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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 89 - (hide annotations) (download)
Sat Feb 24 21:41:27 2007 UTC (7 years, 1 month ago) by nigel
File size: 95469 byte(s)
Load pcre-6.6 into code/trunk.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 nigel 89 Version 6.6 06-Feb-06
5     ---------------------
6    
7     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
8     in pcreposix.h. I have copied the definition from pcre.h.
9    
10     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
11     because pcre.h is no longer a built file.
12    
13     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
14     not normally included in the compiled code.
15    
16    
17 nigel 87 Version 6.5 01-Feb-06
18     ---------------------
19    
20     1. When using the partial match feature with pcre_dfa_exec(), it was not
21     anchoring the second and subsequent partial matches at the new starting
22     point. This could lead to incorrect results. For example, with the pattern
23     /1234/, partially matching against "123" and then "a4" gave a match.
24    
25     2. Changes to pcregrep:
26    
27     (a) All non-match returns from pcre_exec() were being treated as failures
28     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
29     error message is output. Some extra information is given for the
30     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
31     probably the only errors that are likely to be caused by users (by
32     specifying a regex that has nested indefinite repeats, for instance).
33     If there are more than 20 of these errors, pcregrep is abandoned.
34    
35     (b) A binary zero was treated as data while matching, but terminated the
36     output line if it was written out. This has been fixed: binary zeroes
37     are now no different to any other data bytes.
38    
39     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
40     used to set a locale for matching. The --locale=xxxx long option has
41     been added (no short equivalent) to specify a locale explicitly on the
42     pcregrep command, overriding the environment variables.
43    
44     (d) When -B was used with -n, some line numbers in the output were one less
45     than they should have been.
46    
47     (e) Added the -o (--only-matching) option.
48    
49     (f) If -A or -C was used with -c (count only), some lines of context were
50     accidentally printed for the final match.
51    
52     (g) Added the -H (--with-filename) option.
53    
54     (h) The combination of options -rh failed to suppress file names for files
55     that were found from directory arguments.
56    
57     (i) Added the -D (--devices) and -d (--directories) options.
58    
59     (j) Added the -F (--fixed-strings) option.
60    
61     (k) Allow "-" to be used as a file name for -f as well as for a data file.
62    
63     (l) Added the --colo(u)r option.
64    
65     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
66     is not present by default.
67    
68     3. A nasty bug was discovered in the handling of recursive patterns, that is,
69     items such as (?R) or (?1), when the recursion could match a number of
70     alternatives. If it matched one of the alternatives, but subsequently,
71     outside the recursion, there was a failure, the code tried to back up into
72     the recursion. However, because of the way PCRE is implemented, this is not
73     possible, and the result was an incorrect result from the match.
74    
75     In order to prevent this happening, the specification of recursion has
76     been changed so that all such subpatterns are automatically treated as
77     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
78    
79     4. I had overlooked the fact that, in some locales, there are characters for
80     which isalpha() is true but neither isupper() nor islower() are true. In
81     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
82     and ordfeminine) are like this. This affected the treatment of \w and \W
83     when they appeared in character classes, but not when they appeared outside
84     a character class. The bit map for "word" characters is now created
85     separately from the results of isalnum() instead of just taking it from the
86     upper, lower, and digit maps. (Plus the underscore character, of course.)
87    
88     5. The above bug also affected the handling of POSIX character classes such as
89     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
90     permanent tables. Instead, the bit maps for such a class were previously
91     created as the appropriate unions of the upper, lower, and digit bitmaps.
92     Now they are created by subtraction from the [[:word:]] class, which has
93     its own bitmap.
94    
95     6. The [[:blank:]] character class matches horizontal, but not vertical space.
96     It is created by subtracting the vertical space characters (\x09, \x0a,
97     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
98     subtraction was done in the overall bitmap for a character class, meaning
99     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
100     be recognized. This bug has been fixed.
101    
102     7. Patches from the folks at Google:
103    
104     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
105     real life, but is still worth protecting against".
106    
107     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
108     regular expressions".
109    
110     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
111     have it.
112    
113     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
114     "configure" and the latter not, in order to fix a problem somebody had
115     with compiling the Arg class on HP-UX.
116    
117     (e) Improve the error-handling of the C++ wrapper a little bit.
118    
119     (f) New tests for checking recursion limiting.
120    
121     8. The pcre_memmove() function, which is used only if the environment does not
122     have a standard memmove() function (and is therefore rarely compiled),
123     contained two bugs: (a) use of int instead of size_t, and (b) it was not
124     returning a result (though PCRE never actually uses the result).
125    
126     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
127     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
128     returned instead of calling malloc() with an overflowing number that would
129     most likely cause subsequent chaos.
130    
131     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
132    
133     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
134     with this option is matched, the nmatch and pmatch options of regexec() are
135     ignored.
136    
137     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
138     provided in case anyone wants to the the POSIX interface with UTF-8
139     strings.
140    
141     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
142     C++ linking (needed for some HP-UX environments).
143    
144     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
145     (unused parameter) and in the pcre_printint() function (omitted "default"
146     switch label when the default is to do nothing).
147    
148     15. Added some code to make it possible, when PCRE is compiled as a C++
149     library, to replace subject pointers for pcre_exec() with a smart pointer
150     class, thus making it possible to process discontinuous strings.
151    
152     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
153     much the same function. They were added by different people who were trying
154     to make PCRE easy to compile on non-Unix systems. It has been suggested
155     that PCRE_EXPORT be abolished now that there is more automatic apparatus
156     for compiling on Windows systems. I have therefore replaced it with
157     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
158     defaults to "extern" for C or "extern C" for C++, which works fine on
159     Unix-like systems. It is now possible to override the value of PCRE_DATA_
160     SCOPE with something explicit in config.h. In addition:
161    
162     (a) pcreposix.h still had just "extern" instead of either of these macros;
163     I have replaced it with PCRE_DATA_SCOPE.
164    
165     (b) Functions such as _pcre_xclass(), which are internal to the library,
166     but external in the C sense, all had PCRE_EXPORT in their definitions.
167     This is apparently wrong for the Windows case, so I have removed it.
168     (It makes no difference on Unix-like systems.)
169    
170     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
171     of recursive calls to match(). This is different to MATCH_LIMIT because
172     that limits the total number of calls to match(), not all of which increase
173     the depth of recursion. Limiting the recursion depth limits the amount of
174     stack (or heap if NO_RECURSE is set) that is used. The default can be set
175     when PCRE is compiled, and changed at run time. A patch from Google adds
176     this functionality to the C++ interface.
177    
178     18. Changes to the handling of Unicode character properties:
179    
180     (a) Updated the table to Unicode 4.1.0.
181    
182     (b) Recognize characters that are not in the table as "Cn" (undefined).
183    
184     (c) I revised the way the table is implemented to a much improved format
185     which includes recognition of ranges. It now supports the ranges that
186     are defined in UnicodeData.txt, and it also amalgamates other
187     characters into ranges. This has reduced the number of entries in the
188     table from around 16,000 to around 3,000, thus reducing its size
189     considerably. I realized I did not need to use a tree structure after
190     all - a binary chop search is just as efficient. Having reduced the
191     number of entries, I extended their size from 6 bytes to 8 bytes to
192     allow for more data.
193    
194     (d) Added support for Unicode script names via properties such as \p{Han}.
195    
196     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
197     matching that character.
198    
199     20. When matching a repeated Unicode property with a minimum greater than zero,
200     (for example \pL{2,}), PCRE could look past the end of the subject if it
201     reached it while seeking the minimum number of characters. This could
202     happen only if some of the characters were more than one byte long, because
203     there is a check for at least the minimum number of bytes.
204    
205     21. Refactored the implementation of \p and \P so as to be more general, to
206     allow for more different types of property in future. This has changed the
207     compiled form incompatibly. Anybody with saved compiled patterns that use
208     \p or \P will have to recompile them.
209    
210     22. Added "Any" and "L&" to the supported property types.
211    
212     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
213     but give a compile time error if the value is greater than 0xff.
214    
215     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
216     accidentally not being installed or uninstalled.
217    
218     25. The pcre.h file was built from pcre.h.in, but the only changes that were
219     made were to insert the current release number. This seemed silly, because
220     it made things harder for people building PCRE on systems that don't run
221     "configure". I have turned pcre.h into a distributed file, no longer built
222     by "configure", with the version identification directly included. There is
223     no longer a pcre.h.in file.
224    
225     However, this change necessitated a change to the pcre-config script as
226     well. It is built from pcre-config.in, and one of the substitutions was the
227     release number. I have updated configure.ac so that ./configure now finds
228     the release number by grepping pcre.h.
229    
230     26. Added the ability to run the tests under valgrind.
231    
232    
233 nigel 85 Version 6.4 05-Sep-05
234     ---------------------
235    
236     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
237     "--" to be printed when multiple files were scanned, even when none of the
238     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
239     consider it to be a bug, and have restored the previous behaviour.
240    
241     2. A couple of code tidies to get rid of compiler warnings.
242    
243     3. The pcretest program used to cheat by referring to symbols in the library
244     whose names begin with _pcre_. These are internal symbols that are not
245     really supposed to be visible externally, and in some environments it is
246     possible to suppress them. The cheating is now confined to including
247     certain files from the library's source, which is a bit cleaner.
