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

Diff of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 19 by nigel, Sat Feb 24 21:38:33 2007 UTC revision 71 by nigel, Sat Feb 24 21:40:24 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 4.4 13-Aug-03
5    ---------------------
6    
7     1. In UTF-8 mode, a character class containing characters with values between
8        127 and 255 was not handled correctly if the compiled pattern was studied.
9        In fixing this, I have also improved the studying algorithm for such
10        classes (slightly).
11    
12     2. Three internal functions had redundant arguments passed to them. Removal
13        might give a very teeny performance improvement.
14    
15     3. Documentation bug: the value of the capture_top field in a callout is *one
16        more than* the number of the hightest numbered captured substring.
17    
18     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
19        in incorrectly linking with a previously installed version. They now link
20        explicitly with libpcre.la.
21    
22     5. configure.in no longer needs to recognize Cygwin specially.
23    
24     6. A problem in pcre.in for Windows platforms is fixed.
25    
26     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
27        pcretest, it used to include the size of the study block as part of its
28        output. Unfortunately, the structure contains a field that has a different
29        size on different hardware architectures. This meant that the tests that
30        showed this size failed. As the block is currently always of a fixed size,
31        this information isn't actually particularly useful in pcretest output, so
32        I have just removed it.
33    
34     8. Three pre-processor statements accidentally did not start in column 1.
35        Sadly, there are *still* compilers around that complain, even though
36        standard C has not required this for well over a decade. Sigh.
37    
38     9. In pcretest, the code for checking callouts passed small integers in the
39        callout_data field, which is a void * field. However, some picky compilers
40        complained about the casts involved for this on 64-bit systems. Now
41        pcretest passes the address of the small integer instead, which should get
42        rid of the warnings.
43    
44    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
45        both compile and run time, and gives an error if an invalid UTF-8 sequence
46        is found. There is a option for disabling this check in cases where the
47        string is known to be correct and/or the maximum performance is wanted.
48    
49    11. In response to a bug report, I changed one line in Makefile.in from
50    
51            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
52        to
53            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
54    
55        to look similar to other lines, but I have no way of telling whether this
56        is the right thing to do, as I do not use Windows. No doubt I'll get told
57        if it's wrong...
58    
59    
60    Version 4.3 21-May-03
61    ---------------------
62    
63    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
64       Makefile.
65    
66    2. Some refactoring to improve the quality of the code:
67    
68       (i)   The utf8_table... variables are now declared "const".
69    
70       (ii)  The code for \cx, which used the "case flipping" table to upper case
71             lower case letters, now just substracts 32. This is ASCII-specific,
72             but the whole concept of \cx is ASCII-specific, so it seems
73             reasonable.
74    
75       (iii) PCRE was using its character types table to recognize decimal and
76             hexadecimal digits in the pattern. This is silly, because it handles
77             only 0-9, a-f, and A-F, but the character types table is locale-
78             specific, which means strange things might happen. A private
79             table is now used for this - though it costs 256 bytes, a table is
80             much faster than multiple explicit tests. Of course, the standard
81             character types table is still used for matching digits in subject
82             strings against \d.
83    
84       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
85             ending in _t are). So I've renamed it as ESC_tee.
86    
87    3. The first argument for regexec() in the POSIX wrapper should have been
88       defined as "const".
89    
90    4. Changed pcretest to use malloc() for its buffers so that they can be
91       Electric Fenced for debugging.
92    
93    5. There were several places in the code where, in UTF-8 mode, PCRE would try
94       to read one or more bytes before the start of the subject string. Often this
95       had no effect on PCRE's behaviour, but in some circumstances it could
96       provoke a segmentation fault.
97    
98    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
99       to try to read one or more bytes before the start of the subject string.
100    
101    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
102       UTF-8 support could misbehave in various ways if the subject string
103       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
104       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
105       back over UTF-8 characters.)
106    
107    
108    Version 4.2 14-Apr-03
109    ---------------------
110    
111    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
112    
113    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
114         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
115         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
116         [WIN_PREFIX]: new variable, "cyg" for Cygwin
117         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
118           and BUILD_EXEEXT
119         Note: automatic setting of the BUILD variables is not yet working
120         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
121           compile-time but not at link-time
122         [LINK]: use for linking executables only
123         make different versions for Windows and non-Windows
124         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
125           libraries
126         [LINK_FOR_BUILD]: new variable
127         [OBJEXT]: use throughout
128         [EXEEXT]: use throughout
129         <winshared>: new target
130         <wininstall>: new target
131         <dftables.o>: use native compiler
132         <dftables>: use native linker
133         <install>: handle Windows platform correctly
134         <clean>: ditto
135         <check>: ditto
136         copy DLL to top builddir before testing
137    
138       As part of these changes, -no-undefined was removed again. This was reported
139       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
140       in any case.
