1 Revision history for Variable-Magic
3 0.56 2015-03-11 15:15 UTC
4 + Fix : [RT #101410] : Install fails in blead
5 Even though the change that caused this error was reverted from
6 blead, lvalue uses of ERRSV have been removed from this module
7 so that there will be no breakage when this change is possibly
8 reintroduced in the following months.
9 Thanks Dave Rolsky for reporting.
10 + Tst : $ENV{$Config{ldlibpthname}} is now preserved on all platforms,
11 which will address failures of t/17-ctl.t with unusual
12 compilers (like icc) that link all their compiled objects to
14 + Tst : The global destruction test is now only run on perl 5.13.4 and
15 higher, and only if either Perl::Destruct::Level is installed
16 or PERL_DESTRUCT_LEVEL is set and the perl is a debugging perl.
17 This will solve rare crashes of t/15-self.t on perl 5.13.3 and
20 0.55 2014-10-20 23:45 UTC
21 + Fix : [RT #99319] : t/18-opinfo.t: test failure due to change in
23 The new METHOP op class added in perl 5.21.5 is now supported.
24 Thanks James E Keenan for reporting.
26 0.54 2014-09-22 17:30 UTC
27 + Add : The new constant VMG_COMPAT_CODE_COPY_CLONE evaluates to true
28 if your perl calls 'copy' magic when a magical code prototype
29 is cloned, which is currently the case for perl 5.17.0 and
31 + Fix : [RT #90205] : copy magic on subs puts raw CV in $_[3]
32 $_[3] will now contain a reference to the cloned code when
33 'copy' magic is called for a coderef.
34 Thanks Lukas Mai for reporting.
35 + Fix : t/35-stash.t has been taught about perl 5.21.4.
36 + Fix : Tests using run_perl() in t/17-ctl.t will no longer fail on
39 0.53 2013-09-01 17:50 UTC
40 This is a maintenance release. The code contains no functional change.
41 Satisfied users of version 0.52 can skip this update.
42 + Fix : [RT #86338] : typo fix.
43 Thanks dsteinbrunner@pobox.com for the patch.
44 + Tst : Author tests are no longer bundled with this distribution.
45 They are only made available to authors in the git repository.
47 0.52 2012-11-05 02:30 UTC
48 + Add : The new constant VMG_COMPAT_SCALAR_NOLEN evaluates to true
49 when your perl does not call 'len' magic for scalars, which
50 is the case for perl 5.17.4 and above.
52 + Tst : t/35-stash.t has been taught about perl 5.17.4.
53 + Tst : Author tests overhaul.
55 0.51 2012-08-18 15:00 UTC
56 + Fix : It is now safe to call dispell() from inside 'free', 'copy' and
57 'uvar' callbacks to dispell the magic currently in use.
58 Thanks Clinton Gormley for reporting.
59 + Fix : Exceptions thrown from inside a 'free' callback are now always
60 consistently propagated outside of the callback. They used to
61 be lost when the 'free' callback was invoked at the end of an
63 + Fix : The 'reset RMG flag' workaroundn used to allow wizards with
64 both 'uvar' and 'clear' magics to be cast onto a hash, has been
67 0.50 2012-06-24 23:00 UTC
68 + Fix : Less memory is leaked when a wizard is freed during global
69 destruction, or when an exception is thrown from a 'free'
71 + Fix : [RT #77991] : t/17-ctl.t fails on perl 5.14 and 5.16.
72 This was actually an issue with ActivePerl, and this test
73 has learned to cope with it.
74 Thanks Gisle Aas for reporting.
75 + Tst : t/35-stash.t has been taught about perl 5.17.1.
76 + Doc : Many clarifications.
78 0.49 2012-06-05 21:40 UTC
79 This is a maintenance release. The code contains no functional change.
80 Satisfied users of version 0.48 can skip this update.
81 + Fix : [RT #77644] : t/17_ctl.t fails in test 44 since 5.17.0.
82 This test has been taught about perl 5.17.0.
83 Thanks Reini Urban for reporting.
84 + Tst : t/99-kwalitee.t will be skipped when only problematic versions
85 of its dependencies are available.
87 0.48 2012-02-17 23:40 UTC
88 + Add : You can now pass a reference to undef as the magic callback in
89 order to install a no-op callback.
90 Thanks Florian Ragwitz for the suggestion.
92 0.47 2011-10-27 16:55 UTC
93 + Add : The new constant VMG_COMPAT_HASH_DELETE_NOUVAR_VOID evaluates
94 to true when "delete $hash{key}" does not call 'delete' uvar
95 magic in void context.
