7 1. Internal change: literal characters are no longer packed up into items
8 containing multiple characters in a single byte-string. Each character
9 is now matched using a separate opcode. However, there may be more than one
10 byte in the character in UTF-8 mode.
12 2. The pcre_callout_block structure has two new fields: pattern_position and
13 next_item_length. These contain the offset in the pattern to the next match
14 item, and its length, respectively.
16 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
17 insertion of callouts before each pattern item. Added the /C option to
18 pcretest to make use of this.
20 4. On the advice of a Windows user, the lines
22 #if defined(_WIN32) || defined(WIN32)
23 _setmode( _fileno( stdout ), 0x8000 );
24 #endif /* defined(_WIN32) || defined(WIN32) */
26 have been added to the source of pcretest. This apparently does useful
27 magic in relation to line terminators.
29 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
30 for the benefit of those environments where the "b" makes a difference.
32 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
33 to know about it. I have put a hack into configure.in that adds in code
34 to set GCC=yes if CC=icc. This seems to end up at a point in the
35 generated configure script that is early enough to affect the setting of
36 compiler options, which is what is needed, but I have no means of testing
37 whether it really works. (The user who reported this had patched the
38 generated configure script, which of course I cannot do.)
40 LATER: After change 22 below (new libtool files), the configure script
41 seems to know about icc (and also ecc). Therefore, I have commented out
42 this hack in configure.in.
44 7. Added support for pkg-config (2 patches were sent in).
46 8. Negated POSIX character classes that used a combination of internal tables
47 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
48 [[:^ascii]]. Typically, they would match almost any characters. The other
49 POSIX classes were not broken in this way.
51 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
52 to find the match, as PCRE was deluded into thinking that the match had to
53 start at the start point or following a newline. The same bug applied to
54 patterns with negative forward assertions or any backward assertions
55 preceding ".*" at the start, unless the pattern required a fixed first
56 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
58 10. In UTF-8 mode, when moving forwards in the subject after a failed match
59 starting at the last subject character, bytes beyond the end of the subject
62 11. Renamed the variable "class" as "classbits" to make life easier for C++
63 users. (Previously there was a macro definition, but it apparently wasn't
66 12. Added the new field "tables" to the extra data so that tables can be passed
67 in at exec time, or the internal tables can be re-selected. This allows
68 a compiled regex to be saved and re-used at a later time by a different
69 program that might have everything at different addresses.
71 13. Modified the pcre-config script so that, when run on Solaris, it shows a
72 -R library as well as a -L library.
74 14. The debugging options of pcretest (-d on the command line or D on a
75 pattern) showed incorrect output for anything following an extended class
76 that contained multibyte characters and which was followed by a quantifier.
78 15. Added optional support for general category Unicode character properties
79 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
80 support. It adds about 90K to the size of the library. The meanings of the
81 inbuilt class escapes such as \d and \s have NOT been changed.
83 16. Updated pcredemo.c to include calls to free() to release the memory for the
86 17. The generated file chartables.c was being created in the source directory
87 instead of in the building directory. This caused the build to fail if the
88 source directory was different from the building directory, and was
91 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
92 file. No doubt somebody will tell me if they don't make sense... Also added
93 Dan Mooney's comments about building on OpenVMS.
95 19. Added support for partial matching via the PCRE_PARTIAL option for
96 pcre_exec() and the \P data escape in pcretest.
98 20. Extended pcretest with 3 new pattern features:
100 (i) A pattern option of the form ">rest-of-line" causes pcretest to
101 write the compiled pattern to the file whose name is "rest-of-line".
102 This is a straight binary dump of the data, with the saved pointer to
103 the character tables forced to be NULL. The study data, if any, is
104 written too. After writing, pcretest reads a new pattern.
106 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
107 compiled pattern from the given file. There must not be any
108 occurrences of "<" in the file name (pretty unlikely); if there are,
109 pcretest will instead treat the initial "<" as a pattern delimiter.
110 After reading in the pattern, pcretest goes on to read data lines as
113 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
114 and 16-bit fields in a compiled pattern, to simulate a pattern that
115 was compiled on a host of opposite endianness.
117 21. The pcre-exec() function can now cope with patterns that were compiled on
118 hosts of opposite endianness, with this restriction:
120 As for any compiled expression that is saved and used later, the tables
121 pointer field cannot be preserved; the extra_data field in the arguments
122 to pcre_exec() should be used to pass in a tables address if a value
123 other than the default internal tables were used at compile time.
125 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
126 now diagnosed as an error. Previously, most of the time, a negative number
127 would have been treated as zero, but if in addition "ovector" was passed as
128 NULL, a crash could occur.
130 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
131 new versions from the libtool 1.5 distribution (the last one is a copy of
132 a file called libtool.m4). This seems to have fixed the need to patch
133 "configure" to support Darwin 1.3 (which I used to do). However, I still
134 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
137 24. Changed the PCRE licence to be the more standard "BSD" licence.
140 Version 4.5 01-Dec-03
141 ---------------------
143 1. There has been some re-arrangement of the code for the match() function so
144 that it can be compiled in a version that does not call itself recursively.
145 Instead, it keeps those local variables that need separate instances for
146 each "recursion" in a frame on the heap, and gets/frees frames whenever it
147 needs to "recurse". Keeping track of where control must go is done by means
148 of setjmp/longjmp. The whole thing is implemented by a set of macros that
149 hide most of the details from the main code, and operates only if
150 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
151 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
154 To make it easier for callers to provide specially tailored get/free
155 functions for this usage, two new functions, pcre_stack_malloc, and
156 pcre_stack_free, are used. They are always called in strict stacking order,
157 and the size of block requested is always the same.
159 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
160 PCRE has been compiled to use the stack or the heap for recursion. The
161 -C option of pcretest uses this to show which version is compiled.
163 A new data escape \S, is added to pcretest; it causes the amounts of store
164 obtained and freed by both kinds of malloc/free at match time to be added
167 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
168 what's available on my current Linux desktop machine.
170 3. When matching a UTF-8 string, the test for a valid string at the start has
171 been extended. If start_offset is not zero, PCRE now checks that it points
172 to a byte that is the start of a UTF-8 character. If not, it returns
173 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
174 this is necessary because there may be backward assertions in the pattern.
