default this
1572     is defined to be empty.
1573     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1574     that it reads its output file name as an argument instead of writing
1575     to the standard output and assuming this can be redirected.
1576    
1577     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1578     class containing characters with values greater than 255, PCRE compilation
1579     went into a loop.
1580    
1581     12. A recursive reference to a subpattern that was within another subpattern
1582     that had a minimum quantifier of zero caused PCRE to crash. For example,
1583     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1584     recursion. If the recursively-called subpattern itself had a zero repeat,
1585     that was OK.
1586    
1587     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1588     buffer into which it was copied (for escape processing) was still set at
1589     1024, so long lines caused crashes.
1590    
1591     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1592     "internal error: code overflow...". This applied to any character class
1593     that was followed by a possessive quantifier.
1594    
1595     15. Modified the Makefile to add libpcre.la as a prerequisite for
1596     libpcreposix.la because I was told this is needed for a parallel build to
1597     work.
1598    
1599     16. If a pattern that contained .* following optional items at the start was
1600     studied, the wrong optimizing data was generated, leading to matching
1601     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1602     matching string must start with a or b or c. The correct conclusion for
1603     this pattern is that a match can start with any character.
1604    
1605    
1606 nigel 71 Version 4.4 13-Aug-03
1607     ---------------------
1608    
1609     1. In UTF-8 mode, a character class containing characters with values between
1610     127 and 255 was not handled correctly if the compiled pattern was studied.
1611     In fixing this, I have also improved the studying algorithm for such
1612     classes (slightly).
1613    
1614     2. Three internal functions had redundant arguments passed to them. Removal
1615     might give a very teeny performance improvement.
1616    
1617     3. Documentation bug: the value of the capture_top field in a callout is *one
1618     more than* the number of the hightest numbered captured substring.
1619    
1620     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1621     in incorrectly linking with a previously installed version. They now link
1622     explicitly with libpcre.la.
1623    
1624     5. configure.in no longer needs to recognize Cygwin specially.
1625    
1626     6. A problem in pcre.in for Windows platforms is fixed.
1627    
1628     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1629     pcretest, it used to include the size of the study block as part of its
1630     output. Unfortunately, the structure contains a field that has a different
1631     size on different hardware architectures. This meant that the tests that
1632     showed this size failed. As the block is currently always of a fixed size,
1633     this information isn't actually particularly useful in pcretest output, so
1634     I have just removed it.
1635    
1636     8. Three pre-processor statements accidentally did not start in column 1.
1637     Sadly, there are *still* compilers around that complain, even though
1638     standard C has not required this for well over a decade. Sigh.
1639    
1640     9. In pcretest, the code for checking callouts passed small integers in the
1641     callout_data field, which is a void * field. However, some picky compilers
1642     complained about the casts involved for this on 64-bit systems. Now
1643     pcretest passes the address of the small integer instead, which should get
1644     rid of the warnings.
1645    
1646     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1647     both compile and run time, and gives an error if an invalid UTF-8 sequence
1648     is found. There is a option for disabling this check in cases where the
1649     string is known to be correct and/or the maximum performance is wanted.
1650    
1651     11. In response to a bug report, I changed one line in Makefile.in from
1652    
1653     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1654     to
1655     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1656    
1657     to look similar to other lines, but I have no way of telling whether this
1658     is the right thing to do, as I do not use Windows. No doubt I'll get told
1659     if it's wrong...
1660    
1661    
1662 nigel 69 Version 4.3 21-May-03
1663     ---------------------
1664    
1665     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1666     Makefile.
1667    
1668     2. Some refactoring to improve the quality of the code:
1669    
1670     (i) The utf8_table... variables are now declared "const".
1671    
1672     (ii) The code for \cx, which used the "case flipping" table to upper case
1673     lower case letters, now just substracts 32. This is ASCII-specific,
1674     but the whole concept of \cx is ASCII-specific, so it seems
1675     reasonable.
1676    
1677     (iii) PCRE was using its character types table to recognize decimal and
1678     hexadecimal digits in the pattern. This is silly, because it handles
1679     only 0-9, a-f, and A-F, but the character types table is locale-
1680     specific, which means strange things might happen. A private
1681     table is now used for this - though it costs 256 bytes, a table is
1682     much faster than multiple explicit tests. Of course, the standard
1683     character types table is still used for matching digits in subject
1684     strings against \d.
1685    
1686     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1687     ending in _t are). So I've renamed it as ESC_tee.
1688    
1689     3. The first argument for regexec() in the POSIX wrapper should have been
1690     defined as "const".
1691    
1692     4. Changed pcretest to use malloc() for its buffers so that they can be
1693     Electric Fenced for debugging.
1694    
1695     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1696     to read one or more bytes before the start of the subject string. Often this
1697     had no effect on PCRE's behaviour, but in some circumstances it could
1698     provoke a segmentation fault.
1699    
1700     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1701     to try to read one or more bytes before the start of the subject string.
1702    
1703     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1704     UTF-8 support could misbehave in various ways if the subject string
1705     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1706     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1707     back over UTF-8 characters.)
1708    
1709    
1710 nigel 67 Version 4.2 14-Apr-03
1711     ---------------------
1712    
1713     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1714    
1715     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1716     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1717     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1718     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1719     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1720     and BUILD_EXEEXT
1721     Note: automatic setting of the BUILD variables is not yet working
1722     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1723     compile-time but not at link-time
1724     [LINK]: use for linking executables only
1725     make different versions for Windows and non-Windows
1726     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1727     libraries
1728     [LINK_FOR_BUILD]: new variable
1729     [OBJEXT]: use throughout
1730     [EXEEXT]: use throughout
1731     <winshared>: new target
1732     <wininstall>: new target
1733     <dftables.o>: use native compiler
1734     <dftables>: use native linker
1735     <install>: handle Windows platform correctly
1736     <clean>: ditto
1737     <check>: ditto
1738     copy DLL to top builddir before testing
1739    
1740     As part of these changes, -no-undefined was removed again. This was reported
1741     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1742     in any case.