96 + Chg : The MAGIC tokens created by this module no longer use the
97 mg_private member for storing a "magical" signature.
98 + Fix : Triggering magic in a thread on a variable cloned from the main
99 interpreter, and when the wizard already went out of scope,
100 will not segfault anymore.
101 + Opt : The pointer table is no longer needed to ensure thread safety.
102 The size of the object code is about 8% smaller for threaded
104 + Tst : Threads tests will not fail anymore if resources constraints
105 prevent the system from creating all the required threads.
107 0.46 2011-01-23 16:45 UTC
108 + Fix : [RT #64866] : Assertion failure with perl 5.13.9.
109 The real problem was that dispell() wasn't properly resetting
110 the magical sv flags, which became visible in the test suite
111 only with perl 5.13.9.
112 Thanks Joshua ben Jore for reporting.
114 0.45 2010-11-21 23:15 UTC
115 This is a maintenance release. The code contains no functional change.
116 Users of 0.44 can skip this update.
117 + Doc : C++ compilers are officially NOT supported.
118 + Doc : The minimum perl 5.10.0 requirement for uvar magic has been
120 Thanks Peter Rabbitson for pointing this out and contributing a
122 + Tst : Tune for perl 5.13.7.
123 + Tst : Capture::Tiny will be used in t/17-ctl.t if and only if it can
124 capture a simple run.
126 0.44 2010-09-24 19:10 UTC
127 + Fix : Broken linkage on Windows with gcc 3.4, which appears in
128 particular when using ActivePerl's default compiler suite.
129 For those setups, the Variable::Magic shared library will now
130 be linked against the perl dll directly (instead of the import
131 library). This (should) fix RT #51483.
132 Thanks Christian Walde for helping to reproduce this failure
134 + Rem : Support for development perls from the 5.11 branch but older
135 than the 5.11.0 release was removed. This could cause more
136 recent setups to fail.
137 + Tst : Threads tests are now only run on perl 5.13.4 and higher.
138 They could segfault randomly because of what seems to be an
139 internal bug of Perl, which has been addressed in 5.13.4.
140 There is also an environment variable that allows you to
141 forcefully run those tests, but it should be set only for
142 author testing and not for end users.
144 0.43 2010-06-25 23:35 UTC
145 + Add : The new constant VMG_COMPAT_GLOB_GET tells you whether get magic
146 is called for globs. It's true starting perl 5.13.2.
147 + Chg : All callbacks are now called within an eval-like context.
148 Only free callbacks used to be called that way.
149 + Fix : Some exceptions thrown from a free callback could be lost.
150 + Fix : Croak messages could sometimes be repeated several times.
151 + Fix : t/41-clone.t segfaulting with perl 5.13.2.
153 0.42 2010-05-19 00:15 UTC
154 This is a maintenance release. The code contains no functional change.
155 Users of 0.41 can skip this update.
156 + Fix : Test failures with perl 5.13.
157 + Tst : Improve coverage.
159 0.41 2010-03-15 17:35 UTC
160 + Doc : Tweaks and fixups.
162 + Fix : Compatibility with the soon-to-be-released perl 5.12.0.
163 + Fix : Correctly propagate the errors thrown when variable destruction
164 happens at compile-time and not from inside eval STRING.
165 Thanks Florian Ragwitz and Ash Berlin for reporting.
167 0.40 2010-01-06 23:20 UTC
168 + Fix : Possible memory miswrites when passing data arguments to cast().
169 + Fix : Minor C portability tweaks.
171 0.39 2009-12-01 00:05 UTC
172 + Add : You can use a function name as a callback by passing a string
173 reference to wizard() instead of a code reference.
174 + Fix : Compatiblity with perl 5.11.1 and 5.11.2.
175 + Fix : Scalars stored into the data slot no longer leak.
176 + Fix : Thread destruction should not segfault anymore.
177 + Opt : As a result of removing the deprecated features, less memory is
178 used for both threaded (a pointer table is used instead of a
179 hash) and non-threaded (no global structure is needed anymore)
181 + Rem : As advertised in the precendent version, all the signature
182 related features were removed. This includes the 'sig' option
183 to wizard(), the getsig() and gensig() functions, and the SIG_*
186 0.38 2009-10-04 16:10 UTC
187 + Dep : All the signature-related features are DEPRECATED and will be
188 removed in december 2009. This includes the 'sig' option to
189 wizard(), the getsig() and gensig() functions, and the SIG_*
191 Afaik those features were never used in real life, and they make
192 the XS code slow and brittle. If you want your wizards to be
193 globally available, you'll have to handle that yourself by
194 storing them in a hash.