175 When matching the same subject several times, it may save resources to use
176 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
178 4. The code for checking the validity of UTF-8 strings has been tightened so
179 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
180 containing "overlong sequences".
182 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
183 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
184 should have been "&", but it just so happened that all the cases this let
185 through by mistake were picked up later in the function.
187 6. I had used a variable called "isblank" - this is a C99 function, causing
188 some compilers to warn. To avoid this, I renamed it (as "blankclass").
190 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
191 prompting; (b) run "./pcretest /dev/null" at the start of the test script
192 so the version is shown; (c) stop "make test" echoing "./RunTest".
194 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
196 9. The prototype for memmove() for systems that don't have it was using
197 size_t, but the inclusion of the header that defines size_t was later. I've
198 moved the #includes for the C headers earlier to avoid this.
200 10. Added some adjustments to the code to make it easier to compiler on certain
203 (a) Some "const" qualifiers were missing.
204 (b) Added the macro EXPORT before all exported functions; by default this
205 is defined to be empty.
206 (c) Changed the dftables auxiliary program (that builds chartables.c) so
207 that it reads its output file name as an argument instead of writing
208 to the standard output and assuming this can be redirected.
210 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
211 class containing characters with values greater than 255, PCRE compilation
214 12. A recursive reference to a subpattern that was within another subpattern
215 that had a minimum quantifier of zero caused PCRE to crash. For example,
216 (x(y(?2))z)? provoked this bug with a subject that got as far as the
217 recursion. If the recursively-called subpattern itself had a zero repeat,
220 13. In pcretest, the buffer for reading a data line was set at 30K, but the
221 buffer into which it was copied (for escape processing) was still set at
222 1024, so long lines caused crashes.
224 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
225 "internal error: code overflow...". This applied to any character class
226 that was followed by a possessive quantifier.
228 15. Modified the Makefile to add libpcre.la as a prerequisite for
229 libpcreposix.la because I was told this is needed for a parallel build to
232 16. If a pattern that contained .* following optional items at the start was
233 studied, the wrong optimizing data was generated, leading to matching
234 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
235 matching string must start with a or b or c. The correct conclusion for
236 this pattern is that a match can start with any character.
239 Version 4.4 13-Aug-03
240 ---------------------
242 1. In UTF-8 mode, a character class containing characters with values between
243 127 and 255 was not handled correctly if the compiled pattern was studied.
244 In fixing this, I have also improved the studying algorithm for such
247 2. Three internal functions had redundant arguments passed to them. Removal
248 might give a very teeny performance improvement.
250 3. Documentation bug: the value of the capture_top field in a callout is *one
251 more than* the number of the hightest numbered captured substring.
253 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
254 in incorrectly linking with a previously installed version. They now link
255 explicitly with libpcre.la.
257 5. configure.in no longer needs to recognize Cygwin specially.
259 6. A problem in pcre.in for Windows platforms is fixed.
261 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
262 pcretest, it used to include the size of the study block as part of its
263 output. Unfortunately, the structure contains a field that has a different
264 size on different hardware architectures. This meant that the tests that
265 showed this size failed. As the block is currently always of a fixed size,
266 this information isn't actually particularly useful in pcretest output, so
267 I have just removed it.
269 8. Three pre-processor statements accidentally did not start in column 1.
270 Sadly, there are *still* compilers around that complain, even though
271 standard C has not required this for well over a decade. Sigh.
273 9. In pcretest, the code for checking callouts passed small integers in the
274 callout_data field, which is a void * field. However, some picky compilers
275 complained about the casts involved for this on 64-bit systems. Now
276 pcretest passes the address of the small integer instead, which should get
279 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
280 both compile and run time, and gives an error if an invalid UTF-8 sequence
281 is found. There is a option for disabling this check in cases where the
282 string is known to be correct and/or the maximum performance is wanted.
284 11. In response to a bug report, I changed one line in Makefile.in from
286 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
288 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
290 to look similar to other lines, but I have no way of telling whether this
291 is the right thing to do, as I do not use Windows. No doubt I'll get told
295 Version 4.3 21-May-03
296 ---------------------
298 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
301 2. Some refactoring to improve the quality of the code:
303 (i) The utf8_table... variables are now declared "const".
305 (ii) The code for \cx, which used the "case flipping" table to upper case
306 lower case letters, now just substracts 32. This is ASCII-specific,
307 but the whole concept of \cx is ASCII-specific, so it seems
310 (iii) PCRE was using its character types table to recognize decimal and
311 hexadecimal digits in the pattern. This is silly, because it handles
312 only 0-9, a-f, and A-F, but the character types table is locale-
313 specific, which means strange things might happen. A private
314 table is now used for this - though it costs 256 bytes, a table is
315 much faster than multiple explicit tests. Of course, the standard
316 character types table is still used for matching digits in subject
319 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
320 ending in _t are). So I've renamed it as ESC_tee.
322 3. The first argument for regexec() in the POSIX wrapper should have been
325 4. Changed pcretest to use malloc() for its buffers so that they can be
326 Electric Fenced for debugging.
328 5. There were several places in the code where, in UTF-8 mode, PCRE would try
329 to read one or more bytes before the start of the subject string. Often this
330 had no effect on PCRE's behaviour, but in some circumstances it could
331 provoke a segmentation fault.
333 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
334 to try to read one or more bytes before the start of the subject string.
336 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
337 UTF-8 support could misbehave in various ways if the subject string
338 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
339 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
340 back over UTF-8 characters.)
343 Version 4.2 14-Apr-03
344 ---------------------
346 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
348 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
349 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
350 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
351 [WIN_PREFIX]: new variable, "cyg" for Cygwin
352 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
354 Note: automatic setting of the BUILD variables is not yet working
355 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
356 compile-time but not at link-time
357 [LINK]: use for linking executables only
358 make different versions for Windows and non-Windows
359 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
361 [LINK_FOR_BUILD]: new variable
362 [OBJEXT]: use throughout
363 [EXEEXT]: use throughout
364 <winshared>: new target
365 <wininstall>: new target
366 <dftables.o>: use native compiler
367 <dftables>: use native linker
368 <install>: handle Windows platform correctly
371 copy DLL to top builddir before testing
373 As part of these changes, -no-undefined was removed again. This was reported
374 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
377 3. Some tidies to get rid of compiler warnings:
379 . In the match_data structure, match_limit was an unsigned long int, whereas
380 match_call_count was an int. I've made them both unsigned long ints.
