1 Fri May 17 22:48:59 BST 2002 Nicholas Clark <nick@ccl4.org>
5 Version 2.0, binary format 2.5 (but writes format 2.4 on pre 5.7.3)
7 The perl5 porters have decided to make sure that Storable still
8 builds on pre-5.8 perls, and make the 5.8 version available on CPAN.
9 The VERSION is now 2.0, and it passes all tests on 5.005_03, 5.6.1
10 and 5.6.1 with threads. On 5.6.0 t/downgrade.t fails tests 34 and 37,
11 due to a bug in 5.6.0 - upgrade to 5.6.1.
13 Jarkko and I have collated the list of changes the perl5 porters have
14 from the perl5 Changes file:
16 - data features of upcoming perl 5.8.0 are supported: Unicode hash
17 keys (Unicode hash values have been supported since Storable 1.0.1)
18 and "restricted hashes" (readonly hashes and hash entries)
19 - a newer version of perl can now be used to serialize data which is
20 not supported in earlier perls: Storable will attempt to do the
21 right thing for as long as possible, croaking only when safe data
22 conversion simply isn't possible. Alternatively earlier perls can
23 opt to have a lossy downgrade data instead of croaking
24 - when built with perls pre 5.7.3 this Storable writes out files
25 with binary format 2.4, the same format as Storable 1.0.8 onwards.
26 This should mean that this Storable will inter-operate seamlessly
27 with any Storable 1.0.8 or newer on perls pre 5.7.3
28 - dclone() now works with empty string scalar objects
29 - retrieving of large hashes is now more efficient
30 - more routines autosplit out of the main module, so Storable should
31 load slightly more quickly
32 - better documentation
33 - the internal context objects are now freed explicitly, rather than
34 relying on thread or process exit
35 - bugs fixed in debugging trace code affecting builds made with 64 bit
37 - code tidy-ups to allow clean compiles with more warning options
38 turned on avoid problems with $@ getting corrupted on 5.005_03 if
39 Carp wasn't already loaded
40 - added &show_file_magic, so you can add to /etc/magic and teach
41 Unix's file command about Storable files
43 We plan to keep Storable on CPAN in sync with the Perl core, so
44 if you encounter bugs or other problems building or using Storable,
45 please let us know at perl5-porters@perl.org
48 Sat Dec 1 14:37:54 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
52 This is the LAST maintenance release of the Storable module.
53 Indeed, Storable is now part of perl 5.8, and will be maintained
54 as part of Perl. The CPAN module will remain available there
55 for people running pre-5.8 perls.
57 Avoid requiring Fcntl upfront, useful to embedded runtimes.
58 Use an eval {} for testing, instead of making Storable.pm
59 simply fail its compilation in the BEGIN block.
61 store_fd() will now correctly autoflush file if needed.
63 Tue Aug 28 23:53:20 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
67 Fixed truncation race with lock_retrieve() in lock_store().
68 The file has to be truncated only once the exclusive lock is held.
70 Removed spurious debugging messages in .xs file.
72 Sun Jul 1 13:27:32 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
76 Systematically use "=over 4" for POD linters.
77 Apparently, POD linters are much stricter than would
78 otherwise be needed, but that's OK.
80 Fixed memory corruption on croaks during thaw(). Thanks
81 to Claudio Garcia for reproducing this bug and providing the
82 code to exercise it. Added test cases for this bug, adapted
85 Made code compile cleanly with -Wall (from Jarkko Hietaniemi).
87 Changed tagnum and classnum from I32 to IV in context. Also
90 Thu Mar 15 01:22:32 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
94 Last version was wrongly compiling with assertions on, due
95 to an edit glitch. That did not cause any problem (apart from
96 a slight performance loss) excepted on Win* platforms, where the
97 assertion code does not compile.
99 Sat Feb 17 13:37:37 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
105 Forgot to increase version number at previous patch (there were
106 two of them, which is why we jump from 1.0.8 to 1.0.10).
108 Sat Feb 17 13:35:00 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
112 Version 1.0.8, binary format 2.4.
114 Fixed incorrect error message.
116 Now bless objects ASAP at retrieve time, which is meant to fix
119 * Indirect references to overloaded object were not able to
120 restore overloading if the object was not blessed yet,
121 which was possible since blessing occurred only after the
124 * Storable hooks asking for serialization of blessed ref could
125 get un-blessed refs at retrieval time, for the very same
128 The fix implemented here was suggested by Nick Ing-Simmons.
130 Added support for blessed ref to tied structures. This is the
131 cause for the binary format change.
