ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 253 - (hide annotations) (download)
Mon Sep 17 10:51:30 2007 UTC (7 years, 7 months ago) by ph10
File size: 141388 byte(s)
Added checks for ANY and ANYCRLF to pcrecpp.cc where previously it checked only 
for CRLF.

1 nigel 3 ChangeLog for PCRE
2     ------------------
4 ph10 230 Version 7.4 10-Sep-07
5     ---------------------
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.
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 230
29 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
31     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
32 ph10 233 compatibility, even though it is no longer used.
33 ph10 243
34     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
35 ph10 248 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
36     windows.h file is present (where different names are used).
37 ph10 243
38     8. Changed all #include <config.h> to #include "config.h". There were also
39 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
40 ph10 243
41 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
42 ph10 243 sequence off the lines that it output.
44     10. It was pointed out to me that arrays of string pointers cause lots of
45     relocations when a shared library is dynamically loaded. A technique of
46     using a single long string with a table of offsets can drastically reduce
47     these. I have refactored PCRE in four places to do this. The result is
48 ph10 240 dramatic:
49 ph10 243
50 ph10 240 Originally: 290
51     After changing UCP table: 187
52 ph10 243 After changing error message table: 43
53 ph10 240 After changing table of "verbs" 36
54     After changing table of Posix names 22
55 ph10 243
56 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
57 ph10 247
58 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
59 ph10 247 unicode-properties was also set.
61 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
62 ph10 231
63 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
64     checked only for CRLF.
65 ph10 233
66 ph10 253
67 ph10 228 Version 7.3 28-Aug-07
68 ph10 157 ---------------------
70 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
71     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
72     brackets there is not right, since it causes compilers to look for an
73     installed pcre.h, not the version that is in the source that is being
74     compiled (which of course may be different). I have changed it back to:
75 ph10 197
76 ph10 189 #include "pcre.h"
77 ph10 197
78     I have a vague recollection that the change was concerned with compiling in
79     different directories, but in the new build system, that is taken care of
80     by the VPATH setting the Makefile.
82 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
83     when the subject happened to end in the byte 0x85 (e.g. if the last
84     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
85     characters but of course it shouldn't be taken as a newline when it is part
86     of another character. The bug was that, for an unlimited repeat of . in
87     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
88     characters when looking for a newline.
89 ph10 197
90     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
92     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
93     in debug output.
95 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
96     long printing in the pcrecpp unittest when running under MinGW.
97 ph10 189
98 ph10 197 6. ESC_K was left out of the EBCDIC table.
99 ph10 189
100 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
101     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
102     limit also applies to "virtual nesting" when a pattern is recursive, and in
103     this case 1000 isn't so big. I have been able to remove this limit at the
104     expense of backing off one optimization in certain circumstances. Normally,
105     when pcre_exec() would call its internal match() function recursively and
106     immediately return the result unconditionally, it uses a "tail recursion"
107     feature to save stack. However, when a subpattern that can match an empty
108     string has an unlimited repetition quantifier, it no longer makes this
109     optimization. That gives it a stack frame in which to save the data for
110     checking that an empty string has been matched. Previously this was taken
111     from the 1000-entry workspace that had been reserved. So now there is no
112     explicit limit, but more stack is used.
113 ph10 189
114 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
115     syntax that is required for Windows, and which was going wrong for the
116     pcreposix and pcrecpp parts of the library. These were overlooked when this
117     problem was solved for the main library.
118 ph10 197
119 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
120     the size of patterns that contain repeated groups with explicit upper
121     limits. As the maximum quantifier is 65535, the maximum group length was
122     set at 30,000 so that the product of these two numbers did not overflow a
123     32-bit integer. However, it turns out that people want to use groups that
124     are longer than 30,000 bytes (though not repeat them that many times).
125     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
126     made it possible to implement the integer overflow checks in a much more
127     dynamic way, which I have now done. The artificial limitation on group
128     length has been removed - we now have only the limit on the total length of
129     the compiled pattern, which depends on the LINK_SIZE setting.
130 ph10 208
131     10. Fixed a bug in the documentation for get/copy named substring when
132     duplicate names are permitted. If none of the named substrings are set, the
133     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
134     empty string.
136     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
137     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
138     because the ] is interpreted as the first data character and the
139     terminating ] is not found. PCRE has been made compatible with Perl in this
140     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
141     cause memory overwriting.
143 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
144     string has been matched (to stop an infinite loop). It was not recognizing
145 ph10 208 a conditional subpattern that could match an empty string if that
146 ph10 206 subpattern was within another subpattern. For example, it looped when
147 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
148 ph10 206 condition was not nested. This bug has been fixed.
149 ph10 208
150 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
151     past the start of the subject in the presence of bytes with the top bit
152     set, for example "\x8aBCD".
153 ph10 211
154 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
155     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
156 ph10 200
157 ph10 211 14. Optimized (?!) to (*FAIL).
158 ph10 202
159 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
160     This restricts code points to be within the range 0 to 0x10FFFF, excluding
161     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
162     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
163     does: it's just the validity check that is more restrictive.
164 ph10 220
165     16. Inserted checks for integer overflows during escape sequence (backslash)
166     processing, and also fixed erroneous offset values for syntax errors during
167     backslash processing.
169 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
170 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
172 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
173     caused an overrun.
174 ph10 220
175     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
176     something other than just ASCII characters) inside a group that had an
177     unlimited repeat caused a loop at compile time (while checking to see
178     whether the group could match an empty string).
180 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
181     [\P{Any}] did so. (Error in the code for printing property names.)
182 ph10 210
183 ph10 220 21. An orphan \E inside a character class could cause a crash.
185     22. A repeated capturing bracket such as (A)? could cause a wild memory
186 ph10 218 reference during compilation.
187 ph10 220
188     23. There are several functions in pcre_compile() that scan along a compiled
189     expression for various reasons (e.g. to see if it's fixed length for look
190 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
191 ph10 220 present in the pattern. These operators have additional parameters compared
192     with \d, etc, and these were not being taken into account when moving along
193 ph10 218 the compiled data. Specifically:
194 ph10 220
195     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
196     length.
198     (b) An item such as \pL+ within a repeated group could cause crashes or
199 ph10 218 loops.
200 ph10 220
201 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
202 ph10 220 "reference to non-existent subpattern" error.
204 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
206 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
207 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
208 ph10 211
209 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
210 ph10 218
211 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
212 ph10 223 character were causing crashes (broken optimization).
213 ph10 227
214 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
215 ph10 227 \p or \P) caused a compile-time loop.
217 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
218     break. For example, the unstudied pattern [\r\n]A does not match the string
219     "\r\nA" because change 7.0/46 below moves the current point on by two
220     characters after failing to match at the start. However, the pattern \nA
221 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
222     the same is true. There doesn't seem any very clean way out of this, but
223     what I have chosen to do makes the common cases work: PCRE now takes note
224     of whether there can be an explicit match for \r or \n anywhere in the
225     pattern, and if so, 7.0/46 no longer applies. As part of this change,
226     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
227     pattern has explicit CR or LF references.
228 ph10 222
229 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
230 ph10 223
231 ph10 227
232 ph10 189 Version 7.2 19-Jun-07
233     ---------------------
235 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
236     which is apparently normally available under Windows.
237 ph10 159
238     2. Re-jig the pcregrep tests with different newline settings in an attempt
239     to make them independent of the local environment's newline setting.
240 ph10 157
241 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
242 ph10 165
243 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
244     was not 2, because the output contained actual offsets. The recent new
245 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
246     usable with all link sizes.
248 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
249     stack recursion. This gives a massive performance boost under BSD, but just
250 ph10 165 a small improvement under Linux. However, it saves one field in the frame
251 ph10 164 in all cases.
252 ph10 172
253 ph10 166 6. Added more features from the forthcoming Perl 5.10:
254 ph10 172
255 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
256     recursion call. It refers to the nth most recently opened parentheses.
257 ph10 172
258 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
259 ph10 172 to be opened parentheses.
261     (c) Conditions that refer to capturing parentheses can be specified
262 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
263 ph10 172
264 ph10 168 (d) \K resets the start of the current match so that everything before
265 ph10 172 is not part of it.
267 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
268 ph10 172
269 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
270 ph10 172 reference syntax.
271 ph10 182
272 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
273 ph10 182 alternative starts with the same number.
274 ph10 172
275 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
277 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
280     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
281     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
282 ph10 170 for detecting groups that can match an empty string.
283 ph10 159
284 ph10 172 9. A pattern with a very large number of alternatives (more than several
285     hundred) was running out of internal workspace during the pre-compile
286     phase, where pcre_compile() figures out how much memory will be needed. A
287     bit of new cunning has reduced the workspace needed for groups with
288     alternatives. The 1000-alternative test pattern now uses 12 bytes of
289     workspace instead of running out of the 4096 that are available.
290 ph10 182
291 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
292 ph10 172
293 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
294     The report of the bug said:
295 ph10 182
296 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
297     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
298 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
299 ph10 185
300 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
301 ph10 185 it matched the wrong number of bytes.
302 ph10 172
303 ph10 179
304 ph10 155 Version 7.1 24-Apr-07
305 ph10 98 ---------------------
307 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
308 ph10 122 that is more "standard", making use of automake and other Autotools. There
309 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
310     on this.
311 ph10 111
312     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
313     for recursive directory scanning broke on some systems because the files
314     are not scanned in any specific order and on different systems the order
315     was different. A call to "sort" has been inserted into RunGrepTest for the
316     approprate test as a short-term fix. In the longer term there may be an
317 ph10 100 alternative.
318 ph10 111
319 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
320 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
321     people do their own conversions for various reasons). The problems
322     concerned the use of low-level troff macros .br and .in. I have therefore
323     removed all such uses from the man pages (some were redundant, some could
324 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
325     HTML has been updated to handle .nf/.fi and to complain if it encounters
326     .br or .in.
327 ph10 111
328 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
329 ph10 123 arranged for config.h to be included in the distribution, with the name
330 ph10 111 config.h.generic, for the benefit of those who have to compile without
331     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
333     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
334     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
335 ph10 127 makevp.bat for the current PCRE, using the additional files
336 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
337 ph10 111
338     6. A Windows user reported a minor discrepancy with test 2, which turned out
339     to be caused by a trailing space on an input line that had got lost in his
340 ph10 102 copy. The trailing space was an accident, so I've just removed it.
341 ph10 111
342 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
343 ph10 111 that is needed.
345 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
346     as "const" (a) because they are and (b) because it helps the PHP
347     maintainers who have recently made a script to detect big data structures
348 ph10 111 in the php code that should be moved to the .rodata section. I remembered
349     to update Builducptable as well, so it won't revert if ucptable.h is ever
350 ph10 105 re-created.
351 ph10 111
352     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
353 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
354 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
355     support is not required. This saves 1.5-2K of code, which is important in
356 ph10 107 some applications.
357 ph10 111
358 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
359 ph10 111 so as not to refer to the tables, even though these functions will never be
360     called when UTF-8 support is disabled. Otherwise there are problems with a
361     shared library.
363 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
365     (a) It was defining its arguments as char * instead of void *.
367 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
368     a long time ago when I wrote it, but is no longer the case.
370 ph10 118 The emulated memove() is provided for those environments that have neither
371 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
372 ph10 118 is clearly not the case, as these two bugs were recently reported.
373 ph10 123
374 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
375 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
376     pages, and it removes trailing spaces from listed files. It also creates
377     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
378     case, it wraps all the #defines with #ifndefs. This script should be run
379 ph10 111 before "make dist".
380 ph10 123
381 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
382     with Unicode property support.
383 ph10 123
384     (a) For a maximizing quantifier, if the two different cases of the
385     character were of different lengths in their UTF-8 codings (there are
386     some cases like this - I found 11), and the matching function had to
387 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
388     were both the same length.
389 ph10 123
390     (b) When PCRE was configured to use the heap rather than the stack for
391     recursion during matching, it was not correctly preserving the data for
392     the other case of a UTF-8 character when checking ahead for a match
393     while processing a minimizing repeat. If the check also involved
394 ph10 115 matching a wide character, but failed, corruption could cause an
395     erroneous result when trying to check for a repeat of the original
396     character.