382 . In pcretest the fact that a const uschar * doesn't automatically cast to
383 a void * provoked a warning.
385 . Turning on some more compiler warnings threw up some "shadow" variables
386 and a few more missing casts.
388 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
389 option, a class that contained a single character with a value between 128
390 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
392 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
393 option, a class that contained several characters, but with at least one
394 whose value was between 128 and 255 caused PCRE to crash.
397 Version 4.1 12-Mar-03
398 ---------------------
400 1. Compiling with gcc -pedantic found a couple of places where casts were
401 needed, and a string in dftables.c that was longer than standard compilers are
404 2. Compiling with Sun's compiler found a few more places where the code could
405 be tidied up in order to avoid warnings.
407 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
408 first of these names is deprecated in the latest Autoconf in favour of the name
409 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
410 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
411 analogy I have changed it to CFLAGS_FOR_BUILD.
413 4. Added -no-undefined to the linking command in the Makefile, because this is
414 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
415 linking step for the pcreposix library.
417 5. PCRE was failing to diagnose the case of two named groups with the same
420 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
421 literal character that is needed in the subject for a match, and scans along to
422 ensure that it is present before embarking on the full matching process. This
423 saves time in cases of nested unlimited repeats that are never going to match.
424 Problem: the scan can take a lot of time if the subject is very long (e.g.
425 megabytes), thus penalizing straightforward matches. It is now done only if the
426 amount of subject to be scanned is less than 1000 bytes.
428 7. A lesser problem with the same optimization is that it was recording the
429 first character of an anchored pattern as "needed", thus provoking a search
430 right along the subject, even when the first match of the pattern was going to
431 fail. The "needed" character is now not set for anchored patterns, unless it
432 follows something in the pattern that is of non-fixed length. Thus, it still
433 fulfils its original purpose of finding quick non-matches in cases of nested
434 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
437 Version 4.0 17-Feb-03
438 ---------------------
440 1. If a comment in an extended regex that started immediately after a meta-item
441 extended to the end of string, PCRE compiled incorrect data. This could lead to
442 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
444 2. Moved to autoconf 2.53 and libtool 1.4.2.
446 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
447 the special perltest8 script is no longer needed - all the tests can be run
448 from a single perltest script.
450 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
451 by \s. It has now been removed in PCRE. This means it isn't recognized as
452 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
453 class [:space:] *does* include VT, thereby creating a mess.
455 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
458 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
459 its new features into the main test script, reducing the number of scripts.
461 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
462 were backward compatible, and made the (?i) apply to the whole pattern, as if
463 /i were given. Now it behaves more logically, and applies the option setting
464 only to what follows. PCRE has been changed to follow suit. However, if it
465 finds options settings right at the start of the pattern, it extracts them into
466 the global options, as before. Thus, they show up in the info data.
468 8. Added support for the \Q...\E escape sequence. Characters in between are
469 treated as literals. This is slightly different from Perl in that $ and @ are
470 also handled as literals inside the quotes. In Perl, they will cause variable
471 interpolation. Note the following examples:
473 Pattern PCRE matches Perl matches
475 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
476 \Qabc\$xyz\E abc\$xyz abc\$xyz
477 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
479 For compatibility with Perl, \Q...\E sequences are recognized inside character
480 classes as well as outside them.
482 9. Re-organized 3 code statements in pcretest to avoid "overflow in
483 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
484 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
485 signed/unsigned warnings.
487 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
488 option for pcretest, so I've replaced it by a simple function that does just
491 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
494 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
495 Java package. This provides some syntactic sugar for simple cases of what my
496 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
497 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
498 item, you can use this simplified notation. Note that only makes sense with
499 greedy quantifiers. Consequently, the use of the possessive quantifier forces
500 greediness, whatever the setting of the PCRE_UNGREEDY option.
502 13. A change of greediness default within a pattern was not taking effect at
503 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
504 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
505 was abstracted outside.
507 14. PCRE now supports the \G assertion. It is true when the current matching
508 position is at the start point of the match. This differs from \A when the
509 starting offset is non-zero. Used with the /g option of pcretest (or similar
510 code), it works in the same way as it does for Perl's /g option. If all
511 alternatives of a regex begin with \G, the expression is anchored to the start
512 match position, and the "anchored" flag is set in the compiled expression.
514 15. Some bugs concerning the handling of certain option changes within patterns
515 have been fixed. These applied to options other than (?ims). For example,
516 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
517 been the other way round. Some of this was related to change 7 above.
519 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
520 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
521 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
522 POSIX classes only within a class (e.g. /[[:alpha:]]/).
524 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
525 mode. Unlike ".", it always matches newline, whatever the setting of
526 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
527 assertions. Perl allows it, but it doesn't (in general) work because it can't
528 calculate the length of the lookbehind. At least, that's the case for Perl
529 5.8.0 - I've been told they are going to document that it doesn't work in
532 18. Added an error diagnosis for escapes that PCRE does not support: these are
533 \L, \l, \N, \P, \p, \U, \u, and \X.
535 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
536 reading past the end of the pattern in cases such as /[abcd/.
538 20. PCRE was getting more memory than necessary for patterns with classes that
539 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
541 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
542 compiling PCRE for use with Virtual Pascal.
544 22. Small fix to the Makefile to make it work properly if the build is done
545 outside the source tree.
547 23. Added a new extension: a condition to go with recursion. If a conditional
548 subpattern starts with (?(R) the "true" branch is used if recursion has
549 happened, whereas the "false" branch is used only at the top level.
551 24. When there was a very long string of literal characters (over 255 bytes
552 without UTF support, over 250 bytes with UTF support), the computation of how
553 much memory was required could be incorrect, leading to segfaults or other
556 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
557 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
558 there was a subsequent back reference to those brackets. This meant that, for
559 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
560 possible to check for precisely this case. All we can do is abandon the
561 optimization if .* occurs inside capturing brackets when there are any back
562 references whatsoever. (See below for a better fix that came later.)
564 26. The handling of the optimization for finding the first character of a
565 non-anchored pattern, and for finding a character that is required later in the
566 match were failing in some cases. This didn't break the matching; it just
567 failed to optimize when it could. The way this is done has been re-implemented.