1743    
1744     3. Some tidies to get rid of compiler warnings:
1745    
1746     . In the match_data structure, match_limit was an unsigned long int, whereas
1747     match_call_count was an int. I've made them both unsigned long ints.
1748    
1749     . In pcretest the fact that a const uschar * doesn't automatically cast to
1750     a void * provoked a warning.
1751    
1752     . Turning on some more compiler warnings threw up some "shadow" variables
1753     and a few more missing casts.
1754    
1755     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1756     option, a class that contained a single character with a value between 128
1757     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1758    
1759     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1760     option, a class that contained several characters, but with at least one
1761     whose value was between 128 and 255 caused PCRE to crash.
1762    
1763    
1764 nigel 65 Version 4.1 12-Mar-03
1765     ---------------------
1766 nigel 63
1767 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1768     needed, and a string in dftables.c that was longer than standard compilers are
1769     required to support.
1770    
1771     2. Compiling with Sun's compiler found a few more places where the code could
1772     be tidied up in order to avoid warnings.
1773    
1774     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1775     first of these names is deprecated in the latest Autoconf in favour of the name
1776     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1777     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1778     analogy I have changed it to CFLAGS_FOR_BUILD.
1779    
1780     4. Added -no-undefined to the linking command in the Makefile, because this is
1781     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1782     linking step for the pcreposix library.
1783    
1784     5. PCRE was failing to diagnose the case of two named groups with the same
1785     name.
1786    
1787     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1788     literal character that is needed in the subject for a match, and scans along to
1789     ensure that it is present before embarking on the full matching process. This
1790     saves time in cases of nested unlimited repeats that are never going to match.
1791     Problem: the scan can take a lot of time if the subject is very long (e.g.
1792     megabytes), thus penalizing straightforward matches. It is now done only if the
1793     amount of subject to be scanned is less than 1000 bytes.
1794    
1795     7. A lesser problem with the same optimization is that it was recording the
1796     first character of an anchored pattern as "needed", thus provoking a search
1797     right along the subject, even when the first match of the pattern was going to
1798     fail. The "needed" character is now not set for anchored patterns, unless it
1799     follows something in the pattern that is of non-fixed length. Thus, it still
1800     fulfils its original purpose of finding quick non-matches in cases of nested
1801     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1802    
1803    
1804     Version 4.0 17-Feb-03
1805     ---------------------
1806    
1807 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1808     extended to the end of string, PCRE compiled incorrect data. This could lead to
1809     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1810    
1811     2. Moved to autoconf 2.53 and libtool 1.4.2.
1812    
1813     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1814     the special perltest8 script is no longer needed - all the tests can be run
1815     from a single perltest script.
1816    
1817     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1818     by \s. It has now been removed in PCRE. This means it isn't recognized as
1819     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1820     class [:space:] *does* include VT, thereby creating a mess.
1821    
1822     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1823     space and tab.
1824    
1825     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1826     its new features into the main test script, reducing the number of scripts.
1827    
1828     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1829     were backward compatible, and made the (?i) apply to the whole pattern, as if
1830     /i were given. Now it behaves more logically, and applies the option setting
1831     only to what follows. PCRE has been changed to follow suit. However, if it
1832     finds options settings right at the start of the pattern, it extracts them into
1833     the global options, as before. Thus, they show up in the info data.
1834    
1835     8. Added support for the \Q...\E escape sequence. Characters in between are
1836     treated as literals. This is slightly different from Perl in that $ and @ are
1837     also handled as literals inside the quotes. In Perl, they will cause variable
1838     interpolation. Note the following examples:
1839    
1840     Pattern PCRE matches Perl matches
1841    
1842     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1843     \Qabc\$xyz\E abc\$xyz abc\$xyz
1844     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1845    
1846     For compatibility with Perl, \Q...\E sequences are recognized inside character
1847     classes as well as outside them.
1848    
1849     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1850     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1851     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1852     signed/unsigned warnings.
1853    
1854     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1855     option for pcretest, so I've replaced it by a simple function that does just
1856     that job.
1857    
1858     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1859     "pcregrep -".
1860    
1861     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1862     Java package. This provides some syntactic sugar for simple cases of what my
1863     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1864     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1865     item, you can use this simplified notation. Note that only makes sense with
1866     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1867     greediness, whatever the setting of the PCRE_UNGREEDY option.
1868    
1869     13. A change of greediness default within a pattern was not taking effect at
1870     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1871     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1872     was abstracted outside.
1873    
1874     14. PCRE now supports the \G assertion. It is true when the current matching
1875     position is at the start point of the match. This differs from \A when the
1876     starting offset is non-zero. Used with the /g option of pcretest (or similar
1877     code), it works in the same way as it does for Perl's /g option. If all
1878     alternatives of a regex begin with \G, the expression is anchored to the start
1879     match position, and the "anchored" flag is set in the compiled expression.
1880    
1881     15. Some bugs concerning the handling of certain option changes within patterns
1882     have been fixed. These applied to options other than (?ims). For example,
1883     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1884     been the other way round. Some of this was related to change 7 above.
1885    
1886     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1887     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1888     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1889     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1890    
1891     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1892     mode. Unlike ".", it always matches newline, whatever the setting of
1893     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1894     assertions. Perl allows it, but it doesn't (in general) work because it can't
1895     calculate the length of the lookbehind. At least, that's the case for Perl
1896     5.8.0 - I've been told they are going to document that it doesn't work in
1897     future.
1898    
1899     18. Added an error diagnosis for escapes that PCRE does not support: these are
1900     \L, \l, \N, \P, \p, \U, \u, and \X.
1901    
1902     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1903     reading past the end of the pattern in cases such as /[abcd/.
1904    
1905     20. PCRE was getting more memory than necessary for patterns with classes that
1906     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1907    
1908     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1909     compiling PCRE for use with Virtual Pascal.
1910    
1911     22. Small fix to the Makefile to make it work properly if the build is done
1912     outside the source tree.