248    
249     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
250     file's purpose clearer.
251    
252     5. Reorganized pcre_ucp_findchar().
253    
254    
255 nigel 83 Version 6.3 15-Aug-05
256     ---------------------
257    
258     1. The file libpcre.pc.in did not have general read permission in the tarball.
259    
260     2. There were some problems when building without C++ support:
261    
262     (a) If C++ support was not built, "make install" and "make test" still
263     tried to test it.
264    
265     (b) There were problems when the value of CXX was explicitly set. Some
266     changes have been made to try to fix these, and ...
267    
268     (c) --disable-cpp can now be used to explicitly disable C++ support.
269    
270     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
271     backslash in a target when C++ was disabled. This confuses some
272     versions of "make", apparently. Using an intermediate variable solves
273     this. (Same for CPP_LOBJ.)
274    
275     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
276     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
277     necessary on certain architectures.
278    
279     4. Added a setting of -export-symbols-regex to the link command to remove
280     those symbols that are exported in the C sense, but actually are local
281     within the library, and not documented. Their names all begin with
282     "_pcre_". This is not a perfect job, because (a) we have to except some
283     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
284     available (and never for static libraries). I have made a note to try to
285     find a way round (a) in the future.
286    
287    
288 nigel 81 Version 6.2 01-Aug-05
289     ---------------------
290    
291     1. There was no test for integer overflow of quantifier values. A construction
292     such as {1111111111111111} would give undefined results. What is worse, if
293     a minimum quantifier for a parenthesized subpattern overflowed and became
294     negative, the calculation of the memory size went wrong. This could have
295     led to memory overwriting.
296    
297     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
298    
299     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
300     operating environments where this matters.
301    
302     4. Applied Giuseppe Maxia's patch to add additional features for controlling
303     PCRE options from within the C++ wrapper.
304    
305     5. Named capturing subpatterns were not being correctly counted when a pattern
306     was compiled. This caused two problems: (a) If there were more than 100
307     such subpatterns, the calculation of the memory needed for the whole
308     compiled pattern went wrong, leading to an overflow error. (b) Numerical
309     back references of the form \12, where the number was greater than 9, were
310     not recognized as back references, even though there were sufficient
311     previous subpatterns.
312    
313     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
314     versions of gcc, e.g. 2.95.4.
315    
316    
317 nigel 79 Version 6.1 21-Jun-05
318     ---------------------
319    
320     1. There was one reference to the variable "posix" in pcretest.c that was not
321     surrounded by "#if !defined NOPOSIX".
322    
323     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
324     the cross-check on the old pcre_info() function, for the benefit of the
325     cut-down version of PCRE that is currently imported into Exim.
326    
327     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
328     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
329     patterns that start (?i) but I don't think that matters. The use of (?i) is
330     just an example; this all applies to the other options as well.
331    
332     4. Since libtool seems to echo the compile commands it is issuing, the output
333     from "make" can be reduced a bit by putting "@" in front of each libtool
334     compile command.
335    
336     5. Patch from the folks at Google for configure.in to be a bit more thorough
337     in checking for a suitable C++ installation before trying to compile the
338     C++ stuff. This should fix a reported problem when a compiler was present,
339     but no suitable headers.
340    
341     6. The man pages all had just "PCRE" as their title. I have changed them to
342     be the relevant file name. I have also arranged that these names are
343     retained in the file doc/pcre.txt, which is a concatenation in text format
344     of all the man pages except the little individual ones for each function.
345    
346     7. The NON-UNIX-USE file had not been updated for the different set of source
347     files that come with release 6. I also added a few comments about the C++
348     wrapper.
349    
350    
351 nigel 77 Version 6.0 07-Jun-05
352     ---------------------
353    
354     1. Some minor internal re-organization to help with my DFA experiments.
355    
356     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
357     didn't matter for the library itself when fully configured, but did matter
358     when compiling without UCP support, or within Exim, where the ucp files are
359     not imported.
360    
361     3. Refactoring of the library code to split up the various functions into
362     different source modules. The addition of the new DFA matching code (see
363     below) to a single monolithic source would have made it really too
364     unwieldy, quite apart from causing all the code to be include in a
365     statically linked application, when only some functions are used. This is
366     relevant even without the DFA addition now that patterns can be compiled in
367     one application and matched in another.
368    
369     The downside of splitting up is that there have to be some external
370     functions and data tables that are used internally in different modules of
371     the library but which are not part of the API. These have all had their
372     names changed to start with "_pcre_" so that they are unlikely to clash
373     with other external names.
374    
375     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
376     a different (DFA) algorithm. Although it is slower than the original
377     function, it does have some advantages for certain types of matching
378     problem.
379    
380     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
381     including restarting after a partial match.
382    
383     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
384     defined when compiling for Windows was sent to me. I have put it into the
385     code, though I have no means of testing or verifying it.
386    
387     7. Added the pcre_refcount() auxiliary function.
388    
389     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
390     match before or at the first newline in the subject string. In pcretest,
391     the /f option on a pattern can be used to set this.
392    
393     9. A repeated \w when used in UTF-8 mode with characters greater than 256
394     would behave wrongly. This has been present in PCRE since release 4.0.
395    
396     10. A number of changes to the pcregrep command:
397    
398     (a) Refactored how -x works; insert ^(...)$ instead of setting
399     PCRE_ANCHORED and checking the length, in preparation for adding
400     something similar for -w.
401    
402     (b) Added the -w (match as a word) option.
403    
404     (c) Refactored the way lines are read and buffered so as to have more
405     than one at a time available.
406    
407     (d) Implemented a pcregrep test script.
408    
409     (e) Added the -M (multiline match) option. This allows patterns to match
410     over several lines of the subject. The buffering ensures that at least
411     8K, or the rest of the document (whichever is the shorter) is available
412     for matching (and similarly the previous 8K for lookbehind assertions).
413    
414     (f) Changed the --help output so that it now says
415    
416     -w, --word-regex(p)
417    
418     instead of two lines, one with "regex" and the other with "regexp"
419     because that confused at least one person since the short forms are the
420     same. (This required a bit of code, as the output is generated
421     automatically from a table. It wasn't just a text change.)
422    
423     (g) -- can be used to terminate pcregrep options if the next thing isn't an
424     option but starts with a hyphen. Could be a pattern or a path name
425     starting with a hyphen, for instance.
426    
427     (h) "-" can be given as a file name to represent stdin.
428    
429     (i) When file names are being printed, "(standard input)" is used for
430     the standard input, for compatibility with GNU grep. Previously
431     "<stdin>" was used.
432    
433     (j) The option --label=xxx can be used to supply a name to be used for
434     stdin when file names are being printed. There is no short form.
435    
436     (k) Re-factored the options decoding logic because we are going to add
437     two more options that take data. Such options can now be given in four
438     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
439    
440     (l) Added the -A, -B, and -C options for requesting that lines of context
441     around matches be printed.
442    
443     (m) Added the -L option to print the names of files that do not contain
444     any matching lines, that is, the complement of -l.
445    
446     (n) The return code is 2 if any file cannot be opened, but pcregrep does
447     continue to scan other files.
448    
449     (o) The -s option was incorrectly implemented. For compatibility with other
450     greps, it now suppresses the error message for a non-existent or non-
451     accessible file (but not the return code). There is a new option called
452     -q that suppresses the output of matching lines, which was what -s was
453     previously doing.
454    
455     (p) Added --include and --exclude options to specify files for inclusion
456     and exclusion when recursing.
457    
458     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
459     Hopefully, it now does.
460    
461     12. Missing cast in pcre_study().
462    
463     13. Added an "uninstall" target to the makefile.
464    
465     14. Replaced "extern" in the function prototypes in Makefile.in with
466     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
467     world, but is set differently for Windows.
468    
469     15. Added a second compiling function called pcre_compile2(). The only
470     difference is that it has an extra argument, which is a pointer to an
471     integer error code. When there is a compile-time failure, this is set
472     non-zero, in addition to the error test pointer being set to point to an
473     error message. The new argument may be NULL if no error number is required
474     (but then you may as well call pcre_compile(), which is now just a
475     wrapper). This facility is provided because some applications need a
476     numeric error indication, but it has also enabled me to tidy up the way
477     compile-time errors are handled in the POSIX wrapper.
478    
479     16. Added VPATH=.libs to the makefile; this should help when building with one
480     prefix path and installing with another. (Or so I'm told by someone who
481     knows more about this stuff than I do.)
482    
483     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
484     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
485     match everything, including newlines. This is not POSIX-compatible, but
486     somebody wanted the feature. From pcretest it can be activated by using
487     both the P and the s flags.
488    
489     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
490    
491     19. libpcre.pc was being incorrectly installed as executable.
492    
493     20. A couple of places in pcretest check for end-of-line by looking for '\n';
494     it now also looks for '\r' so that it will work unmodified on Windows.
495    
496     21. Added Google's contributed C++ wrapper to the distribution.
497    
498     22. Added some untidy missing memory free() calls in pcretest, to keep
499     Electric Fence happy when testing.
500    
501    
502    
503 nigel 75 Version 5.0 13-Sep-04
504     ---------------------
505    
506     1. Internal change: literal characters are no longer packed up into items
507     containing multiple characters in a single byte-string. Each character
508     is now matched using a separate opcode. However, there may be more than one
509     byte in the character in UTF-8 mode.