569 27. Fixed typo in error message for invalid (?R item (it said "(?p").
571 28. Added a new feature that provides some of the functionality that Perl
572 provides with (?{...}). The facility is termed a "callout". The way it is done
573 in PCRE is for the caller to provide an optional function, by setting
574 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
575 global variable. By default it is unset, which disables all calling out. To get
576 the function called, the regex must include (?C) at appropriate points. This
577 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
578 This provides a means of identifying different callout points. When PCRE
579 reaches such a point in the regex, if pcre_callout has been set, the external
580 function is called. It is provided with data in a structure called
581 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
582 matching continues; if it returns a non-zero value, the match at the current
583 point fails. However, backtracking will occur if possible. [This was changed
584 later and other features added - see item 49 below.]
586 29. pcretest is upgraded to test the callout functionality. It provides a
587 callout function that displays information. By default, it shows the start of
588 the match and the current position in the text. There are some new data escapes
589 to vary what happens:
591 \C+ in addition, show current contents of captured substrings
592 \C- do not supply a callout function
593 \C!n return 1 when callout number n is reached
594 \C!n!m return 1 when callout number n is reached for the mth time
596 30. If pcregrep was called with the -l option and just a single file name, it
597 output "<stdin>" if a match was found, instead of the file name.
599 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
600 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
601 pcre_exec(). This saves a malloc/free per call. The default value of
602 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
605 32. The default maximum size of a compiled pattern is 64K. There have been a
606 few cases of people hitting this limit. The code now uses macros to handle the
607 storing of links as offsets within the compiled pattern. It defaults to 2-byte
608 links, but this can be changed to 3 or 4 bytes by --with-link-size when
609 configuring. Tests 2 and 5 work only with 2-byte links because they output
610 debugging information about compiled patterns.
612 33. Internal code re-arrangements:
614 (a) Moved the debugging function for printing out a compiled regex into
615 its own source file (printint.c) and used #include to pull it into
616 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
619 (b) Defined the list of op-code names for debugging as a macro in
620 internal.h so that it is next to the definition of the opcodes.
622 (c) Defined a table of op-code lengths for simpler skipping along compiled
623 code. This is again a macro in internal.h so that it is next to the
624 definition of the opcodes.
626 34. Added support for recursive calls to individual subpatterns, along the
627 lines of Robin Houston's patch (but implemented somewhat differently).
629 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
630 allow it to read and process whole directories in Win32. This code was
631 contributed by Lionel Fourquaux; it has not been tested by me.
633 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
634 used to name a group. Names consist of alphanumerics and underscores, and must
635 be unique. Back references use the syntax (?P=name) and recursive calls use
636 (?P>name) which is a PCRE extension to the Python extension. Groups still have
637 numbers. The function pcre_fullinfo() can be used after compilation to extract
638 a name/number map. There are three relevant calls:
640 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
641 PCRE_INFO_NAMECOUNT yields the number of entries
642 PCRE_INFO_NAMETABLE yields a pointer to the map.
644 The map is a vector of fixed-size entries. The size of each entry depends on
645 the length of the longest name used. The first two bytes of each entry are the
646 group number, most significant byte first. There follows the corresponding
647 name, zero terminated. The names are in alphabetical order.
649 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
650 case instead of 255. Making it the same both with and without UTF-8 support
651 means that the same test output works with both.
653 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
654 calling malloc() with a zero argument.
656 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
657 optimization. I've improved things by keeping a bitmap of backreferences with
658 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
659 fact referenced, the optimization can be applied. It is unlikely that a
660 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
661 the match to follow \n) will appear inside brackets with a number greater than
662 31, but if it does, any back reference > 31 suppresses the optimization.
664 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
665 of disabling numbered capturing parentheses. Any opening parenthesis that is
666 not followed by ? behaves as if it were followed by ?: but named parentheses
667 can still be used for capturing (and they will acquire numbers in the usual
670 41. Redesigned the return codes from the match() function into yes/no/error so
671 that errors can be passed back from deep inside the nested calls. A malloc
672 failure while inside a recursive subpattern call now causes the
673 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
675 42. It is now possible to set a limit on the number of times the match()
676 function is called in a call to pcre_exec(). This facility makes it possible to
677 limit the amount of recursion and backtracking, though not in a directly
678 obvious way, because the match() function is used in a number of different
679 circumstances. The count starts from zero for each position in the subject
680 string (for non-anchored patterns). The default limit is, for compatibility, a
681 large number, namely 10 000 000. You can change this in two ways:
683 (a) When configuring PCRE before making, you can use --with-match-limit=n
684 to set a default value for the compiled library.
686 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
687 a different value is set. See 45 below.
689 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
691 43. Added a new function pcre_config(int, void *) to enable run-time extraction
692 of things that can be changed at compile time. The first argument specifies
693 what is wanted and the second points to where the information is to be placed.
694 The current list of available information is:
698 The output is an integer that is set to one if UTF-8 support is available;
699 otherwise it is set to zero.
703 The output is an integer that it set to the value of the code that is used for
704 newline. It is either LF (10) or CR (13).
706 PCRE_CONFIG_LINK_SIZE
708 The output is an integer that contains the number of bytes used for internal
709 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
711 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
713 The output is an integer that contains the threshold above which the POSIX
714 interface uses malloc() for output vectors. See item 31 above.
716 PCRE_CONFIG_MATCH_LIMIT
718 The output is an unsigned integer that contains the default limit of the number
719 of match() calls in a pcre_exec() execution. See 42 above.
721 44. pcretest has been upgraded by the addition of the -C option. This causes it
722 to extract all the available output from the new pcre_config() function, and to
723 output it. The program then exits immediately.
725 45. A need has arisen to pass over additional data with calls to pcre_exec() in
726 order to support additional features. One way would have been to define
727 pcre_exec2() (for example) with extra arguments, but this would not have been
728 extensible, and would also have required all calls to the original function to
729 be mapped to the new one. Instead, I have chosen to extend the mechanism that
730 is used for passing in "extra" data from pcre_study().