397 ph10 123
398 ph10 116 13. Some tidying changes to the testing mechanism:
399 ph10 98
400 ph10 116 (a) The RunTest script now detects the internal link size and whether there
401     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
402 ph10 123 values substituted by "configure". (The RunGrepTest script already did
403     this for UTF-8.) The configure.ac script no longer substitutes the
404     relevant variables.
406 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
407     with length and offset values. This means that the output is different
408     for different internal link sizes. Test 2 is skipped for link sizes
409     other than 2 because of this, bypassing the problem. Unfortunately,
410     there was also a test in test 3 (the locale tests) that used /B and
411 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
412     I have added a new /Z option to pcretest that replaces the length and
413     offset values with spaces. This is now used to make test 3 independent
414 ph10 122 of link size. (Test 2 will be tidied up later.)
415 ph10 123
416     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
417 ph10 122 segmentation fault instead of returning the appropriate error message.
418 ph10 134
419 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
420 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
421     This doesn't seem right; it now treats the CRLF combination as the line
422     ending, and so does not match in that case. It's only a pattern such as ^$
423     that would hit this one: something like ^ABC$ would have failed after \r
424     and then tried again after \r\n.
426 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
427 ph10 134 in an attempt to make files that differ only in their line terminators
428     compare equal. This works on Linux.
429 ph10 142
430 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
431     as it exited. This is now fixed, thanks to valgrind.
432 ph10 142
433 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
434 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
435     was because its rules for how to advance for /g after matching an empty
436 ph10 143 string at the end of a line did not allow for this case. They now check for
437     it specially.
438 ph10 150
439     20. pcretest is supposed to handle patterns and data of any length, by
440     extending its buffers when necessary. It was getting this wrong when the
441 ph10 147 buffer for a data line had to be extended.
442 ph10 150
443 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
444 ph10 150 CRLF as a newline sequence.
445 ph10 152
446 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
447 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
448     I have nevertheless tidied it up.
449 ph10 111
450 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
451 ph10 123
452 ph10 153 24. Added a man page for pcre-config.
453 ph10 152
454 ph10 153
455 nigel 93 Version 7.0 19-Dec-06
456     ---------------------
458     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
459     moving to gcc 4.1.1.
461     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
462     sys/time.h, which is documented as needed for this function. It doesn't
463     seem to matter on Linux, but it showed up on some releases of OS X.
465     3. It seems that there are systems where bytes whose values are greater than
466     127 match isprint() in the "C" locale. The "C" locale should be the
467     default when a C program starts up. In most systems, only ASCII printing
468     characters match isprint(). This difference caused the output from pcretest
469     to vary, making some of the tests fail. I have changed pcretest so that:
471     (a) When it is outputting text in the compiled version of a pattern, bytes
472     other than 32-126 are always shown as hex escapes.
474     (b) When it is outputting text that is a matched part of a subject string,
475     it does the same, unless a different locale has been set for the match
476     (using the /L modifier). In this case, it uses isprint() to decide.
478     4. Fixed a major bug that caused incorrect computation of the amount of memory
479     required for a compiled pattern when options that changed within the
480     pattern affected the logic of the preliminary scan that determines the
481     length. The relevant options are -x, and -i in UTF-8 mode. The result was
482     that the computed length was too small. The symptoms of this bug were
483     either the PCRE error "internal error: code overflow" from pcre_compile(),
484     or a glibc crash with a message such as "pcretest: free(): invalid next
485     size (fast)". Examples of patterns that provoked this bug (shown in
486     pcretest format) are:
488     /(?-x: )/x
489     /(?x)(?-x: \s*#\s*)/
490     /((?i)[\x{c0}])/8
491     /(?i:[\x{c0}])/8
493     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
494     is now done differently.
496     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
497     wrapper classes; (b) implement a new function in the C++ scanner that is
498     more efficient than the old way of doing things because it avoids levels of
499     recursion in the regex matching; (c) add a paragraph to the documentation
500     for the FullMatch() function.
502     6. The escape sequence \n was being treated as whatever was defined as
503     "newline". Not only was this contrary to the documentation, which states
504     that \n is character 10 (hex 0A), but it also went horribly wrong when
505     "newline" was defined as CRLF. This has been fixed.
507     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
508     was being set to -1 for the "end of line" case (supposedly a value that no
509     character can have). Though this value is never used (the check for end of
510     line is "zero bytes in current character"), it caused compiler complaints.
511     I've changed it to 0xffffffff.
513     8. In pcre_version.c, the version string was being built by a sequence of
514     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
515     string (as it is for production releases) called a macro with an empty
516     argument. The C standard says the result of this is undefined. The gcc
517     compiler treats it as an empty string (which was what was wanted) but it is
518     reported that Visual C gives an error. The source has been hacked around to
519     avoid this problem.
521     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
522     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
523     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
524     of them did).
526     10. Originally, pcretest opened its input and output without "b"; then I was
527     told that "b" was needed in some environments, so it was added for release
528     5.0 to both the input and output. (It makes no difference on Unix-like
529     systems.) Later I was told that it is wrong for the input on Windows. I've
530     now abstracted the modes into two macros, to make it easier to fiddle with
531     them, and removed "b" from the input mode under Windows.
533     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
535     12. Added -help and --help to pcretest as an official way of being reminded
536     of the options.
538     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
539     and pcrecpp.cc because they annoy compilers at high warning levels.
541     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
543     15. Fixed an occurrence of == in configure.ac that should have been = (shell
544     scripts are not C programs :-) and which was not noticed because it works
545     on Linux.
547     16. pcretest is supposed to handle any length of pattern and data line (as one
548     line or as a continued sequence of lines) by extending its input buffer if
549     necessary. This feature was broken for very long pattern lines, leading to
550     a string of junk being passed to pcre_compile() if the pattern was longer
551     than about 50K.
553     17. I have done a major re-factoring of the way pcre_compile() computes the
554     amount of memory needed for a compiled pattern. Previously, there was code
555     that made a preliminary scan of the pattern in order to do this. That was
556     OK when PCRE was new, but as the facilities have expanded, it has become
557     harder and harder to keep it in step with the real compile phase, and there
558     have been a number of bugs (see for example, 4 above). I have now found a
559     cunning way of running the real compile function in a "fake" mode that
560     enables it to compute how much memory it would need, while actually only
561     ever using a few hundred bytes of working memory and without too many
562     tests of the mode. This should make future maintenance and development
563     easier. A side effect of this work is that the limit of 200 on the nesting
564     depth of parentheses has been removed (though this was never a serious
565     limitation, I suspect). However, there is a downside: pcre_compile() now
566     runs more slowly than before (30% or more, depending on the pattern). I
567     hope this isn't a big issue. There is no effect on runtime performance.
569     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
570     newline (only possible for the last line of a file) and it was a
571     pattern that set a locale (followed by /Lsomething), pcretest crashed.
573     19. Added additional timing features to pcretest. (1) The -tm option now times
574     matching only, not compiling. (2) Both -t and -tm can be followed, as a
575     separate command line item, by a number that specifies the number of
576     repeats to use when timing. The default is 50000; this gives better
577     precision, but takes uncomfortably long for very large patterns.
579     20. Extended pcre_study() to be more clever in cases where a branch of a
580     subpattern has no definite first character. For example, (a*|b*)[cd] would
581     previously give no result from pcre_study(). Now it recognizes that the
582     first character must be a, b, c, or d.
584     21. There was an incorrect error "recursive call could loop indefinitely" if
585     a subpattern (or the entire pattern) that was being tested for matching an
586     empty string contained only one non-empty item after a nested subpattern.
587     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
588     incorrectly, because the \d was being skipped in the check.
590     22. The pcretest program now has a new pattern option /B and a command line
591     option -b, which is equivalent to adding /B to every pattern. This causes
592     it to show the compiled bytecode, without the additional information that
593     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
594     is the same as /B/I).
596     23. A new optimization is now able automatically to treat some sequences such
597     as a*b as a*+b. More specifically, if something simple (such as a character
598     or a simple class like \d) has an unlimited quantifier, and is followed by
599     something that cannot possibly match the quantified thing, the quantifier
600     is automatically "possessified".
602     24. A recursive reference to a subpattern whose number was greater than 39
603     went wrong under certain circumstances in UTF-8 mode. This bug could also
604     have affected the operation of pcre_study().
606     25. Realized that a little bit of performance could be had by replacing
607     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
609     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
611     27. Possessive quantifiers such as a++ were previously implemented by turning
612     them into atomic groups such as ($>a+). Now they have their own opcodes,
613     which improves performance. This includes the automatically created ones
614     from 23 above.
616     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
617     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
618     the first matched character to be a colon. This applied both to named and
619     numbered groups.
621     29. The ucpinternal.h header file was missing its idempotency #ifdef.
623     30. I was sent a "project" file called libpcre.a.dev which I understand makes
624     building PCRE on Windows easier, so I have included it in the distribution.
626     31. There is now a check in pcretest against a ridiculously large number being
627     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
628     loop, the loop is abandoned.
630     32. Forward references to subpatterns in conditions such as (?(2)...) where
631     subpattern 2 is defined later cause pcre_compile() to search forwards in
632     the pattern for the relevant set of parentheses. This search went wrong
633     when there were unescaped parentheses in a character class, parentheses
634     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
636     33. "Subroutine" calls and backreferences were previously restricted to
637     referencing subpatterns earlier in the regex. This restriction has now
638     been removed.
640     34. Added a number of extra features that are going to be in Perl 5.10. On the
641     whole, these are just syntactic alternatives for features that PCRE had
642     previously implemented using the Python syntax or my own invention. The
643     other formats are all retained for compatibility.
645     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
646     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
647     also .NET compatible.
649     (b) A recursion or subroutine call to a named group can now be defined as
650     (?&name) as well as (?P>name).
652     (c) A backreference to a named group can now be defined as \k<name> or
653     \k'name' as well as (?P=name). The new forms, as well as being in Perl
654     5.10, are also .NET compatible.
656     (d) A conditional reference to a named group can now use the syntax
657     (?(<name>) or (?('name') as well as (?(name).
659     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
660     groups (named and numbered) that are never evaluated inline, but can be
661     called as "subroutines" from elsewhere. In effect, the DEFINE condition
662     is always false. There may be only one alternative in such a group.
664     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
665     as the simple (?(R). The condition is true only if the most recent
666     recursion is that of the given number or name. It does not search out
667     through the entire recursion stack.
669     (g) The escape \gN or \g{N} has been added, where N is a positive or
670     negative number, specifying an absolute or relative reference.
672     35. Tidied to get rid of some further signed/unsigned compiler warnings and
673     some "unreachable code" warnings.
675     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
676     things, this adds five new scripts.
678     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
679     There were also incompatibilities regarding the handling of \Q..\E inside
680     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
681     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
683     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
684     matches an empty string, and forcibly breaks the loop. There were bugs in
685     this code in non-simple cases. For a pattern such as ^(a()*)* matched
686     against aaaa the result was just "a" rather than "aaaa", for example. Two
687     separate and independent bugs (that affected different cases) have been
688     fixed.
690     39. Refactored the code to abolish the use of different opcodes for small
691     capturing bracket numbers. This is a tidy that I avoided doing when I
692     removed the limit on the number of capturing brackets for 3.5 back in 2001.
693     The new approach is not only tidier, it makes it possible to reduce the
694     memory needed to fix the previous bug (38).
696     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
697     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
698     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
699     mode.
701     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
702     report.
704     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
705     copy construction and assignment in the C++ wrapper.
707     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
708     couple of bugs that could have given wrong results in the "--newline=crlf"
709     case.
711     44. Added a number of casts and did some reorganization of signed/unsigned int
712     variables following suggestions from Dair Grant. Also renamed the variable
713     "this" as "item" because it is a C++ keyword.
715     45. Arranged for dftables to add
717     #include "pcre_internal.h"
719     to pcre_chartables.c because without it, gcc 4.x may remove the array
720     definition from the final binary if PCRE is built into a static library and
721     dead code stripping is activated.
723     46. For an unanchored pattern, if a match attempt fails at the start of a
724     newline sequence, and the newline setting is CRLF or ANY, and the next two
725     characters are CRLF, advance by two characters instead of one.
