1 Revision history for Variable-Magic
3 0.52 2012-11-05 02:30 UTC
4 + Add : The new constant VMG_COMPAT_SCALAR_NOLEN evaluates to true
5 when your perl does not call 'len' magic for scalars, which
6 is the case for perl 5.17.4 and above.
8 + Tst : t/35-stash.t has been taught about perl 5.17.4.
9 + Tst : Author tests overhaul.
11 0.51 2012-08-18 15:00 UTC
12 + Fix : It is now safe to call dispell() from inside 'free', 'copy' and
13 'uvar' callbacks to dispell the magic currently in use.
14 Thanks Clinton Gormley for reporting.
15 + Fix : Exceptions thrown from inside a 'free' callback are now always
16 consistently propagated outside of the callback. They used to
17 be lost when the 'free' callback was invoked at the end of an
19 + Fix : The 'reset RMG flag' workaroundn used to allow wizards with
20 both 'uvar' and 'clear' magics to be cast onto a hash, has been
23 0.50 2012-06-24 23:00 UTC
24 + Fix : Less memory is leaked when a wizard is freed during global
25 destruction, or when an exception is thrown from a 'free'
27 + Fix : [RT #77991] : t/17-ctl.t fails on perl 5.14 and 5.16.
28 This was actually an issue with ActivePerl, and this test
29 has learned to cope with it.
30 Thanks Gisle Aas for reporting.
31 + Tst : t/35-stash.t has been taught about perl 5.17.1.
32 + Doc : Many clarifications.
34 0.49 2012-06-05 21:40 UTC
35 This is a maintenance release. The code contains no functional change.
36 Satisfied users of version 0.48 can skip this update.
37 + Fix : [RT #77644] : t/17_ctl.t fails in test 44 since 5.17.0.
38 This test has been taught about perl 5.17.0.
39 Thanks Reini Urban for reporting.
40 + Tst : t/99-kwalitee.t will be skipped when only problematic versions
41 of its dependencies are available.
43 0.48 2012-02-17 23:40 UTC
44 + Add : You can now pass a reference to undef as the magic callback in
45 order to install a no-op callback.
46 Thanks Florian Ragwitz for the suggestion.
48 0.47 2011-10-27 16:55 UTC
49 + Add : The new constant VMG_COMPAT_HASH_DELETE_NOUVAR_VOID evaluates
50 to true when "delete $hash{key}" does not call 'delete' uvar
51 magic in void context.
52 + Chg : The MAGIC tokens created by this module no longer use the
53 mg_private member for storing a "magical" signature.
54 + Fix : Triggering magic in a thread on a variable cloned from the main
55 interpreter, and when the wizard already went out of scope,
56 will not segfault anymore.
57 + Opt : The pointer table is no longer needed to ensure thread safety.
58 The size of the object code is about 8% smaller for threaded
60 + Tst : Threads tests will not fail anymore if resources constraints
61 prevent the system from creating all the required threads.
63 0.46 2011-01-23 16:45 UTC
64 + Fix : [RT #64866] : Assertion failure with perl 5.13.9.
65 The real problem was that dispell() wasn't properly resetting
66 the magical sv flags, which became visible in the test suite
67 only with perl 5.13.9.
68 Thanks Joshua ben Jore for reporting.
70 0.45 2010-11-21 23:15 UTC
71 This is a maintenance release. The code contains no functional change.
72 Users of 0.44 can skip this update.
73 + Doc : C++ compilers are officially NOT supported.
74 + Doc : The minimum perl 5.10.0 requirement for uvar magic has been
76 Thanks Peter Rabbitson for pointing this out and contributing a
78 + Tst : Tune for perl 5.13.7.
79 + Tst : Capture::Tiny will be used in t/17-ctl.t if and only if it can
82 0.44 2010-09-24 19:10 UTC
83 + Fix : Broken linkage on Windows with gcc 3.4, which appears in
84 particular when using ActivePerl's default compiler suite.
85 For those setups, the Variable::Magic shared library will now
86 be linked against the perl dll directly (instead of the import
87 library). This (should) fix RT #51483.