732 The pcre_extra structure is now exposed and defined in pcre.h. It currently
733 contains the following fields:
735 flags a bitmap indicating which of the following fields are set
736 study_data opaque data from pcre_study()
737 match_limit a way of specifying a limit on match() calls for a specific
739 callout_data data for callouts (see 49 below)
741 The flag bits are also defined in pcre.h, and are
743 PCRE_EXTRA_STUDY_DATA
744 PCRE_EXTRA_MATCH_LIMIT
745 PCRE_EXTRA_CALLOUT_DATA
747 The pcre_study() function now returns one of these new pcre_extra blocks, with
748 the actual study data pointed to by the study_data field, and the
749 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
750 before. That is, this change is entirely upwards-compatible and requires no
751 change to existing code.
753 If you want to pass in additional data to pcre_exec(), you can either place it
754 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
757 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
758 data string contains the escape sequence \M, pcretest calls pcre_exec() several
759 times with different match limits, until it finds the minimum value needed for
760 pcre_exec() to complete. The value is then output. This can be instructive; for
761 most simple matches the number is quite small, but for pathological cases it
762 gets very large very quickly.
764 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
765 returns the size of the data block pointed to by the study_data field in a
766 pcre_extra block, that is, the value that was passed as the argument to
767 pcre_malloc() when PCRE was getting memory in which to place the information
768 created by pcre_study(). The fourth argument should point to a size_t variable.
769 pcretest has been extended so that this information is shown after a successful
770 pcre_study() call when information about the compiled regex is being displayed.
772 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
773 because what follows is always an absolute path. (Later: it turns out that this
774 is more than cosmetic for MinGW, because it doesn't like empty path
777 49. Some changes have been made to the callout feature (see 28 above):
779 (i) A callout function now has three choices for what it returns:
781 0 => success, carry on matching
782 > 0 => failure at this point, but backtrack if possible
783 < 0 => serious error, return this value from pcre_exec()
785 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
786 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
787 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
788 use by callout functions. It will never be used by PCRE itself.
790 (ii) The pcre_extra structure (see 45 above) has a void * field called
791 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
792 pcre_callout_block structure has a field of the same name. The contents of
793 the field passed in the pcre_extra structure are passed to the callout
794 function in the corresponding field in the callout block. This makes it
795 easier to use the same callout-containing regex from multiple threads. For
796 testing, the pcretest program has a new data escape
798 \C*n pass the number n (may be negative) as callout_data
800 If the callout function in pcretest receives a non-zero value as
801 callout_data, it returns that value.
803 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
804 there were some redundant $(CFLAGS) in commands that are now specified as
805 $(LINK), which already includes $(CFLAGS).
807 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
808 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
809 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
810 one-byte characters throughout. Note that case-insensitive matching applies
811 only to characters whose values are less than 256. PCRE doesn't support the
812 notion of cases for higher-valued characters.
814 (i) A character class whose characters are all within 0-255 is handled as
815 a bit map, and the map is inverted for negative classes. Previously, a
816 character > 255 always failed to match such a class; however it should
817 match if the class was a negative one (e.g. [^ab]). This has been fixed.
819 (ii) A negated character class with a single character < 255 is coded as
820 "not this character" (OP_NOT). This wasn't working properly when the test
821 character was multibyte, either singly or repeated.
823 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
824 mode, for example: \x{100}{2,3}.
826 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
827 singly or repeated) now correctly test multibyte characters. However,
828 PCRE doesn't recognize any characters with values greater than 255 as
829 digits, spaces, or word characters. Such characters always match \D, \S,
830 and \W, and never match \d, \s, or \w.
832 (v) Classes may now contain characters and character ranges with values
833 greater than 255. For example: [ab\x{100}-\x{400}].
835 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
838 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
839 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
840 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
843 53. The single man page has become too large. I have therefore split it up into
844 a number of separate man pages. These also give rise to individual HTML pages;
845 these are now put in a separate directory, and there is an index.html page that
846 lists them all. Some hyperlinking between the pages has been installed.
848 54. Added convenience functions for handling named capturing parentheses.
850 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
851 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
852 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
855 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
856 calling configure. These values are used when compiling the dftables.c program
857 which is run to generate the source of the default character tables. They
858 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
859 you will need to set these values.
861 57. Updated the building process for Windows DLL, as provided by Fred Cox.
864 Version 3.9 02-Jan-02
865 ---------------------
867 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
869 2. If --disable-static was given, the building process failed when trying to
870 build pcretest and pcregrep. (For some reason it was using libtool to compile
871 them, which is not right, as they aren't part of the library.)
874 Version 3.8 18-Dec-01
875 ---------------------
877 1. The experimental UTF-8 code was completely screwed up. It was packing the
878 bytes in the wrong order. How dumb can you get?
881 Version 3.7 29-Oct-01
882 ---------------------
884 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
885 This caused pcretest, when used on the test data, to segfault. Unfortunately,
886 this didn't happen under Solaris 8, where I normally test things.
888 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
889 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
890 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
891 this fix an hour or so after the initial 3.7 release.)
894 Version 3.6 23-Oct-01
895 ---------------------
897 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
898 offsets passed as NULL with zero offset count.
900 2. The config.guess and config.sub files had not been updated when I moved to
904 Version 3.5 15-Aug-01
905 ---------------------
907 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
910 2. By using declared but undefined structures, we can avoid using "void"
911 definitions in pcre.h while keeping the internal definitions of the structures
914 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
915 user point of view, this means that both static and shared libraries are built
916 by default, but this can be individually controlled. More of the work of
917 handling this static/shared cases is now inside libtool instead of PCRE's make
920 4. The pcretest utility is now installed along with pcregrep because it is
921 useful for users (to test regexs) and by doing this, it automatically gets
922 relinked by libtool. The documentation has been turned into a man page, so
923 there are now .1, .txt, and .html versions in /doc.
925 5. Upgrades to pcregrep:
926 (i) Added long-form option names like gnu grep.
927 (ii) Added --help to list all options with an explanatory phrase.
928 (iii) Added -r, --recursive to recurse into sub-directories.
929 (iv) Added -f, --file to read patterns from a file.
931 6. pcre_exec() was referring to its "code" argument before testing that
932 argument for NULL (and giving an error if it was NULL).
934 7. Upgraded Makefile.in to allow for compiling in a different directory from
935 the source directory.
937 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
938 options bits, the pointer it was passed was to an int instead of to an unsigned
939 long int. This mattered only on 64-bit systems.
