ViewVC logotype

Contents of /code/tags/pcre-6.0/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 78 - (show annotations)
Sat Feb 24 21:40:47 2007 UTC (13 years, 10 months ago) by nigel
File size: 78753 byte(s)
Tag code/trunk as code/tags/pcre-6.0.
1 ChangeLog for PCRE
2 ------------------
4 Version 6.0 07-Jun-05
5 ---------------------
7 1. Some minor internal re-organization to help with my DFA experiments.
9 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
10 didn't matter for the library itself when fully configured, but did matter
11 when compiling without UCP support, or within Exim, where the ucp files are
12 not imported.
14 3. Refactoring of the library code to split up the various functions into
15 different source modules. The addition of the new DFA matching code (see
16 below) to a single monolithic source would have made it really too
17 unwieldy, quite apart from causing all the code to be include in a
18 statically linked application, when only some functions are used. This is
19 relevant even without the DFA addition now that patterns can be compiled in
20 one application and matched in another.
22 The downside of splitting up is that there have to be some external
23 functions and data tables that are used internally in different modules of
24 the library but which are not part of the API. These have all had their
25 names changed to start with "_pcre_" so that they are unlikely to clash
26 with other external names.
28 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
29 a different (DFA) algorithm. Although it is slower than the original
30 function, it does have some advantages for certain types of matching
31 problem.
33 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
34 including restarting after a partial match.
36 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
37 defined when compiling for Windows was sent to me. I have put it into the
38 code, though I have no means of testing or verifying it.
40 7. Added the pcre_refcount() auxiliary function.
42 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
43 match before or at the first newline in the subject string. In pcretest,
44 the /f option on a pattern can be used to set this.
46 9. A repeated \w when used in UTF-8 mode with characters greater than 256
47 would behave wrongly. This has been present in PCRE since release 4.0.
49 10. A number of changes to the pcregrep command:
51 (a) Refactored how -x works; insert ^(...)$ instead of setting
52 PCRE_ANCHORED and checking the length, in preparation for adding
53 something similar for -w.
55 (b) Added the -w (match as a word) option.
57 (c) Refactored the way lines are read and buffered so as to have more
58 than one at a time available.
60 (d) Implemented a pcregrep test script.
62 (e) Added the -M (multiline match) option. This allows patterns to match
63 over several lines of the subject. The buffering ensures that at least
64 8K, or the rest of the document (whichever is the shorter) is available
65 for matching (and similarly the previous 8K for lookbehind assertions).
67 (f) Changed the --help output so that it now says
69 -w, --word-regex(p)
71 instead of two lines, one with "regex" and the other with "regexp"
72 because that confused at least one person since the short forms are the
73 same. (This required a bit of code, as the output is generated
74 automatically from a table. It wasn't just a text change.)
76 (g) -- can be used to terminate pcregrep options if the next thing isn't an
77 option but starts with a hyphen. Could be a pattern or a path name
78 starting with a hyphen, for instance.
80 (h) "-" can be given as a file name to represent stdin.
82 (i) When file names are being printed, "(standard input)" is used for
83 the standard input, for compatibility with GNU grep. Previously
84 "<stdin>" was used.
86 (j) The option --label=xxx can be used to supply a name to be used for
87 stdin when file names are being printed. There is no short form.
89 (k) Re-factored the options decoding logic because we are going to add
90 two more options that take data. Such options can now be given in four
91 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
93 (l) Added the -A, -B, and -C options for requesting that lines of context
94 around matches be printed.
96 (m) Added the -L option to print the names of files that do not contain
97 any matching lines, that is, the complement of -l.
99 (n) The return code is 2 if any file cannot be opened, but pcregrep does
100 continue to scan other files.
102 (o) The -s option was incorrectly implemented. For compatibility with other
103 greps, it now suppresses the error message for a non-existent or non-
104 accessible file (but not the return code). There is a new option called
105 -q that suppresses the output of matching lines, which was what -s was
106 previously doing.
108 (p) Added --include and --exclude options to specify files for inclusion
109 and exclusion when recursing.
111 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
112 Hopefully, it now does.
114 12. Missing cast in pcre_study().
116 13. Added an "uninstall" target to the makefile.
118 14. Replaced "extern" in the function prototypes in Makefile.in with
119 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
120 world, but is set differently for Windows.
122 15. Added a second compiling function called pcre_compile2(). The only
123 difference is that it has an extra argument, which is a pointer to an
124 integer error code. When there is a compile-time failure, this is set
125 non-zero, in addition to the error test pointer being set to point to an
126 error message. The new argument may be NULL if no error number is required
127 (but then you may as well call pcre_compile(), which is now just a
128 wrapper). This facility is provided because some applications need a
129 numeric error indication, but it has also enabled me to tidy up the way
130 compile-time errors are handled in the POSIX wrapper.
132 16. Added VPATH=.libs to the makefile; this should help when building with one
133 prefix path and installing with another. (Or so I'm told by someone who
134 knows more about this stuff than I do.)
136 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
137 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
138 match everything, including newlines. This is not POSIX-compatible, but
139 somebody wanted the feature. From pcretest it can be activated by using
140 both the P and the s flags.
142 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
144 19. libpcre.pc was being incorrectly installed as executable.
146 20. A couple of places in pcretest check for end-of-line by looking for '\n';
147 it now also looks for '\r' so that it will work unmodified on Windows.
149 21. Added Google's contributed C++ wrapper to the distribution.
151 22. Added some untidy missing memory free() calls in pcretest, to keep
152 Electric Fence happy when testing.
156 Version 5.0 13-Sep-04
157 ---------------------
159 1. Internal change: literal characters are no longer packed up into items
160 containing multiple characters in a single byte-string. Each character
161 is now matched using a separate opcode. However, there may be more than one
162 byte in the character in UTF-8 mode.
164 2. The pcre_callout_block structure has two new fields: pattern_position and
165 next_item_length. These contain the offset in the pattern to the next match
166 item, and its length, respectively.
168 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
169 insertion of callouts before each pattern item. Added the /C option to
170 pcretest to make use of this.
172 4. On the advice of a Windows user, the lines
174 #if defined(_WIN32) || defined(WIN32)
175 _setmode( _fileno( stdout ), 0x8000 );
176 #endif /* defined(_WIN32) || defined(WIN32) */
178 have been added to the source of pcretest. This apparently does useful
179 magic in relation to line terminators.
181 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
182 for the benefit of those environments where the "b" makes a difference.
184 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
185 to know about it. I have put a hack into configure.in that adds in code
186 to set GCC=yes if CC=icc. This seems to end up at a point in the
187 generated configure script that is early enough to affect the setting of
188 compiler options, which is what is needed, but I have no means of testing
189 whether it really works. (The user who reported this had patched the
190 generated configure script, which of course I cannot do.)
192 LATER: After change 22 below (new libtool files), the configure script
193 seems to know about icc (and also ecc). Therefore, I have commented out
194 this hack in configure.in.
196 7. Added support for pkg-config (2 patches were sent in).