88 Thanks Christian Walde for helping to reproduce this failure
90 + Rem : Support for development perls from the 5.11 branch but older
91 than the 5.11.0 release was removed. This could cause more
92 recent setups to fail.
93 + Tst : Threads tests are now only run on perl 5.13.4 and higher.
94 They could segfault randomly because of what seems to be an
95 internal bug of Perl, which has been addressed in 5.13.4.
96 There is also an environment variable that allows you to
97 forcefully run those tests, but it should be set only for
98 author testing and not for end users.
100 0.43 2010-06-25 23:35 UTC
101 + Add : The new constant VMG_COMPAT_GLOB_GET tells you whether get magic
102 is called for globs. It's true starting perl 5.13.2.
103 + Chg : All callbacks are now called within an eval-like context.
104 Only free callbacks used to be called that way.
105 + Fix : Some exceptions thrown from a free callback could be lost.
106 + Fix : Croak messages could sometimes be repeated several times.
107 + Fix : t/41-clone.t segfaulting with perl 5.13.2.
109 0.42 2010-05-19 00:15 UTC
110 This is a maintenance release. The code contains no functional change.
111 Users of 0.41 can skip this update.
112 + Fix : Test failures with perl 5.13.
113 + Tst : Improve coverage.
115 0.41 2010-03-15 17:35 UTC
116 + Doc : Tweaks and fixups.
118 + Fix : Compatibility with the soon-to-be-released perl 5.12.0.
119 + Fix : Correctly propagate the errors thrown when variable destruction
120 happens at compile-time and not from inside eval STRING.
121 Thanks Florian Ragwitz and Ash Berlin for reporting.
123 0.40 2010-01-06 23:20 UTC
124 + Fix : Possible memory miswrites when passing data arguments to cast().
125 + Fix : Minor C portability tweaks.
127 0.39 2009-12-01 00:05 UTC
128 + Add : You can use a function name as a callback by passing a string
129 reference to wizard() instead of a code reference.
130 + Fix : Compatiblity with perl 5.11.1 and 5.11.2.
131 + Fix : Scalars stored into the data slot no longer leak.
132 + Fix : Thread destruction should not segfault anymore.
133 + Opt : As a result of removing the deprecated features, less memory is
134 used for both threaded (a pointer table is used instead of a
135 hash) and non-threaded (no global structure is needed anymore)
137 + Rem : As advertised in the precendent version, all the signature
138 related features were removed. This includes the 'sig' option
139 to wizard(), the getsig() and gensig() functions, and the SIG_*
142 0.38 2009-10-04 16:10 UTC
143 + Dep : All the signature-related features are DEPRECATED and will be
144 removed in december 2009. This includes the 'sig' option to
145 wizard(), the getsig() and gensig() functions, and the SIG_*
147 Afaik those features were never used in real life, and they make
148 the XS code slow and brittle. If you want your wizards to be
149 globally available, you'll have to handle that yourself by
150 storing them in a hash.
151 + Doc : More examples in a new COOKBOOK section.
152 + Fix : Compatiblity with the 5.11.0 release. The new compatibility
153 constant VMG_COMPAT_ARRAY_PUSH_NOLEN_VOID was added to cover
155 + Fix : Work around Kwalitee test misfailures.
157 0.37 2009-08-25 15:25 UTC
158 + Add : The new constant VMG_FORKSAFE can be tested to know whether the
159 module will behave nicely when fork()ing. It's currently always
160 true except on Windows where you need perl 5.10.1 for it to be
162 + Doc : Nits and clarifications.
164 0.36 2009-07-05 16:30 UTC
165 + Chg : getdata() now returns an empty list when no magic is present.
166 It used to return undef but did not croak as stated in the doc.
167 Thanks Matt S. Trout for pointing this out.
168 + Fix : Building and tests with ActiveState Perl 5.8 build >= 822.
170 0.35 2009-05-15 20:50 UTC
171 This is a maintenance release. Users of 0.34 can skip this update.
172 + Chg : uvar magic used to be enabled for 5.9.5. It now requires 5.10 or
174 + Fix : Silence compilation warnings for cxinc().
