ViewVC logotype

Contents of /code/trunk/maint/README

Parent Directory Parent Directory | Revision Log Revision Log

Revision 984 - (hide annotations) (download)
Fri Jul 6 09:58:13 2012 UTC (2 years, 8 months ago) by ph10
File size: 16646 byte(s)
Final tidies for 8.31.

2 ph10 507 ===========================
3 ph10 122
4 ph10 181 The files in the "maint" directory of the PCRE source contain data, scripts,
5 ph10 122 and programs that are used for the maintenance of PCRE, but which do not form
6 ph10 181 part of the PCRE distribution tarballs. This document describes these files and
7 ph10 122 also contains some notes for maintainers. Its contents are:
8 ph10 97
9 ph10 122 Files in the maint directory
10     Updating to a new Unicode release
11     Preparing for a PCRE release
12     Making a PCRE release
13 ph10 181 Long-term ideas (wish list)
14 ph10 122
16     Files in the maint directory
17 ph10 507 ============================
18 ph10 122
19 ph10 507 ---------------- This file is now OBSOLETE and no longer used ----------------
20 ph10 129 Builducptable A Perl script that creates the contents of the ucptable.h file
21     from two Unicode data files, which themselves are downloaded
22     from the Unicode web site. Run this script in the "maint"
23     directory.
24 ph10 507 ---------------- This file is now OBSOLETE and no longer used ----------------
25 ph10 181
26 ph10 351 GenerateUtt.py A Python script to generate part of the pcre_tables.c file
27     that contains Unicode script names in a long string with
28 ph10 535 offsets, which is tedious to maintain by hand.
29 ph10 351
30 ph10 129 ManyConfigTests A shell script that runs "configure, make, test" a number of
31     times with different configuration settings.
32 ph10 535
33 ph10 350 MultiStage2.py A Python script that generates the file pcre_ucd.c from three
34     Unicode data tables, which are themselves downloaded from the
35 ph10 535 Unicode web site. Run this script in the "maint" directory.
36 ph10 350 The generated file contains the tables for a 2-stage lookup
37 ph10 535 of Unicode properties.
38 ph10 583
39 ph10 539 pcre_chartables.c.non-standard
40     This is a set of character tables that came from a Windows
41     system. It has characters greater than 128 that are set as
42     spaces, amongst other things. I kept it so that it can be
43 ph10 583 used for testing from time to time.
44 ph10 181
45 ph10 454 README This file.
47 ph10 535 Unicode.tables The files in this directory, DerivedGeneralCategory.txt,
48 ph10 350 Scripts.txt and UnicodeData.txt, were downloaded from the
49     Unicode web site. They contain information about Unicode
50     characters and scripts.
51 ph10 181
52 ph10 350 ucptest.c A short C program for testing the Unicode property macros
53     that do lookups in the pcre_ucd.c data, mainly useful after
54     rebuilding the Unicode property table. Compile and run this in
55 ph10 351 the "maint" directory (see comments at its head).
56 ph10 181
57 ph10 129 ucptestdata A directory containing two files, testinput1 and testoutput1,
58     to use in conjunction with the ucptest program.
59 ph10 181
60 ph10 129 utf8.c A short, freestanding C program for converting a Unicode code
61     point into a sequence of bytes in the UTF-8 encoding, and vice
62     versa. If its argument is a hex number such as 0x1234, it
63     outputs a list of the equivalent UTF-8 bytes. If its argument
64     is sequence of concatenated UTF-8 bytes (e.g. e188b4) it
65     treats them as a UTF-8 character and outputs the equivalent
66     code point in hex.
67 ph10 97
68 ph10 181
69 ph10 122 Updating to a new Unicode release
70 ph10 507 =================================
71 ph10 122
72 ph10 181 When there is a new release of Unicode, the files in Unicode.tables must be
73 ph10 454 refreshed from the web site. If the new version of Unicode adds new character
74 ph10 352 scripts, the source file ucp.h and both the MultiStage2.py and the
75 ph10 454 GenerateUtt.py scripts must be edited to add the new names. Then MultiStage2.py
76     can be run to generate a new version of pcre_ucd.c, and GenerateUtt.py can be
77     run to generate the tricky tables for inclusion in pcre_tables.c.