510    
511     2. The pcre_callout_block structure has two new fields: pattern_position and
512     next_item_length. These contain the offset in the pattern to the next match
513     item, and its length, respectively.
514    
515     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
516     insertion of callouts before each pattern item. Added the /C option to
517     pcretest to make use of this.
518    
519     4. On the advice of a Windows user, the lines
520    
521     #if defined(_WIN32) || defined(WIN32)
522     _setmode( _fileno( stdout ), 0x8000 );
523     #endif /* defined(_WIN32) || defined(WIN32) */
524    
525     have been added to the source of pcretest. This apparently does useful
526     magic in relation to line terminators.
527    
528     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
529     for the benefit of those environments where the "b" makes a difference.
530    
531     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
532     to know about it. I have put a hack into configure.in that adds in code
533     to set GCC=yes if CC=icc. This seems to end up at a point in the
534     generated configure script that is early enough to affect the setting of
535     compiler options, which is what is needed, but I have no means of testing
536     whether it really works. (The user who reported this had patched the
537     generated configure script, which of course I cannot do.)
538    
539     LATER: After change 22 below (new libtool files), the configure script
540     seems to know about icc (and also ecc). Therefore, I have commented out
541     this hack in configure.in.
542    
543     7. Added support for pkg-config (2 patches were sent in).
544    
545     8. Negated POSIX character classes that used a combination of internal tables
546     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
547     [[:^ascii]]. Typically, they would match almost any characters. The other
548     POSIX classes were not broken in this way.
549    
550     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
551     to find the match, as PCRE was deluded into thinking that the match had to
552     start at the start point or following a newline. The same bug applied to
553     patterns with negative forward assertions or any backward assertions
554     preceding ".*" at the start, unless the pattern required a fixed first
555     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
556    
557     10. In UTF-8 mode, when moving forwards in the subject after a failed match
558     starting at the last subject character, bytes beyond the end of the subject
559     string were read.
560    
561     11. Renamed the variable "class" as "classbits" to make life easier for C++
562     users. (Previously there was a macro definition, but it apparently wasn't
563     enough.)
564    
565     12. Added the new field "tables" to the extra data so that tables can be passed
566     in at exec time, or the internal tables can be re-selected. This allows
567     a compiled regex to be saved and re-used at a later time by a different
568     program that might have everything at different addresses.
569    
570     13. Modified the pcre-config script so that, when run on Solaris, it shows a
571     -R library as well as a -L library.
572    
573     14. The debugging options of pcretest (-d on the command line or D on a
574     pattern) showed incorrect output for anything following an extended class
575     that contained multibyte characters and which was followed by a quantifier.
576    
577     15. Added optional support for general category Unicode character properties
578     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
579     support. It adds about 90K to the size of the library. The meanings of the
580     inbuilt class escapes such as \d and \s have NOT been changed.
581    
582     16. Updated pcredemo.c to include calls to free() to release the memory for the
583     compiled pattern.
584    
585     17. The generated file chartables.c was being created in the source directory
586     instead of in the building directory. This caused the build to fail if the
587     source directory was different from the building directory, and was
588     read-only.
589    
590     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
591     file. No doubt somebody will tell me if they don't make sense... Also added
592     Dan Mooney's comments about building on OpenVMS.
593    
594     19. Added support for partial matching via the PCRE_PARTIAL option for
595     pcre_exec() and the \P data escape in pcretest.
596    
597     20. Extended pcretest with 3 new pattern features:
598    
599     (i) A pattern option of the form ">rest-of-line" causes pcretest to
600     write the compiled pattern to the file whose name is "rest-of-line".
601     This is a straight binary dump of the data, with the saved pointer to
602     the character tables forced to be NULL. The study data, if any, is
603     written too. After writing, pcretest reads a new pattern.
604    
605     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
606     compiled pattern from the given file. There must not be any
607     occurrences of "<" in the file name (pretty unlikely); if there are,
608     pcretest will instead treat the initial "<" as a pattern delimiter.
609     After reading in the pattern, pcretest goes on to read data lines as
610     usual.
611    
612     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
613     and 16-bit fields in a compiled pattern, to simulate a pattern that
614     was compiled on a host of opposite endianness.
615    
616     21. The pcre-exec() function can now cope with patterns that were compiled on
617     hosts of opposite endianness, with this restriction:
618    
619     As for any compiled expression that is saved and used later, the tables
620     pointer field cannot be preserved; the extra_data field in the arguments
621     to pcre_exec() should be used to pass in a tables address if a value
622     other than the default internal tables were used at compile time.
623    
624     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
625     now diagnosed as an error. Previously, most of the time, a negative number
626     would have been treated as zero, but if in addition "ovector" was passed as
627     NULL, a crash could occur.
628    
629     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
630     new versions from the libtool 1.5 distribution (the last one is a copy of
631     a file called libtool.m4). This seems to have fixed the need to patch
632     "configure" to support Darwin 1.3 (which I used to do). However, I still
633     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
634     workstation).
635    
636     24. Changed the PCRE licence to be the more standard "BSD" licence.
637    
638    
639 nigel 73 Version 4.5 01-Dec-03
640     ---------------------
641    
642     1. There has been some re-arrangement of the code for the match() function so
643     that it can be compiled in a version that does not call itself recursively.
644     Instead, it keeps those local variables that need separate instances for
645     each "recursion" in a frame on the heap, and gets/frees frames whenever it
646     needs to "recurse". Keeping track of where control must go is done by means
647     of setjmp/longjmp. The whole thing is implemented by a set of macros that
648     hide most of the details from the main code, and operates only if
649     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
650     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
651     operating.
652    
653     To make it easier for callers to provide specially tailored get/free
654     functions for this usage, two new functions, pcre_stack_malloc, and
655     pcre_stack_free, are used. They are always called in strict stacking order,
656     and the size of block requested is always the same.
657    
658     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
659     PCRE has been compiled to use the stack or the heap for recursion. The
660     -C option of pcretest uses this to show which version is compiled.
661    
662     A new data escape \S, is added to pcretest; it causes the amounts of store
663     obtained and freed by both kinds of malloc/free at match time to be added
664     to the output.
665    
666     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
667     what's available on my current Linux desktop machine.
668    
669     3. When matching a UTF-8 string, the test for a valid string at the start has
670     been extended. If start_offset is not zero, PCRE now checks that it points
671     to a byte that is the start of a UTF-8 character. If not, it returns
672     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
673     this is necessary because there may be backward assertions in the pattern.
674     When matching the same subject several times, it may save resources to use
675     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
676    
677     4. The code for checking the validity of UTF-8 strings has been tightened so
678     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
679     containing "overlong sequences".
680    
681     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
682     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
683     should have been "&", but it just so happened that all the cases this let
684     through by mistake were picked up later in the function.
685    
686     6. I had used a variable called "isblank" - this is a C99 function, causing
687     some compilers to warn. To avoid this, I renamed it (as "blankclass").
688    
689     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
690     prompting; (b) run "./pcretest /dev/null" at the start of the test script
691     so the version is shown; (c) stop "make test" echoing "./RunTest".
692    
693     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
694    
695     9. The prototype for memmove() for systems that don't have it was using
696     size_t, but the inclusion of the header that defines size_t was later. I've
697     moved the #includes for the C headers earlier to avoid this.
698    
699     10. Added some adjustments to the code to make it easier to compiler on certain
700     special systems:
701    
702     (a) Some "const" qualifiers were missing.
703     (b) Added the macro EXPORT before all exported functions; by default this
704     is defined to be empty.
705     (c) Changed the dftables auxiliary program (that builds chartables.c) so
706     that it reads its output file name as an argument instead of writing
707     to the standard output and assuming this can be redirected.
708    
709     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
710     class containing characters with values greater than 255, PCRE compilation
711     went into a loop.
712    
713     12. A recursive reference to a subpattern that was within another subpattern
714     that had a minimum quantifier of zero caused PCRE to crash. For example,
715     (x(y(?2))z)? provoked this bug with a subject that got as far as the
716     recursion. If the recursively-called subpattern itself had a zero repeat,
717     that was OK.
718    
719     13. In pcretest, the buffer for reading a data line was set at 30K, but the
720     buffer into which it was copied (for escape processing) was still set at
721     1024, so long lines caused crashes.
722    
723     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
724     "internal error: code overflow...". This applied to any character class
725     that was followed by a possessive quantifier.
726    
727     15. Modified the Makefile to add libpcre.la as a prerequisite for
728     libpcreposix.la because I was told this is needed for a parallel build to
729     work.
730    
731     16. If a pattern that contained .* following optional items at the start was
732     studied, the wrong optimizing data was generated, leading to matching
733     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
734     matching string must start with a or b or c. The correct conclusion for
735     this pattern is that a match can start with any character.
736    
737    
738 nigel 71 Version 4.4 13-Aug-03
739     ---------------------
740    
741     1. In UTF-8 mode, a character class containing characters with values between
742     127 and 255 was not handled correctly if the compiled pattern was studied.
743     In fixing this, I have also improved the studying algorithm for such
744     classes (slightly).