175 + Rem : The workaround introduced in 0.31 for the "Unknown errors" issue
176 was removed, as regression tests show that 0.34 provides a
178 + Tst : Skip t/40-threads.t on 5.8.x with old versions of threads and
181 0.34 2009-04-19 16:55 UTC
182 + Fix : Destruction of wizards in require.
183 + Fix : "panic: restartop" in 5.10 when dieing at compile time triggers
186 0.33 2009-03-26 00:00 UTC
187 + Chg : cast() and getsig() now croak too when an invalid signature is
189 + Fix : It was possible to generate with gensig() the same signature
190 twice before actually defining the wizards.
191 + Tst : More stash tests.
192 + Tst : Test when the magic takes place in some cases.
193 + Tst : Improved coverage for the op object feature.
195 0.32 2009-03-01 13:45 UTC
196 + Chg : dispell() and getdata() now croak when an invalid signature is
198 + Doc : More examples and a brand new synopsis.
199 + Fix : The signature is no longer stored in the mg_private member of
200 the MAGIC struct. This fixes possible clashes with magics from
202 + Fix : op info objects weren't blessed into the right class.
203 + Fix : Races when initializing the op names cache under threads.
205 0.31 2009-02-19 23:50 UTC
206 + Doc : A somewhat better introduction. The FUNCTION section was moved
207 before CONSTANTS and HISTORY since it's more important. And
208 'assignation' is really 'assignment', says Ricardo. :)
209 + Fix : The op_info features are now thread safe.
210 + Fix : Segfaults that occured when trying to get the op name during
212 + Fix : Segfaults and wrong "Unknown error" exceptions that happened
213 when dieing in require caused a free callback to fire.
214 Thanks Florian Ragwitz for reporting.
216 0.30 2009-02-12 17:05 UTC
217 + Add : You can now get in all the callbacks either the name or a B::OP
218 object representing the current op if you pass the new 'op_info'
219 option to wizard() with respective values VMG_OP_INFO_NAME and
221 + Doc : Document magic on array/hashes values.
222 + Fix : Completely skip free callbacks during global destruction.
223 + Tst : Tests now pass again on 5.8.3 and lower (they were broken since
225 + Tst : Test magic on symbol table and on array/hash values.
227 0.29 2009-02-08 11:10 UTC
228 + Add : Set $_[2] to the default length for len magic on scalars.
229 + Chg : Perl 5.8 is required.
230 + Chg : The CLONE method will no longer be defined for non-threaded
233 + Fix : The logic for turning off the GMAGICAL flag on hashes when no
234 uvar magic is involved could do nasty things on tied hashes.
235 + Upd : More resources in META.yml.
237 0.28 2009-01-24 17:05 UTC
238 + Add : The 'copy_key' option to wizard(), that enables the safe edition
239 of $_[2] in uvar callbacks to redirect actions to another key.
240 + Fix : Exception objects thrown from callbacks at compile time were not
242 + Fix : free callbacks triggered during global destruction could lead to
243 segfaults because the wizard was already freed.
244 + Fix : Refcount of scalars with free callback.
245 + Fix : Segfaults when using get or uvar magic simultaneously with clear
247 + Doc : The documentation about what the callbacks take and should
248 return is now somewhat clearer.
250 0.27 2009-01-18 23:40 UTC
251 + Fix : Offset of 1 with len magic on scalars.
252 + Fix : Segfaults that seem to happen when one croak in a callback at
253 compile time (thanks Ash Berlin for the testcase).
254 + Upd : Resources in META.yml (Florian Ragwitz)
255 + Tst : Large cleanup resulting in a wider set of tested behaviours.
257 0.26 2008-11-29 22:05 UTC
258 + Fix : Leaks of cloned coderefs that access lexicals.
259 + Fix : Building with maint-5.8 and blead (thanks Andreas Koenig).
261 0.25 2008-11-12 19:40 UTC
262 + Fix : Compatibility with 5.8.9.