78 ph10 122
79 ph10 491 If MultiStage2.py gives the error "ValueError: list.index(x): x not in list",
80 ph10 535 the cause is usually a missing (or misspelt) name in the list of scripts. I
81     couldn't find a straightforward list of scripts on the Unicode site, but
82     there's a useful Wikipedia page that list them, and notes the Unicode version
83 ph10 491 in which they were introduced:
85     http://en.wikipedia.org/wiki/Unicode_scripts#Table_of_Unicode_scripts
87 ph10 454 The ucptest program can be compiled and used to check that the new tables in
88     pcre_ucd.c work properly, using the data files in ucptestdata to check a number
89 ph10 491 of test characters. The source file ucptest.c must be updated whenever new
90     Unicode script names are added.
91 ph10 122
92 ph10 535 Note also that both the pcresyntax.3 and pcrepattern.3 man pages contain lists
93 ph10 491 of Unicode script names.
94 ph10 351
95 ph10 491
96 ph10 122 Preparing for a PCRE release
97 ph10 507 ============================
98 ph10 122
99     This section contains a checklist of things that I consult before building a
100     distribution for a new release.
102 ph10 454 . Ensure that the version number and version date are correct in configure.ac.
103 ph10 535
104 ph10 981 . Update the library version numbers in configure.ac according to the rules
105     given below.
107 ph10 292 . If new build options have been added, ensure that they are added to the CMake
108 ph10 654 files as well as to the autoconf files. The relevant files are CMakeLists.txt
109     and config-cmake.h.in. After making a release tarball, test it out with CMake
110     if there have been changes here.
111 ph10 122
112     . Run ./autogen.sh to ensure everything is up-to-date.
114 ph10 181 . Compile and test with many different config options, and combinations of
115 ph10 733 options. Also, test with valgrind by running "RunTest valgrind" and
116     "RunGrepTest valgrind" (which takes quite a long time). The script
117     maint/ManyConfigTests now encapsulates this testing. It runs tests with
118     different configurations, and it also runs some of them with valgrind, all of
119     which can take quite some time.
120 ph10 181
121 ph10 874 . Run perltest.pl on the test data for tests 1, 4, and 6. The output
122 ph10 654 should match the PCRE test output, apart from the version identification at
123     the start of each test. The other tests are not Perl-compatible (they use
124     various PCRE-specific features or options).
125 ph10 122
126 ph10 583 . It is possible to test with the emulated memmove() function by undefining
127     HAVE_MEMMOVE and HAVE_BCOPY in config.h, though I do not do this often. You
128     may see a number of "pcre_memmove defined but not used" warnings for the
129     modules in which there is no call to memmove(). These can be ignored.
130 ph10 122
131 ph10 984 . Documentation: check AUTHORS, ChangeLog (check version and date), LICENCE,
132     NEWS (check version and date), NON-AUTOTOOLS-BUILD, and README. Many of these
133     won't need changing, but over the long term things do change.
134 ph10 181
135 ph10 583 . I used to test new releases myself on a number of different operating
136     systems, using different compilers as well. For example, on Solaris it is
137     helpful to test using Sun's cc compiler as a change from gcc. Adding
138     -xarch=v9 to the cc options does a 64-bit test, but it also needs -S 64 for
139     pcretest to increase the stack size for test 2. Since I retired I can no
140     longer do this, but instead I rely on putting out release candidates for
141     folks on the pcre-dev list to test.
142 ph10 981
143 ph10 122
144 ph10 981 Updating version info for libtool
145     =================================
146 ph10 122
147 ph10 981 This set of rules for updating library version information came from a web page
148     whose URL I have forgotten. The version information consists of three parts:
149     (current, revision, age).
151     1. Start with version information of 0:0:0 for each libtool library.
153     2. Update the version information only immediately before a public release of
154     your software. More frequent updates are unnecessary, and only guarantee
155     that the current interface number gets larger faster.
157     3. If the library source code has changed at all since the last update, then
158     increment revision; c:r:a becomes c:r+1:a.
160     4. If any interfaces have been added, removed, or changed since the last
161     update, increment current, and set revision to 0.
163     5. If any interfaces have been added since the last public release, then
164     increment age.
166     6. If any interfaces have been removed or changed since the last public
167     release, then set age to 0.
169     The following explanation may help in understanding the above rules a bit
170     better. Consider that there are three possible kinds of reaction from users to
171     changes in a shared library:
173     1. Programs using the previous version may use the new version as a drop-in
174     replacement, and programs using the new version can also work with the
175     previous one. In other words, no recompiling nor relinking is needed. In
176     this case, increment revision only, don't touch current or age.
178     2. Programs using the previous version may use the new version as a drop-in
179     replacement, but programs using the new version may use APIs not present in
180     the previous one. In other words, a program linking against the new version
181     may fail if linked against the old version at run time. In this case, set
182     revision to 0, increment current and age.