745    
746     2. Three internal functions had redundant arguments passed to them. Removal
747     might give a very teeny performance improvement.
748    
749     3. Documentation bug: the value of the capture_top field in a callout is *one
750     more than* the number of the hightest numbered captured substring.
751    
752     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
753     in incorrectly linking with a previously installed version. They now link
754     explicitly with libpcre.la.
755    
756     5. configure.in no longer needs to recognize Cygwin specially.
757    
758     6. A problem in pcre.in for Windows platforms is fixed.
759    
760     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
761     pcretest, it used to include the size of the study block as part of its
762     output. Unfortunately, the structure contains a field that has a different
763     size on different hardware architectures. This meant that the tests that
764     showed this size failed. As the block is currently always of a fixed size,
765     this information isn't actually particularly useful in pcretest output, so
766     I have just removed it.
767    
768     8. Three pre-processor statements accidentally did not start in column 1.
769     Sadly, there are *still* compilers around that complain, even though
770     standard C has not required this for well over a decade. Sigh.
771    
772     9. In pcretest, the code for checking callouts passed small integers in the
773     callout_data field, which is a void * field. However, some picky compilers
774     complained about the casts involved for this on 64-bit systems. Now
775     pcretest passes the address of the small integer instead, which should get
776     rid of the warnings.
777    
778     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
779     both compile and run time, and gives an error if an invalid UTF-8 sequence
780     is found. There is a option for disabling this check in cases where the
781     string is known to be correct and/or the maximum performance is wanted.
782    
783     11. In response to a bug report, I changed one line in Makefile.in from
784    
785     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
786     to
787     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
788    
789     to look similar to other lines, but I have no way of telling whether this
790     is the right thing to do, as I do not use Windows. No doubt I'll get told
791     if it's wrong...
792    
793    
794 nigel 69 Version 4.3 21-May-03
795     ---------------------
796    
797     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
798     Makefile.
799    
800     2. Some refactoring to improve the quality of the code:
801    
802     (i) The utf8_table... variables are now declared "const".
803    
804     (ii) The code for \cx, which used the "case flipping" table to upper case
805     lower case letters, now just substracts 32. This is ASCII-specific,
806     but the whole concept of \cx is ASCII-specific, so it seems
807     reasonable.
808    
809     (iii) PCRE was using its character types table to recognize decimal and
810     hexadecimal digits in the pattern. This is silly, because it handles
811     only 0-9, a-f, and A-F, but the character types table is locale-
812     specific, which means strange things might happen. A private
813     table is now used for this - though it costs 256 bytes, a table is
814     much faster than multiple explicit tests. Of course, the standard
815     character types table is still used for matching digits in subject
816     strings against \d.
817    
818     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
819     ending in _t are). So I've renamed it as ESC_tee.
820    
821     3. The first argument for regexec() in the POSIX wrapper should have been
822     defined as "const".
823    
824     4. Changed pcretest to use malloc() for its buffers so that they can be
825     Electric Fenced for debugging.
826    
827     5. There were several places in the code where, in UTF-8 mode, PCRE would try
828     to read one or more bytes before the start of the subject string. Often this
829     had no effect on PCRE's behaviour, but in some circumstances it could
830     provoke a segmentation fault.
831    
832     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
833     to try to read one or more bytes before the start of the subject string.
834    
835     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
836     UTF-8 support could misbehave in various ways if the subject string
837     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
838     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
839     back over UTF-8 characters.)
840    
841    
842 nigel 67 Version 4.2 14-Apr-03
843     ---------------------
844    
845     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
846    
847     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
848     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
849     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
850     [WIN_PREFIX]: new variable, "cyg" for Cygwin
851     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
852     and BUILD_EXEEXT
853     Note: automatic setting of the BUILD variables is not yet working
854     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
855     compile-time but not at link-time
856     [LINK]: use for linking executables only
857     make different versions for Windows and non-Windows
858     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
859     libraries
860     [LINK_FOR_BUILD]: new variable
861     [OBJEXT]: use throughout
862     [EXEEXT]: use throughout
863     <winshared>: new target
864     <wininstall>: new target
865     <dftables.o>: use native compiler
866     <dftables>: use native linker
867     <install>: handle Windows platform correctly
868     <clean>: ditto
869     <check>: ditto
870     copy DLL to top builddir before testing
871    
872     As part of these changes, -no-undefined was removed again. This was reported
873     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
874     in any case.
875    
876     3. Some tidies to get rid of compiler warnings:
877    
878     . In the match_data structure, match_limit was an unsigned long int, whereas
879     match_call_count was an int. I've made them both unsigned long ints.
880    
881     . In pcretest the fact that a const uschar * doesn't automatically cast to
882     a void * provoked a warning.
883    
884     . Turning on some more compiler warnings threw up some "shadow" variables
885     and a few more missing casts.
886    
887     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
888     option, a class that contained a single character with a value between 128
889     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
890    
891     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
892     option, a class that contained several characters, but with at least one
893     whose value was between 128 and 255 caused PCRE to crash.
894    
895    
896 nigel 65 Version 4.1 12-Mar-03
897     ---------------------
898 nigel 63
899 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
900     needed, and a string in dftables.c that was longer than standard compilers are
901     required to support.
902    
903     2. Compiling with Sun's compiler found a few more places where the code could
904     be tidied up in order to avoid warnings.
905    
906     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
907     first of these names is deprecated in the latest Autoconf in favour of the name
908     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
909     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
910     analogy I have changed it to CFLAGS_FOR_BUILD.
911    
912     4. Added -no-undefined to the linking command in the Makefile, because this is
913     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
914     linking step for the pcreposix library.
915    
916     5. PCRE was failing to diagnose the case of two named groups with the same
917     name.
918    
919     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
920     literal character that is needed in the subject for a match, and scans along to
921     ensure that it is present before embarking on the full matching process. This
922     saves time in cases of nested unlimited repeats that are never going to match.
923     Problem: the scan can take a lot of time if the subject is very long (e.g.
924     megabytes), thus penalizing straightforward matches. It is now done only if the
925     amount of subject to be scanned is less than 1000 bytes.
926    
927     7. A lesser problem with the same optimization is that it was recording the
928     first character of an anchored pattern as "needed", thus provoking a search
929     right along the subject, even when the first match of the pattern was going to
930     fail. The "needed" character is now not set for anchored patterns, unless it
931     follows something in the pattern that is of non-fixed length. Thus, it still
932     fulfils its original purpose of finding quick non-matches in cases of nested
933     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
934    
935    
936     Version 4.0 17-Feb-03
937     ---------------------
938    
939 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
940     extended to the end of string, PCRE compiled incorrect data. This could lead to
941     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
942    
943     2. Moved to autoconf 2.53 and libtool 1.4.2.
944    
945     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
946     the special perltest8 script is no longer needed - all the tests can be run
947     from a single perltest script.
948    
949     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
950     by \s. It has now been removed in PCRE. This means it isn't recognized as
951     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
952     class [:space:] *does* include VT, thereby creating a mess.
953    
954     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
955     space and tab.
956    
957     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
958     its new features into the main test script, reducing the number of scripts.
959    
960     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
961     were backward compatible, and made the (?i) apply to the whole pattern, as if
962     /i were given. Now it behaves more logically, and applies the option setting
963     only to what follows. PCRE has been changed to follow suit. However, if it
964     finds options settings right at the start of the pattern, it extracts them into
965     the global options, as before. Thus, they show up in the info data.
966    
967     8. Added support for the \Q...\E escape sequence. Characters in between are
968     treated as literals. This is slightly different from Perl in that $ and @ are
969     also handled as literals inside the quotes. In Perl, they will cause variable
970     interpolation. Note the following examples:
971    
972     Pattern PCRE matches Perl matches
973    
974     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
975     \Qabc\$xyz\E abc\$xyz abc\$xyz
976     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
977    
978     For compatibility with Perl, \Q...\E sequences are recognized inside character
979     classes as well as outside them.
980    
981     9. Re-organized 3 code statements in pcretest to avoid "overflow in
982     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
983     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
984     signed/unsigned warnings.
985    
986     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
987     option for pcretest, so I've replaced it by a simple function that does just
988     that job.
989    
990     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
991     "pcregrep -".
992    
993     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
994     Java package. This provides some syntactic sugar for simple cases of what my
995     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
996     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
997     item, you can use this simplified notation. Note that only makes sense with
998     greedy quantifiers. Consequently, the use of the possessive quantifier forces
999     greediness, whatever the setting of the PCRE_UNGREEDY option.
1000    
1001     13. A change of greediness default within a pattern was not taking effect at
1002     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1003     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1004     was abstracted outside.
1005    
1006     14. PCRE now supports the \G assertion. It is true when the current matching
1007     position is at the start point of the match. This differs from \A when the
1008     starting offset is non-zero. Used with the /g option of pcretest (or similar
1009     code), it works in the same way as it does for Perl's /g option. If all
1010     alternatives of a regex begin with \G, the expression is anchored to the start
1011     match position, and the "anchored" flag is set in the compiled expression.
1012    
1013     15. Some bugs concerning the handling of certain option changes within patterns
1014     have been fixed. These applied to options other than (?ims). For example,
1015     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1016     been the other way round. Some of this was related to change 7 above.
