1 Revision history for Variable-Magic
3 0.41 2010-03-15 17:35 UTC
4 + Doc : Tweaks and fixups.
6 + Fix : Compatibility with the soon-to-be-released perl 5.12.0.
7 + Fix : Correctly propagate the errors thrown when variable destruction
8 happens at compile-time and not from inside eval STRING.
9 Thanks Florian Ragwitz and Ash Berlin for reporting.
11 0.40 2010-01-06 23:20 UTC
12 + Fix : Possible memory miswrites when passing data arguments to cast().
13 + Fix : Minor C portability tweaks.
15 0.39 2009-12-01 00:05 UTC
16 + Add : You can use a function name as a callback by passing a string
17 reference to wizard() instead of a code reference.
18 + Fix : Compatiblity with perl 5.11.1 and 5.11.2.
19 + Fix : Scalars stored into the data slot no longer leak.
20 + Fix : Thread destruction should not segfault anymore.
21 + Opt : As a result of removing the deprecated features, less memory is
22 used for both threaded (a pointer table is used instead of a
23 hash) and non-threaded (no global structure is needed anymore)
25 + Rem : As advertised in the precendent version, all the signature
26 related features were removed. This includes the 'sig' option
27 to wizard(), the getsig() and gensig() functions, and the SIG_*
30 0.38 2009-10-04 16:10 UTC
31 + Dep : All the signature-related features are DEPRECATED and will be
32 removed in december 2009. This includes the 'sig' option to
33 wizard(), the getsig() and gensig() functions, and the SIG_*
35 Afaik those features were never used in real life, and they make
36 the XS code slow and brittle. If you want your wizards to be
37 globally available, you'll have to handle that yourself by
38 storing them in a hash.
39 + Doc : More examples in a new COOKBOOK section.
40 + Fix : Compatiblity with the 5.11.0 release. The new compatibility
41 constant VMG_COMPAT_ARRAY_PUSH_NOLEN_VOID was added to cover
43 + Fix : Work around Kwalitee test misfailures.
45 0.37 2009-08-25 15:25 UTC
46 + Add : The new constant VMG_FORKSAFE can be tested to know whether the
47 module will behave nicely when fork()ing. It's currently always
48 true except on Windows where you need perl 5.10.1 for it to be
50 + Doc : Nits and clarifications.
52 0.36 2009-07-05 16:30 UTC
53 + Chg : getdata() now returns an empty list when no magic is present.
54 It used to return undef but did not croak as stated in the doc.
55 Thanks Matt S. Trout for pointing this out.
56 + Fix : Building and tests with ActiveState Perl 5.8 build >= 822.
58 0.35 2009-05-15 20:50 UTC
59 This is a maintenance release. Users of 0.34 can skip this update.
60 + Chg : uvar magic used to be enabled for 5.9.5. It now requires 5.10 or
62 + Fix : Silence compilation warnings for cxinc().
63 + Rem : The workaround introduced in 0.31 for the "Unknown errors" issue
64 was removed, as regression tests show that 0.34 provides a
66 + Tst : Skip t/40-threads.t on 5.8.x with old versions of threads and
69 0.34 2009-04-19 16:55 UTC
70 + Fix : Destruction of wizards in require.
71 + Fix : "panic: restartop" in 5.10 when dieing at compile time triggers
74 0.33 2009-03-26 00:00 UTC
75 + Chg : cast() and getsig() now croak too when an invalid signature is
77 + Fix : It was possible to generate with gensig() the same signature
78 twice before actually defining the wizards.
79 + Tst : More stash tests.
80 + Tst : Test when the magic takes place in some cases.
81 + Tst : Improved coverage for the op object feature.
83 0.32 2009-03-01 13:45 UTC
84 + Chg : dispell() and getdata() now croak when an invalid signature is
86 + Doc : More examples and a brand new synopsis.
87 + Fix : The signature is no longer stored in the mg_private member of
88 the MAGIC struct. This fixes possible clashes with magics from
90 + Fix : op info objects weren't blessed into the right class.
91 + Fix : Races when initializing the op names cache under threads.
93 0.31 2009-02-19 23:50 UTC
94 + Doc : A somewhat better introduction. The FUNCTION section was moved
95 before CONSTANTS and HISTORY since it's more important. And
96 'assignation' is really 'assignment', says Ricardo. :)
97 + Fix : The op_info features are now thread safe.
98 + Fix : Segfaults that occured when trying to get the op name during
100 + Fix : Segfaults and wrong "Unknown error" exceptions that happened
101 when dieing in require caused a free callback to fire.
102 Thanks Florian Ragwitz for reporting.
104 0.30 2009-02-12 17:05 UTC
105 + Add : You can now get in all the callbacks either the name or a B::OP
106 object representing the current op if you pass the new 'op_info'
107 option to wizard() with respective values VMG_OP_INFO_NAME and
109 + Doc : Document magic on array/hashes values.
110 + Fix : Completely skip free callbacks during global destruction.