941 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
942 generated) instead of pcre.in, which it its source. Also made the same change
943 in several of the .c files.
945 10. A new release of gcc defines printf() as a macro, which broke pcretest
946 because it had an ifdef in the middle of a string argument for printf(). Fixed
947 by using separate calls to printf().
949 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
950 script, to force use of CR or LF instead of \n in the source. On non-Unix
951 systems, the value can be set in config.h.
953 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
954 absolute limit. Changed the text of the error message to make this clear, and
955 likewise updated the man page.
957 13. The limit of 99 on the number of capturing subpatterns has been removed.
958 The new limit is 65535, which I hope will not be a "real" limit.
961 Version 3.4 22-Aug-00
962 ---------------------
964 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
966 2. Diagnose condition (?(0) as an error instead of crashing on matching.
969 Version 3.3 01-Aug-00
970 ---------------------
972 1. If an octal character was given, but the value was greater than \377, it
973 was not getting masked to the least significant bits, as documented. This could
974 lead to crashes in some systems.
976 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
977 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
979 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
980 These just pass their arguments on to (pcre_free)(), but they are provided
981 because some uses of PCRE bind it to non-C systems that can call its functions,
982 but cannot call free() or pcre_free() directly.
984 4. Add "make test" as a synonym for "make check". Corrected some comments in
987 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
990 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
991 command called pgrep for grepping around the active processes.
993 7. Added the beginnings of support for UTF-8 character strings.
995 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
996 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
997 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
998 out for the ar command.)
1001 Version 3.2 12-May-00
1002 ---------------------
1004 This is purely a bug fixing release.
1006 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1007 of ZA. This was just one example of several cases that could provoke this bug,
1008 which was introduced by change 9 of version 2.00. The code for breaking
1009 infinite loops after an iteration that matches an empty string was't working
1012 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1013 when matched against abbab (for example). After matching an empty string, it
1014 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1015 caused it to match further down the string than it should.
1017 3. The code contained an inclusion of sys/types.h. It isn't clear why this
1018 was there because it doesn't seem to be needed, and it causes trouble on some
1019 systems, as it is not a Standard C header. It has been removed.
1021 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1022 were reported on the Macintosh. The changes were from
1024 while ((c = *(++ptr)) != 0 && c != '\n');
1026 while ((c = *(++ptr)) != 0 && c != '\n') ;
1028 Totally extraordinary, but if that's what it takes...
1030 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1031 available. Added HAVE_BCOPY and an autoconf test for it; if neither
1032 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1033 assumes the way PCRE uses memmove() (always moving upwards).
1035 6. PCRE is being used in one environment where strchr() is not available. There
1036 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1040 Version 3.1 09-Feb-00
1041 ---------------------
1043 The only change in this release is the fixing of some bugs in Makefile.in for
1044 the "install" target:
1046 (1) It was failing to install pcreposix.h.
1048 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
1051 Version 3.0 01-Feb-00
1052 ---------------------
1054 1. Add support for the /+ modifier to perltest (to output $` like it does in
1057 2. Add support for the /g modifier to perltest.
1059 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
1060 matches null strings.
1062 4. Fix perltest so that it doesn't do unwanted things when fed an empty
1063 pattern. Perl treats empty patterns specially - it reuses the most recent
1064 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
1067 5. The POSIX interface was broken in that it was just handing over the POSIX
1068 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
1069 required a bigger vector, with some working space on the end. This means that
1070 the POSIX wrapper now has to get and free some memory, and copy the results.
1072 6. Added some simple autoconf support, placing the test data and the
1073 documentation in separate directories, re-organizing some of the
1074 information files, and making it build pcre-config (a GNU standard). Also added
1075 libtool support for building PCRE as a shared library, which is now the
1078 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
1079 09 are not valid octal constants. Single digits will be used for minor values
1082 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
1083 existing programs that set these in the POSIX interface can use PCRE without
1086 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
1087 return all that pcre_info() returns, plus additional data. The pcre_info()
1088 function is retained for compatibility, but is considered to be obsolete.
1090 10. Added experimental recursion feature (?R) to handle one common case that
1091 Perl 5.6 will be able to do with (?p{...}).
1093 11. Added support for POSIX character classes like [:alpha:], which Perl is
1097 Version 2.08 31-Aug-99
1098 ----------------------
1100 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
1101 trying to match at the startoffset position, but instead was moving forward to
1102 the next newline as if a previous match had failed.
1104 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
1105 and could get into a loop if a null string was matched other than at the start
1108 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
1109 be distinguished at compile time, and for completeness also added PCRE_DATE.
1111 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
1112 in GnuWin32 environments.
1115 Version 2.07 29-Jul-99
1116 ----------------------
1118 1. The documentation is now supplied in plain text form and HTML as well as in
1119 the form of man page sources.
1121 2. C++ compilers don't like assigning (void *) values to other pointer types.
1122 In particular this affects malloc(). Although there is no problem in Standard
1123 C, I've put in casts to keep C++ compilers happy.
1125 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
1126 should be (const char *).
1128 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
1129 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
1130 However, I haven't made this a standard facility. The documentation doesn't
1131 mention it, and the Makefile doesn't support it.
1133 5. The Makefile now contains an "install" target, with editable destinations at
1134 the top of the file. The pcretest program is not installed.
1136 6. pgrep -V now gives the PCRE version number and date.
1138 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1139 causing the entire string to be ignored, instead of just the last character.
1141 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1142 non-matching string, it can take a very, very long time, even for strings of
1143 quite modest length, because of the nested recursion. PCRE now does better in
1144 some of these cases. It does this by remembering the last required literal
1145 character in the pattern, and pre-searching the subject to ensure it is present
1146 before running the real match. In other words, it applies a heuristic to detect
1147 some types of certain failure quickly, and in the above example, if presented
1148 with a string that has no trailing " it gives "no match" very quickly.
1150 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1151 other alternatives are tried instead.
1154 Version 2.06 09-Jun-99
1155 ----------------------
1157 1. Change pcretest's output for amount of store used to show just the code
1158 space, because the remainder (the data block) varies in size between 32-bit and
1161 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1162 start matching at. This allows lookbehinds to work when searching for multiple
1163 occurrences in a string.