1913    
1914     23. Added a new extension: a condition to go with recursion. If a conditional
1915     subpattern starts with (?(R) the "true" branch is used if recursion has
1916     happened, whereas the "false" branch is used only at the top level.
1917    
1918     24. When there was a very long string of literal characters (over 255 bytes
1919     without UTF support, over 250 bytes with UTF support), the computation of how
1920     much memory was required could be incorrect, leading to segfaults or other
1921     strange effects.
1922    
1923     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1924     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1925     there was a subsequent back reference to those brackets. This meant that, for
1926     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1927     possible to check for precisely this case. All we can do is abandon the
1928     optimization if .* occurs inside capturing brackets when there are any back
1929     references whatsoever. (See below for a better fix that came later.)
1930    
1931     26. The handling of the optimization for finding the first character of a
1932     non-anchored pattern, and for finding a character that is required later in the
1933     match were failing in some cases. This didn't break the matching; it just
1934     failed to optimize when it could. The way this is done has been re-implemented.
1935    
1936     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1937    
1938     28. Added a new feature that provides some of the functionality that Perl
1939     provides with (?{...}). The facility is termed a "callout". The way it is done
1940     in PCRE is for the caller to provide an optional function, by setting
1941     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1942     global variable. By default it is unset, which disables all calling out. To get
1943     the function called, the regex must include (?C) at appropriate points. This
1944     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1945     This provides a means of identifying different callout points. When PCRE
1946     reaches such a point in the regex, if pcre_callout has been set, the external
1947     function is called. It is provided with data in a structure called
1948     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1949     matching continues; if it returns a non-zero value, the match at the current
1950     point fails. However, backtracking will occur if possible. [This was changed
1951     later and other features added - see item 49 below.]
1952    
1953     29. pcretest is upgraded to test the callout functionality. It provides a
1954     callout function that displays information. By default, it shows the start of
1955     the match and the current position in the text. There are some new data escapes
1956     to vary what happens:
1957    
1958     \C+ in addition, show current contents of captured substrings
1959     \C- do not supply a callout function
1960     \C!n return 1 when callout number n is reached
1961     \C!n!m return 1 when callout number n is reached for the mth time
1962    
1963     30. If pcregrep was called with the -l option and just a single file name, it
1964     output "<stdin>" if a match was found, instead of the file name.
1965    
1966     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1967     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1968     pcre_exec(). This saves a malloc/free per call. The default value of
1969     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1970     when configuring.
1971    
1972     32. The default maximum size of a compiled pattern is 64K. There have been a
1973     few cases of people hitting this limit. The code now uses macros to handle the
1974     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1975     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1976     configuring. Tests 2 and 5 work only with 2-byte links because they output
1977     debugging information about compiled patterns.
1978    
1979     33. Internal code re-arrangements:
1980    
1981     (a) Moved the debugging function for printing out a compiled regex into
1982     its own source file (printint.c) and used #include to pull it into
1983     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1984     separate copies.
1985    
1986     (b) Defined the list of op-code names for debugging as a macro in
1987     internal.h so that it is next to the definition of the opcodes.
1988    
1989     (c) Defined a table of op-code lengths for simpler skipping along compiled
1990     code. This is again a macro in internal.h so that it is next to the
1991     definition of the opcodes.
1992    
1993     34. Added support for recursive calls to individual subpatterns, along the
1994     lines of Robin Houston's patch (but implemented somewhat differently).
1995    
1996     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1997     allow it to read and process whole directories in Win32. This code was
1998     contributed by Lionel Fourquaux; it has not been tested by me.
1999    
2000     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2001     used to name a group. Names consist of alphanumerics and underscores, and must
2002     be unique. Back references use the syntax (?P=name) and recursive calls use
2003     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2004     numbers. The function pcre_fullinfo() can be used after compilation to extract
2005     a name/number map. There are three relevant calls:
2006    
2007     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2008     PCRE_INFO_NAMECOUNT yields the number of entries
2009     PCRE_INFO_NAMETABLE yields a pointer to the map.
2010    
2011     The map is a vector of fixed-size entries. The size of each entry depends on
2012     the length of the longest name used. The first two bytes of each entry are the
2013     group number, most significant byte first. There follows the corresponding
2014     name, zero terminated. The names are in alphabetical order.
2015    
2016     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2017     case instead of 255. Making it the same both with and without UTF-8 support
2018     means that the same test output works with both.
2019    
2020     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2021     calling malloc() with a zero argument.
2022    
2023     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2024     optimization. I've improved things by keeping a bitmap of backreferences with
2025     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2026     fact referenced, the optimization can be applied. It is unlikely that a
2027     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2028     the match to follow \n) will appear inside brackets with a number greater than
2029     31, but if it does, any back reference > 31 suppresses the optimization.
2030    
2031     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2032     of disabling numbered capturing parentheses. Any opening parenthesis that is
2033     not followed by ? behaves as if it were followed by ?: but named parentheses
2034     can still be used for capturing (and they will acquire numbers in the usual
2035     way).
2036    
2037     41. Redesigned the return codes from the match() function into yes/no/error so
2038     that errors can be passed back from deep inside the nested calls. A malloc
2039     failure while inside a recursive subpattern call now causes the
2040     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2041    
2042     42. It is now possible to set a limit on the number of times the match()
2043     function is called in a call to pcre_exec(). This facility makes it possible to
2044     limit the amount of recursion and backtracking, though not in a directly
2045     obvious way, because the match() function is used in a number of different
2046     circumstances. The count starts from zero for each position in the subject
2047     string (for non-anchored patterns). The default limit is, for compatibility, a
2048     large number, namely 10 000 000. You can change this in two ways:
2049    
2050     (a) When configuring PCRE before making, you can use --with-match-limit=n
2051     to set a default value for the compiled library.
2052    
2053     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2054     a different value is set. See 45 below.
2055    
2056     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2057    
2058     43. Added a new function pcre_config(int, void *) to enable run-time extraction
2059     of things that can be changed at compile time. The first argument specifies
2060     what is wanted and the second points to where the information is to be placed.