198 8. Negated POSIX character classes that used a combination of internal tables
199 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
200 [[:^ascii]]. Typically, they would match almost any characters. The other
201 POSIX classes were not broken in this way.
203 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
204 to find the match, as PCRE was deluded into thinking that the match had to
205 start at the start point or following a newline. The same bug applied to
206 patterns with negative forward assertions or any backward assertions
207 preceding ".*" at the start, unless the pattern required a fixed first
208 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
210 10. In UTF-8 mode, when moving forwards in the subject after a failed match
211 starting at the last subject character, bytes beyond the end of the subject
212 string were read.
214 11. Renamed the variable "class" as "classbits" to make life easier for C++
215 users. (Previously there was a macro definition, but it apparently wasn't
216 enough.)
218 12. Added the new field "tables" to the extra data so that tables can be passed
219 in at exec time, or the internal tables can be re-selected. This allows
220 a compiled regex to be saved and re-used at a later time by a different
221 program that might have everything at different addresses.
223 13. Modified the pcre-config script so that, when run on Solaris, it shows a
224 -R library as well as a -L library.
226 14. The debugging options of pcretest (-d on the command line or D on a
227 pattern) showed incorrect output for anything following an extended class
228 that contained multibyte characters and which was followed by a quantifier.
230 15. Added optional support for general category Unicode character properties
231 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
232 support. It adds about 90K to the size of the library. The meanings of the
233 inbuilt class escapes such as \d and \s have NOT been changed.
235 16. Updated pcredemo.c to include calls to free() to release the memory for the
236 compiled pattern.
238 17. The generated file chartables.c was being created in the source directory
239 instead of in the building directory. This caused the build to fail if the
240 source directory was different from the building directory, and was
241 read-only.
243 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
244 file. No doubt somebody will tell me if they don't make sense... Also added
245 Dan Mooney's comments about building on OpenVMS.
247 19. Added support for partial matching via the PCRE_PARTIAL option for
248 pcre_exec() and the \P data escape in pcretest.
250 20. Extended pcretest with 3 new pattern features:
252 (i) A pattern option of the form ">rest-of-line" causes pcretest to
253 write the compiled pattern to the file whose name is "rest-of-line".
254 This is a straight binary dump of the data, with the saved pointer to
255 the character tables forced to be NULL. The study data, if any, is
256 written too. After writing, pcretest reads a new pattern.
258 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
259 compiled pattern from the given file. There must not be any
260 occurrences of "<" in the file name (pretty unlikely); if there are,
261 pcretest will instead treat the initial "<" as a pattern delimiter.
262 After reading in the pattern, pcretest goes on to read data lines as
263 usual.
265 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
266 and 16-bit fields in a compiled pattern, to simulate a pattern that
267 was compiled on a host of opposite endianness.
269 21. The pcre-exec() function can now cope with patterns that were compiled on
270 hosts of opposite endianness, with this restriction:
272 As for any compiled expression that is saved and used later, the tables
273 pointer field cannot be preserved; the extra_data field in the arguments
274 to pcre_exec() should be used to pass in a tables address if a value
275 other than the default internal tables were used at compile time.
277 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
278 now diagnosed as an error. Previously, most of the time, a negative number
279 would have been treated as zero, but if in addition "ovector" was passed as
280 NULL, a crash could occur.
282 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
283 new versions from the libtool 1.5 distribution (the last one is a copy of
284 a file called libtool.m4). This seems to have fixed the need to patch
285 "configure" to support Darwin 1.3 (which I used to do). However, I still
286 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
287 workstation).
289 24. Changed the PCRE licence to be the more standard "BSD" licence.
292 Version 4.5 01-Dec-03
293 ---------------------
295 1. There has been some re-arrangement of the code for the match() function so
296 that it can be compiled in a version that does not call itself recursively.
297 Instead, it keeps those local variables that need separate instances for
298 each "recursion" in a frame on the heap, and gets/frees frames whenever it
299 needs to "recurse". Keeping track of where control must go is done by means
300 of setjmp/longjmp. The whole thing is implemented by a set of macros that
301 hide most of the details from the main code, and operates only if
302 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
303 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
304 operating.
306 To make it easier for callers to provide specially tailored get/free
307 functions for this usage, two new functions, pcre_stack_malloc, and
308 pcre_stack_free, are used. They are always called in strict stacking order,
309 and the size of block requested is always the same.
311 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
312 PCRE has been compiled to use the stack or the heap for recursion. The
313 -C option of pcretest uses this to show which version is compiled.
315 A new data escape \S, is added to pcretest; it causes the amounts of store
316 obtained and freed by both kinds of malloc/free at match time to be added
317 to the output.
319 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
320 what's available on my current Linux desktop machine.
322 3. When matching a UTF-8 string, the test for a valid string at the start has
323 been extended. If start_offset is not zero, PCRE now checks that it points
324 to a byte that is the start of a UTF-8 character. If not, it returns
325 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
326 this is necessary because there may be backward assertions in the pattern.
327 When matching the same subject several times, it may save resources to use
328 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
330 4. The code for checking the validity of UTF-8 strings has been tightened so
331 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
332 containing "overlong sequences".
334 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
335 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
336 should have been "&", but it just so happened that all the cases this let
337 through by mistake were picked up later in the function.
339 6. I had used a variable called "isblank" - this is a C99 function, causing
340 some compilers to warn. To avoid this, I renamed it (as "blankclass").
342 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
343 prompting; (b) run "./pcretest /dev/null" at the start of the test script
344 so the version is shown; (c) stop "make test" echoing "./RunTest".
346 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
348 9. The prototype for memmove() for systems that don't have it was using
349 size_t, but the inclusion of the header that defines size_t was later. I've
350 moved the #includes for the C headers earlier to avoid this.
352 10. Added some adjustments to the code to make it easier to compiler on certain
353 special systems:
355 (a) Some "const" qualifiers were missing.
356 (b) Added the macro EXPORT before all exported functions; by default this
357 is defined to be empty.
358 (c) Changed the dftables auxiliary program (that builds chartables.c) so
359 that it reads its output file name as an argument instead of writing
360 to the standard output and assuming this can be redirected.
362 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
363 class containing characters with values greater than 255, PCRE compilation
364 went into a loop.
366 12. A recursive reference to a subpattern that was within another subpattern
367 that had a minimum quantifier of zero caused PCRE to crash. For example,
368 (x(y(?2))z)? provoked this bug with a subject that got as far as the
369 recursion. If the recursively-called subpattern itself had a zero repeat,
370 that was OK.
372 13. In pcretest, the buffer for reading a data line was set at 30K, but the
373 buffer into which it was copied (for escape processing) was still set at
374 1024, so long lines caused crashes.
376 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
377 "internal error: code overflow...". This applied to any character class
378 that was followed by a possessive quantifier.
380 15. Modified the Makefile to add libpcre.la as a prerequisite for
381 libpcreposix.la because I was told this is needed for a parallel build to
382 work.
384 16. If a pattern that contained .* following optional items at the start was
385 studied, the wrong optimizing data was generated, leading to matching
386 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
387 matching string must start with a or b or c. The correct conclusion for
388 this pattern is that a match can start with any character.