141    
142    3. Some tidies to get rid of compiler warnings:
143    
144       . In the match_data structure, match_limit was an unsigned long int, whereas
145         match_call_count was an int. I've made them both unsigned long ints.
146    
147       . In pcretest the fact that a const uschar * doesn't automatically cast to
148         a void * provoked a warning.
149    
150       . Turning on some more compiler warnings threw up some "shadow" variables
151         and a few more missing casts.
152    
153    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
154       option, a class that contained a single character with a value between 128
155       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
156    
157    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
158       option, a class that contained several characters, but with at least one
159       whose value was between 128 and 255 caused PCRE to crash.
160    
161    
162    Version 4.1 12-Mar-03
163    ---------------------
164    
165    1. Compiling with gcc -pedantic found a couple of places where casts were
166    needed, and a string in dftables.c that was longer than standard compilers are
167    required to support.
168    
169    2. Compiling with Sun's compiler found a few more places where the code could
170    be tidied up in order to avoid warnings.
171    
172    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
173    first of these names is deprecated in the latest Autoconf in favour of the name
174    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
175    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
176    analogy I have changed it to CFLAGS_FOR_BUILD.
177    
178    4. Added -no-undefined to the linking command in the Makefile, because this is
179    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
180    linking step for the pcreposix library.
181    
182    5. PCRE was failing to diagnose the case of two named groups with the same
183    name.
184    
185    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
186    literal character that is needed in the subject for a match, and scans along to
187    ensure that it is present before embarking on the full matching process. This
188    saves time in cases of nested unlimited repeats that are never going to match.
189    Problem: the scan can take a lot of time if the subject is very long (e.g.
190    megabytes), thus penalizing straightforward matches. It is now done only if the
191    amount of subject to be scanned is less than 1000 bytes.
192    
193    7. A lesser problem with the same optimization is that it was recording the
194    first character of an anchored pattern as "needed", thus provoking a search
195    right along the subject, even when the first match of the pattern was going to
196    fail. The "needed" character is now not set for anchored patterns, unless it
197    follows something in the pattern that is of non-fixed length. Thus, it still
198    fulfils its original purpose of finding quick non-matches in cases of nested
199    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
200    
201    
202    Version 4.0 17-Feb-03
203    ---------------------
204    
205    1. If a comment in an extended regex that started immediately after a meta-item
206    extended to the end of string, PCRE compiled incorrect data. This could lead to
207    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
208    
209    2. Moved to autoconf 2.53 and libtool 1.4.2.
210    
211    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
212    the special perltest8 script is no longer needed - all the tests can be run
213    from a single perltest script.
214    
215    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
216    by \s. It has now been removed in PCRE. This means it isn't recognized as
217    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
218    class [:space:] *does* include VT, thereby creating a mess.
219    
220    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
221    space and tab.
222    
223    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
224    its new features into the main test script, reducing the number of scripts.
225    
226    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
227    were backward compatible, and made the (?i) apply to the whole pattern, as if
228    /i were given. Now it behaves more logically, and applies the option setting
229    only to what follows. PCRE has been changed to follow suit. However, if it
230    finds options settings right at the start of the pattern, it extracts them into
231    the global options, as before. Thus, they show up in the info data.
232    
233    8. Added support for the \Q...\E escape sequence. Characters in between are
234    treated as literals. This is slightly different from Perl in that $ and @ are
235    also handled as literals inside the quotes. In Perl, they will cause variable
236    interpolation. Note the following examples:
237    
238        Pattern            PCRE matches      Perl matches
239    
240        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
241        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
242        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
243    
244    For compatibility with Perl, \Q...\E sequences are recognized inside character
245    classes as well as outside them.
246    
247    9. Re-organized 3 code statements in pcretest to avoid "overflow in
248    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
249    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
250    signed/unsigned warnings.
251    
252    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
253    option for pcretest, so I've replaced it by a simple function that does just
254    that job.
255    
256    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
257    "pcregrep -".
258    
259    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
260    Java package. This provides some syntactic sugar for simple cases of what my
261    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
262    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
263    item, you can use this simplified notation. Note that only makes sense with
264    greedy quantifiers. Consequently, the use of the possessive quantifier forces
265    greediness, whatever the setting of the PCRE_UNGREEDY option.
266    
267    13. A change of greediness default within a pattern was not taking effect at
268    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
269    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
270    was abstracted outside.