2061     The current list of available information is:
2062    
2063     PCRE_CONFIG_UTF8
2064    
2065     The output is an integer that is set to one if UTF-8 support is available;
2066     otherwise it is set to zero.
2067    
2068     PCRE_CONFIG_NEWLINE
2069    
2070     The output is an integer that it set to the value of the code that is used for
2071     newline. It is either LF (10) or CR (13).
2072    
2073     PCRE_CONFIG_LINK_SIZE
2074    
2075     The output is an integer that contains the number of bytes used for internal
2076     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2077    
2078     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2079    
2080     The output is an integer that contains the threshold above which the POSIX
2081     interface uses malloc() for output vectors. See item 31 above.
2082    
2083     PCRE_CONFIG_MATCH_LIMIT
2084    
2085     The output is an unsigned integer that contains the default limit of the number
2086     of match() calls in a pcre_exec() execution. See 42 above.
2087    
2088     44. pcretest has been upgraded by the addition of the -C option. This causes it
2089     to extract all the available output from the new pcre_config() function, and to
2090     output it. The program then exits immediately.
2091    
2092     45. A need has arisen to pass over additional data with calls to pcre_exec() in
2093     order to support additional features. One way would have been to define
2094     pcre_exec2() (for example) with extra arguments, but this would not have been
2095     extensible, and would also have required all calls to the original function to
2096     be mapped to the new one. Instead, I have chosen to extend the mechanism that
2097     is used for passing in "extra" data from pcre_study().
2098    
2099     The pcre_extra structure is now exposed and defined in pcre.h. It currently
2100     contains the following fields:
2101    
2102     flags a bitmap indicating which of the following fields are set
2103     study_data opaque data from pcre_study()
2104     match_limit a way of specifying a limit on match() calls for a specific
2105     call to pcre_exec()
2106     callout_data data for callouts (see 49 below)
2107    
2108     The flag bits are also defined in pcre.h, and are
2109    
2110     PCRE_EXTRA_STUDY_DATA
2111     PCRE_EXTRA_MATCH_LIMIT
2112     PCRE_EXTRA_CALLOUT_DATA
2113    
2114     The pcre_study() function now returns one of these new pcre_extra blocks, with
2115     the actual study data pointed to by the study_data field, and the
2116     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2117     before. That is, this change is entirely upwards-compatible and requires no
2118     change to existing code.
2119    
2120     If you want to pass in additional data to pcre_exec(), you can either place it
2121     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2122     block.
2123    
2124     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2125     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2126     times with different match limits, until it finds the minimum value needed for
2127     pcre_exec() to complete. The value is then output. This can be instructive; for
2128     most simple matches the number is quite small, but for pathological cases it
2129     gets very large very quickly.
2130    
2131     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2132     returns the size of the data block pointed to by the study_data field in a
2133     pcre_extra block, that is, the value that was passed as the argument to
2134     pcre_malloc() when PCRE was getting memory in which to place the information
2135     created by pcre_study(). The fourth argument should point to a size_t variable.
2136     pcretest has been extended so that this information is shown after a successful
2137     pcre_study() call when information about the compiled regex is being displayed.
2138    
2139     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2140     because what follows is always an absolute path. (Later: it turns out that this
2141     is more than cosmetic for MinGW, because it doesn't like empty path
2142     components.)
2143    
2144     49. Some changes have been made to the callout feature (see 28 above):
2145    
2146     (i) A callout function now has three choices for what it returns:
2147    
2148     0 => success, carry on matching
2149     > 0 => failure at this point, but backtrack if possible
2150     < 0 => serious error, return this value from pcre_exec()
2151    
2152     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2153     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2154     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2155     use by callout functions. It will never be used by PCRE itself.
2156    
2157     (ii) The pcre_extra structure (see 45 above) has a void * field called
2158     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2159     pcre_callout_block structure has a field of the same name. The contents of
2160     the field passed in the pcre_extra structure are passed to the callout
2161     function in the corresponding field in the callout block. This makes it
2162     easier to use the same callout-containing regex from multiple threads. For
2163     testing, the pcretest program has a new data escape
2164    
2165     \C*n pass the number n (may be negative) as callout_data
2166    
2167     If the callout function in pcretest receives a non-zero value as
2168     callout_data, it returns that value.
2169    
2170     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2171     there were some redundant $(CFLAGS) in commands that are now specified as
2172     $(LINK), which already includes $(CFLAGS).
2173    
2174     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2175     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2176     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2177     one-byte characters throughout. Note that case-insensitive matching applies
2178     only to characters whose values are less than 256. PCRE doesn't support the
2179     notion of cases for higher-valued characters.
2180    
2181     (i) A character class whose characters are all within 0-255 is handled as
2182     a bit map, and the map is inverted for negative classes. Previously, a
2183     character > 255 always failed to match such a class; however it should
2184     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2185    
2186     (ii) A negated character class with a single character < 255 is coded as
2187     "not this character" (OP_NOT). This wasn't working properly when the test
2188     character was multibyte, either singly or repeated.
2189    
2190     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2191     mode, for example: \x{100}{2,3}.
2192    
2193     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2194     singly or repeated) now correctly test multibyte characters. However,
2195     PCRE doesn't recognize any characters with values greater than 255 as
2196     digits, spaces, or word characters. Such characters always match \D, \S,
2197     and \W, and never match \d, \s, or \w.
2198    
2199     (v) Classes may now contain characters and character ranges with values
2200     greater than 255. For example: [ab\x{100}-\x{400}].
2201    
2202     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2203     PCRE in UTF-8 mode.
2204    
2205     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2206     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2207     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2208     value.)
2209    
2210     53. The single man page has become too large. I have therefore split it up into
2211     a number of separate man pages. These also give rise to individual HTML pages;
2212     these are now put in a separate directory, and there is an index.html page that
2213     lists them all. Some hyperlinking between the pages has been installed.