133 Added EBCDIC version of the compatibility test with 0.6.11,
136 Added tests for the new features, and to make sure the bugs they
137 are meant to fix are indeed fixed.
139 Wed Jan 3 10:43:18 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
143 Removed spurious 'clean' entry in Makefile.PL.
145 Added CAN_FLOCK to determine whether we can flock() or not,
146 by inspecting Perl's configuration parameters, as determined
149 Trace offending package when overloading cannot be restored
152 Made context cleanup safer to avoid dup freeing, mostly in the
153 presence of repeated exceptions during store/retrieve (which can
154 cause memory leaks anyway, so it's just additional safety, not a
157 Sun Nov 5 18:23:48 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
163 Fixed severe "object lost" bug for STORABLE_freeze returns,
164 when refs to lexicals, taken within the hook, were to be
165 serialized by Storable. Enhanced the t/recurse.t test to
166 stress hook a little more with refs to lexicals.
168 Thu Oct 26 19:14:38 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
174 Documented that store() and retrieve() can return undef.
175 That is, the error reporting is not always made via exceptions,
176 as the paragraph on error reporting was implying.
178 Auto requires module of blessed ref when STORABLE_thaw misses.
179 When the Storable engine looks for the STORABLE_thaw hook and
180 does not find it, it now tries to require the package into which
181 the blessed reference is.
183 Just check $^O, in t/lock.t: there's no need to pull the whole
184 Config module for that.
186 Mon Oct 23 20:03:49 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
192 Protected calls to flock() for DOS platform: apparently, the
193 flock/fcnlt emulation is reported to be broken on that
196 Added logcarp emulation if they don't have Log::Agent, since
197 we now use it to carp when lock_store/lock_retrieve is used
200 Fri Sep 29 21:52:29 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
206 Avoid using "tainted" and "dirty" since Perl remaps them via
207 cpp (i.e. #define). This is deeply harmful when threading
208 is enabled. This concerned both the context structure and
209 local variable and argument names. Brrr..., scary!
211 Thu Sep 28 23:46:39 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
217 Fixed spelling in README.
219 Added lock_store, lock_nstore, and lock_retrieve (advisory locking)
220 after a proposal from Erik Haugan <erik@solbors.no>.
222 Perls before 5.004_04 lack newSVpvn, added remapping in XS.
224 Fixed stupid typo in the t/utf8.t test.
226 Sun Sep 17 18:51:10 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
230 Version 1.0.1, binary format 2.3.
232 Documented that doubles are stored stringified by nstore().
234 Added Salvador Ortiz Garcia in CREDITS section, He identified
235 a bug in the store hooks and proposed the right fix: the class
236 id was allocated too soon. His bug case was also added to
237 the regression test suite.
239 Now only taint retrieved data when source was tainted. A bug
240 discovered by Marc Lehmann.
242 Added support for UTF-8 strings, a contribution of Marc Lehmann.
243 This is normally only activated in post-5.6 perls.
245 Thu Aug 31 23:06:06 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
247 First official release Storable 1.0, for inclusion in perl 5.7.0.
248 The license scheme is now compatible with Perl's.
250 Thu Aug 24 01:02:02 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
255 ANSI-fied most of the code, preparing for Perl core integration.
256 The next version of Storable will be 0.8, and will be integrated
257 into the Perl core (development branch).
259 Dispatch tables were moved upfront to relieve some compilers,
260 especially on AIX and Windows platforms.
262 Merged 64-bit fixes from perl5-porters.
264 Mon Aug 14 09:22:04 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
268 Added a refcnt dec in retrieve_tied_key(): sv_magic() increases
269 the refcnt on the mg_ptr as well.
271 Removed spurious dependency to Devel::Peek, which was used for
272 testing only in t/tied_items.t. Thanks to Conrad Heiney
273 <conrad@fringehead.org> for spotting it first.
275 Sun Aug 13 22:12:59 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
279 Marc Lehmann kindly contributed code to add overloading support
280 and to handle references to tied variables.