271    
272    14. PCRE now supports the \G assertion. It is true when the current matching
273    position is at the start point of the match. This differs from \A when the
274    starting offset is non-zero. Used with the /g option of pcretest (or similar
275    code), it works in the same way as it does for Perl's /g option. If all
276    alternatives of a regex begin with \G, the expression is anchored to the start
277    match position, and the "anchored" flag is set in the compiled expression.
278    
279    15. Some bugs concerning the handling of certain option changes within patterns
280    have been fixed. These applied to options other than (?ims). For example,
281    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
282    been the other way round. Some of this was related to change 7 above.
283    
284    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
285    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
286    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
287    POSIX classes only within a class (e.g. /[[:alpha:]]/).
288    
289    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
290    mode. Unlike ".", it always matches newline, whatever the setting of
291    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
292    assertions. Perl allows it, but it doesn't (in general) work because it can't
293    calculate the length of the lookbehind. At least, that's the case for Perl
294    5.8.0 - I've been told they are going to document that it doesn't work in
295    future.
296    
297    18. Added an error diagnosis for escapes that PCRE does not support: these are
298    \L, \l, \N, \P, \p, \U, \u, and \X.
299    
300    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
301    reading past the end of the pattern in cases such as /[abcd/.
302    
303    20. PCRE was getting more memory than necessary for patterns with classes that
304    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
305    
306    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
307    compiling PCRE for use with Virtual Pascal.
308    
309    22. Small fix to the Makefile to make it work properly if the build is done
310    outside the source tree.
311    
312    23. Added a new extension: a condition to go with recursion. If a conditional
313    subpattern starts with (?(R) the "true" branch is used if recursion has
314    happened, whereas the "false" branch is used only at the top level.
315    
316    24. When there was a very long string of literal characters (over 255 bytes
317    without UTF support, over 250 bytes with UTF support), the computation of how
318    much memory was required could be incorrect, leading to segfaults or other
319    strange effects.
320    
321    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
322    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
323    there was a subsequent back reference to those brackets. This meant that, for
324    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
325    possible to check for precisely this case. All we can do is abandon the
326    optimization if .* occurs inside capturing brackets when there are any back
327    references whatsoever. (See below for a better fix that came later.)
328    
329    26. The handling of the optimization for finding the first character of a
330    non-anchored pattern, and for finding a character that is required later in the
331    match were failing in some cases. This didn't break the matching; it just
332    failed to optimize when it could. The way this is done has been re-implemented.
333    
334    27. Fixed typo in error message for invalid (?R item (it said "(?p").
335    
336    28. Added a new feature that provides some of the functionality that Perl
337    provides with (?{...}). The facility is termed a "callout". The way it is done
338    in PCRE is for the caller to provide an optional function, by setting
339    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
340    global variable. By default it is unset, which disables all calling out. To get
341    the function called, the regex must include (?C) at appropriate points. This
342    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
343    This provides a means of identifying different callout points. When PCRE
344    reaches such a point in the regex, if pcre_callout has been set, the external
345    function is called. It is provided with data in a structure called
346    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
347    matching continues; if it returns a non-zero value, the match at the current
348    point fails. However, backtracking will occur if possible. [This was changed
349    later and other features added - see item 49 below.]
350    
351    29. pcretest is upgraded to test the callout functionality. It provides a
352    callout function that displays information. By default, it shows the start of
353    the match and the current position in the text. There are some new data escapes
354    to vary what happens:
355    
356        \C+         in addition, show current contents of captured substrings
357        \C-         do not supply a callout function
358        \C!n        return 1 when callout number n is reached
359        \C!n!m      return 1 when callout number n is reached for the mth time
360    
361    30. If pcregrep was called with the -l option and just a single file name, it
362    output "<stdin>" if a match was found, instead of the file name.
363    
364    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
365    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
366    pcre_exec(). This saves a malloc/free per call. The default value of
367    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
368    when configuring.
369    
370    32. The default maximum size of a compiled pattern is 64K. There have been a
371    few cases of people hitting this limit. The code now uses macros to handle the
372    storing of links as offsets within the compiled pattern. It defaults to 2-byte
373    links, but this can be changed to 3 or 4 bytes by --with-link-size when
374    configuring. Tests 2 and 5 work only with 2-byte links because they output
375    debugging information about compiled patterns.
376    
377    33. Internal code re-arrangements:
378    
379    (a) Moved the debugging function for printing out a compiled regex into
380        its own source file (printint.c) and used #include to pull it into
381        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
382        separate copies.
383    
384    (b) Defined the list of op-code names for debugging as a macro in
385        internal.h so that it is next to the definition of the opcodes.