728 nigel 91 Version 6.7 04-Jul-06
729     ---------------------
731     1. In order to handle tests when input lines are enormously long, pcretest has
732     been re-factored so that it automatically extends its buffers when
733     necessary. The code is crude, but this _is_ just a test program. The
734     default size has been increased from 32K to 50K.
736     2. The code in pcre_study() was using the value of the re argument before
737     testing it for NULL. (Of course, in any sensible call of the function, it
738     won't be NULL.)
740     3. The memmove() emulation function in pcre_internal.h, which is used on
741     systems that lack both memmove() and bcopy() - that is, hardly ever -
742     was missing a "static" storage class specifier.
744     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
745     containing an extended class (one that cannot be represented by a bitmap
746     because it contains high-valued characters or Unicode property items, e.g.
747     [\pZ]). Almost always one would set UTF-8 mode when processing such a
748     pattern, but PCRE should not loop if you do not (it no longer does).
749     [Detail: two cases were found: (a) a repeated subpattern containing an
750     extended class; (b) a recursive reference to a subpattern that followed a
751     previous extended class. It wasn't skipping over the extended class
752     correctly when UTF-8 mode was not set.]
754     5. A negated single-character class was not being recognized as fixed-length
755     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
756     compile error "lookbehind assertion is not fixed length".
758     6. The RunPerlTest auxiliary script was showing an unexpected difference
759     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
760     write a Perl script that can interpret lines of an input file either as
761     byte characters or as UTF-8, which is what "perltest" was being required to
762     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
763     can't do is switch easily at run time between having the "use utf8;" pragma
764     or not. In the end, I fudged it by using the RunPerlTest script to insert
765     "use utf8;" explicitly for the UTF-8 tests.
767     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
768     the end of the subject string, contrary to the documentation and to what
769     Perl does. This was true of both matching functions. Now it matches only at
770     the start of the subject and immediately after *internal* newlines.
772     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
773     a pointer to an int instead of a pointer to an unsigned long int. This
774     caused problems on 64-bit systems.
776     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
777     instance of the 'standard' template library not being so standard".
779     10. There was no check on the number of named subpatterns nor the maximum
780     length of a subpattern name. The product of these values is used to compute
781     the size of the memory block for a compiled pattern. By supplying a very
782     long subpattern name and a large number of named subpatterns, the size
783     computation could be caused to overflow. This is now prevented by limiting
784     the length of names to 32 characters, and the number of named subpatterns
785     to 10,000.
787     11. Subpatterns that are repeated with specific counts have to be replicated in
788     the compiled pattern. The size of memory for this was computed from the
789     length of the subpattern and the repeat count. The latter is limited to
790     65535, but there was no limit on the former, meaning that integer overflow
791     could in principle occur. The compiled length of a repeated subpattern is
792     now limited to 30,000 bytes in order to prevent this.
794     12. Added the optional facility to have named substrings with the same name.
796     13. Added the ability to use a named substring as a condition, using the
797     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
798     are numbers (not recommended). Forward references are permitted.
800     14. Added forward references in named backreferences (if you see what I mean).
802     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
803     pattern could run off the end of the subject. For example, the pattern
804     "(?s)(.{1,5})"8 did this with the subject "ab".
806     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
807     PCRE_CASELESS was set when matching characters that were quantified with ?
808     or *.
810     17. A character class other than a single negated character that had a minimum
811     but no maximum quantifier - for example [ab]{6,} - was not handled
812     correctly by pce_dfa_exec(). It would match only one character.
814     18. A valid (though odd) pattern that looked like a POSIX character
815     class but used an invalid character after [ (for example [[,abc,]]) caused
816     pcre_compile() to give the error "Failed: internal error: code overflow" or
817     in some cases to crash with a glibc free() error. This could even happen if
818     the pattern terminated after [[ but there just happened to be a sequence of
819     letters, a binary zero, and a closing ] in the memory that followed.
821     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
822     over the years. Originally (before any Unicode support), just the bottom 8
823     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
824     output from "man perlunicode" includes this:
826     The regular expression compiler produces polymorphic opcodes. That
827     is, the pattern adapts to the data and automatically switches to
828     the Unicode character scheme when presented with Unicode data--or
829     instead uses a traditional byte scheme when presented with byte
830     data.
832     Sadly, a wide octal escape does not cause a switch, and in a string with
833     no other multibyte characters, these octal escapes are treated as before.
834     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
835     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
836     Unicode string.
838     I have not perpetrated such confusion in PCRE. Up till now, it took just
839     the bottom 8 bits, as in old Perl. I have now made octal escapes with
840     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
841     translate to the appropriate multibyte character.
843     29. Applied some refactoring to reduce the number of warnings from Microsoft
844     and Borland compilers. This has included removing the fudge introduced
845     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
846     a warning about an unused variable.
848     21. PCRE has not included VT (character 0x0b) in the set of whitespace
849     characters since release 4.0, because Perl (from release 5.004) does not.
850     [Or at least, is documented not to: some releases seem to be in conflict
851     with the documentation.] However, when a pattern was studied with
852     pcre_study() and all its branches started with \s, PCRE still included VT
853     as a possible starting character. Of course, this did no harm; it just
854     caused an unnecessary match attempt.
856     22. Removed a now-redundant internal flag bit that recorded the fact that case
857     dependency changed within the pattern. This was once needed for "required
858     byte" processing, but is no longer used. This recovers a now-scarce options
859     bit. Also moved the least significant internal flag bit to the most-
860     significant bit of the word, which was not previously used (hangover from
861     the days when it was an int rather than a uint) to free up another bit for
862     the future.
864     23. Added support for CRLF line endings as well as CR and LF. As well as the
865     default being selectable at build time, it can now be changed at runtime
866     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
867     specify that it is scanning data with non-default line endings.
869     24. Changed the definition of CXXLINK to make it agree with the definition of
870     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
872     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
873     recursions. This makes a big different to stack usage for some patterns.
875     26. If a subpattern containing a named recursion or subroutine reference such
876     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
877     the space required for the compiled pattern went wrong and gave too small a
878     value. Depending on the environment, this could lead to "Failed: internal
879     error: code overflow at offset 49" or "glibc detected double free or
880     corruption" errors.
882     27. Applied patches from Google (a) to support the new newline modes and (b) to
883     advance over multibyte UTF-8 characters in GlobalReplace.
885     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
886     difference for some implementation of PCRE in some Windows version.
888     29. Added some extra testing facilities to pcretest:
890     \q<number> in a data line sets the "match limit" value
891     \Q<number> in a data line sets the "match recursion limt" value
892     -S <number> sets the stack size, where <number> is in megabytes
894     The -S option isn't available for Windows.
897 nigel 89 Version 6.6 06-Feb-06
898     ---------------------
900     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
901     in pcreposix.h. I have copied the definition from pcre.h.
903     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
904     because pcre.h is no longer a built file.
906     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
907     not normally included in the compiled code.
910 nigel 87 Version 6.5 01-Feb-06
911     ---------------------
913     1. When using the partial match feature with pcre_dfa_exec(), it was not
914     anchoring the second and subsequent partial matches at the new starting
915     point. This could lead to incorrect results. For example, with the pattern
916     /1234/, partially matching against "123" and then "a4" gave a match.
918     2. Changes to pcregrep:
920     (a) All non-match returns from pcre_exec() were being treated as failures
921     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
922     error message is output. Some extra information is given for the
924     probably the only errors that are likely to be caused by users (by
925     specifying a regex that has nested indefinite repeats, for instance).
926     If there are more than 20 of these errors, pcregrep is abandoned.
928     (b) A binary zero was treated as data while matching, but terminated the
929     output line if it was written out. This has been fixed: binary zeroes
930     are now no different to any other data bytes.
932     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
933     used to set a locale for matching. The --locale=xxxx long option has
934     been added (no short equivalent) to specify a locale explicitly on the
935     pcregrep command, overriding the environment variables.
937     (d) When -B was used with -n, some line numbers in the output were one less
938     than they should have been.
940     (e) Added the -o (--only-matching) option.
942     (f) If -A or -C was used with -c (count only), some lines of context were
943     accidentally printed for the final match.
945     (g) Added the -H (--with-filename) option.
947     (h) The combination of options -rh failed to suppress file names for files
948     that were found from directory arguments.
950     (i) Added the -D (--devices) and -d (--directories) options.
952     (j) Added the -F (--fixed-strings) option.
954     (k) Allow "-" to be used as a file name for -f as well as for a data file.
956     (l) Added the --colo(u)r option.
958     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
959     is not present by default.
961     3. A nasty bug was discovered in the handling of recursive patterns, that is,
962     items such as (?R) or (?1), when the recursion could match a number of
963     alternatives. If it matched one of the alternatives, but subsequently,
964     outside the recursion, there was a failure, the code tried to back up into
965     the recursion. However, because of the way PCRE is implemented, this is not
966     possible, and the result was an incorrect result from the match.
968     In order to prevent this happening, the specification of recursion has
969     been changed so that all such subpatterns are automatically treated as
970     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
972     4. I had overlooked the fact that, in some locales, there are characters for
973     which isalpha() is true but neither isupper() nor islower() are true. In
974     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
975     and ordfeminine) are like this. This affected the treatment of \w and \W
976     when they appeared in character classes, but not when they appeared outside
977     a character class. The bit map for "word" characters is now created
978     separately from the results of isalnum() instead of just taking it from the
979     upper, lower, and digit maps. (Plus the underscore character, of course.)
981     5. The above bug also affected the handling of POSIX character classes such as
982     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
983     permanent tables. Instead, the bit maps for such a class were previously
984     created as the appropriate unions of the upper, lower, and digit bitmaps.
985     Now they are created by subtraction from the [[:word:]] class, which has
986     its own bitmap.
988     6. The [[:blank:]] character class matches horizontal, but not vertical space.
989     It is created by subtracting the vertical space characters (\x09, \x0a,
990     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
991     subtraction was done in the overall bitmap for a character class, meaning
992     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
993     be recognized. This bug has been fixed.
995     7. Patches from the folks at Google:
997     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
998     real life, but is still worth protecting against".
1000     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1001     regular expressions".
1003     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1004     have it.
1006     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1007     "configure" and the latter not, in order to fix a problem somebody had
1008     with compiling the Arg class on HP-UX.
1010     (e) Improve the error-handling of the C++ wrapper a little bit.
1012     (f) New tests for checking recursion limiting.
1014     8. The pcre_memmove() function, which is used only if the environment does not
1015     have a standard memmove() function (and is therefore rarely compiled),
1016     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1017     returning a result (though PCRE never actually uses the result).
1019     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1020     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1021     returned instead of calling malloc() with an overflowing number that would
1022     most likely cause subsequent chaos.
1024     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1026     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1027     with this option is matched, the nmatch and pmatch options of regexec() are
1028     ignored.
1030     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1031     provided in case anyone wants to the the POSIX interface with UTF-8
1032     strings.
1034     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1035     C++ linking (needed for some HP-UX environments).
1037     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1038     (unused parameter) and in the pcre_printint() function (omitted "default"
1039     switch label when the default is to do nothing).
1041     15. Added some code to make it possible, when PCRE is compiled as a C++
1042     library, to replace subject pointers for pcre_exec() with a smart pointer
1043     class, thus making it possible to process discontinuous strings.
1045     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1046     much the same function. They were added by different people who were trying
1047     to make PCRE easy to compile on non-Unix systems. It has been suggested
1048     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1049     for compiling on Windows systems. I have therefore replaced it with
1050     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1051     defaults to "extern" for C or "extern C" for C++, which works fine on
1052     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1053     SCOPE with something explicit in config.h. In addition:
1055     (a) pcreposix.h still had just "extern" instead of either of these macros;
1056     I have replaced it with PCRE_DATA_SCOPE.
1058     (b) Functions such as _pcre_xclass(), which are internal to the library,
1059     but external in the C sense, all had PCRE_EXPORT in their definitions.
1060     This is apparently wrong for the Windows case, so I have removed it.
1061     (It makes no difference on Unix-like systems.)
1063     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1064     of recursive calls to match(). This is different to MATCH_LIMIT because
1065     that limits the total number of calls to match(), not all of which increase
1066     the depth of recursion. Limiting the recursion depth limits the amount of
1067     stack (or heap if NO_RECURSE is set) that is used. The default can be set
1068     when PCRE is compiled, and changed at run time. A patch from Google adds
1069     this functionality to the C++ interface.
