1 Tue Aug 28 23:53:20 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
5 Fixed truncation race with lock_retrieve() in lock_store().
6 The file has to be truncated only once the exclusive lock is held.
8 Removed spurious debugging messages in .xs file.
10 Sun Jul 1 13:27:32 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
14 Systematically use "=over 4" for POD linters.
15 Apparently, POD linters are much stricter than would
16 otherwise be needed, but that's OK.
18 Fixed memory corruption on croaks during thaw(). Thanks
19 to Claudio Garcia for reproducing this bug and providing the
20 code to exercise it. Added test cases for this bug, adapted
23 Made code compile cleanly with -Wall (from Jarkko Hietaniemi).
25 Changed tagnum and classnum from I32 to IV in context. Also
28 Thu Mar 15 01:22:32 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
32 Last version was wrongly compiling with assertions on, due
33 to an edit glitch. That did not cause any problem (apart from
34 a slight performance loss) excepted on Win* platforms, where the
35 assertion code does not compile.
37 Sat Feb 17 13:37:37 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
43 Forgot to increase version number at previous patch (there were
44 two of them, which is why we jump from 1.0.8 to 1.0.10).
46 Sat Feb 17 13:35:00 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
50 Version 1.0.8, binary format 2.4.
52 Fixed incorrect error message.
54 Now bless objects ASAP at retrieve time, which is meant to fix
57 * Indirect references to overloaded object were not able to
58 restore overloading if the object was not blessed yet,
59 which was possible since blessing occurred only after the
62 * Storable hooks asking for serialization of blessed ref could
63 get un-blessed refs at retrieval time, for the very same
66 The fix implemented here was suggested by Nick Ing-Simmons.
68 Added support for blessed ref to tied structures. This is the
69 cause for the binary format change.
71 Added EBCDIC version of the compatibility test with 0.6.11,
74 Added tests for the new features, and to make sure the bugs they
75 are meant to fix are indeed fixed.
77 Wed Jan 3 10:43:18 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
81 Removed spurious 'clean' entry in Makefile.PL.
83 Added CAN_FLOCK to determine whether we can flock() or not,
84 by inspecting Perl's configuration parameters, as determined
87 Trace offending package when overloading cannot be restored
90 Made context cleanup safer to avoid dup freeing, mostly in the
91 presence of repeated exceptions during store/retrieve (which can
92 cause memory leaks anyway, so it's just additional safety, not a
95 Sun Nov 5 18:23:48 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
101 Fixed severe "object lost" bug for STORABLE_freeze returns,
102 when refs to lexicals, taken within the hook, were to be
103 serialized by Storable. Enhanced the t/recurse.t test to
104 stress hook a little more with refs to lexicals.
106 Thu Oct 26 19:14:38 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
112 Documented that store() and retrieve() can return undef.
113 That is, the error reporting is not always made via exceptions,
114 as the paragraph on error reporting was implying.
116 Auto requires module of blessed ref when STORABLE_thaw misses.
117 When the Storable engine looks for the STORABLE_thaw hook and
118 does not find it, it now tries to require the package into which
119 the blessed reference is.
121 Just check $^O, in t/lock.t: there's no need to pull the whole
122 Config module for that.
124 Mon Oct 23 20:03:49 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
130 Protected calls to flock() for DOS platform: apparently, the
131 flock/fcnlt emulation is reported to be broken on that
134 Added logcarp emulation if they don't have Log::Agent, since
135 we now use it to carp when lock_store/lock_retrieve is used
138 Fri Sep 29 21:52:29 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
144 Avoid using "tainted" and "dirty" since Perl remaps them via
145 cpp (i.e. #define). This is deeply harmful when threading
146 is enabled. This concerned both the context structure and
147 local variable and argument names. Brrr..., scary!
149 Thu Sep 28 23:46:39 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
155 Fixed spelling in README.
157 Added lock_store, lock_nstore, and lock_retrieve (advisory locking)
158 after a proposal from Erik Haugan <erik@solbors.no>.
160 Perls before 5.004_04 lack newSVpvn, added remapping in XS.
162 Fixed stupid typo in the t/utf8.t test.
164 Sun Sep 17 18:51:10 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
168 Version 1.0.1, binary format 2.3.
170 Documented that doubles are stored stringified by nstore().
172 Added Salvador Ortiz Garcia in CREDITS section, He identified
173 a bug in the store hooks and proposed the right fix: the class
174 id was allocated too soon. His bug case was also added to
175 the regression test suite.
177 Now only taint retrieved data when source was tainted. A bug
178 discovered by Marc Lehmann.
180 Added support for UTF-8 strings, a contribution of Marc Lehmann.
181 This is normally only activated in post-5.6 perls.
183 Thu Aug 31 23:06:06 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
185 First official release Storable 1.0, for inclusion in perl 5.7.0.
186 The license scheme is now compatible with Perl's.
188 Thu Aug 24 01:02:02 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
193 ANSI-fied most of the code, preparing for Perl core integration.
194 The next version of Storable will be 0.8, and will be integrated
195 into the Perl core (development branch).