1165 3. Added additional options to pcretest for testing multiple occurrences:
1167 /+ outputs the rest of the string that follows a match
1168 /g loops for multiple occurrences, using the new startoffset argument
1169 /G loops for multiple occurrences by passing an incremented pointer
1171 4. PCRE wasn't doing the "first character" optimization for patterns starting
1172 with \b or \B, though it was doing it for other lookbehind assertions. That is,
1173 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1174 the letter 'x'. On long subject strings, this gives a significant speed-up.
1177 Version 2.05 21-Apr-99
1178 ----------------------
1180 1. Changed the type of magic_number from int to long int so that it works
1181 properly on 16-bit systems.
1183 2. Fixed a bug which caused patterns starting with .* not to work correctly
1184 when the subject string contained newline characters. PCRE was assuming
1185 anchoring for such patterns in all cases, which is not correct because .* will
1186 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1187 DOTALL is set at top level; otherwise it knows that patterns starting with .*
1188 must be retried after every newline in the subject.
1191 Version 2.04 18-Feb-99
1192 ----------------------
1194 1. For parenthesized subpatterns with repeats whose minimum was zero, the
1195 computation of the store needed to hold the pattern was incorrect (too large).
1196 If such patterns were nested a few deep, this could multiply and become a real
1199 2. Added /M option to pcretest to show the memory requirement of a specific
1200 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1202 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1203 compiled in such a way that the backtracking after subsequent failure was
1204 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1205 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1208 Version 2.03 02-Feb-99
1209 ----------------------
1211 1. Fixed typo and small mistake in man page.
1213 2. Added 4th condition (GPL supersedes if conflict) and created separate
1214 LICENCE file containing the conditions.
1216 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1217 Perl, that is the internal \ allows the delimiter to be included in the
1218 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1219 the final delimiter, add \ to the end of the pattern (to test the error).
1221 4. Added the convenience functions for extracting substrings after a successful
1222 match. Updated pcretest to make it able to test these functions.
1225 Version 2.02 14-Jan-99
1226 ----------------------
1228 1. Initialized the working variables associated with each extraction so that
1229 their saving and restoring doesn't refer to uninitialized store.
1231 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1232 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1235 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1236 calls, and wasn't printing the correct value for compiling calls. Increased the
1237 default value of LOOPREPEAT, and the number of significant figures in the
1240 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1242 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1243 a building problem on Windows NT with a FAT file system.
1246 Version 2.01 21-Oct-98
1247 ----------------------
1249 1. Changed the API for pcre_compile() to allow for the provision of a pointer
1250 to character tables built by pcre_maketables() in the current locale. If NULL
1251 is passed, the default tables are used.
1254 Version 2.00 24-Sep-98
1255 ----------------------
1257 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1260 2. Allow quantification of (?>) groups, and make it work correctly.
1262 3. The first character computation wasn't working for (?>) groups.
1264 4. Correct the implementation of \Z (it is permitted to match on the \n at the
1265 end of the subject) and add 5.005's \z, which really does match only at the
1266 very end of the subject.
1268 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1270 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1271 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1272 localized options. All options to pcre_study() were also removed.
1274 7. Add other new features from 5.005:
1276 $(?<= positive lookbehind
1277 $(?<! negative lookbehind
1278 (?imsx-imsx) added the unsetting capability
1279 such a setting is global if at outer level; local otherwise
1280 (?imsx-imsx:) non-capturing groups with option setting
1281 (?(cond)re|re) conditional pattern matching
1283 A backreference to itself in a repeated group matches the previous
1286 8. General tidying up of studying (both automatic and via "study")
1287 consequential on the addition of new assertions.
1289 9. As in 5.005, unlimited repeated groups that could match an empty substring
1290 are no longer faulted at compile time. Instead, the loop is forcibly broken at
1291 runtime if any iteration does actually match an empty substring.
1293 10. Include the RunTest script in the distribution.
1295 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1296 discrepancies, some of which were old and were also with respect to 5.004. They
1297 have now been fixed.
1300 Version 1.09 28-Apr-98
1301 ----------------------
1303 1. A negated single character class followed by a quantifier with a minimum
1304 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
1305 program crashes, or just wrong answers. This did not apply to negated classes
1306 containing more than one character, or to minima other than one.
1309 Version 1.08 27-Mar-98
1310 ----------------------
1312 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
1314 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
1315 latter must appear before anything that relies on it in the pattern.
1318 Version 1.07 16-Feb-98
1319 ----------------------
1321 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
1322 repeat of a potentially empty string).
1325 Version 1.06 23-Jan-98
1326 ----------------------
1328 1. Added Markus Oberhumer's little patches for C++.
1330 2. Literal strings longer than 255 characters were broken.
1333 Version 1.05 23-Dec-97
1334 ----------------------
1336 1. Negated character classes containing more than one character were failing if
1337 PCRE_CASELESS was set at run time.
1340 Version 1.04 19-Dec-97
1341 ----------------------
1343 1. Corrected the man page, where some "const" qualifiers had been omitted.
1345 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
1348 3. Fixed memory leak which occurred when a regex with back references was
1349 matched with an offsets vector that wasn't big enough. The temporary memory
1350 that is used in this case wasn't being freed if the match failed.
1352 4. Tidied pcretest to ensure it frees memory that it gets.
1354 5. Temporary memory was being obtained in the case where the passed offsets
1355 vector was exactly big enough.
1357 6. Corrected definition of offsetof() from change 5 below.
1359 7. I had screwed up change 6 below and broken the rules for the use of
1360 setjmp(). Now fixed.
1363 Version 1.03 18-Dec-97
1364 ----------------------
1366 1. A erroneous regex with a missing opening parenthesis was correctly
1367 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
1370 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
1371 it was reported that one broken compiler failed on the former because "code" is
1372 also an independent variable.
1374 3. The erroneous regex a[]b caused an array overrun reference.
1376 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
1377 fail on data ending with that character. (It was going on too far, and checking
1378 the next character, typically a binary zero.) This was specific to the
1379 optimized code for single-character negative classes.
1381 5. Added a contributed patch from the TIN world which does the following:
1383 + Add an undef for memmove, in case the the system defines a macro for it.