1071     18. Changes to the handling of Unicode character properties:
1073     (a) Updated the table to Unicode 4.1.0.
1075     (b) Recognize characters that are not in the table as "Cn" (undefined).
1077     (c) I revised the way the table is implemented to a much improved format
1078     which includes recognition of ranges. It now supports the ranges that
1079     are defined in UnicodeData.txt, and it also amalgamates other
1080     characters into ranges. This has reduced the number of entries in the
1081     table from around 16,000 to around 3,000, thus reducing its size
1082     considerably. I realized I did not need to use a tree structure after
1083     all - a binary chop search is just as efficient. Having reduced the
1084     number of entries, I extended their size from 6 bytes to 8 bytes to
1085     allow for more data.
1087     (d) Added support for Unicode script names via properties such as \p{Han}.
1089     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1090     matching that character.
1092     20. When matching a repeated Unicode property with a minimum greater than zero,
1093     (for example \pL{2,}), PCRE could look past the end of the subject if it
1094     reached it while seeking the minimum number of characters. This could
1095     happen only if some of the characters were more than one byte long, because
1096     there is a check for at least the minimum number of bytes.
1098     21. Refactored the implementation of \p and \P so as to be more general, to
1099     allow for more different types of property in future. This has changed the
1100     compiled form incompatibly. Anybody with saved compiled patterns that use
1101     \p or \P will have to recompile them.
1103     22. Added "Any" and "L&" to the supported property types.
1105     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1106     but give a compile time error if the value is greater than 0xff.
1108     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1109     accidentally not being installed or uninstalled.
1111     25. The pcre.h file was built from pcre.h.in, but the only changes that were
1112     made were to insert the current release number. This seemed silly, because
1113     it made things harder for people building PCRE on systems that don't run
1114     "configure". I have turned pcre.h into a distributed file, no longer built
1115     by "configure", with the version identification directly included. There is
1116     no longer a pcre.h.in file.
1118     However, this change necessitated a change to the pcre-config script as
1119     well. It is built from pcre-config.in, and one of the substitutions was the
1120     release number. I have updated configure.ac so that ./configure now finds
1121     the release number by grepping pcre.h.
1123     26. Added the ability to run the tests under valgrind.
1126 nigel 85 Version 6.4 05-Sep-05
1127     ---------------------
1129     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1130     "--" to be printed when multiple files were scanned, even when none of the
1131     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1132     consider it to be a bug, and have restored the previous behaviour.
1134     2. A couple of code tidies to get rid of compiler warnings.
1136     3. The pcretest program used to cheat by referring to symbols in the library
1137     whose names begin with _pcre_. These are internal symbols that are not
1138     really supposed to be visible externally, and in some environments it is
1139     possible to suppress them. The cheating is now confined to including
1140     certain files from the library's source, which is a bit cleaner.
1142     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1143     file's purpose clearer.
1145     5. Reorganized pcre_ucp_findchar().
1148 nigel 83 Version 6.3 15-Aug-05
1149     ---------------------
1151     1. The file libpcre.pc.in did not have general read permission in the tarball.
1153     2. There were some problems when building without C++ support:
1155     (a) If C++ support was not built, "make install" and "make test" still
1156     tried to test it.
1158     (b) There were problems when the value of CXX was explicitly set. Some
1159     changes have been made to try to fix these, and ...
1161     (c) --disable-cpp can now be used to explicitly disable C++ support.
1163     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1164     backslash in a target when C++ was disabled. This confuses some
1165     versions of "make", apparently. Using an intermediate variable solves
1166     this. (Same for CPP_LOBJ.)
1168     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1169     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1170     necessary on certain architectures.
1172     4. Added a setting of -export-symbols-regex to the link command to remove
1173     those symbols that are exported in the C sense, but actually are local
1174     within the library, and not documented. Their names all begin with
1175     "_pcre_". This is not a perfect job, because (a) we have to except some
1176     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1177     available (and never for static libraries). I have made a note to try to
1178     find a way round (a) in the future.
1181 nigel 81 Version 6.2 01-Aug-05
1182     ---------------------
1184     1. There was no test for integer overflow of quantifier values. A construction
1185     such as {1111111111111111} would give undefined results. What is worse, if
1186     a minimum quantifier for a parenthesized subpattern overflowed and became
1187     negative, the calculation of the memory size went wrong. This could have
1188     led to memory overwriting.
1190     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1192     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1193     operating environments where this matters.
1195     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1196     PCRE options from within the C++ wrapper.
1198     5. Named capturing subpatterns were not being correctly counted when a pattern
1199     was compiled. This caused two problems: (a) If there were more than 100
1200     such subpatterns, the calculation of the memory needed for the whole
1201     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1202     back references of the form \12, where the number was greater than 9, were
1203     not recognized as back references, even though there were sufficient
1204     previous subpatterns.
1206     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1207     versions of gcc, e.g. 2.95.4.
1210 nigel 79 Version 6.1 21-Jun-05
1211     ---------------------
1213     1. There was one reference to the variable "posix" in pcretest.c that was not
1214     surrounded by "#if !defined NOPOSIX".
1216     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1217     the cross-check on the old pcre_info() function, for the benefit of the
1218     cut-down version of PCRE that is currently imported into Exim.
1220     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1221     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1222     patterns that start (?i) but I don't think that matters. The use of (?i) is
1223     just an example; this all applies to the other options as well.
1225     4. Since libtool seems to echo the compile commands it is issuing, the output
1226     from "make" can be reduced a bit by putting "@" in front of each libtool
1227     compile command.
1229     5. Patch from the folks at Google for configure.in to be a bit more thorough
1230     in checking for a suitable C++ installation before trying to compile the
1231     C++ stuff. This should fix a reported problem when a compiler was present,
1232     but no suitable headers.
1234     6. The man pages all had just "PCRE" as their title. I have changed them to
1235     be the relevant file name. I have also arranged that these names are
1236     retained in the file doc/pcre.txt, which is a concatenation in text format
1237     of all the man pages except the little individual ones for each function.
1239     7. The NON-UNIX-USE file had not been updated for the different set of source
1240     files that come with release 6. I also added a few comments about the C++
1241     wrapper.
1244 nigel 77 Version 6.0 07-Jun-05
1245     ---------------------
1247     1. Some minor internal re-organization to help with my DFA experiments.
1249     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1250     didn't matter for the library itself when fully configured, but did matter
1251     when compiling without UCP support, or within Exim, where the ucp files are
1252     not imported.
1254     3. Refactoring of the library code to split up the various functions into
1255     different source modules. The addition of the new DFA matching code (see
1256     below) to a single monolithic source would have made it really too
1257     unwieldy, quite apart from causing all the code to be include in a
1258     statically linked application, when only some functions are used. This is
1259     relevant even without the DFA addition now that patterns can be compiled in
1260     one application and matched in another.
1262     The downside of splitting up is that there have to be some external
1263     functions and data tables that are used internally in different modules of
1264     the library but which are not part of the API. These have all had their
1265     names changed to start with "_pcre_" so that they are unlikely to clash
1266     with other external names.
1268     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1269     a different (DFA) algorithm. Although it is slower than the original
1270     function, it does have some advantages for certain types of matching
1271     problem.
1273     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1274     including restarting after a partial match.
1276     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1277     defined when compiling for Windows was sent to me. I have put it into the
1278     code, though I have no means of testing or verifying it.
1280     7. Added the pcre_refcount() auxiliary function.
1282     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1283     match before or at the first newline in the subject string. In pcretest,
1284     the /f option on a pattern can be used to set this.
1286     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1287     would behave wrongly. This has been present in PCRE since release 4.0.
1289     10. A number of changes to the pcregrep command:
1291     (a) Refactored how -x works; insert ^(...)$ instead of setting
1292     PCRE_ANCHORED and checking the length, in preparation for adding
1293     something similar for -w.
1295     (b) Added the -w (match as a word) option.
1297     (c) Refactored the way lines are read and buffered so as to have more
1298     than one at a time available.
1300     (d) Implemented a pcregrep test script.
1302     (e) Added the -M (multiline match) option. This allows patterns to match
1303     over several lines of the subject. The buffering ensures that at least
1304     8K, or the rest of the document (whichever is the shorter) is available
1305     for matching (and similarly the previous 8K for lookbehind assertions).
1307     (f) Changed the --help output so that it now says
1309     -w, --word-regex(p)
1311     instead of two lines, one with "regex" and the other with "regexp"
1312     because that confused at least one person since the short forms are the
1313     same. (This required a bit of code, as the output is generated
1314     automatically from a table. It wasn't just a text change.)
1316     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1317     option but starts with a hyphen. Could be a pattern or a path name
1318     starting with a hyphen, for instance.
1320     (h) "-" can be given as a file name to represent stdin.
1322     (i) When file names are being printed, "(standard input)" is used for
1323     the standard input, for compatibility with GNU grep. Previously
1324     "<stdin>" was used.
1326     (j) The option --label=xxx can be used to supply a name to be used for
1327     stdin when file names are being printed. There is no short form.
1329     (k) Re-factored the options decoding logic because we are going to add
1330     two more options that take data. Such options can now be given in four
1331     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1333     (l) Added the -A, -B, and -C options for requesting that lines of context
1334     around matches be printed.
1336     (m) Added the -L option to print the names of files that do not contain
1337     any matching lines, that is, the complement of -l.
1339     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1340     continue to scan other files.
1342     (o) The -s option was incorrectly implemented. For compatibility with other
1343     greps, it now suppresses the error message for a non-existent or non-
1344     accessible file (but not the return code). There is a new option called
1345     -q that suppresses the output of matching lines, which was what -s was
1346     previously doing.
1348     (p) Added --include and --exclude options to specify files for inclusion
1349     and exclusion when recursing.
1351     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1352     Hopefully, it now does.
1354     12. Missing cast in pcre_study().
1356     13. Added an "uninstall" target to the makefile.
1358     14. Replaced "extern" in the function prototypes in Makefile.in with
1359     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1360     world, but is set differently for Windows.
1362     15. Added a second compiling function called pcre_compile2(). The only
1363     difference is that it has an extra argument, which is a pointer to an
1364     integer error code. When there is a compile-time failure, this is set
1365     non-zero, in addition to the error test pointer being set to point to an
1366     error message. The new argument may be NULL if no error number is required
1367     (but then you may as well call pcre_compile(), which is now just a
1368     wrapper). This facility is provided because some applications need a
1369     numeric error indication, but it has also enabled me to tidy up the way
1370     compile-time errors are handled in the POSIX wrapper.
1372     16. Added VPATH=.libs to the makefile; this should help when building with one
1373     prefix path and installing with another. (Or so I'm told by someone who
1374     knows more about this stuff than I do.)
1376     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1377     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1378     match everything, including newlines. This is not POSIX-compatible, but
1379     somebody wanted the feature. From pcretest it can be activated by using
1380     both the P and the s flags.
1382     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1384     19. libpcre.pc was being incorrectly installed as executable.
1386     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1387     it now also looks for '\r' so that it will work unmodified on Windows.
1389     21. Added Google's contributed C++ wrapper to the distribution.
1391     22. Added some untidy missing memory free() calls in pcretest, to keep
1392     Electric Fence happy when testing.
1396 nigel 75 Version 5.0 13-Sep-04
1397     ---------------------
1399     1. Internal change: literal characters are no longer packed up into items
1400     containing multiple characters in a single byte-string. Each character
1401     is now matched using a separate opcode. However, there may be more than one
1402     byte in the character in UTF-8 mode.
1404     2. The pcre_callout_block structure has two new fields: pattern_position and
1405     next_item_length. These contain the offset in the pattern to the next match
1406     item, and its length, respectively.
1408     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1409     insertion of callouts before each pattern item. Added the /C option to
1410     pcretest to make use of this.
1412     4. On the advice of a Windows user, the lines
1414     #if defined(_WIN32) || defined(WIN32)
1415     _setmode( _fileno( stdout ), 0x8000 );
1416     #endif /* defined(_WIN32) || defined(WIN32) */
1418     have been added to the source of pcretest. This apparently does useful
1419     magic in relation to line terminators.
1421     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1422     for the benefit of those environments where the "b" makes a difference.