197 Dispatch tables were moved upfront to relieve some compilers,
198 especially on AIX and Windows platforms.
200 Merged 64-bit fixes from perl5-porters.
202 Mon Aug 14 09:22:04 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
206 Added a refcnt dec in retrieve_tied_key(): sv_magic() increases
207 the refcnt on the mg_ptr as well.
209 Removed spurious dependency to Devel::Peek, which was used for
210 testing only in t/tied_items.t. Thanks to Conrad Heiney
211 <conrad@fringehead.org> for spotting it first.
213 Sun Aug 13 22:12:59 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
217 Marc Lehmann kindly contributed code to add overloading support
218 and to handle references to tied variables.
220 Rewrote leading blurb about compatibility to make it clearer what
221 "backward compatibility" is about: when I say 0.7 is backward
222 compatible with 0.6, it means the revision 0.7 can read files
225 Mention new Clone(3) extension in SEE ALSO.
227 Was wrongly optimizing for "undef" values in hashes by not
228 fully recursing: as a result, tied "undef" values were incorrectly
231 Sun Jul 30 12:59:17 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
233 First revision of Storable 0.7.
235 The serializing format is new, known as version 2.0. It is fully
236 backward compatible with 0.6. Earlier formats are deprecated and
237 have not even been tested: next version will drop pre-0.6 format.
239 Changes since 0.6@11:
241 - Moved interface to the "beta" status. Some tiny parts are still
242 subject to change, but nothing important enough to warrant an "alpha"
245 - Slightly reduced the size of the Storable image by factorizing
246 object class names and removing final object storage notification due
247 to a redesign of the blessed object storing.
249 - Classes can now redefine how they wish their instances to be serialized
250 and/or deep cloned. Serializing hooks are written in Perl code.
252 - The engine is now fully re-entrant.
254 Sun Apr 2 23:47:50 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
258 Added provision to detect more recent binary formats, since
259 the new upcoming Storable-0.7 will use a different format.
260 In order to prevent attempting the de-serialization of newer
261 formats by older versions, I'm adding this now to the 0.6 series.
263 I'm expecting this revision to be the last of the 0.6 series.
264 Unless it does not work with perl 5.6, which I don't use yet,
265 and therefore against which I cannot test.
267 Wed Mar 29 19:55:21 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
271 Added note about format incompatibilities with old versions
272 (i.e. pre 0.5@9 formats, which cannot be understood as there
273 was no versionning information in the file by then).
275 Protect all $@ variables when eval {} used, to avoid corrupting
276 it when store/retrieve is called within an exception handler.
278 Mistakenly included "patchlevel.h" instead of <patchlevel.h>,
279 preventing Perl's patchlevel from being included, which is
280 needed starting from 5.6.
282 Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
286 Fixed shared "undef" bug in hashes, which did not remain shared
287 through store/retrieve.
289 Thu Feb 10 19:48:16 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
293 added last_op_in_netorder() predicate
294 documented last_op_in_netorder()
295 added tests for the new last_op_in_netorder() predicate
297 Wed Oct 20 19:07:36 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
301 Forgot to update VERSION
303 Tue Oct 19 21:25:02 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
307 Added mention of japanese translation for the manual page.
309 Fixed typo in macro that made threaded code not compilable,
310 especially on Win32 platforms.
312 Changed detection of older perls (pre-5.005) by testing PATCHLEVEL
313 directly instead of relying on internal symbols.
315 Tue Sep 14 22:13:28 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
319 Integrated "thread-safe" patch from Murray Nesbitt.
320 Note that this may not be very efficient for threaded code,
321 see comment in the code.
323 Try to avoid compilation warning on 64-bit CPUs. Can't test it,
324 since I don't have access to such machines.
326 Mon Jul 12 14:37:19 METDST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
330 changed my e-mail to pobox.
332 mentionned it is not thread-safe.
334 updated version number.
336 uses new internal PL_* naming convention.
338 Fri Jul 3 13:38:16 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
342 Updated benchmark figures due to recent optimizations done in
343 store(): tagnums are now stored as-is in the hash table, so
344 no surrounding SV is created. And the "shared keys" mode for
345 hash table was turned off.
347 Fixed backward compatibility (wrt 0.5@9) for retrieval of
348 blessed refs. That old version did something wrong, but the
349 bugfix prevented correct retrieval of the old format.
351 Mon Jun 22 11:00:48 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
355 Changed benchmark figures.
357 Adjust refcnt of tied objects after calling sv_magic() to avoid
358 memory leaks. Contributed by Jeff Gresham.
360 Fri Jun 12 11:50:04 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
364 Added workaround for persistent LVALUE-ness in perl5.004. All
365 scalars tagged as being an lvalue are handled as if they were
366 not an lvalue at all. Added test for that LVALUE bug workaround.