111 + Tst : Tests now pass again on 5.8.3 and lower (they were broken since
113 + Tst : Test magic on symbol table and on array/hash values.
115 0.29 2009-02-08 11:10 UTC
116 + Add : Set $_[2] to the default length for len magic on scalars.
117 + Chg : Perl 5.8 is required.
118 + Chg : The CLONE method will no longer be defined for non-threaded
121 + Fix : The logic for turning off the GMAGICAL flag on hashes when no
122 uvar magic is involved could do nasty things on tied hashes.
123 + Upd : More resources in META.yml.
125 0.28 2009-01-24 17:05 UTC
126 + Add : The 'copy_key' option to wizard(), that enables the safe edition
127 of $_[2] in uvar callbacks to redirect actions to another key.
128 + Fix : Exception objects thrown from callbacks at compile time were not
130 + Fix : free callbacks triggered during global destruction could lead to
131 segfaults because the wizard was already freed.
132 + Fix : Refcount of scalars with free callback.
133 + Fix : Segfaults when using get or uvar magic simultaneously with clear
135 + Doc : The documentation about what the callbacks take and should
136 return is now somewhat clearer.
138 0.27 2009-01-18 23:40 UTC
139 + Fix : Offset of 1 with len magic on scalars.
140 + Fix : Segfaults that seem to happen when one croak in a callback at
141 compile time (thanks Ash Berlin for the testcase).
142 + Upd : Resources in META.yml (Florian Ragwitz)
143 + Tst : Large cleanup resulting in a wider set of tested behaviours.
145 0.26 2008-11-29 22:05 UTC
146 + Fix : Leaks of cloned coderefs that access lexicals.
147 + Fix : Building with maint-5.8 and blead (thanks Andreas Koenig).
149 0.25 2008-11-12 19:40 UTC
150 + Fix : Compatibility with 5.8.9.
151 + Fix : Old Pod::Coverage don't ignore CLONE.
152 + Upd : META.yml spec updated to 1.4.
154 0.24 2008-10-12 14:55 UTC
155 + Fix : Really fix it.
157 0.23 2008-10-11 17:25 UTC
158 + Fix : Building on perls with multiplicity but not ithreads.
160 0.22 2008-09-29 19:40 UTC
161 + Doc : Explain how magic differs from overloading.
162 + Fix : Disable thread safety for perl 5.8 on Windows. If you want it
163 enabled, reliable feedback is welcome.
165 0.21_02 2008-09-28 23:10 UTC
166 + Fix : Hide previously deleted magic tokens in vmg_svt_free (Yuval
169 0.21_01 2008-09-22 13:25 UTC
170 + Chg : Shrink the context data.
171 + Fix : Enable thread safety features for perls that just lack
173 + Tst : Output threads and threads::shared versions in threads tests.
175 0.20 2008-09-20 23:35 UTC
176 + Fix : The module is now thread safe. You can test if it had been
177 built with thread safety features enabled through the
178 VMG_THREADSAFE constant.
179 + Fix : Correct backtrace when wizard() croaks.
180 + Fix : 'cast $a, undef' should throw an 'invalid wizard' error instead
181 of an 'invalid signature'.
183 0.19 2008-07-05 15:05 UTC
184 + Add : The patchlevel with which the module has been built is now
185 exported with the VMG_PERL_PATCHLEVEL constant.
186 + Chg : Version checking macros overhaul.
187 + Fix : 'uvar' magic should only be enabled in maint-5.10 and higher
188 since it's never going to be backported to 5.8.
189 + Fix : Likewise, the copy callback API change should only be enabled in
191 + Fix : Instead of those ackward fake refcount increments in the wizard
192 destructor, flag the wizard with SVf_BREAK and force temps
193 cleanup. This should fix the wizard destruction for all perls,
194 and solves the leak introduced by the previous solution.
196 0.18 2008-07-04 12:20 UTC
197 + Fix : Enable uvar magic when patchlevel >= 28419.
198 + Fix : Set VMG_COMPAT_ARRAY_PUSH_NOLEN for Windows ActiveState Perl
201 0.17 2008-05-11 09:05 UTC
202 + Fix : 5.8.4 and lower don't have mPUSHi. Gah. Yeah, I know, I should
205 0.16 2008-05-10 22:05 UTC
206 + Add : The samples/copy.pl script.
207 + Chg : The sv_magical() + vmg_mg_magical() combo was simplified into
209 + Tst : t/33-code.t was testing scalars, not code.
210 + Tst : is() and like() are better than ok().
211 + Tst : 100% coverage reached.
213 0.15 2008-04-11 18:25 UTC
214 + Chg : Factor vmg_cb_call{1,2,3}() into one function.
215 + Fix : len magic is no longer called when taking the length() of a
216 magical scalar since p32969. The VMG_COMPAT_SCALAR_LENGTH_NOLEN
217 constant was added to cover this.
218 + Tst : More tests for t/22-len.t.