391 Version 4.4 13-Aug-03
392 ---------------------
394 1. In UTF-8 mode, a character class containing characters with values between
395 127 and 255 was not handled correctly if the compiled pattern was studied.
396 In fixing this, I have also improved the studying algorithm for such
397 classes (slightly).
399 2. Three internal functions had redundant arguments passed to them. Removal
400 might give a very teeny performance improvement.
402 3. Documentation bug: the value of the capture_top field in a callout is *one
403 more than* the number of the hightest numbered captured substring.
405 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
406 in incorrectly linking with a previously installed version. They now link
407 explicitly with libpcre.la.
409 5. configure.in no longer needs to recognize Cygwin specially.
411 6. A problem in pcre.in for Windows platforms is fixed.
413 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
414 pcretest, it used to include the size of the study block as part of its
415 output. Unfortunately, the structure contains a field that has a different
416 size on different hardware architectures. This meant that the tests that
417 showed this size failed. As the block is currently always of a fixed size,
418 this information isn't actually particularly useful in pcretest output, so
419 I have just removed it.
421 8. Three pre-processor statements accidentally did not start in column 1.
422 Sadly, there are *still* compilers around that complain, even though
423 standard C has not required this for well over a decade. Sigh.
425 9. In pcretest, the code for checking callouts passed small integers in the
426 callout_data field, which is a void * field. However, some picky compilers
427 complained about the casts involved for this on 64-bit systems. Now
428 pcretest passes the address of the small integer instead, which should get
429 rid of the warnings.
431 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
432 both compile and run time, and gives an error if an invalid UTF-8 sequence
433 is found. There is a option for disabling this check in cases where the
434 string is known to be correct and/or the maximum performance is wanted.
436 11. In response to a bug report, I changed one line in Makefile.in from
438 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
439 to
440 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
442 to look similar to other lines, but I have no way of telling whether this
443 is the right thing to do, as I do not use Windows. No doubt I'll get told
444 if it's wrong...
447 Version 4.3 21-May-03
448 ---------------------
450 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
451 Makefile.
453 2. Some refactoring to improve the quality of the code:
455 (i) The utf8_table... variables are now declared "const".
457 (ii) The code for \cx, which used the "case flipping" table to upper case
458 lower case letters, now just substracts 32. This is ASCII-specific,
459 but the whole concept of \cx is ASCII-specific, so it seems
460 reasonable.
462 (iii) PCRE was using its character types table to recognize decimal and
463 hexadecimal digits in the pattern. This is silly, because it handles
464 only 0-9, a-f, and A-F, but the character types table is locale-
465 specific, which means strange things might happen. A private
466 table is now used for this - though it costs 256 bytes, a table is
467 much faster than multiple explicit tests. Of course, the standard
468 character types table is still used for matching digits in subject
469 strings against \d.
471 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
472 ending in _t are). So I've renamed it as ESC_tee.
474 3. The first argument for regexec() in the POSIX wrapper should have been
475 defined as "const".
477 4. Changed pcretest to use malloc() for its buffers so that they can be
478 Electric Fenced for debugging.
480 5. There were several places in the code where, in UTF-8 mode, PCRE would try
481 to read one or more bytes before the start of the subject string. Often this
482 had no effect on PCRE's behaviour, but in some circumstances it could
483 provoke a segmentation fault.
485 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
486 to try to read one or more bytes before the start of the subject string.
488 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
489 UTF-8 support could misbehave in various ways if the subject string
490 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
491 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
492 back over UTF-8 characters.)
495 Version 4.2 14-Apr-03
496 ---------------------
498 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
500 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
501 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
502 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
503 [WIN_PREFIX]: new variable, "cyg" for Cygwin
504 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
506 Note: automatic setting of the BUILD variables is not yet working
507 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
508 compile-time but not at link-time
509 [LINK]: use for linking executables only
510 make different versions for Windows and non-Windows
511 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
512 libraries
513 [LINK_FOR_BUILD]: new variable
514 [OBJEXT]: use throughout
515 [EXEEXT]: use throughout
516 <winshared>: new target
517 <wininstall>: new target
518 <dftables.o>: use native compiler
519 <dftables>: use native linker
520 <install>: handle Windows platform correctly
521 <clean>: ditto
522 <check>: ditto
523 copy DLL to top builddir before testing
525 As part of these changes, -no-undefined was removed again. This was reported
526 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
527 in any case.
529 3. Some tidies to get rid of compiler warnings:
531 . In the match_data structure, match_limit was an unsigned long int, whereas
532 match_call_count was an int. I've made them both unsigned long ints.
534 . In pcretest the fact that a const uschar * doesn't automatically cast to
535 a void * provoked a warning.
537 . Turning on some more compiler warnings threw up some "shadow" variables
538 and a few more missing casts.
540 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
541 option, a class that contained a single character with a value between 128
542 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
544 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
545 option, a class that contained several characters, but with at least one
546 whose value was between 128 and 255 caused PCRE to crash.
549 Version 4.1 12-Mar-03
550 ---------------------
552 1. Compiling with gcc -pedantic found a couple of places where casts were
553 needed, and a string in dftables.c that was longer than standard compilers are
554 required to support.
556 2. Compiling with Sun's compiler found a few more places where the code could
557 be tidied up in order to avoid warnings.
559 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
560 first of these names is deprecated in the latest Autoconf in favour of the name
561 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
562 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
563 analogy I have changed it to CFLAGS_FOR_BUILD.
565 4. Added -no-undefined to the linking command in the Makefile, because this is
566 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
567 linking step for the pcreposix library.
569 5. PCRE was failing to diagnose the case of two named groups with the same
570 name.
572 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
573 literal character that is needed in the subject for a match, and scans along to
574 ensure that it is present before embarking on the full matching process. This
575 saves time in cases of nested unlimited repeats that are never going to match.
576 Problem: the scan can take a lot of time if the subject is very long (e.g.
577 megabytes), thus penalizing straightforward matches. It is now done only if the
578 amount of subject to be scanned is less than 1000 bytes.
580 7. A lesser problem with the same optimization is that it was recording the
581 first character of an anchored pattern as "needed", thus provoking a search
582 right along the subject, even when the first match of the pattern was going to
583 fail. The "needed" character is now not set for anchored patterns, unless it
584 follows something in the pattern that is of non-fixed length. Thus, it still
585 fulfils its original purpose of finding quick non-matches in cases of nested
586 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
589 Version 4.0 17-Feb-03
590 ---------------------
592 1. If a comment in an extended regex that started immediately after a meta-item
593 extended to the end of string, PCRE compiled incorrect data. This could lead to
594 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
596 2. Moved to autoconf 2.53 and libtool 1.4.2.
598 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
599 the special perltest8 script is no longer needed - all the tests can be run
600 from a single perltest script.
602 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
603 by \s. It has now been removed in PCRE. This means it isn't recognized as
604 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
605 class [:space:] *does* include VT, thereby creating a mess.
607 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
608 space and tab.
610 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
611 its new features into the main test script, reducing the number of scripts.
613 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
614 were backward compatible, and made the (?i) apply to the whole pattern, as if
615 /i were given. Now it behaves more logically, and applies the option setting
616 only to what follows. PCRE has been changed to follow suit. However, if it
617 finds options settings right at the start of the pattern, it extracts them into
618 the global options, as before. Thus, they show up in the info data.