195 + Doc : More examples in a new COOKBOOK section.
196 + Fix : Compatiblity with the 5.11.0 release. The new compatibility
197 constant VMG_COMPAT_ARRAY_PUSH_NOLEN_VOID was added to cover
199 + Fix : Work around Kwalitee test misfailures.
201 0.37 2009-08-25 15:25 UTC
202 + Add : The new constant VMG_FORKSAFE can be tested to know whether the
203 module will behave nicely when fork()ing. It's currently always
204 true except on Windows where you need perl 5.10.1 for it to be
206 + Doc : Nits and clarifications.
208 0.36 2009-07-05 16:30 UTC
209 + Chg : getdata() now returns an empty list when no magic is present.
210 It used to return undef but did not croak as stated in the doc.
211 Thanks Matt S. Trout for pointing this out.
212 + Fix : Building and tests with ActiveState Perl 5.8 build >= 822.
214 0.35 2009-05-15 20:50 UTC
215 This is a maintenance release. Users of 0.34 can skip this update.
216 + Chg : uvar magic used to be enabled for 5.9.5. It now requires 5.10 or
218 + Fix : Silence compilation warnings for cxinc().
219 + Rem : The workaround introduced in 0.31 for the "Unknown errors" issue
220 was removed, as regression tests show that 0.34 provides a
222 + Tst : Skip t/40-threads.t on 5.8.x with old versions of threads and
225 0.34 2009-04-19 16:55 UTC
226 + Fix : Destruction of wizards in require.
227 + Fix : "panic: restartop" in 5.10 when dieing at compile time triggers
230 0.33 2009-03-26 00:00 UTC
231 + Chg : cast() and getsig() now croak too when an invalid signature is
233 + Fix : It was possible to generate with gensig() the same signature
234 twice before actually defining the wizards.
235 + Tst : More stash tests.
236 + Tst : Test when the magic takes place in some cases.
237 + Tst : Improved coverage for the op object feature.
239 0.32 2009-03-01 13:45 UTC
240 + Chg : dispell() and getdata() now croak when an invalid signature is
242 + Doc : More examples and a brand new synopsis.
243 + Fix : The signature is no longer stored in the mg_private member of
244 the MAGIC struct. This fixes possible clashes with magics from
246 + Fix : op info objects weren't blessed into the right class.
247 + Fix : Races when initializing the op names cache under threads.
249 0.31 2009-02-19 23:50 UTC
250 + Doc : A somewhat better introduction. The FUNCTION section was moved
251 before CONSTANTS and HISTORY since it's more important. And
252 'assignation' is really 'assignment', says Ricardo. :)
253 + Fix : The op_info features are now thread safe.
254 + Fix : Segfaults that occured when trying to get the op name during
256 + Fix : Segfaults and wrong "Unknown error" exceptions that happened
257 when dieing in require caused a free callback to fire.
258 Thanks Florian Ragwitz for reporting.
260 0.30 2009-02-12 17:05 UTC
261 + Add : You can now get in all the callbacks either the name or a B::OP
262 object representing the current op if you pass the new 'op_info'
263 option to wizard() with respective values VMG_OP_INFO_NAME and
265 + Doc : Document magic on array/hashes values.
266 + Fix : Completely skip free callbacks during global destruction.
267 + Tst : Tests now pass again on 5.8.3 and lower (they were broken since
269 + Tst : Test magic on symbol table and on array/hash values.
271 0.29 2009-02-08 11:10 UTC
272 + Add : Set $_[2] to the default length for len magic on scalars.
273 + Chg : Perl 5.8 is required.
274 + Chg : The CLONE method will no longer be defined for non-threaded
277 + Fix : The logic for turning off the GMAGICAL flag on hashes when no
278 uvar magic is involved could do nasty things on tied hashes.
279 + Upd : More resources in META.yml.
281 0.28 2009-01-24 17:05 UTC
282 + Add : The 'copy_key' option to wizard(), that enables the safe edition
283 of $_[2] in uvar callbacks to redirect actions to another key.
284 + Fix : Exception objects thrown from callbacks at compile time were not
286 + Fix : free callbacks triggered during global destruction could lead to
287 segfaults because the wizard was already freed.
288 + Fix : Refcount of scalars with free callback.
