1 Revision history for Variable-Magic
3 0.43 2010-06-25 23:35 UTC
4 + Add : The new constant VMG_COMPAT_GLOB_GET tells you whether get magic
5 is called for globs. It's true starting perl 5.13.2.
6 + Chg : All callbacks are now called within an eval-like context.
7 Only free callbacks used to be called that way.
8 + Fix : Some exceptions thrown from a free callback could be lost.
9 + Fix : Croak messages could sometimes be repeated several times.
10 + Fix : t/41-clone.t segfaulting with perl 5.13.2.
12 0.42 2010-05-19 00:15 UTC
13 This is a maintenance release. The code contains no functional change.
14 Users of 0.41 can skip this update.
15 + Fix : Test failures with perl 5.13.
16 + Tst : Improve coverage.
18 0.41 2010-03-15 17:35 UTC
19 + Doc : Tweaks and fixups.
21 + Fix : Compatibility with the soon-to-be-released perl 5.12.0.
22 + Fix : Correctly propagate the errors thrown when variable destruction
23 happens at compile-time and not from inside eval STRING.
24 Thanks Florian Ragwitz and Ash Berlin for reporting.
26 0.40 2010-01-06 23:20 UTC
27 + Fix : Possible memory miswrites when passing data arguments to cast().
28 + Fix : Minor C portability tweaks.
30 0.39 2009-12-01 00:05 UTC
31 + Add : You can use a function name as a callback by passing a string
32 reference to wizard() instead of a code reference.
33 + Fix : Compatiblity with perl 5.11.1 and 5.11.2.
34 + Fix : Scalars stored into the data slot no longer leak.
35 + Fix : Thread destruction should not segfault anymore.
36 + Opt : As a result of removing the deprecated features, less memory is
37 used for both threaded (a pointer table is used instead of a
38 hash) and non-threaded (no global structure is needed anymore)
40 + Rem : As advertised in the precendent version, all the signature
41 related features were removed. This includes the 'sig' option
42 to wizard(), the getsig() and gensig() functions, and the SIG_*
45 0.38 2009-10-04 16:10 UTC
46 + Dep : All the signature-related features are DEPRECATED and will be
47 removed in december 2009. This includes the 'sig' option to
48 wizard(), the getsig() and gensig() functions, and the SIG_*
50 Afaik those features were never used in real life, and they make
51 the XS code slow and brittle. If you want your wizards to be
52 globally available, you'll have to handle that yourself by
53 storing them in a hash.
54 + Doc : More examples in a new COOKBOOK section.
55 + Fix : Compatiblity with the 5.11.0 release. The new compatibility
56 constant VMG_COMPAT_ARRAY_PUSH_NOLEN_VOID was added to cover
58 + Fix : Work around Kwalitee test misfailures.
60 0.37 2009-08-25 15:25 UTC
61 + Add : The new constant VMG_FORKSAFE can be tested to know whether the
62 module will behave nicely when fork()ing. It's currently always
63 true except on Windows where you need perl 5.10.1 for it to be
65 + Doc : Nits and clarifications.
67 0.36 2009-07-05 16:30 UTC
68 + Chg : getdata() now returns an empty list when no magic is present.
69 It used to return undef but did not croak as stated in the doc.
70 Thanks Matt S. Trout for pointing this out.
71 + Fix : Building and tests with ActiveState Perl 5.8 build >= 822.
73 0.35 2009-05-15 20:50 UTC
74 This is a maintenance release. Users of 0.34 can skip this update.
75 + Chg : uvar magic used to be enabled for 5.9.5. It now requires 5.10 or
77 + Fix : Silence compilation warnings for cxinc().
78 + Rem : The workaround introduced in 0.31 for the "Unknown errors" issue
79 was removed, as regression tests show that 0.34 provides a
81 + Tst : Skip t/40-threads.t on 5.8.x with old versions of threads and
84 0.34 2009-04-19 16:55 UTC
85 + Fix : Destruction of wizards in require.
86 + Fix : "panic: restartop" in 5.10 when dieing at compile time triggers
89 0.33 2009-03-26 00:00 UTC
90 + Chg : cast() and getsig() now croak too when an invalid signature is
92 + Fix : It was possible to generate with gensig() the same signature
93 twice before actually defining the wizards.
94 + Tst : More stash tests.
95 + Tst : Test when the magic takes place in some cases.
96 + Tst : Improved coverage for the op object feature.
98 0.32 2009-03-01 13:45 UTC
99 + Chg : dispell() and getdata() now croak when an invalid signature is
101 + Doc : More examples and a brand new synopsis.
102 + Fix : The signature is no longer stored in the mg_private member of
103 the MAGIC struct. This fixes possible clashes with magics from
105 + Fix : op info objects weren't blessed into the right class.
106 + Fix : Races when initializing the op names cache under threads.
108 0.31 2009-02-19 23:50 UTC
109 + Doc : A somewhat better introduction. The FUNCTION section was moved
110 before CONSTANTS and HISTORY since it's more important. And
111 'assignation' is really 'assignment', says Ricardo. :)
112 + Fix : The op_info features are now thread safe.
113 + Fix : Segfaults that occured when trying to get the op name during
115 + Fix : Segfaults and wrong "Unknown error" exceptions that happened
116 when dieing in require caused a free callback to fire.