620 8. Added support for the \Q...\E escape sequence. Characters in between are
621 treated as literals. This is slightly different from Perl in that $ and @ are
622 also handled as literals inside the quotes. In Perl, they will cause variable
623 interpolation. Note the following examples:
625 Pattern PCRE matches Perl matches
627 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
628 \Qabc\$xyz\E abc\$xyz abc\$xyz
629 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
631 For compatibility with Perl, \Q...\E sequences are recognized inside character
632 classes as well as outside them.
634 9. Re-organized 3 code statements in pcretest to avoid "overflow in
635 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
636 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
637 signed/unsigned warnings.
639 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
640 option for pcretest, so I've replaced it by a simple function that does just
641 that job.
643 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
644 "pcregrep -".
646 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
647 Java package. This provides some syntactic sugar for simple cases of what my
648 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
649 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
650 item, you can use this simplified notation. Note that only makes sense with
651 greedy quantifiers. Consequently, the use of the possessive quantifier forces
652 greediness, whatever the setting of the PCRE_UNGREEDY option.
654 13. A change of greediness default within a pattern was not taking effect at
655 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
656 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
657 was abstracted outside.
659 14. PCRE now supports the \G assertion. It is true when the current matching
660 position is at the start point of the match. This differs from \A when the
661 starting offset is non-zero. Used with the /g option of pcretest (or similar
662 code), it works in the same way as it does for Perl's /g option. If all
663 alternatives of a regex begin with \G, the expression is anchored to the start
664 match position, and the "anchored" flag is set in the compiled expression.
666 15. Some bugs concerning the handling of certain option changes within patterns
667 have been fixed. These applied to options other than (?ims). For example,
668 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
669 been the other way round. Some of this was related to change 7 above.
671 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
672 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
673 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
674 POSIX classes only within a class (e.g. /[[:alpha:]]/).
676 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
677 mode. Unlike ".", it always matches newline, whatever the setting of
678 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
679 assertions. Perl allows it, but it doesn't (in general) work because it can't
680 calculate the length of the lookbehind. At least, that's the case for Perl
681 5.8.0 - I've been told they are going to document that it doesn't work in
682 future.
684 18. Added an error diagnosis for escapes that PCRE does not support: these are
685 \L, \l, \N, \P, \p, \U, \u, and \X.
687 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
688 reading past the end of the pattern in cases such as /[abcd/.
690 20. PCRE was getting more memory than necessary for patterns with classes that
691 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
693 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
694 compiling PCRE for use with Virtual Pascal.
696 22. Small fix to the Makefile to make it work properly if the build is done
697 outside the source tree.
699 23. Added a new extension: a condition to go with recursion. If a conditional
700 subpattern starts with (?(R) the "true" branch is used if recursion has
701 happened, whereas the "false" branch is used only at the top level.
703 24. When there was a very long string of literal characters (over 255 bytes
704 without UTF support, over 250 bytes with UTF support), the computation of how
705 much memory was required could be incorrect, leading to segfaults or other
706 strange effects.
708 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
709 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
710 there was a subsequent back reference to those brackets. This meant that, for
711 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
712 possible to check for precisely this case. All we can do is abandon the
713 optimization if .* occurs inside capturing brackets when there are any back
714 references whatsoever. (See below for a better fix that came later.)
716 26. The handling of the optimization for finding the first character of a
717 non-anchored pattern, and for finding a character that is required later in the
718 match were failing in some cases. This didn't break the matching; it just
719 failed to optimize when it could. The way this is done has been re-implemented.
721 27. Fixed typo in error message for invalid (?R item (it said "(?p").
723 28. Added a new feature that provides some of the functionality that Perl
724 provides with (?{...}). The facility is termed a "callout". The way it is done
725 in PCRE is for the caller to provide an optional function, by setting
726 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
727 global variable. By default it is unset, which disables all calling out. To get
728 the function called, the regex must include (?C) at appropriate points. This
729 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
730 This provides a means of identifying different callout points. When PCRE
731 reaches such a point in the regex, if pcre_callout has been set, the external
732 function is called. It is provided with data in a structure called
733 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
734 matching continues; if it returns a non-zero value, the match at the current
735 point fails. However, backtracking will occur if possible. [This was changed
736 later and other features added - see item 49 below.]
738 29. pcretest is upgraded to test the callout functionality. It provides a
739 callout function that displays information. By default, it shows the start of
740 the match and the current position in the text. There are some new data escapes
741 to vary what happens:
743 \C+ in addition, show current contents of captured substrings
744 \C- do not supply a callout function
745 \C!n return 1 when callout number n is reached
746 \C!n!m return 1 when callout number n is reached for the mth time
748 30. If pcregrep was called with the -l option and just a single file name, it
749 output "<stdin>" if a match was found, instead of the file name.
751 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
752 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
753 pcre_exec(). This saves a malloc/free per call. The default value of
754 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
755 when configuring.
757 32. The default maximum size of a compiled pattern is 64K. There have been a
758 few cases of people hitting this limit. The code now uses macros to handle the
759 storing of links as offsets within the compiled pattern. It defaults to 2-byte
760 links, but this can be changed to 3 or 4 bytes by --with-link-size when
761 configuring. Tests 2 and 5 work only with 2-byte links because they output
762 debugging information about compiled patterns.
764 33. Internal code re-arrangements:
766 (a) Moved the debugging function for printing out a compiled regex into
767 its own source file (printint.c) and used #include to pull it into
768 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
769 separate copies.
771 (b) Defined the list of op-code names for debugging as a macro in
772 internal.h so that it is next to the definition of the opcodes.
774 (c) Defined a table of op-code lengths for simpler skipping along compiled
775 code. This is again a macro in internal.h so that it is next to the
776 definition of the opcodes.
778 34. Added support for recursive calls to individual subpatterns, along the
779 lines of Robin Houston's patch (but implemented somewhat differently).
781 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
782 allow it to read and process whole directories in Win32. This code was
783 contributed by Lionel Fourquaux; it has not been tested by me.
785 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
786 used to name a group. Names consist of alphanumerics and underscores, and must
787 be unique. Back references use the syntax (?P=name) and recursive calls use
788 (?P>name) which is a PCRE extension to the Python extension. Groups still have
789 numbers. The function pcre_fullinfo() can be used after compilation to extract
790 a name/number map. There are three relevant calls:
792 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
793 PCRE_INFO_NAMECOUNT yields the number of entries
794 PCRE_INFO_NAMETABLE yields a pointer to the map.
796 The map is a vector of fixed-size entries. The size of each entry depends on
797 the length of the longest name used. The first two bytes of each entry are the
798 group number, most significant byte first. There follows the corresponding
799 name, zero terminated. The names are in alphabetical order.
801 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
802 case instead of 255. Making it the same both with and without UTF-8 support
803 means that the same test output works with both.
805 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
806 calling malloc() with a zero argument.