282 Rewrote leading blurb about compatibility to make it clearer what
283 "backward compatibility" is about: when I say 0.7 is backward
284 compatible with 0.6, it means the revision 0.7 can read files
287 Mention new Clone(3) extension in SEE ALSO.
289 Was wrongly optimizing for "undef" values in hashes by not
290 fully recursing: as a result, tied "undef" values were incorrectly
293 Sun Jul 30 12:59:17 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
295 First revision of Storable 0.7.
297 The serializing format is new, known as version 2.0. It is fully
298 backward compatible with 0.6. Earlier formats are deprecated and
299 have not even been tested: next version will drop pre-0.6 format.
301 Changes since 0.6@11:
303 - Moved interface to the "beta" status. Some tiny parts are still
304 subject to change, but nothing important enough to warrant an "alpha"
307 - Slightly reduced the size of the Storable image by factorizing
308 object class names and removing final object storage notification due
309 to a redesign of the blessed object storing.
311 - Classes can now redefine how they wish their instances to be serialized
312 and/or deep cloned. Serializing hooks are written in Perl code.
314 - The engine is now fully re-entrant.
316 Sun Apr 2 23:47:50 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
320 Added provision to detect more recent binary formats, since
321 the new upcoming Storable-0.7 will use a different format.
322 In order to prevent attempting the de-serialization of newer
323 formats by older versions, I'm adding this now to the 0.6 series.
325 I'm expecting this revision to be the last of the 0.6 series.
326 Unless it does not work with perl 5.6, which I don't use yet,
327 and therefore against which I cannot test.
329 Wed Mar 29 19:55:21 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
333 Added note about format incompatibilities with old versions
334 (i.e. pre 0.5@9 formats, which cannot be understood as there
335 was no versionning information in the file by then).
337 Protect all $@ variables when eval {} used, to avoid corrupting
338 it when store/retrieve is called within an exception handler.
340 Mistakenly included "patchlevel.h" instead of <patchlevel.h>,
341 preventing Perl's patchlevel from being included, which is
342 needed starting from 5.6.
344 Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
348 Fixed shared "undef" bug in hashes, which did not remain shared
349 through store/retrieve.
351 Thu Feb 10 19:48:16 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
355 added last_op_in_netorder() predicate
356 documented last_op_in_netorder()
357 added tests for the new last_op_in_netorder() predicate
359 Wed Oct 20 19:07:36 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
363 Forgot to update VERSION
365 Tue Oct 19 21:25:02 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
369 Added mention of japanese translation for the manual page.
371 Fixed typo in macro that made threaded code not compilable,
372 especially on Win32 platforms.
374 Changed detection of older perls (pre-5.005) by testing PATCHLEVEL
375 directly instead of relying on internal symbols.
377 Tue Sep 14 22:13:28 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
381 Integrated "thread-safe" patch from Murray Nesbitt.
382 Note that this may not be very efficient for threaded code,
383 see comment in the code.
385 Try to avoid compilation warning on 64-bit CPUs. Can't test it,
386 since I don't have access to such machines.
388 Mon Jul 12 14:37:19 METDST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
392 changed my e-mail to pobox.
394 mentionned it is not thread-safe.
396 updated version number.
398 uses new internal PL_* naming convention.
400 Fri Jul 3 13:38:16 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
404 Updated benchmark figures due to recent optimizations done in
405 store(): tagnums are now stored as-is in the hash table, so
406 no surrounding SV is created. And the "shared keys" mode for
407 hash table was turned off.
409 Fixed backward compatibility (wrt 0.5@9) for retrieval of
410 blessed refs. That old version did something wrong, but the
411 bugfix prevented correct retrieval of the old format.
413 Mon Jun 22 11:00:48 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
417 Changed benchmark figures.
419 Adjust refcnt of tied objects after calling sv_magic() to avoid
420 memory leaks. Contributed by Jeff Gresham.
422 Fri Jun 12 11:50:04 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
426 Added workaround for persistent LVALUE-ness in perl5.004. All
427 scalars tagged as being an lvalue are handled as if they were
428 not an lvalue at all. Added test for that LVALUE bug workaround.
430 Now handles Perl immortal scalars explicitely, by storing &sv_yes
431 as such, explicitely.
433 Retrieval of non-immortal undef cannot be shared. Previous
434 version was over-optimizing by not creating a separate SV for
435 all undefined scalars seen.
437 Thu Jun 4 17:21:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
441 Baseline for Storable-0.6@0.
443 This version introduces a binary incompatibility in the generated
444 binary image, which is more compact than older ones by approximatively
445 15%, depending on the exact degree of sharing in your structures.