2214    
2215     54. Added convenience functions for handling named capturing parentheses.
2216    
2217     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2218     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2219     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2220     are faulted.
2221    
2222     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2223     calling configure. These values are used when compiling the dftables.c program
2224     which is run to generate the source of the default character tables. They
2225     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2226     you will need to set these values.
2227    
2228     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2229    
2230    
2231     Version 3.9 02-Jan-02
2232 nigel 61 ---------------------
2233    
2234     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2235    
2236     2. If --disable-static was given, the building process failed when trying to
2237     build pcretest and pcregrep. (For some reason it was using libtool to compile
2238     them, which is not right, as they aren't part of the library.)
2239    
2240    
2241 nigel 59 Version 3.8 18-Dec-01
2242     ---------------------
2243    
2244     1. The experimental UTF-8 code was completely screwed up. It was packing the
2245     bytes in the wrong order. How dumb can you get?
2246    
2247    
2248 nigel 57 Version 3.7 29-Oct-01
2249     ---------------------
2250    
2251     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2252     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2253     this didn't happen under Solaris 8, where I normally test things.
2254    
2255 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2256     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2257     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2258     this fix an hour or so after the initial 3.7 release.)
2259 nigel 57
2260 nigel 59
2261 nigel 55 Version 3.6 23-Oct-01
2262     ---------------------
2263    
2264     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2265     offsets passed as NULL with zero offset count.
2266    
2267     2. The config.guess and config.sub files had not been updated when I moved to
2268     the latest autoconf.
2269    
2270    
2271 nigel 53 Version 3.5 15-Aug-01
2272     ---------------------
2273 nigel 5
2274 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2275     had been forgotten.
2276    
2277     2. By using declared but undefined structures, we can avoid using "void"
2278     definitions in pcre.h while keeping the internal definitions of the structures
2279     private.
2280    
2281     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2282     user point of view, this means that both static and shared libraries are built
2283     by default, but this can be individually controlled. More of the work of
2284     handling this static/shared cases is now inside libtool instead of PCRE's make
2285     file.
2286    
2287     4. The pcretest utility is now installed along with pcregrep because it is
2288     useful for users (to test regexs) and by doing this, it automatically gets
2289     relinked by libtool. The documentation has been turned into a man page, so
2290     there are now .1, .txt, and .html versions in /doc.
2291    
2292     5. Upgrades to pcregrep:
2293     (i) Added long-form option names like gnu grep.
2294     (ii) Added --help to list all options with an explanatory phrase.
2295     (iii) Added -r, --recursive to recurse into sub-directories.
2296     (iv) Added -f, --file to read patterns from a file.
2297    
2298     6. pcre_exec() was referring to its "code" argument before testing that
2299     argument for NULL (and giving an error if it was NULL).
2300    
2301     7. Upgraded Makefile.in to allow for compiling in a different directory from
2302     the source directory.
2303    
2304     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2305     options bits, the pointer it was passed was to an int instead of to an unsigned
2306     long int. This mattered only on 64-bit systems.
2307    
2308     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2309     generated) instead of pcre.in, which it its source. Also made the same change
2310     in several of the .c files.
2311    
2312     10. A new release of gcc defines printf() as a macro, which broke pcretest
2313     because it had an ifdef in the middle of a string argument for printf(). Fixed
2314     by using separate calls to printf().
2315    
2316     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2317     script, to force use of CR or LF instead of \n in the source. On non-Unix
2318     systems, the value can be set in config.h.
2319    
2320     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2321     absolute limit. Changed the text of the error message to make this clear, and
2322     likewise updated the man page.
2323    
2324     13. The limit of 99 on the number of capturing subpatterns has been removed.
2325     The new limit is 65535, which I hope will not be a "real" limit.
2326    
2327    
2328 nigel 51 Version 3.4 22-Aug-00
2329     ---------------------
2330    
2331     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2332    
2333     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2334    
2335    
2336 nigel 49 Version 3.3 01-Aug-00
2337     ---------------------
2338    
2339     1. If an octal character was given, but the value was greater than \377, it
2340     was not getting masked to the least significant bits, as documented. This could
2341     lead to crashes in some systems.
2342    
2343     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2344     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2345    
2346     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2347     These just pass their arguments on to (pcre_free)(), but they are provided
2348     because some uses of PCRE bind it to non-C systems that can call its functions,
2349     but cannot call free() or pcre_free() directly.
2350    
2351     4. Add "make test" as a synonym for "make check". Corrected some comments in
2352     the Makefile.
2353    
2354     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2355     Makefile.
2356    
2357     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2358     command called pgrep for grepping around the active processes.
2359    
2360     7. Added the beginnings of support for UTF-8 character strings.
2361    
2362     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2363     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2364     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2365     out for the ar command.)
2366    
2367    
2368 nigel 47 Version 3.2 12-May-00
2369     ---------------------
2370    
2371     This is purely a bug fixing release.
2372    
2373     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2374     of ZA. This was just one example of several cases that could provoke this bug,
2375     which was introduced by change 9 of version 2.00. The code for breaking
2376     infinite loops after an iteration that matches an empty string was't working
2377     correctly.
2378    
2379     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2380     when matched against abbab (for example). After matching an empty string, it
2381     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2382     caused it to match further down the string than it should.
2383    
2384     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2385     was there because it doesn't seem to be needed, and it causes trouble on some
2386     systems, as it is not a Standard C header. It has been removed.
2387    
2388     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2389     were reported on the Macintosh. The changes were from
2390    
2391     while ((c = *(++ptr)) != 0 && c != '\n');
2392     to
2393     while ((c = *(++ptr)) != 0 && c != '\n') ;
2394    
2395     Totally extraordinary, but if that's what it takes...
2396    
2397     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2398     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2399     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2400     assumes the way PCRE uses memmove() (always moving upwards).
2401    
2402     6. PCRE is being used in one environment where strchr() is not available. There
2403     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2404     faster code anyway.