808 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
809 optimization. I've improved things by keeping a bitmap of backreferences with
810 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
811 fact referenced, the optimization can be applied. It is unlikely that a
812 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
813 the match to follow \n) will appear inside brackets with a number greater than
814 31, but if it does, any back reference > 31 suppresses the optimization.
816 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
817 of disabling numbered capturing parentheses. Any opening parenthesis that is
818 not followed by ? behaves as if it were followed by ?: but named parentheses
819 can still be used for capturing (and they will acquire numbers in the usual
820 way).
822 41. Redesigned the return codes from the match() function into yes/no/error so
823 that errors can be passed back from deep inside the nested calls. A malloc
824 failure while inside a recursive subpattern call now causes the
825 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
827 42. It is now possible to set a limit on the number of times the match()
828 function is called in a call to pcre_exec(). This facility makes it possible to
829 limit the amount of recursion and backtracking, though not in a directly
830 obvious way, because the match() function is used in a number of different
831 circumstances. The count starts from zero for each position in the subject
832 string (for non-anchored patterns). The default limit is, for compatibility, a
833 large number, namely 10 000 000. You can change this in two ways:
835 (a) When configuring PCRE before making, you can use --with-match-limit=n
836 to set a default value for the compiled library.
838 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
839 a different value is set. See 45 below.
841 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
843 43. Added a new function pcre_config(int, void *) to enable run-time extraction
844 of things that can be changed at compile time. The first argument specifies
845 what is wanted and the second points to where the information is to be placed.
846 The current list of available information is:
850 The output is an integer that is set to one if UTF-8 support is available;
851 otherwise it is set to zero.
855 The output is an integer that it set to the value of the code that is used for
856 newline. It is either LF (10) or CR (13).
860 The output is an integer that contains the number of bytes used for internal
861 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
865 The output is an integer that contains the threshold above which the POSIX
866 interface uses malloc() for output vectors. See item 31 above.
870 The output is an unsigned integer that contains the default limit of the number
871 of match() calls in a pcre_exec() execution. See 42 above.
873 44. pcretest has been upgraded by the addition of the -C option. This causes it
874 to extract all the available output from the new pcre_config() function, and to
875 output it. The program then exits immediately.
877 45. A need has arisen to pass over additional data with calls to pcre_exec() in
878 order to support additional features. One way would have been to define
879 pcre_exec2() (for example) with extra arguments, but this would not have been
880 extensible, and would also have required all calls to the original function to
881 be mapped to the new one. Instead, I have chosen to extend the mechanism that
882 is used for passing in "extra" data from pcre_study().
884 The pcre_extra structure is now exposed and defined in pcre.h. It currently
885 contains the following fields:
887 flags a bitmap indicating which of the following fields are set
888 study_data opaque data from pcre_study()
889 match_limit a way of specifying a limit on match() calls for a specific
890 call to pcre_exec()
891 callout_data data for callouts (see 49 below)
893 The flag bits are also defined in pcre.h, and are
899 The pcre_study() function now returns one of these new pcre_extra blocks, with
900 the actual study data pointed to by the study_data field, and the
901 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
902 before. That is, this change is entirely upwards-compatible and requires no
903 change to existing code.
905 If you want to pass in additional data to pcre_exec(), you can either place it
906 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
907 block.
909 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
910 data string contains the escape sequence \M, pcretest calls pcre_exec() several
911 times with different match limits, until it finds the minimum value needed for
912 pcre_exec() to complete. The value is then output. This can be instructive; for
913 most simple matches the number is quite small, but for pathological cases it
914 gets very large very quickly.
916 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
917 returns the size of the data block pointed to by the study_data field in a
918 pcre_extra block, that is, the value that was passed as the argument to
919 pcre_malloc() when PCRE was getting memory in which to place the information
920 created by pcre_study(). The fourth argument should point to a size_t variable.
921 pcretest has been extended so that this information is shown after a successful
922 pcre_study() call when information about the compiled regex is being displayed.
924 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
925 because what follows is always an absolute path. (Later: it turns out that this
926 is more than cosmetic for MinGW, because it doesn't like empty path
927 components.)
929 49. Some changes have been made to the callout feature (see 28 above):
931 (i) A callout function now has three choices for what it returns:
933 0 => success, carry on matching
934 > 0 => failure at this point, but backtrack if possible
935 < 0 => serious error, return this value from pcre_exec()
937 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
938 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
939 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
940 use by callout functions. It will never be used by PCRE itself.
942 (ii) The pcre_extra structure (see 45 above) has a void * field called
943 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
944 pcre_callout_block structure has a field of the same name. The contents of
945 the field passed in the pcre_extra structure are passed to the callout
946 function in the corresponding field in the callout block. This makes it
947 easier to use the same callout-containing regex from multiple threads. For
948 testing, the pcretest program has a new data escape
950 \C*n pass the number n (may be negative) as callout_data
952 If the callout function in pcretest receives a non-zero value as
953 callout_data, it returns that value.
955 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
956 there were some redundant $(CFLAGS) in commands that are now specified as
957 $(LINK), which already includes $(CFLAGS).
959 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
960 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
961 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
962 one-byte characters throughout. Note that case-insensitive matching applies
963 only to characters whose values are less than 256. PCRE doesn't support the
964 notion of cases for higher-valued characters.
966 (i) A character class whose characters are all within 0-255 is handled as
967 a bit map, and the map is inverted for negative classes. Previously, a
968 character > 255 always failed to match such a class; however it should
969 match if the class was a negative one (e.g. [^ab]). This has been fixed.
971 (ii) A negated character class with a single character < 255 is coded as
972 "not this character" (OP_NOT). This wasn't working properly when the test
973 character was multibyte, either singly or repeated.
975 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
976 mode, for example: \x{100}{2,3}.
978 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
979 singly or repeated) now correctly test multibyte characters. However,
980 PCRE doesn't recognize any characters with values greater than 255 as
981 digits, spaces, or word characters. Such characters always match \D, \S,
982 and \W, and never match \d, \s, or \w.
984 (v) Classes may now contain characters and character ranges with values
985 greater than 255. For example: [ab\x{100}-\x{400}].
987 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
988 PCRE in UTF-8 mode.
990 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
991 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
992 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
993 value.)
995 53. The single man page has become too large. I have therefore split it up into
996 a number of separate man pages. These also give rise to individual HTML pages;
997 these are now put in a separate directory, and there is an index.html page that
998 lists them all. Some hyperlinking between the pages has been installed.
1000 54. Added convenience functions for handling named capturing parentheses.
1002 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1003 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1004 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1005 are faulted.
1007 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1008 calling configure. These values are used when compiling the dftables.c program
1009 which is run to generate the source of the default character tables. They
1010 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1011 you will need to set these values.
1013 57. Updated the building process for Windows DLL, as provided by Fred Cox.
1016 Version 3.9 02-Jan-02
1017 ---------------------
1019 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1021 2. If --disable-static was given, the building process failed when trying to
1022 build pcretest and pcregrep. (For some reason it was using libtool to compile
1023 them, which is not right, as they aren't part of the library.)
1026 Version 3.8 18-Dec-01
1027 ---------------------
1029 1. The experimental UTF-8 code was completely screwed up. It was packing the
1030 bytes in the wrong order. How dumb can you get?