1424     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1425     to know about it. I have put a hack into configure.in that adds in code
1426     to set GCC=yes if CC=icc. This seems to end up at a point in the
1427     generated configure script that is early enough to affect the setting of
1428     compiler options, which is what is needed, but I have no means of testing
1429     whether it really works. (The user who reported this had patched the
1430     generated configure script, which of course I cannot do.)
1432     LATER: After change 22 below (new libtool files), the configure script
1433     seems to know about icc (and also ecc). Therefore, I have commented out
1434     this hack in configure.in.
1436     7. Added support for pkg-config (2 patches were sent in).
1438     8. Negated POSIX character classes that used a combination of internal tables
1439     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1440     [[:^ascii]]. Typically, they would match almost any characters. The other
1441     POSIX classes were not broken in this way.
1443     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1444     to find the match, as PCRE was deluded into thinking that the match had to
1445     start at the start point or following a newline. The same bug applied to
1446     patterns with negative forward assertions or any backward assertions
1447     preceding ".*" at the start, unless the pattern required a fixed first
1448     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1450     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1451     starting at the last subject character, bytes beyond the end of the subject
1452     string were read.
1454     11. Renamed the variable "class" as "classbits" to make life easier for C++
1455     users. (Previously there was a macro definition, but it apparently wasn't
1456     enough.)
1458     12. Added the new field "tables" to the extra data so that tables can be passed
1459     in at exec time, or the internal tables can be re-selected. This allows
1460     a compiled regex to be saved and re-used at a later time by a different
1461     program that might have everything at different addresses.
1463     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1464     -R library as well as a -L library.
1466     14. The debugging options of pcretest (-d on the command line or D on a
1467     pattern) showed incorrect output for anything following an extended class
1468     that contained multibyte characters and which was followed by a quantifier.
1470     15. Added optional support for general category Unicode character properties
1471     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1472     support. It adds about 90K to the size of the library. The meanings of the
1473     inbuilt class escapes such as \d and \s have NOT been changed.
1475     16. Updated pcredemo.c to include calls to free() to release the memory for the
1476     compiled pattern.
1478     17. The generated file chartables.c was being created in the source directory
1479     instead of in the building directory. This caused the build to fail if the
1480     source directory was different from the building directory, and was
1481     read-only.
1483     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1484     file. No doubt somebody will tell me if they don't make sense... Also added
1485     Dan Mooney's comments about building on OpenVMS.
1487     19. Added support for partial matching via the PCRE_PARTIAL option for
1488     pcre_exec() and the \P data escape in pcretest.
1490     20. Extended pcretest with 3 new pattern features:
1492     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1493     write the compiled pattern to the file whose name is "rest-of-line".
1494     This is a straight binary dump of the data, with the saved pointer to
1495     the character tables forced to be NULL. The study data, if any, is
1496     written too. After writing, pcretest reads a new pattern.
1498     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1499     compiled pattern from the given file. There must not be any
1500     occurrences of "<" in the file name (pretty unlikely); if there are,
1501     pcretest will instead treat the initial "<" as a pattern delimiter.
1502     After reading in the pattern, pcretest goes on to read data lines as
1503     usual.
1505     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1506     and 16-bit fields in a compiled pattern, to simulate a pattern that
1507     was compiled on a host of opposite endianness.
1509     21. The pcre-exec() function can now cope with patterns that were compiled on
1510     hosts of opposite endianness, with this restriction:
1512     As for any compiled expression that is saved and used later, the tables
1513     pointer field cannot be preserved; the extra_data field in the arguments
1514     to pcre_exec() should be used to pass in a tables address if a value
1515     other than the default internal tables were used at compile time.
1517     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1518     now diagnosed as an error. Previously, most of the time, a negative number
1519     would have been treated as zero, but if in addition "ovector" was passed as
1520     NULL, a crash could occur.
1522     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1523     new versions from the libtool 1.5 distribution (the last one is a copy of
1524     a file called libtool.m4). This seems to have fixed the need to patch
1525     "configure" to support Darwin 1.3 (which I used to do). However, I still
1526     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1527     workstation).
1529     24. Changed the PCRE licence to be the more standard "BSD" licence.
1532 nigel 73 Version 4.5 01-Dec-03
1533     ---------------------
1535     1. There has been some re-arrangement of the code for the match() function so
1536     that it can be compiled in a version that does not call itself recursively.
1537     Instead, it keeps those local variables that need separate instances for
1538     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1539     needs to "recurse". Keeping track of where control must go is done by means
1540     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1541     hide most of the details from the main code, and operates only if
1542     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1543     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1544     operating.
1546     To make it easier for callers to provide specially tailored get/free
1547     functions for this usage, two new functions, pcre_stack_malloc, and
1548     pcre_stack_free, are used. They are always called in strict stacking order,
1549     and the size of block requested is always the same.
1551     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1552     PCRE has been compiled to use the stack or the heap for recursion. The
1553     -C option of pcretest uses this to show which version is compiled.
1555     A new data escape \S, is added to pcretest; it causes the amounts of store
1556     obtained and freed by both kinds of malloc/free at match time to be added
1557     to the output.
1559     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1560     what's available on my current Linux desktop machine.
1562     3. When matching a UTF-8 string, the test for a valid string at the start has
1563     been extended. If start_offset is not zero, PCRE now checks that it points
1564     to a byte that is the start of a UTF-8 character. If not, it returns
1565     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1566     this is necessary because there may be backward assertions in the pattern.
1567     When matching the same subject several times, it may save resources to use
1568     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1570     4. The code for checking the validity of UTF-8 strings has been tightened so
1571     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1572     containing "overlong sequences".
1574     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1575     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1576     should have been "&", but it just so happened that all the cases this let
1577     through by mistake were picked up later in the function.
1579     6. I had used a variable called "isblank" - this is a C99 function, causing
1580     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1582     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1583     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1584     so the version is shown; (c) stop "make test" echoing "./RunTest".
1586     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1588     9. The prototype for memmove() for systems that don't have it was using
1589     size_t, but the inclusion of the header that defines size_t was later. I've
1590     moved the #includes for the C headers earlier to avoid this.
1592     10. Added some adjustments to the code to make it easier to compiler on certain
1593     special systems:
1595     (a) Some "const" qualifiers were missing.
1596     (b) Added the macro EXPORT before all exported functions; by default this
1597     is defined to be empty.
1598     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1599     that it reads its output file name as an argument instead of writing
1600     to the standard output and assuming this can be redirected.
1602     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1603     class containing characters with values greater than 255, PCRE compilation
1604     went into a loop.
1606     12. A recursive reference to a subpattern that was within another subpattern
1607     that had a minimum quantifier of zero caused PCRE to crash. For example,
1608     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1609     recursion. If the recursively-called subpattern itself had a zero repeat,
1610     that was OK.
1612     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1613     buffer into which it was copied (for escape processing) was still set at
1614     1024, so long lines caused crashes.
1616     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1617     "internal error: code overflow...". This applied to any character class
1618     that was followed by a possessive quantifier.
1620     15. Modified the Makefile to add libpcre.la as a prerequisite for
1621     libpcreposix.la because I was told this is needed for a parallel build to
1622     work.
1624     16. If a pattern that contained .* following optional items at the start was
1625     studied, the wrong optimizing data was generated, leading to matching
1626     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1627     matching string must start with a or b or c. The correct conclusion for
1628     this pattern is that a match can start with any character.
1631 nigel 71 Version 4.4 13-Aug-03
1632     ---------------------
1634     1. In UTF-8 mode, a character class containing characters with values between
1635     127 and 255 was not handled correctly if the compiled pattern was studied.
1636     In fixing this, I have also improved the studying algorithm for such
1637     classes (slightly).
1639     2. Three internal functions had redundant arguments passed to them. Removal
1640     might give a very teeny performance improvement.
1642     3. Documentation bug: the value of the capture_top field in a callout is *one
1643     more than* the number of the hightest numbered captured substring.
1645     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1646     in incorrectly linking with a previously installed version. They now link
1647     explicitly with libpcre.la.
1649     5. configure.in no longer needs to recognize Cygwin specially.
1651     6. A problem in pcre.in for Windows platforms is fixed.
1653     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1654     pcretest, it used to include the size of the study block as part of its
1655     output. Unfortunately, the structure contains a field that has a different
1656     size on different hardware architectures. This meant that the tests that
1657     showed this size failed. As the block is currently always of a fixed size,
1658     this information isn't actually particularly useful in pcretest output, so
1659     I have just removed it.
1661     8. Three pre-processor statements accidentally did not start in column 1.
1662     Sadly, there are *still* compilers around that complain, even though
1663     standard C has not required this for well over a decade. Sigh.
1665     9. In pcretest, the code for checking callouts passed small integers in the
1666     callout_data field, which is a void * field. However, some picky compilers
1667     complained about the casts involved for this on 64-bit systems. Now
1668     pcretest passes the address of the small integer instead, which should get
1669     rid of the warnings.
1671     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1672     both compile and run time, and gives an error if an invalid UTF-8 sequence
1673     is found. There is a option for disabling this check in cases where the
1674     string is known to be correct and/or the maximum performance is wanted.
1676     11. In response to a bug report, I changed one line in Makefile.in from
1678     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1679     to
1680     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1682     to look similar to other lines, but I have no way of telling whether this
1683     is the right thing to do, as I do not use Windows. No doubt I'll get told
1684     if it's wrong...
1687 nigel 69 Version 4.3 21-May-03
1688     ---------------------
1690     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1691     Makefile.
1693     2. Some refactoring to improve the quality of the code:
1695     (i) The utf8_table... variables are now declared "const".
1697     (ii) The code for \cx, which used the "case flipping" table to upper case
1698     lower case letters, now just substracts 32. This is ASCII-specific,
1699     but the whole concept of \cx is ASCII-specific, so it seems
1700     reasonable.
1702     (iii) PCRE was using its character types table to recognize decimal and
1703     hexadecimal digits in the pattern. This is silly, because it handles
1704     only 0-9, a-f, and A-F, but the character types table is locale-
1705     specific, which means strange things might happen. A private
1706     table is now used for this - though it costs 256 bytes, a table is
1707     much faster than multiple explicit tests. Of course, the standard
1708     character types table is still used for matching digits in subject
1709     strings against \d.
1711     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1712     ending in _t are). So I've renamed it as ESC_tee.
1714     3. The first argument for regexec() in the POSIX wrapper should have been
1715     defined as "const".
1717     4. Changed pcretest to use malloc() for its buffers so that they can be
1718     Electric Fenced for debugging.
1720     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1721     to read one or more bytes before the start of the subject string. Often this
1722     had no effect on PCRE's behaviour, but in some circumstances it could
1723     provoke a segmentation fault.
1725     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1726     to try to read one or more bytes before the start of the subject string.
1728     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1729     UTF-8 support could misbehave in various ways if the subject string
1730     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1731     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1732     back over UTF-8 characters.)
1735 nigel 67 Version 4.2 14-Apr-03
1736     ---------------------
1738     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1740     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1741     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1742     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1743     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1744     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1745     and BUILD_EXEEXT
1746     Note: automatic setting of the BUILD variables is not yet working
1747     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1748     compile-time but not at link-time
1749     [LINK]: use for linking executables only
1750     make different versions for Windows and non-Windows
1751     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1752     libraries
1753     [LINK_FOR_BUILD]: new variable
1754     [OBJEXT]: use throughout
1755     [EXEEXT]: use throughout
1756     <winshared>: new target
1757     <wininstall>: new target
1758     <dftables.o>: use native compiler
1759     <dftables>: use native linker
1760     <install>: handle Windows platform correctly
1761     <clean>: ditto
1762     <check>: ditto
1763     copy DLL to top builddir before testing
1765     As part of these changes, -no-undefined was removed again. This was reported
1766     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1767     in any case.
1769     3. Some tidies to get rid of compiler warnings:
1771     . In the match_data structure, match_limit was an unsigned long int, whereas
1772     match_call_count was an int. I've made them both unsigned long ints.
1774     . In pcretest the fact that a const uschar * doesn't automatically cast to
1775     a void * provoked a warning.
1777     . Turning on some more compiler warnings threw up some "shadow" variables
1778     and a few more missing casts.
1780     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1781     option, a class that contained a single character with a value between 128
1782     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1784     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1785     option, a class that contained several characters, but with at least one
1786     whose value was between 128 and 255 caused PCRE to crash.