447 The good news is that your older images can still be retrieved with
448 this version, i.e. backward compatibility is preserved. This version
449 of Storable can only generate new binaries however.
451 Another good news is that the retrieval of data structure is
452 significantly quicker than before, because a Perl array is used
453 instead of a hash table to keep track of retrieved objects, and
454 also because the image being smaller, less I/O function calls are
457 Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
461 Version number now got from Storable.pm directly.
463 Fixed overzealous sv_type() optimization, which would make
464 Storable fail when faced with an "upgraded" SV to the PVIV
465 or PVNV kind containing a reference.
467 Thu Apr 30 15:11:30 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
471 Extended the SYNOPSIS section to give quick overview of the
472 routines and their signature.
474 Optimized sv_type() to avoid flags checking when not needed, i.e.
475 when their type makes it impossible for them to be refs or tied.
476 This slightly increases throughput by a few percents when refs
477 and tied variables are marginal occurrences in your data.
479 Stubs for XS now use OutputStream and InputStream file types to
480 make it work when the given file is actually a socket. Perl
481 makes a distinction for sockets in its internal I/O structures
482 by having both a read and a write structure, whereas plain files
485 Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
489 Thanks to a contribution from Benjamin A. Holzman, Storable is now
490 able to correctly serialize tied SVs, i.e. tied arrays, hashes
493 Thu Apr 9 18:07:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
497 I said SvPOK() had changed to SvPOKp(), but that was a lie...
499 Wed Apr 8 13:14:29 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
503 Wrote sizeof(SV *) instead of sizeof(I32) when portable, which
504 in effect mangled the object tags and prevented portability
505 accross 32/64 bit architectures!
507 Wed Mar 25 14:57:02 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
511 Added code example for store_fd() and retrieve_fd() in the
512 man page, to emphasize that file descriptors must be passed as
513 globs, not as plain strings.
515 Cannot use SV addresses as tag when using nstore() on LP64. This
516 was the cause of problems when creating a storable image on an
517 LP64 machine and retrieving it on an ILP32 system, which is
518 exactly what nstore() is meant for...
520 However, we continue to use SV addresses as tags for plain store(),
521 because benchamarking shows that it saves up to 8% of the store
522 time, and store() is meant to be fast at the expense of lack
525 This means there will be approximately an 8% degradation of
526 performance for nstore(), but it's now working as expected.
527 That cost may vary on your machine of course, since it is
528 solely caused by the memory allocation overhead used to create
529 unique SV tags for each distinct stored SV.
531 Tue Jan 20 09:21:53 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
535 Don't use any '_' in version number.
537 Tue Jan 13 17:51:50 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
541 Updated version number.
543 added binmode() calls for systems where it matters.
545 Be sure to pass globs, not plain file strings, to C routines,
546 so that Storable can be used under the Perl debugger.
548 Wed Nov 5 10:53:22 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
552 Fix memory leaks on seen hash table and returned SV refs.
554 Storable did not work properly when tainting enabled.
556 Fixed "Allocation too large" messages in freeze/thaw and added.
557 proper regression test in t/freeze.t.
559 Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
563 Updated version number
565 Added freeze/thaw interface and dclone.
567 Fri May 16 10:45:47 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
571 Forgot that AutoLoader does not export its own AUTOLOAD.
574 use AutoLoader 'AUTOLOAD';
576 but that would not be backward compatible. So the export is
579 Tue Mar 25 11:21:32 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
583 Empty scalar strings are now "defined" at retrieval time.
585 New test to ensure an empty string is defined when retrieved.
587 Thu Feb 27 16:32:44 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
591 Updated version number
593 Declare VERSION as being used
595 Fixed a typo in the PerlIO_putc remapping.
596 PerlIO_read and perlIO_write inverted size/nb_items.
597 (only relevant for pre-perl5.004 versions)
599 Thu Feb 27 15:58:31 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
603 Updated version number
605 Added VERSION identification
607 Allow build with perl5.003, which is ante perlIO time
609 Mon Jan 13 17:53:18 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
615 Wed Jan 22 15:19:56 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
619 Updated version number in Makefile.PL.
621 Added "thanks to" section to README.
623 Documented new forgive_me variable.
627 Added forgive_me support to allow store() of data structures
628 containing non-storable items like CODE refs.