289 + Fix : Segfaults when using get or uvar magic simultaneously with clear
291 + Doc : The documentation about what the callbacks take and should
292 return is now somewhat clearer.
294 0.27 2009-01-18 23:40 UTC
295 + Fix : Offset of 1 with len magic on scalars.
296 + Fix : Segfaults that seem to happen when one croak in a callback at
297 compile time (thanks Ash Berlin for the testcase).
298 + Upd : Resources in META.yml (Florian Ragwitz)
299 + Tst : Large cleanup resulting in a wider set of tested behaviours.
301 0.26 2008-11-29 22:05 UTC
302 + Fix : Leaks of cloned coderefs that access lexicals.
303 + Fix : Building with maint-5.8 and blead (thanks Andreas Koenig).
305 0.25 2008-11-12 19:40 UTC
306 + Fix : Compatibility with 5.8.9.
307 + Fix : Old Pod::Coverage don't ignore CLONE.
308 + Upd : META.yml spec updated to 1.4.
310 0.24 2008-10-12 14:55 UTC
311 + Fix : Really fix it.
313 0.23 2008-10-11 17:25 UTC
314 + Fix : Building on perls with multiplicity but not ithreads.
316 0.22 2008-09-29 19:40 UTC
317 + Doc : Explain how magic differs from overloading.
318 + Fix : Disable thread safety for perl 5.8 on Windows. If you want it
319 enabled, reliable feedback is welcome.
321 0.21_02 2008-09-28 23:10 UTC
322 + Fix : Hide previously deleted magic tokens in vmg_svt_free (Yuval
325 0.21_01 2008-09-22 13:25 UTC
326 + Chg : Shrink the context data.
327 + Fix : Enable thread safety features for perls that just lack
329 + Tst : Output threads and threads::shared versions in threads tests.
331 0.20 2008-09-20 23:35 UTC
332 + Fix : The module is now thread safe. You can test if it had been
333 built with thread safety features enabled through the
334 VMG_THREADSAFE constant.
335 + Fix : Correct backtrace when wizard() croaks.
336 + Fix : 'cast $a, undef' should throw an 'invalid wizard' error instead
337 of an 'invalid signature'.
339 0.19 2008-07-05 15:05 UTC
340 + Add : The patchlevel with which the module has been built is now
341 exported with the VMG_PERL_PATCHLEVEL constant.
342 + Chg : Version checking macros overhaul.
343 + Fix : 'uvar' magic should only be enabled in maint-5.10 and higher
344 since it's never going to be backported to 5.8.
345 + Fix : Likewise, the copy callback API change should only be enabled in
347 + Fix : Instead of those ackward fake refcount increments in the wizard
348 destructor, flag the wizard with SVf_BREAK and force temps
349 cleanup. This should fix the wizard destruction for all perls,
350 and solves the leak introduced by the previous solution.
352 0.18 2008-07-04 12:20 UTC
353 + Fix : Enable uvar magic when patchlevel >= 28419.
354 + Fix : Set VMG_COMPAT_ARRAY_PUSH_NOLEN for Windows ActiveState Perl
357 0.17 2008-05-11 09:05 UTC
358 + Fix : 5.8.4 and lower don't have mPUSHi. Gah. Yeah, I know, I should
361 0.16 2008-05-10 22:05 UTC
362 + Add : The samples/copy.pl script.
363 + Chg : The sv_magical() + vmg_mg_magical() combo was simplified into
365 + Tst : t/33-code.t was testing scalars, not code.
366 + Tst : is() and like() are better than ok().
367 + Tst : 100% coverage reached.
369 0.15 2008-04-11 18:25 UTC
370 + Chg : Factor vmg_cb_call{1,2,3}() into one function.
371 + Fix : len magic is no longer called when taking the length() of a
372 magical scalar since p32969. The VMG_COMPAT_SCALAR_LENGTH_NOLEN
373 constant was added to cover this.
374 + Tst : More tests for t/22-len.t.
376 0.14 2008-03-24 12:35 UTC
377 + Fix : t/16-huf.t failures on Solaris and FreeBSD caused by not
378 updating mg->mg_ptr after Renew-ing it on dispell.
379 + Fix : Undefining MGf_DUP caused the wizard's magic to be wrongly
380 initialized. MGf_DUP now appears as true from the user point of
381 view, but the dup callback isn't actually set.
382 + Fix : Warnings with blead due to copy callbacks now taking an I32 (API
384 + Fix : vmg_svt_val() and vmg_uvar_del() aren't required if no uvar
386 + Tst : Tests now output the version of optional modules loaded.