1789 nigel 65 Version 4.1 12-Mar-03
1790     ---------------------
1791 nigel 63
1792 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1793     needed, and a string in dftables.c that was longer than standard compilers are
1794     required to support.
1796     2. Compiling with Sun's compiler found a few more places where the code could
1797     be tidied up in order to avoid warnings.
1799     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1800     first of these names is deprecated in the latest Autoconf in favour of the name
1801     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1802     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1803     analogy I have changed it to CFLAGS_FOR_BUILD.
1805     4. Added -no-undefined to the linking command in the Makefile, because this is
1806     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1807     linking step for the pcreposix library.
1809     5. PCRE was failing to diagnose the case of two named groups with the same
1810     name.
1812     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1813     literal character that is needed in the subject for a match, and scans along to
1814     ensure that it is present before embarking on the full matching process. This
1815     saves time in cases of nested unlimited repeats that are never going to match.
1816     Problem: the scan can take a lot of time if the subject is very long (e.g.
1817     megabytes), thus penalizing straightforward matches. It is now done only if the
1818     amount of subject to be scanned is less than 1000 bytes.
1820     7. A lesser problem with the same optimization is that it was recording the
1821     first character of an anchored pattern as "needed", thus provoking a search
1822     right along the subject, even when the first match of the pattern was going to
1823     fail. The "needed" character is now not set for anchored patterns, unless it
1824     follows something in the pattern that is of non-fixed length. Thus, it still
1825     fulfils its original purpose of finding quick non-matches in cases of nested
1826     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1829     Version 4.0 17-Feb-03
1830     ---------------------
1832 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1833     extended to the end of string, PCRE compiled incorrect data. This could lead to
1834     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1836     2. Moved to autoconf 2.53 and libtool 1.4.2.
1838     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1839     the special perltest8 script is no longer needed - all the tests can be run
1840     from a single perltest script.
1842     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1843     by \s. It has now been removed in PCRE. This means it isn't recognized as
1844     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1845     class [:space:] *does* include VT, thereby creating a mess.
1847     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1848     space and tab.
1850     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1851     its new features into the main test script, reducing the number of scripts.
1853     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1854     were backward compatible, and made the (?i) apply to the whole pattern, as if
1855     /i were given. Now it behaves more logically, and applies the option setting
1856     only to what follows. PCRE has been changed to follow suit. However, if it
1857     finds options settings right at the start of the pattern, it extracts them into
1858     the global options, as before. Thus, they show up in the info data.
1860     8. Added support for the \Q...\E escape sequence. Characters in between are
1861     treated as literals. This is slightly different from Perl in that $ and @ are
1862     also handled as literals inside the quotes. In Perl, they will cause variable
1863     interpolation. Note the following examples:
1865     Pattern PCRE matches Perl matches
1867     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1868     \Qabc\$xyz\E abc\$xyz abc\$xyz
1869     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1871     For compatibility with Perl, \Q...\E sequences are recognized inside character
1872     classes as well as outside them.
1874     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1875     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1876     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1877     signed/unsigned warnings.
1879     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1880     option for pcretest, so I've replaced it by a simple function that does just
1881     that job.
1883     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1884     "pcregrep -".
1886     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1887     Java package. This provides some syntactic sugar for simple cases of what my
1888     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1889     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1890     item, you can use this simplified notation. Note that only makes sense with
1891     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1892     greediness, whatever the setting of the PCRE_UNGREEDY option.
1894     13. A change of greediness default within a pattern was not taking effect at
1895     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1896     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1897     was abstracted outside.
1899     14. PCRE now supports the \G assertion. It is true when the current matching
1900     position is at the start point of the match. This differs from \A when the
1901     starting offset is non-zero. Used with the /g option of pcretest (or similar
1902     code), it works in the same way as it does for Perl's /g option. If all
1903     alternatives of a regex begin with \G, the expression is anchored to the start
1904     match position, and the "anchored" flag is set in the compiled expression.
1906     15. Some bugs concerning the handling of certain option changes within patterns
1907     have been fixed. These applied to options other than (?ims). For example,
1908     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1909     been the other way round. Some of this was related to change 7 above.
1911     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1912     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1913     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1914     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1916     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1917     mode. Unlike ".", it always matches newline, whatever the setting of
1918     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1919     assertions. Perl allows it, but it doesn't (in general) work because it can't
1920     calculate the length of the lookbehind. At least, that's the case for Perl
1921     5.8.0 - I've been told they are going to document that it doesn't work in
1922     future.
1924     18. Added an error diagnosis for escapes that PCRE does not support: these are
1925     \L, \l, \N, \P, \p, \U, \u, and \X.
1927     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1928     reading past the end of the pattern in cases such as /[abcd/.
1930     20. PCRE was getting more memory than necessary for patterns with classes that
1931     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1933     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1934     compiling PCRE for use with Virtual Pascal.
1936     22. Small fix to the Makefile to make it work properly if the build is done
1937     outside the source tree.
1939     23. Added a new extension: a condition to go with recursion. If a conditional
1940     subpattern starts with (?(R) the "true" branch is used if recursion has
1941     happened, whereas the "false" branch is used only at the top level.
1943     24. When there was a very long string of literal characters (over 255 bytes
1944     without UTF support, over 250 bytes with UTF support), the computation of how
1945     much memory was required could be incorrect, leading to segfaults or other
1946     strange effects.
1948     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1949     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1950     there was a subsequent back reference to those brackets. This meant that, for
1951     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1952     possible to check for precisely this case. All we can do is abandon the
1953     optimization if .* occurs inside capturing brackets when there are any back
1954     references whatsoever. (See below for a better fix that came later.)
1956     26. The handling of the optimization for finding the first character of a
1957     non-anchored pattern, and for finding a character that is required later in the
1958     match were failing in some cases. This didn't break the matching; it just
1959     failed to optimize when it could. The way this is done has been re-implemented.
1961     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1963     28. Added a new feature that provides some of the functionality that Perl
1964     provides with (?{...}). The facility is termed a "callout". The way it is done
1965     in PCRE is for the caller to provide an optional function, by setting
1966     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1967     global variable. By default it is unset, which disables all calling out. To get
1968     the function called, the regex must include (?C) at appropriate points. This
1969     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1970     This provides a means of identifying different callout points. When PCRE
1971     reaches such a point in the regex, if pcre_callout has been set, the external
1972     function is called. It is provided with data in a structure called
1973     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1974     matching continues; if it returns a non-zero value, the match at the current
1975     point fails. However, backtracking will occur if possible. [This was changed
1976     later and other features added - see item 49 below.]
1978     29. pcretest is upgraded to test the callout functionality. It provides a
1979     callout function that displays information. By default, it shows the start of
1980     the match and the current position in the text. There are some new data escapes
1981     to vary what happens:
1983     \C+ in addition, show current contents of captured substrings
1984     \C- do not supply a callout function
1985     \C!n return 1 when callout number n is reached
1986     \C!n!m return 1 when callout number n is reached for the mth time
1988     30. If pcregrep was called with the -l option and just a single file name, it
1989     output "<stdin>" if a match was found, instead of the file name.
1991     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1992     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1993     pcre_exec(). This saves a malloc/free per call. The default value of
1994     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1995     when configuring.
1997     32. The default maximum size of a compiled pattern is 64K. There have been a
1998     few cases of people hitting this limit. The code now uses macros to handle the
1999     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2000     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2001     configuring. Tests 2 and 5 work only with 2-byte links because they output
2002     debugging information about compiled patterns.
2004     33. Internal code re-arrangements:
2006     (a) Moved the debugging function for printing out a compiled regex into
2007     its own source file (printint.c) and used #include to pull it into
2008     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2009     separate copies.
2011     (b) Defined the list of op-code names for debugging as a macro in
2012     internal.h so that it is next to the definition of the opcodes.
2014     (c) Defined a table of op-code lengths for simpler skipping along compiled
2015     code. This is again a macro in internal.h so that it is next to the
2016     definition of the opcodes.
2018     34. Added support for recursive calls to individual subpatterns, along the
2019     lines of Robin Houston's patch (but implemented somewhat differently).
2021     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2022     allow it to read and process whole directories in Win32. This code was
2023     contributed by Lionel Fourquaux; it has not been tested by me.
2025     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2026     used to name a group. Names consist of alphanumerics and underscores, and must
2027     be unique. Back references use the syntax (?P=name) and recursive calls use
2028     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2029     numbers. The function pcre_fullinfo() can be used after compilation to extract
2030     a name/number map. There are three relevant calls:
2032     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2033     PCRE_INFO_NAMECOUNT yields the number of entries
2034     PCRE_INFO_NAMETABLE yields a pointer to the map.
2036     The map is a vector of fixed-size entries. The size of each entry depends on
2037     the length of the longest name used. The first two bytes of each entry are the
2038     group number, most significant byte first. There follows the corresponding
2039     name, zero terminated. The names are in alphabetical order.
2041     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2042     case instead of 255. Making it the same both with and without UTF-8 support
2043     means that the same test output works with both.
2045     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2046     calling malloc() with a zero argument.
2048     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2049     optimization. I've improved things by keeping a bitmap of backreferences with
2050     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2051     fact referenced, the optimization can be applied. It is unlikely that a
2052     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2053     the match to follow \n) will appear inside brackets with a number greater than
2054     31, but if it does, any back reference > 31 suppresses the optimization.
2056     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2057     of disabling numbered capturing parentheses. Any opening parenthesis that is
2058     not followed by ? behaves as if it were followed by ?: but named parentheses
2059     can still be used for capturing (and they will acquire numbers in the usual
2060     way).
2062     41. Redesigned the return codes from the match() function into yes/no/error so
2063     that errors can be passed back from deep inside the nested calls. A malloc
2064     failure while inside a recursive subpattern call now causes the
2065     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2067     42. It is now possible to set a limit on the number of times the match()
2068     function is called in a call to pcre_exec(). This facility makes it possible to
2069     limit the amount of recursion and backtracking, though not in a directly
2070     obvious way, because the match() function is used in a number of different
2071     circumstances. The count starts from zero for each position in the subject
2072     string (for non-anchored patterns). The default limit is, for compatibility, a
2073     large number, namely 10 000 000. You can change this in two ways:
2075     (a) When configuring PCRE before making, you can use --with-match-limit=n
2076     to set a default value for the compiled library.
2078     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2079     a different value is set. See 45 below.
2081     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2083     43. Added a new function pcre_config(int, void *) to enable run-time extraction
2084     of things that can be changed at compile time. The first argument specifies
2085     what is wanted and the second points to where the information is to be placed.
2086     The current list of available information is:
2090     The output is an integer that is set to one if UTF-8 support is available;
2091     otherwise it is set to zero.
2095     The output is an integer that it set to the value of the code that is used for
2096     newline. It is either LF (10) or CR (13).
2100     The output is an integer that contains the number of bytes used for internal
2101     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2105     The output is an integer that contains the threshold above which the POSIX
2106     interface uses malloc() for output vectors. See item 31 above.
2110     The output is an unsigned integer that contains the default limit of the number
2111     of match() calls in a pcre_exec() execution. See 42 above.
2113     44. pcretest has been upgraded by the addition of the -C option. This causes it
2114     to extract all the available output from the new pcre_config() function, and to
2115     output it. The program then exits immediately.
2117     45. A need has arisen to pass over additional data with calls to pcre_exec() in
2118     order to support additional features. One way would have been to define
2119     pcre_exec2() (for example) with extra arguments, but this would not have been
2120     extensible, and would also have required all calls to the original function to
2121     be mapped to the new one. Instead, I have chosen to extend the mechanism that
2122     is used for passing in "extra" data from pcre_study().
2124     The pcre_extra structure is now exposed and defined in pcre.h. It currently
2125     contains the following fields:
2127     flags a bitmap indicating which of the following fields are set
2128     study_data opaque data from pcre_study()
2129     match_limit a way of specifying a limit on match() calls for a specific
2130     call to pcre_exec()
2131     callout_data data for callouts (see 49 below)
2133     The flag bits are also defined in pcre.h, and are
2139     The pcre_study() function now returns one of these new pcre_extra blocks, with
2140     the actual study data pointed to by the study_data field, and the
2141     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2142     before. That is, this change is entirely upwards-compatible and requires no
2143     change to existing code.