263 + Fix : Old Pod::Coverage don't ignore CLONE.
264 + Upd : META.yml spec updated to 1.4.
266 0.24 2008-10-12 14:55 UTC
267 + Fix : Really fix it.
269 0.23 2008-10-11 17:25 UTC
270 + Fix : Building on perls with multiplicity but not ithreads.
272 0.22 2008-09-29 19:40 UTC
273 + Doc : Explain how magic differs from overloading.
274 + Fix : Disable thread safety for perl 5.8 on Windows. If you want it
275 enabled, reliable feedback is welcome.
277 0.21_02 2008-09-28 23:10 UTC
278 + Fix : Hide previously deleted magic tokens in vmg_svt_free (Yuval
281 0.21_01 2008-09-22 13:25 UTC
282 + Chg : Shrink the context data.
283 + Fix : Enable thread safety features for perls that just lack
285 + Tst : Output threads and threads::shared versions in threads tests.
287 0.20 2008-09-20 23:35 UTC
288 + Fix : The module is now thread safe. You can test if it had been
289 built with thread safety features enabled through the
290 VMG_THREADSAFE constant.
291 + Fix : Correct backtrace when wizard() croaks.
292 + Fix : 'cast $a, undef' should throw an 'invalid wizard' error instead
293 of an 'invalid signature'.
295 0.19 2008-07-05 15:05 UTC
296 + Add : The patchlevel with which the module has been built is now
297 exported with the VMG_PERL_PATCHLEVEL constant.
298 + Chg : Version checking macros overhaul.
299 + Fix : 'uvar' magic should only be enabled in maint-5.10 and higher
300 since it's never going to be backported to 5.8.
301 + Fix : Likewise, the copy callback API change should only be enabled in
303 + Fix : Instead of those ackward fake refcount increments in the wizard
304 destructor, flag the wizard with SVf_BREAK and force temps
305 cleanup. This should fix the wizard destruction for all perls,
306 and solves the leak introduced by the previous solution.
308 0.18 2008-07-04 12:20 UTC
309 + Fix : Enable uvar magic when patchlevel >= 28419.
310 + Fix : Set VMG_COMPAT_ARRAY_PUSH_NOLEN for Windows ActiveState Perl
313 0.17 2008-05-11 09:05 UTC
314 + Fix : 5.8.4 and lower don't have mPUSHi. Gah. Yeah, I know, I should
317 0.16 2008-05-10 22:05 UTC
318 + Add : The samples/copy.pl script.
319 + Chg : The sv_magical() + vmg_mg_magical() combo was simplified into
321 + Tst : t/33-code.t was testing scalars, not code.
322 + Tst : is() and like() are better than ok().
323 + Tst : 100% coverage reached.
325 0.15 2008-04-11 18:25 UTC
326 + Chg : Factor vmg_cb_call{1,2,3}() into one function.
327 + Fix : len magic is no longer called when taking the length() of a
328 magical scalar since p32969. The VMG_COMPAT_SCALAR_LENGTH_NOLEN
329 constant was added to cover this.
330 + Tst : More tests for t/22-len.t.
332 0.14 2008-03-24 12:35 UTC
333 + Fix : t/16-huf.t failures on Solaris and FreeBSD caused by not
334 updating mg->mg_ptr after Renew-ing it on dispell.
335 + Fix : Undefining MGf_DUP caused the wizard's magic to be wrongly
336 initialized. MGf_DUP now appears as true from the user point of
337 view, but the dup callback isn't actually set.
338 + Fix : Warnings with blead due to copy callbacks now taking an I32 (API
340 + Fix : vmg_svt_val() and vmg_uvar_del() aren't required if no uvar
342 + Tst : Tests now output the version of optional modules loaded.
344 0.13 2008-03-19 14:35 UTC
345 + Doc : Link to coverage report.
346 + Fix : Correct dependencies listing in META.yml.
347 + Tst : Improved test coverage.
348 + Tst : Print the patchlevel as a comment.
349 + Tst : Use is() where it's relevant.
350 + Tst : t/16-huf.t now really tests interaction with H::U::FH.