386    
387    (c) Defined a table of op-code lengths for simpler skipping along compiled
388        code. This is again a macro in internal.h so that it is next to the
389        definition of the opcodes.
390    
391    34. Added support for recursive calls to individual subpatterns, along the
392    lines of Robin Houston's patch (but implemented somewhat differently).
393    
394    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
395    allow it to read and process whole directories in Win32. This code was
396    contributed by Lionel Fourquaux; it has not been tested by me.
397    
398    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
399    used to name a group. Names consist of alphanumerics and underscores, and must
400    be unique. Back references use the syntax (?P=name) and recursive calls use
401    (?P>name) which is a PCRE extension to the Python extension. Groups still have
402    numbers. The function pcre_fullinfo() can be used after compilation to extract
403    a name/number map. There are three relevant calls:
404    
405      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
406      PCRE_INFO_NAMECOUNT            yields the number of entries
407      PCRE_INFO_NAMETABLE            yields a pointer to the map.
408    
409    The map is a vector of fixed-size entries. The size of each entry depends on
410    the length of the longest name used. The first two bytes of each entry are the
411    group number, most significant byte first. There follows the corresponding
412    name, zero terminated. The names are in alphabetical order.
413    
414    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
415    case instead of 255. Making it the same both with and without UTF-8 support
416    means that the same test output works with both.
417    
418    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
419    calling malloc() with a zero argument.
420    
421    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
422    optimization. I've improved things by keeping a bitmap of backreferences with
423    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
424    fact referenced, the optimization can be applied. It is unlikely that a
425    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
426    the match to follow \n) will appear inside brackets with a number greater than
427    31, but if it does, any back reference > 31 suppresses the optimization.
428    
429    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
430    of disabling numbered capturing parentheses. Any opening parenthesis that is
431    not followed by ? behaves as if it were followed by ?: but named parentheses
432    can still be used for capturing (and they will acquire numbers in the usual
433    way).
434    
435    41. Redesigned the return codes from the match() function into yes/no/error so
436    that errors can be passed back from deep inside the nested calls. A malloc
437    failure while inside a recursive subpattern call now causes the
438    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
439    
440    42. It is now possible to set a limit on the number of times the match()
441    function is called in a call to pcre_exec(). This facility makes it possible to
442    limit the amount of recursion and backtracking, though not in a directly
443    obvious way, because the match() function is used in a number of different
444    circumstances. The count starts from zero for each position in the subject
445    string (for non-anchored patterns). The default limit is, for compatibility, a
446    large number, namely 10 000 000. You can change this in two ways:
447    
448    (a) When configuring PCRE before making, you can use --with-match-limit=n
449        to set a default value for the compiled library.
450    
451    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
452        a different value is set. See 45 below.
453    
454    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
455    
456    43. Added a new function pcre_config(int, void *) to enable run-time extraction
457    of things that can be changed at compile time. The first argument specifies
458    what is wanted and the second points to where the information is to be placed.
459    The current list of available information is:
460    
461      PCRE_CONFIG_UTF8
462    
463    The output is an integer that is set to one if UTF-8 support is available;
464    otherwise it is set to zero.
465    
466      PCRE_CONFIG_NEWLINE
467    
468    The output is an integer that it set to the value of the code that is used for
469    newline. It is either LF (10) or CR (13).
470    
471      PCRE_CONFIG_LINK_SIZE
472    
473    The output is an integer that contains the number of bytes used for internal
474    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
475    
476      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
477    
478    The output is an integer that contains the threshold above which the POSIX
479    interface uses malloc() for output vectors. See item 31 above.
480    
481      PCRE_CONFIG_MATCH_LIMIT
482    
483    The output is an unsigned integer that contains the default limit of the number
484    of match() calls in a pcre_exec() execution. See 42 above.
485    
486    44. pcretest has been upgraded by the addition of the -C option. This causes it
487    to extract all the available output from the new pcre_config() function, and to
488    output it. The program then exits immediately.
489    
490    45. A need has arisen to pass over additional data with calls to pcre_exec() in
491    order to support additional features. One way would have been to define
492    pcre_exec2() (for example) with extra arguments, but this would not have been
493    extensible, and would also have required all calls to the original function to
494    be mapped to the new one. Instead, I have chosen to extend the mechanism that
495    is used for passing in "extra" data from pcre_study().