1385 + Add a definition of offsetof(), in case there isn't one. (I don't know
1386 the reason behind this - offsetof() is part of the ANSI standard - but
1389 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
1390 most of the places where whitespace preceded '#'. I have given up and
1391 allowed the remaining 2 cases to be at the margin.
1393 + Rename some variables in pcre to eliminate shadowing. This seems very
1394 pedantic, but does no harm, of course.
1396 6. Moved the call to setjmp() into its own function, to get rid of warnings
1397 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
1399 7. Constructs such as \d{8,} were compiling into the equivalent of
1400 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
1401 outcome, but in this particular case used more store than had been allocated,
1402 which caused the bug to be discovered because it threw up an internal error.
1404 8. The debugging code in both pcre and pcretest for outputting the compiled
1405 form of a regex was going wrong in the case of back references followed by
1406 curly-bracketed repeats.
1409 Version 1.02 12-Dec-97
1410 ----------------------
1412 1. Typos in pcre.3 and comments in the source fixed.
1414 2. Applied a contributed patch to get rid of places where it used to remove
1415 'const' from variables, and fixed some signed/unsigned and uninitialized
1418 3. Added the "runtest" target to Makefile.
1420 4. Set default compiler flag to -O2 rather than just -O.
1423 Version 1.01 19-Nov-97
1424 ----------------------
1426 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
1427 like /([ab]*)*/, that is, for classes with more than one character in them.
1429 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
1430 as /((?>a*))*/ (a PCRE_EXTRA facility).
1433 Version 1.00 18-Nov-97
1434 ----------------------
1436 1. Added compile-time macros to support systems such as SunOS4 which don't have
1437 memmove() or strerror() but have other things that can be used instead.
1439 2. Arranged that "make clean" removes the executables.
1442 Version 0.99 27-Oct-97
1443 ----------------------
1445 1. Fixed bug in code for optimizing classes with only one character. It was
1446 initializing a 32-byte map regardless, which could cause it to run off the end
1447 of the memory it had got.
1449 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1452 Version 0.98 22-Oct-97
1453 ----------------------
1455 1. Fixed bug in code for handling temporary memory usage when there are more
1456 back references than supplied space in the ovector. This could cause segfaults.
1459 Version 0.97 21-Oct-97
1460 ----------------------
1462 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1464 2. Optimized negated single characters not to use a bit map.
1466 3. Brought error texts together as macro definitions; clarified some of them;
1467 fixed one that was wrong - it said "range out of order" when it meant "invalid
1470 4. Changed some char * arguments to const char *.
1472 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1474 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1478 Version 0.96 16-Oct-97
1479 ----------------------
1481 1. Added a simple "pgrep" utility to the distribution.
1483 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1484 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1485 where "ddd" means "one or more decimal digits".
1487 3. Fixed serious bug. If a pattern had a back reference, but the call to
1488 pcre_exec() didn't supply a large enough ovector to record the related
1489 identifying subpattern, the match always failed. PCRE now remembers the number
1490 of the largest back reference, and gets some temporary memory in which to save
1491 the offsets during matching if necessary, in order to ensure that
1492 backreferences always work.
1494 4. Increased the compatibility with Perl in a number of ways:
1496 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1497 to request this handling. The option can be set at compile or exec time.
1499 (b) $ matches before a terminating newline by default; an option
1500 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
1501 mode). The option can be set at compile or exec time.
1503 (c) The handling of \ followed by a digit other than 0 is now supposed to be
1504 the same as Perl's. If the decimal number it represents is less than 10
1505 or there aren't that many previous left capturing parentheses, an octal
1506 escape is read. Inside a character class, it's always an octal escape,
1507 even if it is a single digit.
1509 (d) An escaped but undefined alphabetic character is taken as a literal,
1510 unless PCRE_EXTRA is set. Currently this just reserves the remaining
1513 (e) {0} is now permitted. (The previous item is removed from the compiled
1516 5. Changed all the names of code files so that the basic parts are no longer
1517 than 10 characters, and abolished the teeny "globals.c" file.
1519 6. Changed the handling of character classes; they are now done with a 32-byte
1522 7. Added the -d and /D options to pcretest to make it possible to look at the
1523 internals of compilation without having to recompile pcre.
1526 Version 0.95 23-Sep-97
1527 ----------------------
1529 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
1530 \x20 at the start of a run of normal characters. These were being treated as
1531 real characters, instead of the source characters being re-checked.
1534 Version 0.94 18-Sep-97
1535 ----------------------
1537 1. The functions are now thread-safe, with the caveat that the global variables
1538 containing pointers to malloc() and free() or alternative functions are the
1539 same for all threads.
1541 2. Get pcre_study() to generate a bitmap of initial characters for non-
1542 anchored patterns when this is possible, and use it if passed to pcre_exec().
1545 Version 0.93 15-Sep-97
1546 ----------------------
1548 1. /(b)|(:+)/ was computing an incorrect first character.
1550 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
1551 but not actually doing anything yet.
1553 3. Treat "-" characters in classes that cannot be part of ranges as literals,
1554 as Perl does (e.g. [-az] or [az-]).
1556 4. Set the anchored flag if a branch starts with .* or .*? because that tests
1557 all possible positions.
1559 5. Split up into different modules to avoid including unneeded functions in a
1560 compiled binary. However, compile and exec are still in one module. The "study"
1561 function is split off.
1563 6. The character tables are now in a separate module whose source is generated
1564 by an auxiliary program - but can then be edited by hand if required. There are
1565 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
1566 toupper() in the code.
1568 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
1569 make them global. Abolish the function for setting them, as the caller can now
1573 Version 0.92 11-Sep-97
1574 ----------------------
1576 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
1577 (e.g. /a{1,3}/) was broken (I mis-optimized it).
1579 2. Caseless matching was not working in character classes if the characters in
1580 the pattern were in upper case.
1582 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
1584 4. Make PCRE_ANCHORED public and accept as a compile option.
1586 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
1587 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
1590 6. Give an error if bad option bits passed at compile or run time.
1592 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
1593 pcretest to cause it to pass that flag.
1595 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
1596 options, and the first character, if set.
1598 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
1601 Version 0.91 10-Sep-97
1602 ----------------------
1604 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
1605 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
1607 2. PCRE was looping on encountering an indefinitely repeated back reference to
1608 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
1609 Perl does - treats the match as successful.
git://git.exim.org / exim.git / blob