2405    
2406    
2407 nigel 45 Version 3.1 09-Feb-00
2408     ---------------------
2409    
2410     The only change in this release is the fixing of some bugs in Makefile.in for
2411     the "install" target:
2412    
2413     (1) It was failing to install pcreposix.h.
2414    
2415     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2416    
2417    
2418 nigel 43 Version 3.0 01-Feb-00
2419     ---------------------
2420 nigel 41
2421     1. Add support for the /+ modifier to perltest (to output $` like it does in
2422     pcretest).
2423    
2424     2. Add support for the /g modifier to perltest.
2425    
2426     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2427     matches null strings.
2428    
2429     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2430     pattern. Perl treats empty patterns specially - it reuses the most recent
2431     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2432     effect.
2433    
2434     5. The POSIX interface was broken in that it was just handing over the POSIX
2435     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2436     required a bigger vector, with some working space on the end. This means that
2437     the POSIX wrapper now has to get and free some memory, and copy the results.
2438    
2439 nigel 43 6. Added some simple autoconf support, placing the test data and the
2440     documentation in separate directories, re-organizing some of the
2441     information files, and making it build pcre-config (a GNU standard). Also added
2442     libtool support for building PCRE as a shared library, which is now the
2443     default.
2444 nigel 41
2445 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2446     09 are not valid octal constants. Single digits will be used for minor values
2447     less than 10.
2448    
2449     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2450     existing programs that set these in the POSIX interface can use PCRE without
2451     modification.
2452    
2453     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2454     return all that pcre_info() returns, plus additional data. The pcre_info()
2455     function is retained for compatibility, but is considered to be obsolete.
2456    
2457     10. Added experimental recursion feature (?R) to handle one common case that
2458     Perl 5.6 will be able to do with (?p{...}).
2459    
2460     11. Added support for POSIX character classes like [:alpha:], which Perl is
2461     adopting.
2462    
2463    
2464 nigel 39 Version 2.08 31-Aug-99
2465     ----------------------
2466    
2467     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2468     trying to match at the startoffset position, but instead was moving forward to
2469     the next newline as if a previous match had failed.
2470    
2471     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2472     and could get into a loop if a null string was matched other than at the start
2473     of the subject.
2474    
2475     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2476     be distinguished at compile time, and for completeness also added PCRE_DATE.
2477    
2478     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2479     in GnuWin32 environments.
2480    
2481    
2482 nigel 37 Version 2.07 29-Jul-99
2483     ----------------------
2484    
2485     1. The documentation is now supplied in plain text form and HTML as well as in
2486     the form of man page sources.
2487    
2488     2. C++ compilers don't like assigning (void *) values to other pointer types.
2489     In particular this affects malloc(). Although there is no problem in Standard
2490     C, I've put in casts to keep C++ compilers happy.
2491    
2492     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2493     should be (const char *).
2494    
2495     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2496     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2497     However, I haven't made this a standard facility. The documentation doesn't
2498     mention it, and the Makefile doesn't support it.
2499    
2500     5. The Makefile now contains an "install" target, with editable destinations at
2501     the top of the file. The pcretest program is not installed.
2502    
2503     6. pgrep -V now gives the PCRE version number and date.
2504    
2505     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2506     causing the entire string to be ignored, instead of just the last character.
2507    
2508     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2509     non-matching string, it can take a very, very long time, even for strings of
2510     quite modest length, because of the nested recursion. PCRE now does better in
2511     some of these cases. It does this by remembering the last required literal
2512     character in the pattern, and pre-searching the subject to ensure it is present
2513     before running the real match. In other words, it applies a heuristic to detect
2514     some types of certain failure quickly, and in the above example, if presented
2515     with a string that has no trailing " it gives "no match" very quickly.
2516    
2517     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2518     other alternatives are tried instead.
2519    
2520    
2521 nigel 35 Version 2.06 09-Jun-99
2522     ----------------------
2523    
2524     1. Change pcretest's output for amount of store used to show just the code
2525     space, because the remainder (the data block) varies in size between 32-bit and
2526     64-bit systems.
2527    
2528     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2529     start matching at. This allows lookbehinds to work when searching for multiple
2530     occurrences in a string.
2531    
2532     3. Added additional options to pcretest for testing multiple occurrences:
2533    
2534     /+ outputs the rest of the string that follows a match
2535     /g loops for multiple occurrences, using the new startoffset argument
2536     /G loops for multiple occurrences by passing an incremented pointer
2537    
2538     4. PCRE wasn't doing the "first character" optimization for patterns starting
2539     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2540     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2541     the letter 'x'. On long subject strings, this gives a significant speed-up.
2542    
2543    
2544 nigel 33 Version 2.05 21-Apr-99
2545     ----------------------
2546    
2547     1. Changed the type of magic_number from int to long int so that it works
2548     properly on 16-bit systems.
2549    
2550     2. Fixed a bug which caused patterns starting with .* not to work correctly
2551     when the subject string contained newline characters. PCRE was assuming
2552     anchoring for such patterns in all cases, which is not correct because .* will
2553     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2554     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2555     must be retried after every newline in the subject.
2556    
2557    
2558 nigel 31 Version 2.04 18-Feb-99
2559     ----------------------
2560    
2561     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2562     computation of the store needed to hold the pattern was incorrect (too large).
2563     If such patterns were nested a few deep, this could multiply and become a real
2564     problem.
2565    
2566     2. Added /M option to pcretest to show the memory requirement of a specific
2567     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2568    
2569     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2570     compiled in such a way that the backtracking after subsequent failure was
2571     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2572     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2573    
2574    
2575 nigel 29 Version 2.03 02-Feb-99
2576     ----------------------
2577    
2578     1. Fixed typo and small mistake in man page.
2579    
2580 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2581     LICENCE file containing the conditions.
2582 nigel 29
2583     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2584     Perl, that is the internal \ allows the delimiter to be included in the
2585     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2586     the final delimiter, add \ to the end of the pattern (to test the error).