1033 Version 3.7 29-Oct-01
1034 ---------------------
1036 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1037 This caused pcretest, when used on the test data, to segfault. Unfortunately,
1038 this didn't happen under Solaris 8, where I normally test things.
1040 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1041 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1042 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1043 this fix an hour or so after the initial 3.7 release.)
1046 Version 3.6 23-Oct-01
1047 ---------------------
1049 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1050 offsets passed as NULL with zero offset count.
1052 2. The config.guess and config.sub files had not been updated when I moved to
1053 the latest autoconf.
1056 Version 3.5 15-Aug-01
1057 ---------------------
1059 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1060 had been forgotten.
1062 2. By using declared but undefined structures, we can avoid using "void"
1063 definitions in pcre.h while keeping the internal definitions of the structures
1064 private.
1066 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1067 user point of view, this means that both static and shared libraries are built
1068 by default, but this can be individually controlled. More of the work of
1069 handling this static/shared cases is now inside libtool instead of PCRE's make
1070 file.
1072 4. The pcretest utility is now installed along with pcregrep because it is
1073 useful for users (to test regexs) and by doing this, it automatically gets
1074 relinked by libtool. The documentation has been turned into a man page, so
1075 there are now .1, .txt, and .html versions in /doc.
1077 5. Upgrades to pcregrep:
1078 (i) Added long-form option names like gnu grep.
1079 (ii) Added --help to list all options with an explanatory phrase.
1080 (iii) Added -r, --recursive to recurse into sub-directories.
1081 (iv) Added -f, --file to read patterns from a file.
1083 6. pcre_exec() was referring to its "code" argument before testing that
1084 argument for NULL (and giving an error if it was NULL).
1086 7. Upgraded Makefile.in to allow for compiling in a different directory from
1087 the source directory.
1089 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1090 options bits, the pointer it was passed was to an int instead of to an unsigned
1091 long int. This mattered only on 64-bit systems.
1093 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1094 generated) instead of pcre.in, which it its source. Also made the same change
1095 in several of the .c files.
1097 10. A new release of gcc defines printf() as a macro, which broke pcretest
1098 because it had an ifdef in the middle of a string argument for printf(). Fixed
1099 by using separate calls to printf().
1101 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1102 script, to force use of CR or LF instead of \n in the source. On non-Unix
1103 systems, the value can be set in config.h.
1105 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1106 absolute limit. Changed the text of the error message to make this clear, and
1107 likewise updated the man page.
1109 13. The limit of 99 on the number of capturing subpatterns has been removed.
1110 The new limit is 65535, which I hope will not be a "real" limit.
1113 Version 3.4 22-Aug-00
1114 ---------------------
1116 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1118 2. Diagnose condition (?(0) as an error instead of crashing on matching.
1121 Version 3.3 01-Aug-00
1122 ---------------------
1124 1. If an octal character was given, but the value was greater than \377, it
1125 was not getting masked to the least significant bits, as documented. This could
1126 lead to crashes in some systems.
1128 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1129 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1131 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1132 These just pass their arguments on to (pcre_free)(), but they are provided
1133 because some uses of PCRE bind it to non-C systems that can call its functions,
1134 but cannot call free() or pcre_free() directly.
1136 4. Add "make test" as a synonym for "make check". Corrected some comments in
1137 the Makefile.
1139 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1140 Makefile.
1142 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1143 command called pgrep for grepping around the active processes.
1145 7. Added the beginnings of support for UTF-8 character strings.
1147 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1148 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1149 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1150 out for the ar command.)
1153 Version 3.2 12-May-00
1154 ---------------------
1156 This is purely a bug fixing release.
1158 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1159 of ZA. This was just one example of several cases that could provoke this bug,
1160 which was introduced by change 9 of version 2.00. The code for breaking
1161 infinite loops after an iteration that matches an empty string was't working
1162 correctly.
1164 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1165 when matched against abbab (for example). After matching an empty string, it
1166 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1167 caused it to match further down the string than it should.
1169 3. The code contained an inclusion of sys/types.h. It isn't clear why this
1170 was there because it doesn't seem to be needed, and it causes trouble on some
1171 systems, as it is not a Standard C header. It has been removed.
1173 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1174 were reported on the Macintosh. The changes were from
1176 while ((c = *(++ptr)) != 0 && c != '\n');
1177 to
1178 while ((c = *(++ptr)) != 0 && c != '\n') ;
1180 Totally extraordinary, but if that's what it takes...
1182 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1183 available. Added HAVE_BCOPY and an autoconf test for it; if neither
1184 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1185 assumes the way PCRE uses memmove() (always moving upwards).
1187 6. PCRE is being used in one environment where strchr() is not available. There
1188 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1189 faster code anyway.
1192 Version 3.1 09-Feb-00
1193 ---------------------
1195 The only change in this release is the fixing of some bugs in Makefile.in for
1196 the "install" target:
1198 (1) It was failing to install pcreposix.h.
1200 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
1203 Version 3.0 01-Feb-00
1204 ---------------------
1206 1. Add support for the /+ modifier to perltest (to output $` like it does in
1207 pcretest).
1209 2. Add support for the /g modifier to perltest.
1211 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
1212 matches null strings.
1214 4. Fix perltest so that it doesn't do unwanted things when fed an empty
1215 pattern. Perl treats empty patterns specially - it reuses the most recent
1216 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
1217 effect.
1219 5. The POSIX interface was broken in that it was just handing over the POSIX
1220 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
1221 required a bigger vector, with some working space on the end. This means that
1222 the POSIX wrapper now has to get and free some memory, and copy the results.
1224 6. Added some simple autoconf support, placing the test data and the
1225 documentation in separate directories, re-organizing some of the
1226 information files, and making it build pcre-config (a GNU standard). Also added
1227 libtool support for building PCRE as a shared library, which is now the
1228 default.
1230 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
1231 09 are not valid octal constants. Single digits will be used for minor values
1232 less than 10.
1234 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
1235 existing programs that set these in the POSIX interface can use PCRE without
1236 modification.
1238 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
1239 return all that pcre_info() returns, plus additional data. The pcre_info()
1240 function is retained for compatibility, but is considered to be obsolete.
1242 10. Added experimental recursion feature (?R) to handle one common case that
1243 Perl 5.6 will be able to do with (?p{...}).
1245 11. Added support for POSIX character classes like [:alpha:], which Perl is
1246 adopting.
1249 Version 2.08 31-Aug-99
1250 ----------------------
1252 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
1253 trying to match at the startoffset position, but instead was moving forward to
1254 the next newline as if a previous match had failed.
1256 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
1257 and could get into a loop if a null string was matched other than at the start
1258 of the subject.
1260 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
1261 be distinguished at compile time, and for completeness also added PCRE_DATE.
1263 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
1264 in GnuWin32 environments.
1267 Version 2.07 29-Jul-99
1268 ----------------------
1270 1. The documentation is now supplied in plain text form and HTML as well as in
1271 the form of man page sources.
1273 2. C++ compilers don't like assigning (void *) values to other pointer types.