184     3. Programs may need to be changed, recompiled, relinked in order to use the
185     new version. Increment current, set revision and age to 0.
188 ph10 122 Making a PCRE release
189 ph10 507 =====================
190 ph10 122
191     Run PrepareRelease and commit the files that it changes (by removing trailing
192 ph10 583 spaces). The first thing this script does is to run CheckMan on the man pages;
193     if it finds any markup errors, it reports them and then aborts.
194 ph10 122
195 ph10 583 Once PrepareRelease has run clean, run "make distcheck" to create the tarballs
196     and the zipball. Double-check with "svn status", then create an SVN tagged
197     copy:
199 ph10 212 svn copy svn://vcs.exim.org/pcre/code/trunk \
200 ph10 535 svn://vcs.exim.org/pcre/code/tags/pcre-8.xx
201 ph10 212
202 ph10 122 Don't forget to update Freshmeat when the new release is out, and to tell
203 ph10 535 webmaster@pcre.org and the mailing list. Also, update the list of version
204 ph10 507 numbers in Bugzilla (edit products).
205 ph10 122
207     Future ideas (wish list)
208 ph10 507 ========================
209 ph10 122
210     This section records a list of ideas so that they do not get forgotten. They
211 ph10 181 vary enormously in their usefulness and potential for implementation. Some are
212 ph10 122 very sensible; some are rather wacky. Some have been on this list for years;
213     others are relatively new.
215     . Optimization
217 ph10 181 There are always ideas for new optimizations so as to speed up pattern
218     matching. Most of them try to save work by recognizing a non-match without
219 ph10 122 having to scan all the possibilities. These are some that I've recorded:
221     * /((A{0,5}){0,5}){0,5}(something complex)/ on a non-matching string is very
222     slow, though Perl is fast. Can we speed up somehow? Convert to {0,125}?
223 ph10 181 OTOH, this is pathological - the user could easily fix it.
225 ph10 122 * Turn ={4} into ==== ? (for speed). I once did an experiment, and it seems
226     to have little effect, and maybe makes things worse.
227 ph10 181
228     * "Ends with literal string" - note that a single character doesn't gain much
229 ph10 454 over the existing "required byte" (reqbyte) feature that just remembers one
230 ph10 874 data unit.
231 ph10 181
232 ph10 535 * These probably need to go in pcre_study():
233 ph10 181
234 ph10 122 o Remember an initial string rather than just 1 char?
235 ph10 181
236 ph10 874 o A required data unit from alternatives - not just the last unit, but an
237 ph10 122 earlier one if common to all alternatives.
238 ph10 181
239 ph10 122 o Friedl contains other ideas.
240 ph10 535
241     * pcre_study() does not set initial byte flags for Unicode property types
242     such as \p; I don't know how much benefit there would be for, for example,
243     setting the bits for 0-9 and all bytes >= xC0 when a pattern starts with
244     \p{N}.
246     * There is scope for more "auto-possessifying" in connection with \p and \P.
248 ph10 122 . If Perl gets to a consistent state over the settings of capturing sub-
249     patterns inside repeats, see if we can match it. One example of the
250     difference is the matching of /(main(O)?)+/ against mainOmain, where PCRE
251     leaves $2 set. In Perl, it's unset. Changing this in PCRE will be very hard
252     because I think it needs much more state to be remembered.
254 ph10 181 . Perl 6 will be a revolution. Is it a revolution too far for PCRE?
256 ph10 122 . Allow errorptr and erroroffset to be NULL. I don't like this idea.
258     . Line endings:
260     * Option to use NUL as a line terminator in subject strings. This could now
261     be done relatively easily since the extension to support LF, CR, and CRLF.
262 ph10 454 If it is done, a suitable option for pcregrep is also required.
263 ph10 181
264 ph10 122 . Option to provide the pattern with a length instead of with a NUL terminator.
265 ph10 454 This affects quite a few places in the code and is not trivial.
266 ph10 122
267 ph10 181 . Catch SIGSEGV for stack overflows?
268 ph10 122
269     . A feature to suspend a match via a callout was once requested.
271     . Option to convert results into character offsets and character lengths.
273 ph10 181 . Option for pcregrep to scan only the start of a file. I am not keen - this is
274 ph10 122 the job of "head".
275 ph10 181
276     . A (non-Unix) user wanted pcregrep options to (a) list a file name just once,
277 ph10 122 preceded by a blank line, instead of adding it to every matched line, and (b)
278     support --outputfile=name.