368 Now handles Perl immortal scalars explicitely, by storing &sv_yes
369 as such, explicitely.
371 Retrieval of non-immortal undef cannot be shared. Previous
372 version was over-optimizing by not creating a separate SV for
373 all undefined scalars seen.
375 Thu Jun 4 17:21:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
379 Baseline for Storable-0.6@0.
381 This version introduces a binary incompatibility in the generated
382 binary image, which is more compact than older ones by approximatively
383 15%, depending on the exact degree of sharing in your structures.
385 The good news is that your older images can still be retrieved with
386 this version, i.e. backward compatibility is preserved. This version
387 of Storable can only generate new binaries however.
389 Another good news is that the retrieval of data structure is
390 significantly quicker than before, because a Perl array is used
391 instead of a hash table to keep track of retrieved objects, and
392 also because the image being smaller, less I/O function calls are
395 Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
399 Version number now got from Storable.pm directly.
401 Fixed overzealous sv_type() optimization, which would make
402 Storable fail when faced with an "upgraded" SV to the PVIV
403 or PVNV kind containing a reference.
405 Thu Apr 30 15:11:30 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
409 Extended the SYNOPSIS section to give quick overview of the
410 routines and their signature.
412 Optimized sv_type() to avoid flags checking when not needed, i.e.
413 when their type makes it impossible for them to be refs or tied.
414 This slightly increases throughput by a few percents when refs
415 and tied variables are marginal occurrences in your data.
417 Stubs for XS now use OutputStream and InputStream file types to
418 make it work when the given file is actually a socket. Perl
419 makes a distinction for sockets in its internal I/O structures
420 by having both a read and a write structure, whereas plain files
423 Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
427 Thanks to a contribution from Benjamin A. Holzman, Storable is now
428 able to correctly serialize tied SVs, i.e. tied arrays, hashes
431 Thu Apr 9 18:07:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
435 I said SvPOK() had changed to SvPOKp(), but that was a lie...
437 Wed Apr 8 13:14:29 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
441 Wrote sizeof(SV *) instead of sizeof(I32) when portable, which
442 in effect mangled the object tags and prevented portability
443 accross 32/64 bit architectures!
445 Wed Mar 25 14:57:02 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
449 Added code example for store_fd() and retrieve_fd() in the
450 man page, to emphasize that file descriptors must be passed as
451 globs, not as plain strings.
453 Cannot use SV addresses as tag when using nstore() on LP64. This
454 was the cause of problems when creating a storable image on an
455 LP64 machine and retrieving it on an ILP32 system, which is
456 exactly what nstore() is meant for...
458 However, we continue to use SV addresses as tags for plain store(),
459 because benchamarking shows that it saves up to 8% of the store
460 time, and store() is meant to be fast at the expense of lack
463 This means there will be approximately an 8% degradation of
464 performance for nstore(), but it's now working as expected.
465 That cost may vary on your machine of course, since it is
466 solely caused by the memory allocation overhead used to create
467 unique SV tags for each distinct stored SV.
469 Tue Jan 20 09:21:53 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
473 Don't use any '_' in version number.
475 Tue Jan 13 17:51:50 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
479 Updated version number.
481 added binmode() calls for systems where it matters.
483 Be sure to pass globs, not plain file strings, to C routines,
484 so that Storable can be used under the Perl debugger.
486 Wed Nov 5 10:53:22 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
490 Fix memory leaks on seen hash table and returned SV refs.
492 Storable did not work properly when tainting enabled.
494 Fixed "Allocation too large" messages in freeze/thaw and added.
495 proper regression test in t/freeze.t.
497 Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
501 Updated version number
503 Added freeze/thaw interface and dclone.
505 Fri May 16 10:45:47 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
509 Forgot that AutoLoader does not export its own AUTOLOAD.
512 use AutoLoader 'AUTOLOAD';
514 but that would not be backward compatible. So the export is
517 Tue Mar 25 11:21:32 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
521 Empty scalar strings are now "defined" at retrieval time.
523 New test to ensure an empty string is defined when retrieved.
525 Thu Feb 27 16:32:44 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
529 Updated version number
531 Declare VERSION as being used
533 Fixed a typo in the PerlIO_putc remapping.
534 PerlIO_read and perlIO_write inverted size/nb_items.
535 (only relevant for pre-perl5.004 versions)
537 Thu Feb 27 15:58:31 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
541 Updated version number
543 Added VERSION identification
545 Allow build with perl5.003, which is ante perlIO time
547 Mon Jan 13 17:53:18 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
553 Wed Jan 22 15:19:56 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
557 Updated version number in Makefile.PL.
559 Added "thanks to" section to README.
561 Documented new forgive_me variable.
565 Added forgive_me support to allow store() of data structures
566 containing non-storable items like CODE refs.