1274 In particular this affects malloc(). Although there is no problem in Standard
1275 C, I've put in casts to keep C++ compilers happy.
1277 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
1278 should be (const char *).
1280 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
1281 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
1282 However, I haven't made this a standard facility. The documentation doesn't
1283 mention it, and the Makefile doesn't support it.
1285 5. The Makefile now contains an "install" target, with editable destinations at
1286 the top of the file. The pcretest program is not installed.
1288 6. pgrep -V now gives the PCRE version number and date.
1290 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1291 causing the entire string to be ignored, instead of just the last character.
1293 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1294 non-matching string, it can take a very, very long time, even for strings of
1295 quite modest length, because of the nested recursion. PCRE now does better in
1296 some of these cases. It does this by remembering the last required literal
1297 character in the pattern, and pre-searching the subject to ensure it is present
1298 before running the real match. In other words, it applies a heuristic to detect
1299 some types of certain failure quickly, and in the above example, if presented
1300 with a string that has no trailing " it gives "no match" very quickly.
1302 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1303 other alternatives are tried instead.
1306 Version 2.06 09-Jun-99
1307 ----------------------
1309 1. Change pcretest's output for amount of store used to show just the code
1310 space, because the remainder (the data block) varies in size between 32-bit and
1311 64-bit systems.
1313 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1314 start matching at. This allows lookbehinds to work when searching for multiple
1315 occurrences in a string.
1317 3. Added additional options to pcretest for testing multiple occurrences:
1319 /+ outputs the rest of the string that follows a match
1320 /g loops for multiple occurrences, using the new startoffset argument
1321 /G loops for multiple occurrences by passing an incremented pointer
1323 4. PCRE wasn't doing the "first character" optimization for patterns starting
1324 with \b or \B, though it was doing it for other lookbehind assertions. That is,
1325 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1326 the letter 'x'. On long subject strings, this gives a significant speed-up.
1329 Version 2.05 21-Apr-99
1330 ----------------------
1332 1. Changed the type of magic_number from int to long int so that it works
1333 properly on 16-bit systems.
1335 2. Fixed a bug which caused patterns starting with .* not to work correctly
1336 when the subject string contained newline characters. PCRE was assuming
1337 anchoring for such patterns in all cases, which is not correct because .* will
1338 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1339 DOTALL is set at top level; otherwise it knows that patterns starting with .*
1340 must be retried after every newline in the subject.
1343 Version 2.04 18-Feb-99
1344 ----------------------
1346 1. For parenthesized subpatterns with repeats whose minimum was zero, the
1347 computation of the store needed to hold the pattern was incorrect (too large).
1348 If such patterns were nested a few deep, this could multiply and become a real
1349 problem.
1351 2. Added /M option to pcretest to show the memory requirement of a specific
1352 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1354 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1355 compiled in such a way that the backtracking after subsequent failure was
1356 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1357 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1360 Version 2.03 02-Feb-99
1361 ----------------------
1363 1. Fixed typo and small mistake in man page.
1365 2. Added 4th condition (GPL supersedes if conflict) and created separate
1366 LICENCE file containing the conditions.
1368 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1369 Perl, that is the internal \ allows the delimiter to be included in the
1370 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1371 the final delimiter, add \ to the end of the pattern (to test the error).
1373 4. Added the convenience functions for extracting substrings after a successful
1374 match. Updated pcretest to make it able to test these functions.
1377 Version 2.02 14-Jan-99
1378 ----------------------
1380 1. Initialized the working variables associated with each extraction so that
1381 their saving and restoring doesn't refer to uninitialized store.
1383 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1384 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1385 fix the problem.
1387 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1388 calls, and wasn't printing the correct value for compiling calls. Increased the
1389 default value of LOOPREPEAT, and the number of significant figures in the
1390 times.
1392 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1394 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1395 a building problem on Windows NT with a FAT file system.
1398 Version 2.01 21-Oct-98
1399 ----------------------
1401 1. Changed the API for pcre_compile() to allow for the provision of a pointer
1402 to character tables built by pcre_maketables() in the current locale. If NULL
1403 is passed, the default tables are used.
1406 Version 2.00 24-Sep-98
1407 ----------------------
1409 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1410 it any more.
1412 2. Allow quantification of (?>) groups, and make it work correctly.
1414 3. The first character computation wasn't working for (?>) groups.
1416 4. Correct the implementation of \Z (it is permitted to match on the \n at the
1417 end of the subject) and add 5.005's \z, which really does match only at the
1418 very end of the subject.
1420 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1422 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1423 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1424 localized options. All options to pcre_study() were also removed.
1426 7. Add other new features from 5.005:
1428 $(?<= positive lookbehind
1429 $(?<! negative lookbehind
1430 (?imsx-imsx) added the unsetting capability
1431 such a setting is global if at outer level; local otherwise
1432 (?imsx-imsx:) non-capturing groups with option setting
1433 (?(cond)re|re) conditional pattern matching
1435 A backreference to itself in a repeated group matches the previous
1436 captured string.
1438 8. General tidying up of studying (both automatic and via "study")
1439 consequential on the addition of new assertions.
1441 9. As in 5.005, unlimited repeated groups that could match an empty substring
1442 are no longer faulted at compile time. Instead, the loop is forcibly broken at
1443 runtime if any iteration does actually match an empty substring.
1445 10. Include the RunTest script in the distribution.
1447 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1448 discrepancies, some of which were old and were also with respect to 5.004. They
1449 have now been fixed.
1452 Version 1.09 28-Apr-98
1453 ----------------------
1455 1. A negated single character class followed by a quantifier with a minimum
1456 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
1457 program crashes, or just wrong answers. This did not apply to negated classes
1458 containing more than one character, or to minima other than one.
1461 Version 1.08 27-Mar-98
1462 ----------------------
1464 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
1466 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
1467 latter must appear before anything that relies on it in the pattern.
1470 Version 1.07 16-Feb-98
1471 ----------------------
1473 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
1474 repeat of a potentially empty string).
1477 Version 1.06 23-Jan-98
1478 ----------------------
1480 1. Added Markus Oberhumer's little patches for C++.
1482 2. Literal strings longer than 255 characters were broken.
1485 Version 1.05 23-Dec-97
1486 ----------------------
1488 1. Negated character classes containing more than one character were failing if
1489 PCRE_CASELESS was set at run time.
1492 Version 1.04 19-Dec-97
1493 ----------------------
1495 1. Corrected the man page, where some "const" qualifiers had been omitted.
1497 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
1498 input syntax.
1500 3. Fixed memory leak which occurred when a regex with back references was
1501 matched with an offsets vector that wasn't big enough. The temporary memory
1502 that is used in this case wasn't being freed if the match failed.
1504 4. Tidied pcretest to ensure it frees memory that it gets.
1506 5. Temporary memory was being obtained in the case where the passed offsets
1507 vector was exactly big enough.
1509 6. Corrected definition of offsetof() from change 5 below.
1511 7. I had screwed up change 6 below and broken the rules for the use of
1512 setjmp(). Now fixed.
1515 Version 1.03 18-Dec-97
1516 ----------------------
1518 1. A erroneous regex with a missing opening parenthesis was correctly
1519 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
1520 on some systems.