279 ph10 181
280 ph10 454 . Consider making UTF-8 and UCP the default for PCRE n.0 for some n > 8.
281 ph10 874 (And now presumably UTF-16 and UCP for the 16-bit library.)
282 ph10 122
283 ph10 181 . Add a user pointer to pcre_malloc/free functions -- some option would be
284 ph10 122 needed to retain backward compatibility.
285 ph10 181
286 ph10 122 . Define a union for the results from pcre_fullinfo().
288 ph10 181 . Provide a "random access to the subject" facility so that the way in which it
289     is stored is independent of PCRE. For efficiency, it probably isn't possible
290 ph10 122 to switch this dynamically. It would have to be specified when PCRE was
291     compiled. PCRE would then call a function every time it wanted a character.
292 ph10 181
293 ph10 122 . Wild thought: the ability to compile from PCRE's internal byte code to a real
294     FSM and a very fast (third) matcher to process the result. There would be
295     even more restrictions than for pcre_dfa_exec(), however. This is not easy.
296 ph10 874 This is probably obsolete now that we have the JIT support.
297 ph10 181
298 ph10 122 . Should pcretest have some private locale data, to avoid relying on the
299     available locales for the test data, since different OS have different ideas?
300     This won't be as thorough a test, but perhaps that doesn't really matter.
301 ph10 181
302     . pcregrep: add -rs for a sorted recurse? Having to store file names and sort
303 ph10 122 them will of course slow it down.
305 ph10 181 . Someone suggested --disable-callout to save code space when callouts are
306     never wanted. This seems rather marginal.
307 ph10 535
308     . Check names that consist entirely of digits: PCRE allows, but do Perl and
309     Python, etc?
311     . A user suggested a parameter to limit the length of string matched, for
312     example if the parameter is N, the current match should fail if the matched
313     substring exceeds N. This could apply to both match functions. The value
314 ph10 372 could be a new field in the extra block.
315 ph10 535
316 ph10 372 . Callouts with arguments: (?Cn:ARG) for instance.
317 ph10 122
318 ph10 535 . A user is going to supply a patch to generalize the API for user-specific
319 ph10 507 memory allocation so that it is more flexible in threaded environments. This
320 ph10 874 was promised a long time ago, and never appeared. However, this is a live
321     issue not only for threaded environments, but for libraries that use PCRE and
322     want not to be beholden to their caller's memory allocation.
323 ph10 535
324     . Write a wrapper to maintain a structure with specified runtime parameters,
325     such as recurse limit, and pass these to PCRE each time it is called. Also
326 ph10 874 maybe malloc and free. A user sent a prototype. This relates the the previous
327     item.
328 ph10 535
329 ph10 874 . Write a function that generates random matching strings for a compiled regex.
331 ph10 535 . Pcregrep: an option to specify the output line separator, either as a string
332     or select from a fixed list. This is not dead easy, because at the moment it
333 ph10 454 outputs whatever is in the input file.
334 ph10 535
335     . Improve the code for duplicate checking in pcre_dfa_exec(). An incomplete,
336     non-thread-safe patch showed that this can help performance for patterns
337     where there are many alternatives. However, a simple thread-safe
338     implementation that I tried made things worse in many simple cases, so this
339 ph10 454 is not an obviously good thing.
340 ph10 535
341     . Make the longest lookbehind available via pcre_fullinfo(). This is not
342     straightforward because lookbehinds can be nested inside lookbehinds. This
343     case will have to be identified, and the amounts added. This should then give
344     the maximum possible lookbehind length. The reason for wanting this is to
345 ph10 454 help when implementing multi-segment matching using pcre_exec() with partial
346     matching and overlapping segments.
347 ph10 535
348 ph10 454 . PCRE cannot at present distinguish between subpatterns with different names,
349 ph10 535 but the same number (created by the use of ?|). In order to do so, a way of
350 ph10 454 remembering *which* subpattern numbered n matched is needed. Bugzilla #760.
351 ph10 535 Now that (*MARK) has been implemented, it can perhaps be used as a way round
352     this problem.
354     . Instead of having #ifdef HAVE_CONFIG_H in each module, put #include
355 ph10 507 "something" and the the #ifdef appears only in one place, in "something".
356 ph10 454
357 ph10 122 Philip Hazel
358     Email local part: ph10
359     Email domain: cam.ac.uk
360 ph10 981 Last updated: 18 June 2012

ViewVC Help
Powered by ViewVC 1.1.12