2587    
2588     4. Added the convenience functions for extracting substrings after a successful
2589     match. Updated pcretest to make it able to test these functions.
2590    
2591    
2592 nigel 27 Version 2.02 14-Jan-99
2593     ----------------------
2594    
2595     1. Initialized the working variables associated with each extraction so that
2596     their saving and restoring doesn't refer to uninitialized store.
2597    
2598     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2599     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2600     fix the problem.
2601    
2602     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2603     calls, and wasn't printing the correct value for compiling calls. Increased the
2604     default value of LOOPREPEAT, and the number of significant figures in the
2605     times.
2606    
2607     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2608    
2609     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2610     a building problem on Windows NT with a FAT file system.
2611    
2612    
2613 nigel 25 Version 2.01 21-Oct-98
2614     ----------------------
2615    
2616     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2617     to character tables built by pcre_maketables() in the current locale. If NULL
2618     is passed, the default tables are used.
2619    
2620    
2621 nigel 23 Version 2.00 24-Sep-98
2622 nigel 21 ----------------------
2623    
2624 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2625     it any more.
2626    
2627     2. Allow quantification of (?>) groups, and make it work correctly.
2628    
2629     3. The first character computation wasn't working for (?>) groups.
2630    
2631     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2632     end of the subject) and add 5.005's \z, which really does match only at the
2633     very end of the subject.
2634    
2635     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2636    
2637     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2638     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2639     localized options. All options to pcre_study() were also removed.
2640    
2641     7. Add other new features from 5.005:
2642    
2643     $(?<= positive lookbehind
2644     $(?<! negative lookbehind
2645     (?imsx-imsx) added the unsetting capability
2646     such a setting is global if at outer level; local otherwise
2647     (?imsx-imsx:) non-capturing groups with option setting
2648     (?(cond)re|re) conditional pattern matching
2649    
2650     A backreference to itself in a repeated group matches the previous
2651     captured string.
2652    
2653     8. General tidying up of studying (both automatic and via "study")
2654     consequential on the addition of new assertions.
2655    
2656     9. As in 5.005, unlimited repeated groups that could match an empty substring
2657     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2658     runtime if any iteration does actually match an empty substring.
2659    
2660     10. Include the RunTest script in the distribution.
2661    
2662     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2663     discrepancies, some of which were old and were also with respect to 5.004. They
2664     have now been fixed.
2665    
2666    
2667     Version 1.09 28-Apr-98
2668     ----------------------
2669    
2670 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2671     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2672     program crashes, or just wrong answers. This did not apply to negated classes
2673     containing more than one character, or to minima other than one.
2674    
2675    
2676 nigel 19 Version 1.08 27-Mar-98
2677     ----------------------
2678    
2679     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2680    
2681     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2682     latter must appear before anything that relies on it in the pattern.
2683    
2684    
2685 nigel 17 Version 1.07 16-Feb-98
2686     ----------------------
2687    
2688     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2689     repeat of a potentially empty string).
2690    
2691    
2692 nigel 15 Version 1.06 23-Jan-98
2693     ----------------------
2694    
2695     1. Added Markus Oberhumer's little patches for C++.
2696    
2697     2. Literal strings longer than 255 characters were broken.
2698    
2699    
2700 nigel 13 Version 1.05 23-Dec-97
2701     ----------------------
2702    
2703     1. Negated character classes containing more than one character were failing if
2704     PCRE_CASELESS was set at run time.
2705    
2706    
2707 nigel 11 Version 1.04 19-Dec-97
2708     ----------------------
2709    
2710     1. Corrected the man page, where some "const" qualifiers had been omitted.
2711    
2712     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2713     input syntax.
2714    
2715     3. Fixed memory leak which occurred when a regex with back references was
2716     matched with an offsets vector that wasn't big enough. The temporary memory
2717     that is used in this case wasn't being freed if the match failed.
2718    
2719     4. Tidied pcretest to ensure it frees memory that it gets.
2720    
2721     5. Temporary memory was being obtained in the case where the passed offsets
2722     vector was exactly big enough.
2723    
2724     6. Corrected definition of offsetof() from change 5 below.
2725    
2726     7. I had screwed up change 6 below and broken the rules for the use of
2727     setjmp(). Now fixed.
2728    
2729    
2730 nigel 9 Version 1.03 18-Dec-97
2731     ----------------------
2732    
2733     1. A erroneous regex with a missing opening parenthesis was correctly
2734     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2735     on some systems.
2736    
2737     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2738     it was reported that one broken compiler failed on the former because "code" is
2739     also an independent variable.
2740    
2741     3. The erroneous regex a[]b caused an array overrun reference.
2742    
2743     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2744     fail on data ending with that character. (It was going on too far, and checking
2745     the next character, typically a binary zero.) This was specific to the
2746     optimized code for single-character negative classes.
2747    
2748     5. Added a contributed patch from the TIN world which does the following:
2749    
2750     + Add an undef for memmove, in case the the system defines a macro for it.
2751    
2752     + Add a definition of offsetof(), in case there isn't one. (I don't know
2753     the reason behind this - offsetof() is part of the ANSI standard - but
2754     it does no harm).
2755    
2756     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2757     most of the places where whitespace preceded '#'. I have given up and
2758     allowed the remaining 2 cases to be at the margin.
2759    
2760     + Rename some variables in pcre to eliminate shadowing. This seems very
2761     pedantic, but does no harm, of course.
2762    
2763     6. Moved the call to setjmp() into its own function, to get rid of warnings
2764     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2765    
2766     7. Constructs such as \d{8,} were compiling into the equivalent of
2767 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2768 nigel 9 outcome, but in this particular case used more store than had been allocated,
2769     which caused the bug to be discovered because it threw up an internal error.
2770    
2771     8. The debugging code in both pcre and pcretest for outputting the compiled
2772     form of a regex was going wrong in the case of back references followed by
2773     curly-bracketed repeats.
2774    
2775    
2776 nigel 7 Version 1.02 12-Dec-97
2777     ----------------------
2778    
2779     1. Typos in pcre.3 and comments in the source fixed.