496    
497    The pcre_extra structure is now exposed and defined in pcre.h. It currently
498    contains the following fields:
499    
500      flags         a bitmap indicating which of the following fields are set
501      study_data    opaque data from pcre_study()
502      match_limit   a way of specifying a limit on match() calls for a specific
503                      call to pcre_exec()
504      callout_data  data for callouts (see 49 below)
505    
506    The flag bits are also defined in pcre.h, and are
507    
508      PCRE_EXTRA_STUDY_DATA
509      PCRE_EXTRA_MATCH_LIMIT
510      PCRE_EXTRA_CALLOUT_DATA
511    
512    The pcre_study() function now returns one of these new pcre_extra blocks, with
513    the actual study data pointed to by the study_data field, and the
514    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
515    before. That is, this change is entirely upwards-compatible and requires no
516    change to existing code.
517    
518    If you want to pass in additional data to pcre_exec(), you can either place it
519    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
520    block.
521    
522    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
523    data string contains the escape sequence \M, pcretest calls pcre_exec() several
524    times with different match limits, until it finds the minimum value needed for
525    pcre_exec() to complete. The value is then output. This can be instructive; for
526    most simple matches the number is quite small, but for pathological cases it
527    gets very large very quickly.
528    
529    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
530    returns the size of the data block pointed to by the study_data field in a
531    pcre_extra block, that is, the value that was passed as the argument to
532    pcre_malloc() when PCRE was getting memory in which to place the information
533    created by pcre_study(). The fourth argument should point to a size_t variable.
534    pcretest has been extended so that this information is shown after a successful
535    pcre_study() call when information about the compiled regex is being displayed.
536    
537    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
538    because what follows is always an absolute path. (Later: it turns out that this
539    is more than cosmetic for MinGW, because it doesn't like empty path
540    components.)
541    
542    49. Some changes have been made to the callout feature (see 28 above):
543    
544    (i)  A callout function now has three choices for what it returns:
545    
546           0  =>  success, carry on matching
547         > 0  =>  failure at this point, but backtrack if possible
548         < 0  =>  serious error, return this value from pcre_exec()
549    
550         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
551         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
552         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
553         use by callout functions. It will never be used by PCRE itself.
554    
555    (ii) The pcre_extra structure (see 45 above) has a void * field called
556         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
557         pcre_callout_block structure has a field of the same name. The contents of
558         the field passed in the pcre_extra structure are passed to the callout
559         function in the corresponding field in the callout block. This makes it
560         easier to use the same callout-containing regex from multiple threads. For
561         testing, the pcretest program has a new data escape
562    
563           \C*n        pass the number n (may be negative) as callout_data
564    
565         If the callout function in pcretest receives a non-zero value as
566         callout_data, it returns that value.
567    
568    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
569    there were some redundant $(CFLAGS) in commands that are now specified as
570    $(LINK), which already includes $(CFLAGS).
571    
572    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
573    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
574    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
575    one-byte characters throughout. Note that case-insensitive matching applies
576    only to characters whose values are less than 256. PCRE doesn't support the
577    notion of cases for higher-valued characters.
578    
579    (i)   A character class whose characters are all within 0-255 is handled as
580          a bit map, and the map is inverted for negative classes. Previously, a
581          character > 255 always failed to match such a class; however it should
582          match if the class was a negative one (e.g. [^ab]). This has been fixed.
583    
584    (ii)  A negated character class with a single character < 255 is coded as
585          "not this character" (OP_NOT). This wasn't working properly when the test
586          character was multibyte, either singly or repeated.
587    
588    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
589          mode, for example: \x{100}{2,3}.
590    
591    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
592          singly or repeated) now correctly test multibyte characters. However,
593          PCRE doesn't recognize any characters with values greater than 255 as
594          digits, spaces, or word characters. Such characters always match \D, \S,
595          and \W, and never match \d, \s, or \w.
596    
597    (v)   Classes may now contain characters and character ranges with values
598          greater than 255. For example: [ab\x{100}-\x{400}].
599    
600    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
601          PCRE in UTF-8 mode.
602    
603    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
604    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
605    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
606    value.)
607    
608    53. The single man page has become too large. I have therefore split it up into
609    a number of separate man pages. These also give rise to individual HTML pages;
610    these are now put in a separate directory, and there is an index.html page that
611    lists them all. Some hyperlinking between the pages has been installed.
612    
613    54. Added convenience functions for handling named capturing parentheses.
614    
615    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
616    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
617    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
618    are faulted.
619    
620    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
621    calling configure. These values are used when compiling the dftables.c program
622    which is run to generate the source of the default character tables. They
623    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
624    you will need to set these values.
625    
626    57. Updated the building process for Windows DLL, as provided by Fred Cox.
627    
628    
629    Version 3.9 02-Jan-02
630    ---------------------
631    
632    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
633    
634    2. If --disable-static was given, the building process failed when trying to
635    build pcretest and pcregrep. (For some reason it was using libtool to compile
636    them, which is not right, as they aren't part of the library.)