2145     If you want to pass in additional data to pcre_exec(), you can either place it
2146     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2147     block.
2149     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2150     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2151     times with different match limits, until it finds the minimum value needed for
2152     pcre_exec() to complete. The value is then output. This can be instructive; for
2153     most simple matches the number is quite small, but for pathological cases it
2154     gets very large very quickly.
2156     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2157     returns the size of the data block pointed to by the study_data field in a
2158     pcre_extra block, that is, the value that was passed as the argument to
2159     pcre_malloc() when PCRE was getting memory in which to place the information
2160     created by pcre_study(). The fourth argument should point to a size_t variable.
2161     pcretest has been extended so that this information is shown after a successful
2162     pcre_study() call when information about the compiled regex is being displayed.
2164     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2165     because what follows is always an absolute path. (Later: it turns out that this
2166     is more than cosmetic for MinGW, because it doesn't like empty path
2167     components.)
2169     49. Some changes have been made to the callout feature (see 28 above):
2171     (i) A callout function now has three choices for what it returns:
2173     0 => success, carry on matching
2174     > 0 => failure at this point, but backtrack if possible
2175     < 0 => serious error, return this value from pcre_exec()
2177     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2178     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2179     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2180     use by callout functions. It will never be used by PCRE itself.
2182     (ii) The pcre_extra structure (see 45 above) has a void * field called
2183     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2184     pcre_callout_block structure has a field of the same name. The contents of
2185     the field passed in the pcre_extra structure are passed to the callout
2186     function in the corresponding field in the callout block. This makes it
2187     easier to use the same callout-containing regex from multiple threads. For
2188     testing, the pcretest program has a new data escape
2190     \C*n pass the number n (may be negative) as callout_data
2192     If the callout function in pcretest receives a non-zero value as
2193     callout_data, it returns that value.
2195     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2196     there were some redundant $(CFLAGS) in commands that are now specified as
2197     $(LINK), which already includes $(CFLAGS).
2199     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2200     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2201     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2202     one-byte characters throughout. Note that case-insensitive matching applies
2203     only to characters whose values are less than 256. PCRE doesn't support the
2204     notion of cases for higher-valued characters.
2206     (i) A character class whose characters are all within 0-255 is handled as
2207     a bit map, and the map is inverted for negative classes. Previously, a
2208     character > 255 always failed to match such a class; however it should
2209     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2211     (ii) A negated character class with a single character < 255 is coded as
2212     "not this character" (OP_NOT). This wasn't working properly when the test
2213     character was multibyte, either singly or repeated.
2215     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2216     mode, for example: \x{100}{2,3}.
2218     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2219     singly or repeated) now correctly test multibyte characters. However,
2220     PCRE doesn't recognize any characters with values greater than 255 as
2221     digits, spaces, or word characters. Such characters always match \D, \S,
2222     and \W, and never match \d, \s, or \w.
2224     (v) Classes may now contain characters and character ranges with values
2225     greater than 255. For example: [ab\x{100}-\x{400}].
2227     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2228     PCRE in UTF-8 mode.
2230     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2231     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2232     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2233     value.)
2235     53. The single man page has become too large. I have therefore split it up into
2236     a number of separate man pages. These also give rise to individual HTML pages;
2237     these are now put in a separate directory, and there is an index.html page that
2238     lists them all. Some hyperlinking between the pages has been installed.
2240     54. Added convenience functions for handling named capturing parentheses.
2242     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2243     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2244     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2245     are faulted.
2247     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2248     calling configure. These values are used when compiling the dftables.c program
2249     which is run to generate the source of the default character tables. They
2250     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2251     you will need to set these values.
2253     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2256     Version 3.9 02-Jan-02
2257 nigel 61 ---------------------
2259     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2261     2. If --disable-static was given, the building process failed when trying to
2262     build pcretest and pcregrep. (For some reason it was using libtool to compile
2263     them, which is not right, as they aren't part of the library.)
2266 nigel 59 Version 3.8 18-Dec-01
2267     ---------------------
2269     1. The experimental UTF-8 code was completely screwed up. It was packing the
2270     bytes in the wrong order. How dumb can you get?
2273 nigel 57 Version 3.7 29-Oct-01
2274     ---------------------
2276     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2277     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2278     this didn't happen under Solaris 8, where I normally test things.
2280 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2281     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2282     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2283     this fix an hour or so after the initial 3.7 release.)
2284 nigel 57
2285 nigel 59
2286 nigel 55 Version 3.6 23-Oct-01
2287     ---------------------
2289     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2290     offsets passed as NULL with zero offset count.
2292     2. The config.guess and config.sub files had not been updated when I moved to
2293     the latest autoconf.
2296 nigel 53 Version 3.5 15-Aug-01
2297     ---------------------
2298 nigel 5
2299 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2300     had been forgotten.
2302     2. By using declared but undefined structures, we can avoid using "void"
2303     definitions in pcre.h while keeping the internal definitions of the structures
2304     private.
2306     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2307     user point of view, this means that both static and shared libraries are built
2308     by default, but this can be individually controlled. More of the work of
2309     handling this static/shared cases is now inside libtool instead of PCRE's make
2310     file.
2312     4. The pcretest utility is now installed along with pcregrep because it is
2313     useful for users (to test regexs) and by doing this, it automatically gets
2314     relinked by libtool. The documentation has been turned into a man page, so
2315     there are now .1, .txt, and .html versions in /doc.
2317     5. Upgrades to pcregrep:
2318     (i) Added long-form option names like gnu grep.
2319     (ii) Added --help to list all options with an explanatory phrase.
2320     (iii) Added -r, --recursive to recurse into sub-directories.
2321     (iv) Added -f, --file to read patterns from a file.
2323     6. pcre_exec() was referring to its "code" argument before testing that
2324     argument for NULL (and giving an error if it was NULL).
2326     7. Upgraded Makefile.in to allow for compiling in a different directory from
2327     the source directory.
2329     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2330     options bits, the pointer it was passed was to an int instead of to an unsigned
2331     long int. This mattered only on 64-bit systems.
2333     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2334     generated) instead of pcre.in, which it its source. Also made the same change
2335     in several of the .c files.
2337     10. A new release of gcc defines printf() as a macro, which broke pcretest
2338     because it had an ifdef in the middle of a string argument for printf(). Fixed
2339     by using separate calls to printf().
2341     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2342     script, to force use of CR or LF instead of \n in the source. On non-Unix
2343     systems, the value can be set in config.h.
2345     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2346     absolute limit. Changed the text of the error message to make this clear, and
2347     likewise updated the man page.
2349     13. The limit of 99 on the number of capturing subpatterns has been removed.
2350     The new limit is 65535, which I hope will not be a "real" limit.
2353 nigel 51 Version 3.4 22-Aug-00
2354     ---------------------
2356     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2358     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2361 nigel 49 Version 3.3 01-Aug-00
2362     ---------------------
2364     1. If an octal character was given, but the value was greater than \377, it
2365     was not getting masked to the least significant bits, as documented. This could
2366     lead to crashes in some systems.
2368     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2369     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2371     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2372     These just pass their arguments on to (pcre_free)(), but they are provided
2373     because some uses of PCRE bind it to non-C systems that can call its functions,
2374     but cannot call free() or pcre_free() directly.
2376     4. Add "make test" as a synonym for "make check". Corrected some comments in
2377     the Makefile.
2379     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2380     Makefile.
2382     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2383     command called pgrep for grepping around the active processes.
2385     7. Added the beginnings of support for UTF-8 character strings.
2387     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2388     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2389     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2390     out for the ar command.)
2393 nigel 47 Version 3.2 12-May-00
2394     ---------------------
2396     This is purely a bug fixing release.
2398     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2399     of ZA. This was just one example of several cases that could provoke this bug,
2400     which was introduced by change 9 of version 2.00. The code for breaking
2401     infinite loops after an iteration that matches an empty string was't working
2402     correctly.
2404     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2405     when matched against abbab (for example). After matching an empty string, it
2406     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2407     caused it to match further down the string than it should.
2409     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2410     was there because it doesn't seem to be needed, and it causes trouble on some
2411     systems, as it is not a Standard C header. It has been removed.
2413     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2414     were reported on the Macintosh. The changes were from
2416     while ((c = *(++ptr)) != 0 && c != '\n');
2417     to
2418     while ((c = *(++ptr)) != 0 && c != '\n') ;
2420     Totally extraordinary, but if that's what it takes...
2422     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2423     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2424     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2425     assumes the way PCRE uses memmove() (always moving upwards).
2427     6. PCRE is being used in one environment where strchr() is not available. There
2428     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2429     faster code anyway.
2432 nigel 45 Version 3.1 09-Feb-00
2433     ---------------------
2435     The only change in this release is the fixing of some bugs in Makefile.in for
2436     the "install" target:
2438     (1) It was failing to install pcreposix.h.
2440     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2443 nigel 43 Version 3.0 01-Feb-00
2444     ---------------------
2445 nigel 41
2446     1. Add support for the /+ modifier to perltest (to output $` like it does in
2447     pcretest).
2449     2. Add support for the /g modifier to perltest.
2451     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2452     matches null strings.
2454     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2455     pattern. Perl treats empty patterns specially - it reuses the most recent
2456     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2457     effect.
2459     5. The POSIX interface was broken in that it was just handing over the POSIX
2460     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2461     required a bigger vector, with some working space on the end. This means that
2462     the POSIX wrapper now has to get and free some memory, and copy the results.
2464 nigel 43 6. Added some simple autoconf support, placing the test data and the
2465     documentation in separate directories, re-organizing some of the
2466     information files, and making it build pcre-config (a GNU standard). Also added
2467     libtool support for building PCRE as a shared library, which is now the
2468     default.
2469 nigel 41
2470 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2471     09 are not valid octal constants. Single digits will be used for minor values
2472     less than 10.
2474     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2475     existing programs that set these in the POSIX interface can use PCRE without
2476     modification.
2478     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2479     return all that pcre_info() returns, plus additional data. The pcre_info()
2480     function is retained for compatibility, but is considered to be obsolete.
2482     10. Added experimental recursion feature (?R) to handle one common case that
2483     Perl 5.6 will be able to do with (?p{...}).
2485     11. Added support for POSIX character classes like [:alpha:], which Perl is
2486     adopting.
2489 nigel 39 Version 2.08 31-Aug-99
2490     ----------------------
2492     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2493     trying to match at the startoffset position, but instead was moving forward to
2494     the next newline as if a previous match had failed.
2496     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2497     and could get into a loop if a null string was matched other than at the start
2498     of the subject.
2500     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2501     be distinguished at compile time, and for completeness also added PCRE_DATE.
2503     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2504     in GnuWin32 environments.
2507 nigel 37 Version 2.07 29-Jul-99
2508     ----------------------
2510     1. The documentation is now supplied in plain text form and HTML as well as in
2511     the form of man page sources.
2513     2. C++ compilers don't like assigning (void *) values to other pointer types.
2514     In particular this affects malloc(). Although there is no problem in Standard
2515     C, I've put in casts to keep C++ compilers happy.
2517     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2518     should be (const char *).
2520     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2521     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2522     However, I haven't made this a standard facility. The documentation doesn't
2523     mention it, and the Makefile doesn't support it.
2525     5. The Makefile now contains an "install" target, with editable destinations at
2526     the top of the file. The pcretest program is not installed.
2528     6. pgrep -V now gives the PCRE version number and date.
2530     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2531     causing the entire string to be ignored, instead of just the last character.
2533     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2534     non-matching string, it can take a very, very long time, even for strings of
2535     quite modest length, because of the nested recursion. PCRE now does better in
2536     some of these cases. It does this by remembering the last required literal
2537     character in the pattern, and pre-searching the subject to ensure it is present
2538     before running the real match. In other words, it applies a heuristic to detect
2539     some types of certain failure quickly, and in the above example, if presented
2540     with a string that has no trailing " it gives "no match" very quickly.
2542     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2543     other alternatives are tried instead.
2546 nigel 35 Version 2.06 09-Jun-99
2547     ----------------------
2549     1. Change pcretest's output for amount of store used to show just the code
2550     space, because the remainder (the data block) varies in size between 32-bit and
2551     64-bit systems.