1017    
1018     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1019     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1020     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1021     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1022    
1023     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1024     mode. Unlike ".", it always matches newline, whatever the setting of
1025     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1026     assertions. Perl allows it, but it doesn't (in general) work because it can't
1027     calculate the length of the lookbehind. At least, that's the case for Perl
1028     5.8.0 - I've been told they are going to document that it doesn't work in
1029     future.
1030    
1031     18. Added an error diagnosis for escapes that PCRE does not support: these are
1032     \L, \l, \N, \P, \p, \U, \u, and \X.
1033    
1034     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1035     reading past the end of the pattern in cases such as /[abcd/.
1036    
1037     20. PCRE was getting more memory than necessary for patterns with classes that
1038     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1039    
1040     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1041     compiling PCRE for use with Virtual Pascal.
1042    
1043     22. Small fix to the Makefile to make it work properly if the build is done
1044     outside the source tree.
1045    
1046     23. Added a new extension: a condition to go with recursion. If a conditional
1047     subpattern starts with (?(R) the "true" branch is used if recursion has
1048     happened, whereas the "false" branch is used only at the top level.
1049    
1050     24. When there was a very long string of literal characters (over 255 bytes
1051     without UTF support, over 250 bytes with UTF support), the computation of how
1052     much memory was required could be incorrect, leading to segfaults or other
1053     strange effects.
1054    
1055     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1056     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1057     there was a subsequent back reference to those brackets. This meant that, for
1058     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1059     possible to check for precisely this case. All we can do is abandon the
1060     optimization if .* occurs inside capturing brackets when there are any back
1061     references whatsoever. (See below for a better fix that came later.)
1062    
1063     26. The handling of the optimization for finding the first character of a
1064     non-anchored pattern, and for finding a character that is required later in the
1065     match were failing in some cases. This didn't break the matching; it just
1066     failed to optimize when it could. The way this is done has been re-implemented.
1067    
1068     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1069    
1070     28. Added a new feature that provides some of the functionality that Perl
1071     provides with (?{...}). The facility is termed a "callout". The way it is done
1072     in PCRE is for the caller to provide an optional function, by setting
1073     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1074     global variable. By default it is unset, which disables all calling out. To get
1075     the function called, the regex must include (?C) at appropriate points. This
1076     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1077     This provides a means of identifying different callout points. When PCRE
1078     reaches such a point in the regex, if pcre_callout has been set, the external
1079     function is called. It is provided with data in a structure called
1080     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1081     matching continues; if it returns a non-zero value, the match at the current
1082     point fails. However, backtracking will occur if possible. [This was changed
1083     later and other features added - see item 49 below.]
1084    
1085     29. pcretest is upgraded to test the callout functionality. It provides a
1086     callout function that displays information. By default, it shows the start of
1087     the match and the current position in the text. There are some new data escapes
1088     to vary what happens:
1089    
1090     \C+ in addition, show current contents of captured substrings
1091     \C- do not supply a callout function
1092     \C!n return 1 when callout number n is reached
1093     \C!n!m return 1 when callout number n is reached for the mth time
1094    
1095     30. If pcregrep was called with the -l option and just a single file name, it
1096     output "<stdin>" if a match was found, instead of the file name.
1097    
1098     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1099     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1100     pcre_exec(). This saves a malloc/free per call. The default value of
1101     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1102     when configuring.
1103    
1104     32. The default maximum size of a compiled pattern is 64K. There have been a
1105     few cases of people hitting this limit. The code now uses macros to handle the
1106     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1107     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1108     configuring. Tests 2 and 5 work only with 2-byte links because they output
1109     debugging information about compiled patterns.
1110    
1111     33. Internal code re-arrangements:
1112    
1113     (a) Moved the debugging function for printing out a compiled regex into
1114     its own source file (printint.c) and used #include to pull it into
1115     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1116     separate copies.
1117    
1118     (b) Defined the list of op-code names for debugging as a macro in
1119     internal.h so that it is next to the definition of the opcodes.
1120    
1121     (c) Defined a table of op-code lengths for simpler skipping along compiled
1122     code. This is again a macro in internal.h so that it is next to the
1123     definition of the opcodes.
1124    
1125     34. Added support for recursive calls to individual subpatterns, along the
1126     lines of Robin Houston's patch (but implemented somewhat differently).
1127    
1128     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1129     allow it to read and process whole directories in Win32. This code was
1130     contributed by Lionel Fourquaux; it has not been tested by me.
1131    
1132     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1133     used to name a group. Names consist of alphanumerics and underscores, and must
1134     be unique. Back references use the syntax (?P=name) and recursive calls use
1135     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1136     numbers. The function pcre_fullinfo() can be used after compilation to extract
1137     a name/number map. There are three relevant calls:
1138    
1139     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1140     PCRE_INFO_NAMECOUNT yields the number of entries
1141     PCRE_INFO_NAMETABLE yields a pointer to the map.
1142    
1143     The map is a vector of fixed-size entries. The size of each entry depends on
1144     the length of the longest name used. The first two bytes of each entry are the
1145     group number, most significant byte first. There follows the corresponding
1146     name, zero terminated. The names are in alphabetical order.
1147    
1148     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1149     case instead of 255. Making it the same both with and without UTF-8 support
1150     means that the same test output works with both.
1151    
1152     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1153     calling malloc() with a zero argument.
1154    
1155     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1156     optimization. I've improved things by keeping a bitmap of backreferences with
1157     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1158     fact referenced, the optimization can be applied. It is unlikely that a
1159     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1160     the match to follow \n) will appear inside brackets with a number greater than
1161     31, but if it does, any back reference > 31 suppresses the optimization.
1162    
1163     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1164     of disabling numbered capturing parentheses. Any opening parenthesis that is
1165     not followed by ? behaves as if it were followed by ?: but named parentheses
1166     can still be used for capturing (and they will acquire numbers in the usual
1167     way).
1168    
1169     41. Redesigned the return codes from the match() function into yes/no/error so
1170     that errors can be passed back from deep inside the nested calls. A malloc
1171     failure while inside a recursive subpattern call now causes the
1172     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1173    
1174     42. It is now possible to set a limit on the number of times the match()
1175     function is called in a call to pcre_exec(). This facility makes it possible to
1176     limit the amount of recursion and backtracking, though not in a directly
1177     obvious way, because the match() function is used in a number of different
1178     circumstances. The count starts from zero for each position in the subject
1179     string (for non-anchored patterns). The default limit is, for compatibility, a
1180     large number, namely 10 000 000. You can change this in two ways:
1181    
1182     (a) When configuring PCRE before making, you can use --with-match-limit=n
1183     to set a default value for the compiled library.
1184    
1185     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1186     a different value is set. See 45 below.
1187    
1188     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1189    
1190     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1191     of things that can be changed at compile time. The first argument specifies
1192     what is wanted and the second points to where the information is to be placed.
1193     The current list of available information is:
1194    
1195     PCRE_CONFIG_UTF8
1196    
1197     The output is an integer that is set to one if UTF-8 support is available;
1198     otherwise it is set to zero.
1199    
1200     PCRE_CONFIG_NEWLINE
1201    
1202     The output is an integer that it set to the value of the code that is used for
1203     newline. It is either LF (10) or CR (13).
1204    
1205     PCRE_CONFIG_LINK_SIZE
1206    
1207     The output is an integer that contains the number of bytes used for internal
1208     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1209    
1210     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1211    
1212     The output is an integer that contains the threshold above which the POSIX
1213     interface uses malloc() for output vectors. See item 31 above.
1214    
1215     PCRE_CONFIG_MATCH_LIMIT
1216    
1217     The output is an unsigned integer that contains the default limit of the number
1218     of match() calls in a pcre_exec() execution. See 42 above.
1219    
1220     44. pcretest has been upgraded by the addition of the -C option. This causes it
1221     to extract all the available output from the new pcre_config() function, and to
1222     output it. The program then exits immediately.
1223    
1224     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1225     order to support additional features. One way would have been to define
1226     pcre_exec2() (for example) with extra arguments, but this would not have been
1227     extensible, and would also have required all calls to the original function to
1228     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1229     is used for passing in "extra" data from pcre_study().
1230    
1231     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1232     contains the following fields:
1233    
1234     flags a bitmap indicating which of the following fields are set
1235     study_data opaque data from pcre_study()
1236     match_limit a way of specifying a limit on match() calls for a specific
1237     call to pcre_exec()
1238     callout_data data for callouts (see 49 below)
1239    
1240     The flag bits are also defined in pcre.h, and are
1241    
1242     PCRE_EXTRA_STUDY_DATA
1243     PCRE_EXTRA_MATCH_LIMIT
1244     PCRE_EXTRA_CALLOUT_DATA
1245    
1246     The pcre_study() function now returns one of these new pcre_extra blocks, with
1247     the actual study data pointed to by the study_data field, and the
1248     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1249     before. That is, this change is entirely upwards-compatible and requires no
1250     change to existing code.
1251    
1252     If you want to pass in additional data to pcre_exec(), you can either place it
1253     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1254     block.
1255    
1256     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1257     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1258     times with different match limits, until it finds the minimum value needed for
1259     pcre_exec() to complete. The value is then output. This can be instructive; for
1260     most simple matches the number is quite small, but for pathological cases it
1261     gets very large very quickly.