1522 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
1523 it was reported that one broken compiler failed on the former because "code" is
1524 also an independent variable.
1526 3. The erroneous regex a[]b caused an array overrun reference.
1528 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
1529 fail on data ending with that character. (It was going on too far, and checking
1530 the next character, typically a binary zero.) This was specific to the
1531 optimized code for single-character negative classes.
1533 5. Added a contributed patch from the TIN world which does the following:
1535 + Add an undef for memmove, in case the the system defines a macro for it.
1537 + Add a definition of offsetof(), in case there isn't one. (I don't know
1538 the reason behind this - offsetof() is part of the ANSI standard - but
1539 it does no harm).
1541 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
1542 most of the places where whitespace preceded '#'. I have given up and
1543 allowed the remaining 2 cases to be at the margin.
1545 + Rename some variables in pcre to eliminate shadowing. This seems very
1546 pedantic, but does no harm, of course.
1548 6. Moved the call to setjmp() into its own function, to get rid of warnings
1549 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
1551 7. Constructs such as \d{8,} were compiling into the equivalent of
1552 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
1553 outcome, but in this particular case used more store than had been allocated,
1554 which caused the bug to be discovered because it threw up an internal error.
1556 8. The debugging code in both pcre and pcretest for outputting the compiled
1557 form of a regex was going wrong in the case of back references followed by
1558 curly-bracketed repeats.
1561 Version 1.02 12-Dec-97
1562 ----------------------
1564 1. Typos in pcre.3 and comments in the source fixed.
1566 2. Applied a contributed patch to get rid of places where it used to remove
1567 'const' from variables, and fixed some signed/unsigned and uninitialized
1568 variable warnings.
1570 3. Added the "runtest" target to Makefile.
1572 4. Set default compiler flag to -O2 rather than just -O.
1575 Version 1.01 19-Nov-97
1576 ----------------------
1578 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
1579 like /([ab]*)*/, that is, for classes with more than one character in them.
1581 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
1582 as /((?>a*))*/ (a PCRE_EXTRA facility).
1585 Version 1.00 18-Nov-97
1586 ----------------------
1588 1. Added compile-time macros to support systems such as SunOS4 which don't have
1589 memmove() or strerror() but have other things that can be used instead.
1591 2. Arranged that "make clean" removes the executables.
1594 Version 0.99 27-Oct-97
1595 ----------------------
1597 1. Fixed bug in code for optimizing classes with only one character. It was
1598 initializing a 32-byte map regardless, which could cause it to run off the end
1599 of the memory it had got.
1601 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1604 Version 0.98 22-Oct-97
1605 ----------------------
1607 1. Fixed bug in code for handling temporary memory usage when there are more
1608 back references than supplied space in the ovector. This could cause segfaults.
1611 Version 0.97 21-Oct-97
1612 ----------------------
1614 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1616 2. Optimized negated single characters not to use a bit map.
1618 3. Brought error texts together as macro definitions; clarified some of them;
1619 fixed one that was wrong - it said "range out of order" when it meant "invalid
1620 escape sequence".
1622 4. Changed some char * arguments to const char *.
1624 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1626 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1627 pcretest.
1630 Version 0.96 16-Oct-97
1631 ----------------------
1633 1. Added a simple "pgrep" utility to the distribution.
1635 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1636 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1637 where "ddd" means "one or more decimal digits".
1639 3. Fixed serious bug. If a pattern had a back reference, but the call to
1640 pcre_exec() didn't supply a large enough ovector to record the related
1641 identifying subpattern, the match always failed. PCRE now remembers the number
1642 of the largest back reference, and gets some temporary memory in which to save
1643 the offsets during matching if necessary, in order to ensure that
1644 backreferences always work.
1646 4. Increased the compatibility with Perl in a number of ways:
1648 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1649 to request this handling. The option can be set at compile or exec time.
1651 (b) $ matches before a terminating newline by default; an option
1652 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
1653 mode). The option can be set at compile or exec time.
1655 (c) The handling of \ followed by a digit other than 0 is now supposed to be
1656 the same as Perl's. If the decimal number it represents is less than 10
1657 or there aren't that many previous left capturing parentheses, an octal
1658 escape is read. Inside a character class, it's always an octal escape,
1659 even if it is a single digit.
1661 (d) An escaped but undefined alphabetic character is taken as a literal,
1662 unless PCRE_EXTRA is set. Currently this just reserves the remaining
1663 escapes.
1665 (e) {0} is now permitted. (The previous item is removed from the compiled
1666 pattern).
1668 5. Changed all the names of code files so that the basic parts are no longer
1669 than 10 characters, and abolished the teeny "globals.c" file.
1671 6. Changed the handling of character classes; they are now done with a 32-byte
1672 bit map always.
1674 7. Added the -d and /D options to pcretest to make it possible to look at the
1675 internals of compilation without having to recompile pcre.
1678 Version 0.95 23-Sep-97
1679 ----------------------
1681 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
1682 \x20 at the start of a run of normal characters. These were being treated as
1683 real characters, instead of the source characters being re-checked.
1686 Version 0.94 18-Sep-97
1687 ----------------------
1689 1. The functions are now thread-safe, with the caveat that the global variables
1690 containing pointers to malloc() and free() or alternative functions are the
1691 same for all threads.
1693 2. Get pcre_study() to generate a bitmap of initial characters for non-
1694 anchored patterns when this is possible, and use it if passed to pcre_exec().
1697 Version 0.93 15-Sep-97
1698 ----------------------
1700 1. /(b)|(:+)/ was computing an incorrect first character.
1702 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
1703 but not actually doing anything yet.
1705 3. Treat "-" characters in classes that cannot be part of ranges as literals,
1706 as Perl does (e.g. [-az] or [az-]).
1708 4. Set the anchored flag if a branch starts with .* or .*? because that tests
1709 all possible positions.
1711 5. Split up into different modules to avoid including unneeded functions in a
1712 compiled binary. However, compile and exec are still in one module. The "study"
1713 function is split off.
1715 6. The character tables are now in a separate module whose source is generated
1716 by an auxiliary program - but can then be edited by hand if required. There are
1717 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
1718 toupper() in the code.
1720 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
1721 make them global. Abolish the function for setting them, as the caller can now
1722 set them directly.
1725 Version 0.92 11-Sep-97
1726 ----------------------
1728 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
1729 (e.g. /a{1,3}/) was broken (I mis-optimized it).
1731 2. Caseless matching was not working in character classes if the characters in
1732 the pattern were in upper case.
1734 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
1736 4. Make PCRE_ANCHORED public and accept as a compile option.
1738 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
1739 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
1740 pass them.
1742 6. Give an error if bad option bits passed at compile or run time.
1744 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
1745 pcretest to cause it to pass that flag.
1747 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
1748 options, and the first character, if set.
1750 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
1753 Version 0.91 10-Sep-97
1754 ----------------------
1756 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
1757 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
1759 2. PCRE was looping on encountering an indefinitely repeated back reference to
1760 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
1761 Perl does - treats the match as successful.
1763 ****

  ViewVC Help
Powered by ViewVC 1.1.5