388 0.13 2008-03-19 14:35 UTC
389 + Doc : Link to coverage report.
390 + Fix : Correct dependencies listing in META.yml.
391 + Tst : Improved test coverage.
392 + Tst : Print the patchlevel as a comment.
393 + Tst : Use is() where it's relevant.
394 + Tst : t/16-huf.t now really tests interaction with H::U::FH.
396 0.12 2008-02-07 18:15 UTC
397 + Fix : POD error. Thanks to Chris Williams (BinGOs) for the quick
400 0.11 2008-02-07 17:55 UTC
401 + Add : Copy callbacks now receive the current key/index in $_[2].
402 The current value/element is now in $_[3].
403 + Chg : The trigger of copy magic on list assignment with perls greater
404 than 5.10.0 was caused by uvar magic. Hence,
405 VMG_COMPAT_HASH_LISTASSIGN_COPY was removed. Use VMG_UVAR
407 + Fix : Build failures on patched perls.
408 + Tst : Added missing exported symbols to 01-import.t.
410 0.10 2008-02-04 11:30 UTC
411 + Add : New script : samples/vm_vs_tie.pl, that benchmarks our uvar
412 magic versus tied hashes.
413 + Add : The VMG_COMPAT_* constants can be used from userspace to check
414 perl magic abilities.
415 + Fix : Callbacks that returned undef made us croak, breaking the
416 variable behaviour (regression test in 14-callbacks.t).
417 + Fix : uvar callbacks weren't tested for non-NULL-ity before being
418 called (regression test in 28-uvar.t).
419 + Tst : Fix typo in 25-copy.t that prevented Tie::Hash tests to be ran.
421 0.09 2008-02-02 11:30 UTC
422 + Doc : Explicitely say that uvar callbacks are safely ignored for
424 + Doc : Document caveats and fix the usual set of typos.
425 + Fix : vmg_dispell() didn't check if the ext magic were ours when
426 counting wizards that have uvar callbacks, resulting in a
427 possible memory misread.
428 + Fix : getdata() now returns directly the data object, and no longer a
429 copy. This caused a leak.
430 + Tst : Prefix author tests by 9*-.
431 + Tst : New optional author test : 95-portability-files.t, that uses
432 Test::Portability::Files when it's present.
433 + Tst : New test : 14-self.t, that tests application of magic on the
435 + Tst : Move Hash::Util::FieldHash tests out of 11-multiple.t to
438 0.08 2008-02-01 16:55 UTC
439 + Add : copy magic for tied arrays/hashes.
441 + Add : uvar magics : fetch, store, exists, delete for hashes.
444 Internal development versions.
446 0.06 2007-11-20 10:10 UTC
447 + Chg : 5.7.3 is now officially required.
448 + Fix : "data" test failures on 5.8.{0,2}.
449 + Fix : Drand01() vs rand().
451 0.05 2007-11-19 09:10 UTC
452 + Fix : 5.10.0_RC1 compatibility fix.
454 0.04 2007-08-28 12:25 UTC
455 + Chg : s/require (XSLoader)/use $1/.
456 + Fix : Tests are now strict.
457 + Fix : Complete dependencies.
459 0.03 2007-08-01 17:20 UTC
460 + Add : Passing the signature of an already defined magic to wizard()
461 now returns the corresponding magic object.
462 + Add : You can pass the numeric signature as the wizard argument of
463 cast(), dispell() and getdata().
464 + Add : Any argument specified after the wizard (or the signature) in a
465 call to cast() is now passed to the private data constructor in
467 + Chg : $_[0] is now always a reference to the magic variable in all
468 callbacks. The reason for changing from the previous behaviour
469 is that one may want to apply the same magic to a plain scalar
470 and to a scalar reference, and needs a way to distinguish
471 between them in the callback (say, ref()).
472 + Fix : Wizard object destruction used not to free the signature.
474 0.02 2007-07-27 13:50 UTC
475 + Fix : In response to test report 548152 :
476 Newx() and SvMAGIC_set() not present on older perls.
477 + Fix : In response to test report 548275 :
478 Since perl 5.9.5, 'clear' magic is invoked when an array is
479 undefined (bug #43357). Moreover, 'len' magic is no longer
480 called by pushing an element since perl 5.9.3.
481 + Fix : Missing glob test in MANIFEST.
483 0.01 2007-07-25 16:15 UTC
484 First version, released on an unsuspecting world.