1262    
1263     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1264     returns the size of the data block pointed to by the study_data field in a
1265     pcre_extra block, that is, the value that was passed as the argument to
1266     pcre_malloc() when PCRE was getting memory in which to place the information
1267     created by pcre_study(). The fourth argument should point to a size_t variable.
1268     pcretest has been extended so that this information is shown after a successful
1269     pcre_study() call when information about the compiled regex is being displayed.
1270    
1271     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1272     because what follows is always an absolute path. (Later: it turns out that this
1273     is more than cosmetic for MinGW, because it doesn't like empty path
1274     components.)
1275    
1276     49. Some changes have been made to the callout feature (see 28 above):
1277    
1278     (i) A callout function now has three choices for what it returns:
1279    
1280     0 => success, carry on matching
1281     > 0 => failure at this point, but backtrack if possible
1282     < 0 => serious error, return this value from pcre_exec()
1283    
1284     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1285     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1286     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1287     use by callout functions. It will never be used by PCRE itself.
1288    
1289     (ii) The pcre_extra structure (see 45 above) has a void * field called
1290     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1291     pcre_callout_block structure has a field of the same name. The contents of
1292     the field passed in the pcre_extra structure are passed to the callout
1293     function in the corresponding field in the callout block. This makes it
1294     easier to use the same callout-containing regex from multiple threads. For
1295     testing, the pcretest program has a new data escape
1296    
1297     \C*n pass the number n (may be negative) as callout_data
1298    
1299     If the callout function in pcretest receives a non-zero value as
1300     callout_data, it returns that value.
1301    
1302     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1303     there were some redundant $(CFLAGS) in commands that are now specified as
1304     $(LINK), which already includes $(CFLAGS).
1305    
1306     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1307     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1308     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1309     one-byte characters throughout. Note that case-insensitive matching applies
1310     only to characters whose values are less than 256. PCRE doesn't support the
1311     notion of cases for higher-valued characters.
1312    
1313     (i) A character class whose characters are all within 0-255 is handled as
1314     a bit map, and the map is inverted for negative classes. Previously, a
1315     character > 255 always failed to match such a class; however it should
1316     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1317    
1318     (ii) A negated character class with a single character < 255 is coded as
1319     "not this character" (OP_NOT). This wasn't working properly when the test
1320     character was multibyte, either singly or repeated.
1321    
1322     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1323     mode, for example: \x{100}{2,3}.
1324    
1325     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1326     singly or repeated) now correctly test multibyte characters. However,
1327     PCRE doesn't recognize any characters with values greater than 255 as
1328     digits, spaces, or word characters. Such characters always match \D, \S,
1329     and \W, and never match \d, \s, or \w.
1330    
1331     (v) Classes may now contain characters and character ranges with values
1332     greater than 255. For example: [ab\x{100}-\x{400}].
1333    
1334     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1335     PCRE in UTF-8 mode.
1336    
1337     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1338     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1339     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1340     value.)
1341    
1342     53. The single man page has become too large. I have therefore split it up into
1343     a number of separate man pages. These also give rise to individual HTML pages;
1344     these are now put in a separate directory, and there is an index.html page that
1345     lists them all. Some hyperlinking between the pages has been installed.
1346    
1347     54. Added convenience functions for handling named capturing parentheses.
1348    
1349     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1350     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1351     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1352     are faulted.
1353    
1354     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1355     calling configure. These values are used when compiling the dftables.c program
1356     which is run to generate the source of the default character tables. They
1357     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1358     you will need to set these values.
1359    
1360     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1361    
1362    
1363     Version 3.9 02-Jan-02
1364 nigel 61 ---------------------
1365    
1366     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1367    
1368     2. If --disable-static was given, the building process failed when trying to
1369     build pcretest and pcregrep. (For some reason it was using libtool to compile
1370     them, which is not right, as they aren't part of the library.)
1371    
1372    
1373 nigel 59 Version 3.8 18-Dec-01
1374     ---------------------
1375    
1376     1. The experimental UTF-8 code was completely screwed up. It was packing the
1377     bytes in the wrong order. How dumb can you get?
1378    
1379    
1380 nigel 57 Version 3.7 29-Oct-01
1381     ---------------------
1382    
1383     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1384     This caused pcretest, when used on the test data, to segfault. Unfortunately,
1385     this didn't happen under Solaris 8, where I normally test things.
1386    
1387 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1388     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1389     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1390     this fix an hour or so after the initial 3.7 release.)
1391 nigel 57
1392 nigel 59
1393 nigel 55 Version 3.6 23-Oct-01
1394     ---------------------
1395    
1396     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1397     offsets passed as NULL with zero offset count.
1398    
1399     2. The config.guess and config.sub files had not been updated when I moved to
1400     the latest autoconf.
1401    
1402    
1403 nigel 53 Version 3.5 15-Aug-01
1404     ---------------------
1405 nigel 5
1406 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1407     had been forgotten.
1408    
1409     2. By using declared but undefined structures, we can avoid using "void"
1410     definitions in pcre.h while keeping the internal definitions of the structures
1411     private.
1412    
1413     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1414     user point of view, this means that both static and shared libraries are built
1415     by default, but this can be individually controlled. More of the work of
1416     handling this static/shared cases is now inside libtool instead of PCRE's make
1417     file.
1418    
1419     4. The pcretest utility is now installed along with pcregrep because it is
1420     useful for users (to test regexs) and by doing this, it automatically gets
1421     relinked by libtool. The documentation has been turned into a man page, so
1422     there are now .1, .txt, and .html versions in /doc.
1423    
1424     5. Upgrades to pcregrep:
1425     (i) Added long-form option names like gnu grep.
1426     (ii) Added --help to list all options with an explanatory phrase.
1427     (iii) Added -r, --recursive to recurse into sub-directories.
1428     (iv) Added -f, --file to read patterns from a file.
1429    
1430     6. pcre_exec() was referring to its "code" argument before testing that
1431     argument for NULL (and giving an error if it was NULL).
1432    
1433     7. Upgraded Makefile.in to allow for compiling in a different directory from
1434     the source directory.
1435    
1436     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1437     options bits, the pointer it was passed was to an int instead of to an unsigned
1438     long int. This mattered only on 64-bit systems.
1439    
1440     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1441     generated) instead of pcre.in, which it its source. Also made the same change
1442     in several of the .c files.
1443    
1444     10. A new release of gcc defines printf() as a macro, which broke pcretest
1445     because it had an ifdef in the middle of a string argument for printf(). Fixed
1446     by using separate calls to printf().
1447    
1448     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1449     script, to force use of CR or LF instead of \n in the source. On non-Unix
1450     systems, the value can be set in config.h.
1451    
1452     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1453     absolute limit. Changed the text of the error message to make this clear, and
1454     likewise updated the man page.
1455    
1456     13. The limit of 99 on the number of capturing subpatterns has been removed.
1457     The new limit is 65535, which I hope will not be a "real" limit.
1458    
1459    
1460 nigel 51 Version 3.4 22-Aug-00
1461     ---------------------
1462    
1463     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1464    
1465     2. Diagnose condition (?(0) as an error instead of crashing on matching.
1466    
1467    
1468 nigel 49 Version 3.3 01-Aug-00
1469     ---------------------
1470    
1471     1. If an octal character was given, but the value was greater than \377, it
1472     was not getting masked to the least significant bits, as documented. This could
1473     lead to crashes in some systems.
1474    
1475     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1476     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1477    
1478     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1479     These just pass their arguments on to (pcre_free)(), but they are provided
1480     because some uses of PCRE bind it to non-C systems that can call its functions,
1481     but cannot call free() or pcre_free() directly.
1482    
1483     4. Add "make test" as a synonym for "make check". Corrected some comments in
1484     the Makefile.
1485    
1486     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1487     Makefile.
1488    
1489     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1490     command called pgrep for grepping around the active processes.
1491    
1492     7. Added the beginnings of support for UTF-8 character strings.
1493    
1494     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1495     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1496     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1497     out for the ar command.)
1498    
1499    
1500 nigel 47 Version 3.2 12-May-00
1501     ---------------------
1502    
1503     This is purely a bug fixing release.
1504    
1505     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1506     of ZA. This was just one example of several cases that could provoke this bug,
1507     which was introduced by change 9 of version 2.00. The code for breaking
1508     infinite loops after an iteration that matches an empty string was't working
1509     correctly.
1510    
1511     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1512     when matched against abbab (for example). After matching an empty string, it
1513     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1514     caused it to match further down the string than it should.
1515    
1516     3. The code contained an inclusion of sys/types.h. It isn't clear why this
1517     was there because it doesn't seem to be needed, and it causes trouble on some
1518     systems, as it is not a Standard C header. It has been removed.
1519    
1520     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1521     were reported on the Macintosh. The changes were from
1522    
1523     while ((c = *(++ptr)) != 0 && c != '\n');
1524     to
1525     while ((c = *(++ptr)) != 0 && c != '\n') ;
1526    
1527     Totally extraordinary, but if that's what it takes...
1528    
1529     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1530     available. Added HAVE_BCOPY and an autoconf test for it; if neither
1531     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1532     assumes the way PCRE uses memmove() (always moving upwards).