637    
638    
639    Version 3.8 18-Dec-01
640    ---------------------
641    
642    1. The experimental UTF-8 code was completely screwed up. It was packing the
643    bytes in the wrong order. How dumb can you get?
644    
645    
646    Version 3.7 29-Oct-01
647    ---------------------
648    
649    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
650    This caused pcretest, when used on the test data, to segfault. Unfortunately,
651    this didn't happen under Solaris 8, where I normally test things.
652    
653    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
654    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
655    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
656    this fix an hour or so after the initial 3.7 release.)
657    
658    
659    Version 3.6 23-Oct-01
660    ---------------------
661    
662    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
663    offsets passed as NULL with zero offset count.
664    
665    2. The config.guess and config.sub files had not been updated when I moved to
666    the latest autoconf.
667    
668    
669    Version 3.5 15-Aug-01
670    ---------------------
671    
672    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
673    had been forgotten.
674    
675    2. By using declared but undefined structures, we can avoid using "void"
676    definitions in pcre.h while keeping the internal definitions of the structures
677    private.
678    
679    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
680    user point of view, this means that both static and shared libraries are built
681    by default, but this can be individually controlled. More of the work of
682    handling this static/shared cases is now inside libtool instead of PCRE's make
683    file.
684    
685    4. The pcretest utility is now installed along with pcregrep because it is
686    useful for users (to test regexs) and by doing this, it automatically gets
687    relinked by libtool. The documentation has been turned into a man page, so
688    there are now .1, .txt, and .html versions in /doc.
689    
690    5. Upgrades to pcregrep:
691       (i)   Added long-form option names like gnu grep.
692       (ii)  Added --help to list all options with an explanatory phrase.
693       (iii) Added -r, --recursive to recurse into sub-directories.
694       (iv)  Added -f, --file to read patterns from a file.
695    
696    6. pcre_exec() was referring to its "code" argument before testing that
697    argument for NULL (and giving an error if it was NULL).
698    
699    7. Upgraded Makefile.in to allow for compiling in a different directory from
700    the source directory.
701    
702    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
703    options bits, the pointer it was passed was to an int instead of to an unsigned
704    long int. This mattered only on 64-bit systems.
705    
706    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
707    generated) instead of pcre.in, which it its source. Also made the same change
708    in several of the .c files.
709    
710    10. A new release of gcc defines printf() as a macro, which broke pcretest
711    because it had an ifdef in the middle of a string argument for printf(). Fixed
712    by using separate calls to printf().
713    
714    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
715    script, to force use of CR or LF instead of \n in the source. On non-Unix
716    systems, the value can be set in config.h.
717    
718    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
719    absolute limit. Changed the text of the error message to make this clear, and
720    likewise updated the man page.
721    
722    13. The limit of 99 on the number of capturing subpatterns has been removed.
723    The new limit is 65535, which I hope will not be a "real" limit.
724    
725    
726    Version 3.4 22-Aug-00
727    ---------------------
728    
729    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
730    
731    2. Diagnose condition (?(0) as an error instead of crashing on matching.
732    
733    
734    Version 3.3 01-Aug-00
735    ---------------------
736    
737    1. If an octal character was given, but the value was greater than \377, it
738    was not getting masked to the least significant bits, as documented. This could
739    lead to crashes in some systems.
740    
741    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
742    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
743    
744    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
745    These just pass their arguments on to (pcre_free)(), but they are provided
746    because some uses of PCRE bind it to non-C systems that can call its functions,
747    but cannot call free() or pcre_free() directly.
748    
749    4. Add "make test" as a synonym for "make check". Corrected some comments in
750    the Makefile.
751    
752    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
753    Makefile.
754    
755    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
756    command called pgrep for grepping around the active processes.
757    
758    7. Added the beginnings of support for UTF-8 character strings.
759    
760    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
761    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
762    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
763    out for the ar command.)
764    
765    
766    Version 3.2 12-May-00
767    ---------------------
768    
769    This is purely a bug fixing release.
770    
771    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
772    of ZA. This was just one example of several cases that could provoke this bug,
773    which was introduced by change 9 of version 2.00. The code for breaking
774    infinite loops after an iteration that matches an empty string was't working
775    correctly.
776    
777    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
778    when matched against abbab (for example). After matching an empty string, it
779    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
780    caused it to match further down the string than it should.
781    
782    3. The code contained an inclusion of sys/types.h. It isn't clear why this
783    was there because it doesn't seem to be needed, and it causes trouble on some
784    systems, as it is not a Standard C header. It has been removed.