2780    
2781     2. Applied a contributed patch to get rid of places where it used to remove
2782     'const' from variables, and fixed some signed/unsigned and uninitialized
2783     variable warnings.
2784    
2785     3. Added the "runtest" target to Makefile.
2786    
2787     4. Set default compiler flag to -O2 rather than just -O.
2788    
2789    
2790 nigel 5 Version 1.01 19-Nov-97
2791     ----------------------
2792    
2793     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2794     like /([ab]*)*/, that is, for classes with more than one character in them.
2795    
2796     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2797     as /((?>a*))*/ (a PCRE_EXTRA facility).
2798    
2799    
2800     Version 1.00 18-Nov-97
2801     ----------------------
2802    
2803     1. Added compile-time macros to support systems such as SunOS4 which don't have
2804     memmove() or strerror() but have other things that can be used instead.
2805    
2806     2. Arranged that "make clean" removes the executables.
2807    
2808    
2809 nigel 3 Version 0.99 27-Oct-97
2810     ----------------------
2811    
2812     1. Fixed bug in code for optimizing classes with only one character. It was
2813     initializing a 32-byte map regardless, which could cause it to run off the end
2814     of the memory it had got.
2815    
2816     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2817    
2818    
2819     Version 0.98 22-Oct-97
2820     ----------------------
2821    
2822     1. Fixed bug in code for handling temporary memory usage when there are more
2823     back references than supplied space in the ovector. This could cause segfaults.
2824    
2825    
2826     Version 0.97 21-Oct-97
2827     ----------------------
2828    
2829     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2830    
2831     2. Optimized negated single characters not to use a bit map.
2832    
2833     3. Brought error texts together as macro definitions; clarified some of them;
2834     fixed one that was wrong - it said "range out of order" when it meant "invalid
2835     escape sequence".
2836    
2837     4. Changed some char * arguments to const char *.
2838    
2839     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2840    
2841     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2842     pcretest.
2843    
2844    
2845     Version 0.96 16-Oct-97
2846     ----------------------
2847    
2848     1. Added a simple "pgrep" utility to the distribution.
2849    
2850     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2851     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2852     where "ddd" means "one or more decimal digits".
2853    
2854     3. Fixed serious bug. If a pattern had a back reference, but the call to
2855     pcre_exec() didn't supply a large enough ovector to record the related
2856     identifying subpattern, the match always failed. PCRE now remembers the number
2857     of the largest back reference, and gets some temporary memory in which to save
2858     the offsets during matching if necessary, in order to ensure that
2859     backreferences always work.
2860    
2861     4. Increased the compatibility with Perl in a number of ways:
2862    
2863     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2864     to request this handling. The option can be set at compile or exec time.
2865    
2866     (b) $ matches before a terminating newline by default; an option
2867     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2868     mode). The option can be set at compile or exec time.
2869    
2870     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2871     the same as Perl's. If the decimal number it represents is less than 10
2872     or there aren't that many previous left capturing parentheses, an octal
2873     escape is read. Inside a character class, it's always an octal escape,
2874     even if it is a single digit.
2875    
2876     (d) An escaped but undefined alphabetic character is taken as a literal,
2877     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2878     escapes.
2879    
2880     (e) {0} is now permitted. (The previous item is removed from the compiled
2881     pattern).
2882    
2883     5. Changed all the names of code files so that the basic parts are no longer
2884     than 10 characters, and abolished the teeny "globals.c" file.
2885    
2886     6. Changed the handling of character classes; they are now done with a 32-byte
2887     bit map always.
2888    
2889     7. Added the -d and /D options to pcretest to make it possible to look at the
2890     internals of compilation without having to recompile pcre.
2891    
2892    
2893     Version 0.95 23-Sep-97
2894     ----------------------
2895    
2896     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2897     \x20 at the start of a run of normal characters. These were being treated as
2898     real characters, instead of the source characters being re-checked.
2899    
2900    
2901     Version 0.94 18-Sep-97
2902     ----------------------
2903    
2904     1. The functions are now thread-safe, with the caveat that the global variables
2905     containing pointers to malloc() and free() or alternative functions are the
2906     same for all threads.
2907    
2908     2. Get pcre_study() to generate a bitmap of initial characters for non-
2909     anchored patterns when this is possible, and use it if passed to pcre_exec().
2910    
2911    
2912     Version 0.93 15-Sep-97
2913     ----------------------
2914    
2915     1. /(b)|(:+)/ was computing an incorrect first character.
2916    
2917     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2918     but not actually doing anything yet.
2919    
2920     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2921     as Perl does (e.g. [-az] or [az-]).
2922    
2923     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2924     all possible positions.
2925    
2926     5. Split up into different modules to avoid including unneeded functions in a
2927     compiled binary. However, compile and exec are still in one module. The "study"
2928     function is split off.
2929    
2930     6. The character tables are now in a separate module whose source is generated
2931     by an auxiliary program - but can then be edited by hand if required. There are
2932     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2933     toupper() in the code.
2934    
2935     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2936     make them global. Abolish the function for setting them, as the caller can now
2937     set them directly.
2938    
2939    
2940     Version 0.92 11-Sep-97
2941     ----------------------
2942    
2943     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2944     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2945    
2946     2. Caseless matching was not working in character classes if the characters in
2947     the pattern were in upper case.
2948    
2949     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2950    
2951     4. Make PCRE_ANCHORED public and accept as a compile option.
2952    
2953     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2954     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2955     pass them.
2956    
2957     6. Give an error if bad option bits passed at compile or run time.
2958    
2959     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2960     pcretest to cause it to pass that flag.
2961    
2962     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2963     options, and the first character, if set.
2964    
2965     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2966    
2967    
2968     Version 0.91 10-Sep-97
2969     ----------------------
2970    
2971     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2972     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2973    
2974     2. PCRE was looping on encountering an indefinitely repeated back reference to
2975     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2976     Perl does - treats the match as successful.
2977    
2978     ****

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