1533    
1534     6. PCRE is being used in one environment where strchr() is not available. There
1535     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1536     faster code anyway.
1537    
1538    
1539 nigel 45 Version 3.1 09-Feb-00
1540     ---------------------
1541    
1542     The only change in this release is the fixing of some bugs in Makefile.in for
1543     the "install" target:
1544    
1545     (1) It was failing to install pcreposix.h.
1546    
1547     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
1548    
1549    
1550 nigel 43 Version 3.0 01-Feb-00
1551     ---------------------
1552 nigel 41
1553     1. Add support for the /+ modifier to perltest (to output $` like it does in
1554     pcretest).
1555    
1556     2. Add support for the /g modifier to perltest.
1557    
1558     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
1559     matches null strings.
1560    
1561     4. Fix perltest so that it doesn't do unwanted things when fed an empty
1562     pattern. Perl treats empty patterns specially - it reuses the most recent
1563     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
1564     effect.
1565    
1566     5. The POSIX interface was broken in that it was just handing over the POSIX
1567     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
1568     required a bigger vector, with some working space on the end. This means that
1569     the POSIX wrapper now has to get and free some memory, and copy the results.
1570    
1571 nigel 43 6. Added some simple autoconf support, placing the test data and the
1572     documentation in separate directories, re-organizing some of the
1573     information files, and making it build pcre-config (a GNU standard). Also added
1574     libtool support for building PCRE as a shared library, which is now the
1575     default.
1576 nigel 41
1577 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
1578     09 are not valid octal constants. Single digits will be used for minor values
1579     less than 10.
1580    
1581     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
1582     existing programs that set these in the POSIX interface can use PCRE without
1583     modification.
1584    
1585     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
1586     return all that pcre_info() returns, plus additional data. The pcre_info()
1587     function is retained for compatibility, but is considered to be obsolete.
1588    
1589     10. Added experimental recursion feature (?R) to handle one common case that
1590     Perl 5.6 will be able to do with (?p{...}).
1591    
1592     11. Added support for POSIX character classes like [:alpha:], which Perl is
1593     adopting.
1594    
1595    
1596 nigel 39 Version 2.08 31-Aug-99
1597     ----------------------
1598    
1599     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
1600     trying to match at the startoffset position, but instead was moving forward to
1601     the next newline as if a previous match had failed.
1602    
1603     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
1604     and could get into a loop if a null string was matched other than at the start
1605     of the subject.
1606    
1607     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
1608     be distinguished at compile time, and for completeness also added PCRE_DATE.
1609    
1610     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
1611     in GnuWin32 environments.
1612    
1613    
1614 nigel 37 Version 2.07 29-Jul-99
1615     ----------------------
1616    
1617     1. The documentation is now supplied in plain text form and HTML as well as in
1618     the form of man page sources.
1619    
1620     2. C++ compilers don't like assigning (void *) values to other pointer types.
1621     In particular this affects malloc(). Although there is no problem in Standard
1622     C, I've put in casts to keep C++ compilers happy.
1623    
1624     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
1625     should be (const char *).
1626    
1627     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
1628     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
1629     However, I haven't made this a standard facility. The documentation doesn't
1630     mention it, and the Makefile doesn't support it.
1631    
1632     5. The Makefile now contains an "install" target, with editable destinations at
1633     the top of the file. The pcretest program is not installed.
1634    
1635     6. pgrep -V now gives the PCRE version number and date.
1636    
1637     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1638     causing the entire string to be ignored, instead of just the last character.
1639    
1640     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1641     non-matching string, it can take a very, very long time, even for strings of
1642     quite modest length, because of the nested recursion. PCRE now does better in
1643     some of these cases. It does this by remembering the last required literal
1644     character in the pattern, and pre-searching the subject to ensure it is present
1645     before running the real match. In other words, it applies a heuristic to detect
1646     some types of certain failure quickly, and in the above example, if presented
1647     with a string that has no trailing " it gives "no match" very quickly.
1648    
1649     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1650     other alternatives are tried instead.
1651    
1652    
1653 nigel 35 Version 2.06 09-Jun-99
1654     ----------------------
1655    
1656     1. Change pcretest's output for amount of store used to show just the code
1657     space, because the remainder (the data block) varies in size between 32-bit and
1658     64-bit systems.
1659    
1660     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1661     start matching at. This allows lookbehinds to work when searching for multiple
1662     occurrences in a string.
1663    
1664     3. Added additional options to pcretest for testing multiple occurrences:
1665    
1666     /+ outputs the rest of the string that follows a match
1667     /g loops for multiple occurrences, using the new startoffset argument
1668     /G loops for multiple occurrences by passing an incremented pointer
1669    
1670     4. PCRE wasn't doing the "first character" optimization for patterns starting
1671     with \b or \B, though it was doing it for other lookbehind assertions. That is,
1672     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1673     the letter 'x'. On long subject strings, this gives a significant speed-up.
1674    
1675    
1676 nigel 33 Version 2.05 21-Apr-99
1677     ----------------------
1678    
1679     1. Changed the type of magic_number from int to long int so that it works
1680     properly on 16-bit systems.
1681    
1682     2. Fixed a bug which caused patterns starting with .* not to work correctly
1683     when the subject string contained newline characters. PCRE was assuming
1684     anchoring for such patterns in all cases, which is not correct because .* will
1685     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1686     DOTALL is set at top level; otherwise it knows that patterns starting with .*
1687     must be retried after every newline in the subject.
1688    
1689    
1690 nigel 31 Version 2.04 18-Feb-99
1691     ----------------------
1692    
1693     1. For parenthesized subpatterns with repeats whose minimum was zero, the
1694     computation of the store needed to hold the pattern was incorrect (too large).
1695     If such patterns were nested a few deep, this could multiply and become a real
1696     problem.
1697    
1698     2. Added /M option to pcretest to show the memory requirement of a specific
1699     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1700    
1701     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1702     compiled in such a way that the backtracking after subsequent failure was
1703     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1704     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1705    
1706    
1707 nigel 29 Version 2.03 02-Feb-99
1708     ----------------------
1709    
1710     1. Fixed typo and small mistake in man page.
1711    
1712 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
1713     LICENCE file containing the conditions.
1714 nigel 29
1715     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1716     Perl, that is the internal \ allows the delimiter to be included in the
1717     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1718     the final delimiter, add \ to the end of the pattern (to test the error).
1719    
1720     4. Added the convenience functions for extracting substrings after a successful
1721     match. Updated pcretest to make it able to test these functions.
1722    
1723    
1724 nigel 27 Version 2.02 14-Jan-99
1725     ----------------------
1726    
1727     1. Initialized the working variables associated with each extraction so that
1728     their saving and restoring doesn't refer to uninitialized store.
1729    
1730     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1731     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1732     fix the problem.
1733    
1734     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1735     calls, and wasn't printing the correct value for compiling calls. Increased the
1736     default value of LOOPREPEAT, and the number of significant figures in the
1737     times.
1738    
1739     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1740    
1741     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1742     a building problem on Windows NT with a FAT file system.
1743    
1744    
1745 nigel 25 Version 2.01 21-Oct-98
1746     ----------------------
1747    
1748     1. Changed the API for pcre_compile() to allow for the provision of a pointer
1749     to character tables built by pcre_maketables() in the current locale. If NULL
1750     is passed, the default tables are used.
1751    
1752    
1753 nigel 23 Version 2.00 24-Sep-98
1754 nigel 21 ----------------------
1755    
1756 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1757     it any more.
1758    
1759     2. Allow quantification of (?>) groups, and make it work correctly.
1760    
1761     3. The first character computation wasn't working for (?>) groups.
1762    
1763     4. Correct the implementation of \Z (it is permitted to match on the \n at the
1764     end of the subject) and add 5.005's \z, which really does match only at the
1765     very end of the subject.
1766    
1767     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1768    
1769     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1770     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1771     localized options. All options to pcre_study() were also removed.
1772    
1773     7. Add other new features from 5.005:
1774    
1775     $(?<= positive lookbehind
1776     $(?<! negative lookbehind
1777     (?imsx-imsx) added the unsetting capability
1778     such a setting is global if at outer level; local otherwise
1779     (?imsx-imsx:) non-capturing groups with option setting
1780     (?(cond)re|re) conditional pattern matching
1781    
1782     A backreference to itself in a repeated group matches the previous
1783     captured string.
1784    
1785     8. General tidying up of studying (both automatic and via "study")
1786     consequential on the addition of new assertions.
1787    
1788     9. As in 5.005, unlimited repeated groups that could match an empty substring
1789     are no longer faulted at compile time. Instead, the loop is forcibly broken at
1790     runtime if any iteration does actually match an empty substring.
1791    
1792     10. Include the RunTest script in the distribution.
1793    
1794     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1795     discrepancies, some of which were old and were also with respect to 5.004. They
1796     have now been fixed.
1797    
1798    
1799     Version 1.09 28-Apr-98
1800     ----------------------
1801    
1802 nigel 21 1. A negated single character class followed by a quantifier with a minimum
1803     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
1804     program crashes, or just wrong answers. This did not apply to negated classes
1805     containing more than one character, or to minima other than one.
1806    
1807    
1808 nigel 19 Version 1.08 27-Mar-98
1809     ----------------------
1810    
1811     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
1812    
1813     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
1814     latter must appear before anything that relies on it in the pattern.