2553     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2554     start matching at. This allows lookbehinds to work when searching for multiple
2555     occurrences in a string.
2557     3. Added additional options to pcretest for testing multiple occurrences:
2559     /+ outputs the rest of the string that follows a match
2560     /g loops for multiple occurrences, using the new startoffset argument
2561     /G loops for multiple occurrences by passing an incremented pointer
2563     4. PCRE wasn't doing the "first character" optimization for patterns starting
2564     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2565     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2566     the letter 'x'. On long subject strings, this gives a significant speed-up.
2569 nigel 33 Version 2.05 21-Apr-99
2570     ----------------------
2572     1. Changed the type of magic_number from int to long int so that it works
2573     properly on 16-bit systems.
2575     2. Fixed a bug which caused patterns starting with .* not to work correctly
2576     when the subject string contained newline characters. PCRE was assuming
2577     anchoring for such patterns in all cases, which is not correct because .* will
2578     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2579     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2580     must be retried after every newline in the subject.
2583 nigel 31 Version 2.04 18-Feb-99
2584     ----------------------
2586     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2587     computation of the store needed to hold the pattern was incorrect (too large).
2588     If such patterns were nested a few deep, this could multiply and become a real
2589     problem.
2591     2. Added /M option to pcretest to show the memory requirement of a specific
2592     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2594     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2595     compiled in such a way that the backtracking after subsequent failure was
2596     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2597     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2600 nigel 29 Version 2.03 02-Feb-99
2601     ----------------------
2603     1. Fixed typo and small mistake in man page.
2605 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2606     LICENCE file containing the conditions.
2607 nigel 29
2608     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2609     Perl, that is the internal \ allows the delimiter to be included in the
2610     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2611     the final delimiter, add \ to the end of the pattern (to test the error).
2613     4. Added the convenience functions for extracting substrings after a successful
2614     match. Updated pcretest to make it able to test these functions.
2617 nigel 27 Version 2.02 14-Jan-99
2618     ----------------------
2620     1. Initialized the working variables associated with each extraction so that
2621     their saving and restoring doesn't refer to uninitialized store.
2623     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2624     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2625     fix the problem.
2627     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2628     calls, and wasn't printing the correct value for compiling calls. Increased the
2629     default value of LOOPREPEAT, and the number of significant figures in the
2630     times.
2632     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2634     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2635     a building problem on Windows NT with a FAT file system.
2638 nigel 25 Version 2.01 21-Oct-98
2639     ----------------------
2641     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2642     to character tables built by pcre_maketables() in the current locale. If NULL
2643     is passed, the default tables are used.
2646 nigel 23 Version 2.00 24-Sep-98
2647 nigel 21 ----------------------
2649 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2650     it any more.
2652     2. Allow quantification of (?>) groups, and make it work correctly.
2654     3. The first character computation wasn't working for (?>) groups.
2656     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2657     end of the subject) and add 5.005's \z, which really does match only at the
2658     very end of the subject.
2660     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2662     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2663     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2664     localized options. All options to pcre_study() were also removed.
2666     7. Add other new features from 5.005:
2668     $(?<= positive lookbehind
2669     $(?<! negative lookbehind
2670     (?imsx-imsx) added the unsetting capability
2671     such a setting is global if at outer level; local otherwise
2672     (?imsx-imsx:) non-capturing groups with option setting
2673     (?(cond)re|re) conditional pattern matching
2675     A backreference to itself in a repeated group matches the previous
2676     captured string.
2678     8. General tidying up of studying (both automatic and via "study")
2679     consequential on the addition of new assertions.
2681     9. As in 5.005, unlimited repeated groups that could match an empty substring
2682     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2683     runtime if any iteration does actually match an empty substring.
2685     10. Include the RunTest script in the distribution.
2687     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2688     discrepancies, some of which were old and were also with respect to 5.004. They
2689     have now been fixed.
2692     Version 1.09 28-Apr-98
2693     ----------------------
2695 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2696     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2697     program crashes, or just wrong answers. This did not apply to negated classes
2698     containing more than one character, or to minima other than one.
2701 nigel 19 Version 1.08 27-Mar-98
2702     ----------------------
2704     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2706     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2707     latter must appear before anything that relies on it in the pattern.
2710 nigel 17 Version 1.07 16-Feb-98
2711     ----------------------
2713     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2714     repeat of a potentially empty string).
2717 nigel 15 Version 1.06 23-Jan-98
2718     ----------------------
2720     1. Added Markus Oberhumer's little patches for C++.
2722     2. Literal strings longer than 255 characters were broken.
2725 nigel 13 Version 1.05 23-Dec-97
2726     ----------------------
2728     1. Negated character classes containing more than one character were failing if
2729     PCRE_CASELESS was set at run time.
2732 nigel 11 Version 1.04 19-Dec-97
2733     ----------------------
2735     1. Corrected the man page, where some "const" qualifiers had been omitted.
2737     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2738     input syntax.
2740     3. Fixed memory leak which occurred when a regex with back references was
2741     matched with an offsets vector that wasn't big enough. The temporary memory
2742     that is used in this case wasn't being freed if the match failed.
2744     4. Tidied pcretest to ensure it frees memory that it gets.
2746     5. Temporary memory was being obtained in the case where the passed offsets
2747     vector was exactly big enough.
2749     6. Corrected definition of offsetof() from change 5 below.
2751     7. I had screwed up change 6 below and broken the rules for the use of
2752     setjmp(). Now fixed.
2755 nigel 9 Version 1.03 18-Dec-97
2756     ----------------------
2758     1. A erroneous regex with a missing opening parenthesis was correctly
2759     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2760     on some systems.
2762     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2763     it was reported that one broken compiler failed on the former because "code" is
2764     also an independent variable.
2766     3. The erroneous regex a[]b caused an array overrun reference.
2768     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2769     fail on data ending with that character. (It was going on too far, and checking
2770     the next character, typically a binary zero.) This was specific to the
2771     optimized code for single-character negative classes.
2773     5. Added a contributed patch from the TIN world which does the following:
2775     + Add an undef for memmove, in case the the system defines a macro for it.
2777     + Add a definition of offsetof(), in case there isn't one. (I don't know
2778     the reason behind this - offsetof() is part of the ANSI standard - but
2779     it does no harm).
2781     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2782     most of the places where whitespace preceded '#'. I have given up and
2783     allowed the remaining 2 cases to be at the margin.
2785     + Rename some variables in pcre to eliminate shadowing. This seems very
2786     pedantic, but does no harm, of course.
2788     6. Moved the call to setjmp() into its own function, to get rid of warnings
2789     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2791     7. Constructs such as \d{8,} were compiling into the equivalent of
2792 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2793 nigel 9 outcome, but in this particular case used more store than had been allocated,
2794     which caused the bug to be discovered because it threw up an internal error.
2796     8. The debugging code in both pcre and pcretest for outputting the compiled
2797     form of a regex was going wrong in the case of back references followed by
2798     curly-bracketed repeats.
2801 nigel 7 Version 1.02 12-Dec-97
2802     ----------------------
2804     1. Typos in pcre.3 and comments in the source fixed.
2806     2. Applied a contributed patch to get rid of places where it used to remove
2807     'const' from variables, and fixed some signed/unsigned and uninitialized
2808     variable warnings.
2810     3. Added the "runtest" target to Makefile.
2812     4. Set default compiler flag to -O2 rather than just -O.
2815 nigel 5 Version 1.01 19-Nov-97
2816     ----------------------
2818     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2819     like /([ab]*)*/, that is, for classes with more than one character in them.
2821     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2822     as /((?>a*))*/ (a PCRE_EXTRA facility).
2825     Version 1.00 18-Nov-97
2826     ----------------------
2828     1. Added compile-time macros to support systems such as SunOS4 which don't have
2829     memmove() or strerror() but have other things that can be used instead.
2831     2. Arranged that "make clean" removes the executables.
2834 nigel 3 Version 0.99 27-Oct-97
2835     ----------------------
2837     1. Fixed bug in code for optimizing classes with only one character. It was
2838     initializing a 32-byte map regardless, which could cause it to run off the end
2839     of the memory it had got.
2841     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2844     Version 0.98 22-Oct-97
2845     ----------------------
2847     1. Fixed bug in code for handling temporary memory usage when there are more
2848     back references than supplied space in the ovector. This could cause segfaults.
2851     Version 0.97 21-Oct-97
2852     ----------------------
2854     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2856     2. Optimized negated single characters not to use a bit map.
2858     3. Brought error texts together as macro definitions; clarified some of them;
2859     fixed one that was wrong - it said "range out of order" when it meant "invalid
2860     escape sequence".
2862     4. Changed some char * arguments to const char *.
2864     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2866     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2867     pcretest.
2870     Version 0.96 16-Oct-97
2871     ----------------------
2873     1. Added a simple "pgrep" utility to the distribution.
2875     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2876     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2877     where "ddd" means "one or more decimal digits".
2879     3. Fixed serious bug. If a pattern had a back reference, but the call to
2880     pcre_exec() didn't supply a large enough ovector to record the related
2881     identifying subpattern, the match always failed. PCRE now remembers the number
2882     of the largest back reference, and gets some temporary memory in which to save
2883     the offsets during matching if necessary, in order to ensure that
2884     backreferences always work.
2886     4. Increased the compatibility with Perl in a number of ways:
2888     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2889     to request this handling. The option can be set at compile or exec time.
2891     (b) $ matches before a terminating newline by default; an option
2892     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2893     mode). The option can be set at compile or exec time.
2895     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2896     the same as Perl's. If the decimal number it represents is less than 10
2897     or there aren't that many previous left capturing parentheses, an octal
2898     escape is read. Inside a character class, it's always an octal escape,
2899     even if it is a single digit.
2901     (d) An escaped but undefined alphabetic character is taken as a literal,
2902     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2903     escapes.
2905     (e) {0} is now permitted. (The previous item is removed from the compiled
2906     pattern).
2908     5. Changed all the names of code files so that the basic parts are no longer
2909     than 10 characters, and abolished the teeny "globals.c" file.
2911     6. Changed the handling of character classes; they are now done with a 32-byte
2912     bit map always.
2914     7. Added the -d and /D options to pcretest to make it possible to look at the
2915     internals of compilation without having to recompile pcre.
2918     Version 0.95 23-Sep-97
2919     ----------------------
2921     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2922     \x20 at the start of a run of normal characters. These were being treated as
2923     real characters, instead of the source characters being re-checked.
2926     Version 0.94 18-Sep-97
2927     ----------------------
2929     1. The functions are now thread-safe, with the caveat that the global variables
2930     containing pointers to malloc() and free() or alternative functions are the
2931     same for all threads.
2933     2. Get pcre_study() to generate a bitmap of initial characters for non-
2934     anchored patterns when this is possible, and use it if passed to pcre_exec().
2937     Version 0.93 15-Sep-97
2938     ----------------------
2940     1. /(b)|(:+)/ was computing an incorrect first character.
2942     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2943     but not actually doing anything yet.
2945     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2946     as Perl does (e.g. [-az] or [az-]).
2948     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2949     all possible positions.
2951     5. Split up into different modules to avoid including unneeded functions in a
2952     compiled binary. However, compile and exec are still in one module. The "study"
2953     function is split off.
2955     6. The character tables are now in a separate module whose source is generated
2956     by an auxiliary program - but can then be edited by hand if required. There are
2957     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2958     toupper() in the code.
2960     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2961     make them global. Abolish the function for setting them, as the caller can now
2962     set them directly.
2965     Version 0.92 11-Sep-97
2966     ----------------------
2968     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2969     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2971     2. Caseless matching was not working in character classes if the characters in
2972     the pattern were in upper case.
2974     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2976     4. Make PCRE_ANCHORED public and accept as a compile option.
2978     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2979     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2980     pass them.
2982     6. Give an error if bad option bits passed at compile or run time.
2984     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2985     pcretest to cause it to pass that flag.
2987     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2988     options, and the first character, if set.
2990     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2993     Version 0.91 10-Sep-97
2994     ----------------------
2996     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2997     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2999     2. PCRE was looping on encountering an indefinitely repeated back reference to
3000     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3001     Perl does - treats the match as successful.
3003     ****


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

ViewVC Help
Powered by ViewVC 1.1.12