785    
786    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
787    were reported on the Macintosh. The changes were from
788    
789      while ((c = *(++ptr)) != 0 && c != '\n');
790    to
791      while ((c = *(++ptr)) != 0 && c != '\n') ;
792    
793    Totally extraordinary, but if that's what it takes...
794    
795    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
796    available. Added HAVE_BCOPY and an autoconf test for it; if neither
797    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
798    assumes the way PCRE uses memmove() (always moving upwards).
799    
800    6. PCRE is being used in one environment where strchr() is not available. There
801    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
802    faster code anyway.
803    
804    
805    Version 3.1 09-Feb-00
806    ---------------------
807    
808    The only change in this release is the fixing of some bugs in Makefile.in for
809    the "install" target:
810    
811    (1) It was failing to install pcreposix.h.
812    
813    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
814    
815    
816    Version 3.0 01-Feb-00
817    ---------------------
818    
819    1. Add support for the /+ modifier to perltest (to output $` like it does in
820    pcretest).
821    
822    2. Add support for the /g modifier to perltest.
823    
824    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
825    matches null strings.
826    
827    4. Fix perltest so that it doesn't do unwanted things when fed an empty
828    pattern. Perl treats empty patterns specially - it reuses the most recent
829    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
830    effect.
831    
832    5. The POSIX interface was broken in that it was just handing over the POSIX
833    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
834    required a bigger vector, with some working space on the end. This means that
835    the POSIX wrapper now has to get and free some memory, and copy the results.
836    
837    6. Added some simple autoconf support, placing the test data and the
838    documentation in separate directories, re-organizing some of the
839    information files, and making it build pcre-config (a GNU standard). Also added
840    libtool support for building PCRE as a shared library, which is now the
841    default.
842    
843    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
844    09 are not valid octal constants. Single digits will be used for minor values
845    less than 10.
846    
847    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
848    existing programs that set these in the POSIX interface can use PCRE without
849    modification.
850    
851    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
852    return all that pcre_info() returns, plus additional data. The pcre_info()
853    function is retained for compatibility, but is considered to be obsolete.
854    
855    10. Added experimental recursion feature (?R) to handle one common case that
856    Perl 5.6 will be able to do with (?p{...}).
857    
858    11. Added support for POSIX character classes like [:alpha:], which Perl is
859    adopting.
860    
861    
862    Version 2.08 31-Aug-99
863    ----------------------
864    
865    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
866    trying to match at the startoffset position, but instead was moving forward to
867    the next newline as if a previous match had failed.
868    
869    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
870    and could get into a loop if a null string was matched other than at the start
871    of the subject.
872    
873    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
874    be distinguished at compile time, and for completeness also added PCRE_DATE.
875    
876    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
877    in GnuWin32 environments.
878    
879    
880    Version 2.07 29-Jul-99
881    ----------------------
882    
883    1. The documentation is now supplied in plain text form and HTML as well as in
884    the form of man page sources.
885    
886    2. C++ compilers don't like assigning (void *) values to other pointer types.
887    In particular this affects malloc(). Although there is no problem in Standard
888    C, I've put in casts to keep C++ compilers happy.
889    
890    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
891    should be (const char *).
892    
893    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
894    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
895    However, I haven't made this a standard facility. The documentation doesn't
896    mention it, and the Makefile doesn't support it.
897    
898    5. The Makefile now contains an "install" target, with editable destinations at
899    the top of the file. The pcretest program is not installed.
900    
901    6. pgrep -V now gives the PCRE version number and date.
902    
903    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
904    causing the entire string to be ignored, instead of just the last character.
905    
906    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
907    non-matching string, it can take a very, very long time, even for strings of
908    quite modest length, because of the nested recursion. PCRE now does better in
909    some of these cases. It does this by remembering the last required literal
910    character in the pattern, and pre-searching the subject to ensure it is present
911    before running the real match. In other words, it applies a heuristic to detect
912    some types of certain failure quickly, and in the above example, if presented
913    with a string that has no trailing " it gives "no match" very quickly.
914    
915    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
916    other alternatives are tried instead.
917    
918    
919    Version 2.06 09-Jun-99
920    ----------------------
921    
922    1. Change pcretest's output for amount of store used to show just the code
923    space, because the remainder (the data block) varies in size between 32-bit and
924    64-bit systems.
925    
926    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
927    start matching at. This allows lookbehinds to work when searching for multiple
928    occurrences in a string.
929    
930    3. Added additional options to pcretest for testing multiple occurrences:
931    
932       /+   outputs the rest of the string that follows a match
933       /g   loops for multiple occurrences, using the new startoffset argument
934       /G   loops for multiple occurrences by passing an incremented pointer
935    
936    4. PCRE wasn't doing the "first character" optimization for patterns starting
937    with \b or \B, though it was doing it for other lookbehind assertions. That is,
938    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
939    the letter 'x'. On long subject strings, this gives a significant speed-up.