352 0.12 2008-02-07 18:15 UTC
353 + Fix : POD error. Thanks to Chris Williams (BinGOs) for the quick
356 0.11 2008-02-07 17:55 UTC
357 + Add : Copy callbacks now receive the current key/index in $_[2].
358 The current value/element is now in $_[3].
359 + Chg : The trigger of copy magic on list assignment with perls greater
360 than 5.10.0 was caused by uvar magic. Hence,
361 VMG_COMPAT_HASH_LISTASSIGN_COPY was removed. Use VMG_UVAR
363 + Fix : Build failures on patched perls.
364 + Tst : Added missing exported symbols to 01-import.t.
366 0.10 2008-02-04 11:30 UTC
367 + Add : New script : samples/vm_vs_tie.pl, that benchmarks our uvar
368 magic versus tied hashes.
369 + Add : The VMG_COMPAT_* constants can be used from userspace to check
370 perl magic abilities.
371 + Fix : Callbacks that returned undef made us croak, breaking the
372 variable behaviour (regression test in 14-callbacks.t).
373 + Fix : uvar callbacks weren't tested for non-NULL-ity before being
374 called (regression test in 28-uvar.t).
375 + Tst : Fix typo in 25-copy.t that prevented Tie::Hash tests to be ran.
377 0.09 2008-02-02 11:30 UTC
378 + Doc : Explicitely say that uvar callbacks are safely ignored for
380 + Doc : Document caveats and fix the usual set of typos.
381 + Fix : vmg_dispell() didn't check if the ext magic were ours when
382 counting wizards that have uvar callbacks, resulting in a
383 possible memory misread.
384 + Fix : getdata() now returns directly the data object, and no longer a
385 copy. This caused a leak.
386 + Tst : Prefix author tests by 9*-.
387 + Tst : New optional author test : 95-portability-files.t, that uses
388 Test::Portability::Files when it's present.
389 + Tst : New test : 14-self.t, that tests application of magic on the
391 + Tst : Move Hash::Util::FieldHash tests out of 11-multiple.t to
394 0.08 2008-02-01 16:55 UTC
395 + Add : copy magic for tied arrays/hashes.
397 + Add : uvar magics : fetch, store, exists, delete for hashes.
400 Internal development versions.
402 0.06 2007-11-20 10:10 UTC
403 + Chg : 5.7.3 is now officially required.
404 + Fix : "data" test failures on 5.8.{0,2}.
405 + Fix : Drand01() vs rand().
407 0.05 2007-11-19 09:10 UTC
408 + Fix : 5.10.0_RC1 compatibility fix.
410 0.04 2007-08-28 12:25 UTC
411 + Chg : s/require (XSLoader)/use $1/.
412 + Fix : Tests are now strict.
413 + Fix : Complete dependencies.
415 0.03 2007-08-01 17:20 UTC
416 + Add : Passing the signature of an already defined magic to wizard()
417 now returns the corresponding magic object.
418 + Add : You can pass the numeric signature as the wizard argument of
419 cast(), dispell() and getdata().
420 + Add : Any argument specified after the wizard (or the signature) in a
421 call to cast() is now passed to the private data constructor in
423 + Chg : $_[0] is now always a reference to the magic variable in all
424 callbacks. The reason for changing from the previous behaviour
425 is that one may want to apply the same magic to a plain scalar
426 and to a scalar reference, and needs a way to distinguish
427 between them in the callback (say, ref()).
428 + Fix : Wizard object destruction used not to free the signature.
430 0.02 2007-07-27 13:50 UTC
431 + Fix : In response to test report 548152 :
432 Newx() and SvMAGIC_set() not present on older perls.
433 + Fix : In response to test report 548275 :
434 Since perl 5.9.5, 'clear' magic is invoked when an array is
435 undefined (bug #43357). Moreover, 'len' magic is no longer
436 called by pushing an element since perl 5.9.3.
437 + Fix : Missing glob test in MANIFEST.
439 0.01 2007-07-25 16:15 UTC
440 First version, released on an unsuspecting world.