1815    
1816    
1817 nigel 17 Version 1.07 16-Feb-98
1818     ----------------------
1819    
1820     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
1821     repeat of a potentially empty string).
1822    
1823    
1824 nigel 15 Version 1.06 23-Jan-98
1825     ----------------------
1826    
1827     1. Added Markus Oberhumer's little patches for C++.
1828    
1829     2. Literal strings longer than 255 characters were broken.
1830    
1831    
1832 nigel 13 Version 1.05 23-Dec-97
1833     ----------------------
1834    
1835     1. Negated character classes containing more than one character were failing if
1836     PCRE_CASELESS was set at run time.
1837    
1838    
1839 nigel 11 Version 1.04 19-Dec-97
1840     ----------------------
1841    
1842     1. Corrected the man page, where some "const" qualifiers had been omitted.
1843    
1844     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
1845     input syntax.
1846    
1847     3. Fixed memory leak which occurred when a regex with back references was
1848     matched with an offsets vector that wasn't big enough. The temporary memory
1849     that is used in this case wasn't being freed if the match failed.
1850    
1851     4. Tidied pcretest to ensure it frees memory that it gets.
1852    
1853     5. Temporary memory was being obtained in the case where the passed offsets
1854     vector was exactly big enough.
1855    
1856     6. Corrected definition of offsetof() from change 5 below.
1857    
1858     7. I had screwed up change 6 below and broken the rules for the use of
1859     setjmp(). Now fixed.
1860    
1861    
1862 nigel 9 Version 1.03 18-Dec-97
1863     ----------------------
1864    
1865     1. A erroneous regex with a missing opening parenthesis was correctly
1866     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
1867     on some systems.
1868    
1869     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
1870     it was reported that one broken compiler failed on the former because "code" is
1871     also an independent variable.
1872    
1873     3. The erroneous regex a[]b caused an array overrun reference.
1874    
1875     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
1876     fail on data ending with that character. (It was going on too far, and checking
1877     the next character, typically a binary zero.) This was specific to the
1878     optimized code for single-character negative classes.
1879    
1880     5. Added a contributed patch from the TIN world which does the following:
1881    
1882     + Add an undef for memmove, in case the the system defines a macro for it.
1883    
1884     + Add a definition of offsetof(), in case there isn't one. (I don't know
1885     the reason behind this - offsetof() is part of the ANSI standard - but
1886     it does no harm).
1887    
1888     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
1889     most of the places where whitespace preceded '#'. I have given up and
1890     allowed the remaining 2 cases to be at the margin.
1891    
1892     + Rename some variables in pcre to eliminate shadowing. This seems very
1893     pedantic, but does no harm, of course.
1894    
1895     6. Moved the call to setjmp() into its own function, to get rid of warnings
1896     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
1897    
1898     7. Constructs such as \d{8,} were compiling into the equivalent of
1899 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
1900 nigel 9 outcome, but in this particular case used more store than had been allocated,
1901     which caused the bug to be discovered because it threw up an internal error.
1902    
1903     8. The debugging code in both pcre and pcretest for outputting the compiled
1904     form of a regex was going wrong in the case of back references followed by
1905     curly-bracketed repeats.
1906    
1907    
1908 nigel 7 Version 1.02 12-Dec-97
1909     ----------------------
1910    
1911     1. Typos in pcre.3 and comments in the source fixed.
1912    
1913     2. Applied a contributed patch to get rid of places where it used to remove
1914     'const' from variables, and fixed some signed/unsigned and uninitialized
1915     variable warnings.
1916    
1917     3. Added the "runtest" target to Makefile.
1918    
1919     4. Set default compiler flag to -O2 rather than just -O.
1920    
1921    
1922 nigel 5 Version 1.01 19-Nov-97
1923     ----------------------
1924    
1925     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
1926     like /([ab]*)*/, that is, for classes with more than one character in them.
1927    
1928     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
1929     as /((?>a*))*/ (a PCRE_EXTRA facility).
1930    
1931    
1932     Version 1.00 18-Nov-97
1933     ----------------------
1934    
1935     1. Added compile-time macros to support systems such as SunOS4 which don't have
1936     memmove() or strerror() but have other things that can be used instead.
1937    
1938     2. Arranged that "make clean" removes the executables.
1939    
1940    
1941 nigel 3 Version 0.99 27-Oct-97
1942     ----------------------
1943    
1944     1. Fixed bug in code for optimizing classes with only one character. It was
1945     initializing a 32-byte map regardless, which could cause it to run off the end
1946     of the memory it had got.
1947    
1948     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1949    
1950    
1951     Version 0.98 22-Oct-97
1952     ----------------------
1953    
1954     1. Fixed bug in code for handling temporary memory usage when there are more
1955     back references than supplied space in the ovector. This could cause segfaults.
1956    
1957    
1958     Version 0.97 21-Oct-97
1959     ----------------------
1960    
1961     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1962    
1963     2. Optimized negated single characters not to use a bit map.
1964    
1965     3. Brought error texts together as macro definitions; clarified some of them;
1966     fixed one that was wrong - it said "range out of order" when it meant "invalid
1967     escape sequence".
1968    
1969     4. Changed some char * arguments to const char *.
1970    
1971     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1972    
1973     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1974     pcretest.
1975    
1976    
1977     Version 0.96 16-Oct-97
1978     ----------------------
1979    
1980     1. Added a simple "pgrep" utility to the distribution.
1981    
1982     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1983     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1984     where "ddd" means "one or more decimal digits".
1985    
1986     3. Fixed serious bug. If a pattern had a back reference, but the call to
1987     pcre_exec() didn't supply a large enough ovector to record the related
1988     identifying subpattern, the match always failed. PCRE now remembers the number
1989     of the largest back reference, and gets some temporary memory in which to save
1990     the offsets during matching if necessary, in order to ensure that
1991     backreferences always work.
1992    
1993     4. Increased the compatibility with Perl in a number of ways:
1994    
1995     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1996     to request this handling. The option can be set at compile or exec time.
1997    
1998     (b) $ matches before a terminating newline by default; an option
1999     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2000     mode). The option can be set at compile or exec time.
2001    
2002     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2003     the same as Perl's. If the decimal number it represents is less than 10
2004     or there aren't that many previous left capturing parentheses, an octal
2005     escape is read. Inside a character class, it's always an octal escape,
2006     even if it is a single digit.
2007    
2008     (d) An escaped but undefined alphabetic character is taken as a literal,
2009     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2010     escapes.
2011    
2012     (e) {0} is now permitted. (The previous item is removed from the compiled
2013     pattern).
2014    
2015     5. Changed all the names of code files so that the basic parts are no longer
2016     than 10 characters, and abolished the teeny "globals.c" file.
2017    
2018     6. Changed the handling of character classes; they are now done with a 32-byte
2019     bit map always.
2020    
2021     7. Added the -d and /D options to pcretest to make it possible to look at the
2022     internals of compilation without having to recompile pcre.
2023    
2024    
2025     Version 0.95 23-Sep-97
2026     ----------------------
2027    
2028     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2029     \x20 at the start of a run of normal characters. These were being treated as
2030     real characters, instead of the source characters being re-checked.
2031    
2032    
2033     Version 0.94 18-Sep-97
2034     ----------------------
2035    
2036     1. The functions are now thread-safe, with the caveat that the global variables
2037     containing pointers to malloc() and free() or alternative functions are the
2038     same for all threads.
2039    
2040     2. Get pcre_study() to generate a bitmap of initial characters for non-
2041     anchored patterns when this is possible, and use it if passed to pcre_exec().
2042    
2043    
2044     Version 0.93 15-Sep-97
2045     ----------------------
2046    
2047     1. /(b)|(:+)/ was computing an incorrect first character.
2048    
2049     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2050     but not actually doing anything yet.
2051    
2052     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2053     as Perl does (e.g. [-az] or [az-]).
2054    
2055     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2056     all possible positions.
2057    
2058     5. Split up into different modules to avoid including unneeded functions in a
2059     compiled binary. However, compile and exec are still in one module. The "study"
2060     function is split off.
2061    
2062     6. The character tables are now in a separate module whose source is generated
2063     by an auxiliary program - but can then be edited by hand if required. There are
2064     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2065     toupper() in the code.
2066    
2067     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2068     make them global. Abolish the function for setting them, as the caller can now
2069     set them directly.
2070    
2071    
2072     Version 0.92 11-Sep-97
2073     ----------------------
2074    
2075     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2076     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2077    
2078     2. Caseless matching was not working in character classes if the characters in
2079     the pattern were in upper case.
2080    
2081     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2082    
2083     4. Make PCRE_ANCHORED public and accept as a compile option.
2084    
2085     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2086     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2087     pass them.
2088    
2089     6. Give an error if bad option bits passed at compile or run time.
2090    
2091     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2092     pcretest to cause it to pass that flag.
2093    
2094     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2095     options, and the first character, if set.
2096    
2097     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2098    
2099    
2100     Version 0.91 10-Sep-97
2101     ----------------------
2102    
2103     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2104     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2105    
2106     2. PCRE was looping on encountering an indefinitely repeated back reference to
2107     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2108     Perl does - treats the match as successful.
2109    
2110     ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12