940    
941    
942    Version 2.05 21-Apr-99
943    ----------------------
944    
945    1. Changed the type of magic_number from int to long int so that it works
946    properly on 16-bit systems.
947    
948    2. Fixed a bug which caused patterns starting with .* not to work correctly
949    when the subject string contained newline characters. PCRE was assuming
950    anchoring for such patterns in all cases, which is not correct because .* will
951    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
952    DOTALL is set at top level; otherwise it knows that patterns starting with .*
953    must be retried after every newline in the subject.
954    
955    
956    Version 2.04 18-Feb-99
957    ----------------------
958    
959    1. For parenthesized subpatterns with repeats whose minimum was zero, the
960    computation of the store needed to hold the pattern was incorrect (too large).
961    If such patterns were nested a few deep, this could multiply and become a real
962    problem.
963    
964    2. Added /M option to pcretest to show the memory requirement of a specific
965    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
966    
967    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
968    compiled in such a way that the backtracking after subsequent failure was
969    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
970    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
971    
972    
973    Version 2.03 02-Feb-99
974    ----------------------
975    
976    1. Fixed typo and small mistake in man page.
977    
978    2. Added 4th condition (GPL supersedes if conflict) and created separate
979    LICENCE file containing the conditions.
980    
981    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
982    Perl, that is the internal \ allows the delimiter to be included in the
983    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
984    the final delimiter, add \ to the end of the pattern (to test the error).
985    
986    4. Added the convenience functions for extracting substrings after a successful
987    match. Updated pcretest to make it able to test these functions.
988    
989    
990    Version 2.02 14-Jan-99
991    ----------------------
992    
993    1. Initialized the working variables associated with each extraction so that
994    their saving and restoring doesn't refer to uninitialized store.
995    
996    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
997    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
998    fix the problem.
999    
1000    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1001    calls, and wasn't printing the correct value for compiling calls. Increased the
1002    default value of LOOPREPEAT, and the number of significant figures in the
1003    times.
1004    
1005    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1006    
1007    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1008    a building problem on Windows NT with a FAT file system.
1009    
1010    
1011    Version 2.01 21-Oct-98
1012    ----------------------
1013    
1014    1. Changed the API for pcre_compile() to allow for the provision of a pointer
1015    to character tables built by pcre_maketables() in the current locale. If NULL
1016    is passed, the default tables are used.
1017    
1018    
1019    Version 2.00 24-Sep-98
1020    ----------------------
1021    
1022    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1023    it any more.
1024    
1025    2. Allow quantification of (?>) groups, and make it work correctly.
1026    
1027    3. The first character computation wasn't working for (?>) groups.
1028    
1029    4. Correct the implementation of \Z (it is permitted to match on the \n at the
1030    end of the subject) and add 5.005's \z, which really does match only at the
1031    very end of the subject.
1032    
1033    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1034    
1035    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1036    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1037    localized options. All options to pcre_study() were also removed.
1038    
1039    7. Add other new features from 5.005:
1040    
1041       $(?<=           positive lookbehind
1042       $(?<!           negative lookbehind
1043       (?imsx-imsx)    added the unsetting capability
1044                       such a setting is global if at outer level; local otherwise
1045       (?imsx-imsx:)   non-capturing groups with option setting
1046       (?(cond)re|re)  conditional pattern matching
1047    
1048       A backreference to itself in a repeated group matches the previous
1049       captured string.
1050    
1051    8. General tidying up of studying (both automatic and via "study")
1052    consequential on the addition of new assertions.
1053    
1054    9. As in 5.005, unlimited repeated groups that could match an empty substring
1055    are no longer faulted at compile time. Instead, the loop is forcibly broken at
1056    runtime if any iteration does actually match an empty substring.
1057    
1058    10. Include the RunTest script in the distribution.
1059    
1060    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1061    discrepancies, some of which were old and were also with respect to 5.004. They
1062    have now been fixed.
1063    
1064    
1065    Version 1.09 28-Apr-98
1066    ----------------------
1067    
1068    1. A negated single character class followed by a quantifier with a minimum
1069    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
1070    program crashes, or just wrong answers. This did not apply to negated classes
1071    containing more than one character, or to minima other than one.
1072    
1073    
1074  Version 1.08 27-Mar-98  Version 1.08 27-Mar-98
1075  ----------------------  ----------------------

Legend:
Removed from v.19  
changed lines
  Added in v.71

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12