220 0.14 2008-03-24 12:35 UTC
221 + Fix : t/16-huf.t failures on Solaris and FreeBSD caused by not
222 updating mg->mg_ptr after Renew-ing it on dispell.
223 + Fix : Undefining MGf_DUP caused the wizard's magic to be wrongly
224 initialized. MGf_DUP now appears as true from the user point of
225 view, but the dup callback isn't actually set.
226 + Fix : Warnings with blead due to copy callbacks now taking an I32 (API
228 + Fix : vmg_svt_val() and vmg_uvar_del() aren't required if no uvar
230 + Tst : Tests now output the version of optional modules loaded.
232 0.13 2008-03-19 14:35 UTC
233 + Doc : Link to coverage report.
234 + Fix : Correct dependencies listing in META.yml.
235 + Tst : Improved test coverage.
236 + Tst : Print the patchlevel as a comment.
237 + Tst : Use is() where it's relevant.
238 + Tst : t/16-huf.t now really tests interaction with H::U::FH.
240 0.12 2008-02-07 18:15 UTC
241 + Fix : POD error. Thanks to Chris Williams (BinGOs) for the quick
244 0.11 2008-02-07 17:55 UTC
245 + Add : Copy callbacks now receive the current key/index in $_[2].
246 The current value/element is now in $_[3].
247 + Chg : The trigger of copy magic on list assignment with perls greater
248 than 5.10.0 was caused by uvar magic. Hence,
249 VMG_COMPAT_HASH_LISTASSIGN_COPY was removed. Use VMG_UVAR
251 + Fix : Build failures on patched perls.
252 + Tst : Added missing exported symbols to 01-import.t.
254 0.10 2008-02-04 11:30 UTC
255 + Add : New script : samples/vm_vs_tie.pl, that benchmarks our uvar
256 magic versus tied hashes.
257 + Add : The VMG_COMPAT_* constants can be used from userspace to check
258 perl magic abilities.
259 + Fix : Callbacks that returned undef made us croak, breaking the
260 variable behaviour (regression test in 14-callbacks.t).
261 + Fix : uvar callbacks weren't tested for non-NULL-ity before being
262 called (regression test in 28-uvar.t).
263 + Tst : Fix typo in 25-copy.t that prevented Tie::Hash tests to be ran.
265 0.09 2008-02-02 11:30 UTC
266 + Doc : Explicitely say that uvar callbacks are safely ignored for
268 + Doc : Document caveats and fix the usual set of typos.
269 + Fix : vmg_dispell() didn't check if the ext magic were ours when
270 counting wizards that have uvar callbacks, resulting in a
271 possible memory misread.
272 + Fix : getdata() now returns directly the data object, and no longer a
273 copy. This caused a leak.
274 + Tst : Prefix author tests by 9*-.
275 + Tst : New optional author test : 95-portability-files.t, that uses
276 Test::Portability::Files when it's present.
277 + Tst : New test : 14-self.t, that tests application of magic on the
279 + Tst : Move Hash::Util::FieldHash tests out of 11-multiple.t to
282 0.08 2008-02-01 16:55 UTC
283 + Add : copy magic for tied arrays/hashes.
285 + Add : uvar magics : fetch, store, exists, delete for hashes.
288 Internal development versions.
290 0.06 2007-11-20 10:10 UTC
291 + Chg : 5.7.3 is now officially required.
292 + Fix : "data" test failures on 5.8.{0,2}.
293 + Fix : Drand01() vs rand().
295 0.05 2007-11-19 09:10 UTC
296 + Fix : 5.10.0_RC1 compatibility fix.
298 0.04 2007-08-28 12:25 UTC
299 + Chg : s/require (XSLoader)/use $1/.
300 + Fix : Tests are now strict.
301 + Fix : Complete dependencies.
303 0.03 2007-08-01 17:20 UTC
304 + Add : Passing the signature of an already defined magic to wizard()
305 now returns the corresponding magic object.
306 + Add : You can pass the numeric signature as the wizard argument of
307 cast(), dispell() and getdata().
308 + Add : Any argument specified after the wizard (or the signature) in a
309 call to cast() is now passed to the private data constructor in
311 + Chg : $_[0] is now always a reference to the magic variable in all
312 callbacks. The reason for changing from the previous behaviour
313 is that one may want to apply the same magic to a plain scalar
314 and to a scalar reference, and needs a way to distinguish
315 between them in the callback (say, ref()).
316 + Fix : Wizard object destruction used not to free the signature.
318 0.02 2007-07-27 13:50 UTC
319 + Fix : In response to test report 548152 :
320 Newx() and SvMAGIC_set() not present on older perls.
321 + Fix : In response to test report 548275 :
322 Since perl 5.9.5, 'clear' magic is invoked when an array is
323 undefined (bug #43357). Moreover, 'len' magic is no longer
324 called by pushing an element since perl 5.9.3.
325 + Fix : Missing glob test in MANIFEST.
327 0.01 2007-07-25 16:15 UTC
328 First version, released on an unsuspecting world.