117 Thanks Florian Ragwitz for reporting.
119 0.30 2009-02-12 17:05 UTC
120 + Add : You can now get in all the callbacks either the name or a B::OP
121 object representing the current op if you pass the new 'op_info'
122 option to wizard() with respective values VMG_OP_INFO_NAME and
124 + Doc : Document magic on array/hashes values.
125 + Fix : Completely skip free callbacks during global destruction.
126 + Tst : Tests now pass again on 5.8.3 and lower (they were broken since
128 + Tst : Test magic on symbol table and on array/hash values.
130 0.29 2009-02-08 11:10 UTC
131 + Add : Set $_[2] to the default length for len magic on scalars.
132 + Chg : Perl 5.8 is required.
133 + Chg : The CLONE method will no longer be defined for non-threaded
136 + Fix : The logic for turning off the GMAGICAL flag on hashes when no
137 uvar magic is involved could do nasty things on tied hashes.
138 + Upd : More resources in META.yml.
140 0.28 2009-01-24 17:05 UTC
141 + Add : The 'copy_key' option to wizard(), that enables the safe edition
142 of $_[2] in uvar callbacks to redirect actions to another key.
143 + Fix : Exception objects thrown from callbacks at compile time were not
145 + Fix : free callbacks triggered during global destruction could lead to
146 segfaults because the wizard was already freed.
147 + Fix : Refcount of scalars with free callback.
148 + Fix : Segfaults when using get or uvar magic simultaneously with clear
150 + Doc : The documentation about what the callbacks take and should
151 return is now somewhat clearer.
153 0.27 2009-01-18 23:40 UTC
154 + Fix : Offset of 1 with len magic on scalars.
155 + Fix : Segfaults that seem to happen when one croak in a callback at
156 compile time (thanks Ash Berlin for the testcase).
157 + Upd : Resources in META.yml (Florian Ragwitz)
158 + Tst : Large cleanup resulting in a wider set of tested behaviours.
160 0.26 2008-11-29 22:05 UTC
161 + Fix : Leaks of cloned coderefs that access lexicals.
162 + Fix : Building with maint-5.8 and blead (thanks Andreas Koenig).
164 0.25 2008-11-12 19:40 UTC
165 + Fix : Compatibility with 5.8.9.
166 + Fix : Old Pod::Coverage don't ignore CLONE.
167 + Upd : META.yml spec updated to 1.4.
169 0.24 2008-10-12 14:55 UTC
170 + Fix : Really fix it.
172 0.23 2008-10-11 17:25 UTC
173 + Fix : Building on perls with multiplicity but not ithreads.
175 0.22 2008-09-29 19:40 UTC
176 + Doc : Explain how magic differs from overloading.
177 + Fix : Disable thread safety for perl 5.8 on Windows. If you want it
178 enabled, reliable feedback is welcome.
180 0.21_02 2008-09-28 23:10 UTC
181 + Fix : Hide previously deleted magic tokens in vmg_svt_free (Yuval
184 0.21_01 2008-09-22 13:25 UTC
185 + Chg : Shrink the context data.
186 + Fix : Enable thread safety features for perls that just lack
188 + Tst : Output threads and threads::shared versions in threads tests.
190 0.20 2008-09-20 23:35 UTC
191 + Fix : The module is now thread safe. You can test if it had been
192 built with thread safety features enabled through the
193 VMG_THREADSAFE constant.
194 + Fix : Correct backtrace when wizard() croaks.
195 + Fix : 'cast $a, undef' should throw an 'invalid wizard' error instead
196 of an 'invalid signature'.
198 0.19 2008-07-05 15:05 UTC
199 + Add : The patchlevel with which the module has been built is now
200 exported with the VMG_PERL_PATCHLEVEL constant.
201 + Chg : Version checking macros overhaul.
202 + Fix : 'uvar' magic should only be enabled in maint-5.10 and higher
203 since it's never going to be backported to 5.8.
204 + Fix : Likewise, the copy callback API change should only be enabled in
206 + Fix : Instead of those ackward fake refcount increments in the wizard
207 destructor, flag the wizard with SVf_BREAK and force temps
208 cleanup. This should fix the wizard destruction for all perls,
209 and solves the leak introduced by the previous solution.
211 0.18 2008-07-04 12:20 UTC
212 + Fix : Enable uvar magic when patchlevel >= 28419.
213 + Fix : Set VMG_COMPAT_ARRAY_PUSH_NOLEN for Windows ActiveState Perl
216 0.17 2008-05-11 09:05 UTC
217 + Fix : 5.8.4 and lower don't have mPUSHi. Gah. Yeah, I know, I should
220 0.16 2008-05-10 22:05 UTC
221 + Add : The samples/copy.pl script.
222 + Chg : The sv_magical() + vmg_mg_magical() combo was simplified into
224 + Tst : t/33-code.t was testing scalars, not code.
225 + Tst : is() and like() are better than ok().
226 + Tst : 100% coverage reached.
228 0.15 2008-04-11 18:25 UTC
229 + Chg : Factor vmg_cb_call{1,2,3}() into one function.
230 + Fix : len magic is no longer called when taking the length() of a
231 magical scalar since p32969. The VMG_COMPAT_SCALAR_LENGTH_NOLEN
232 constant was added to cover this.
233 + Tst : More tests for t/22-len.t.
235 0.14 2008-03-24 12:35 UTC
236 + Fix : t/16-huf.t failures on Solaris and FreeBSD caused by not
237 updating mg->mg_ptr after Renew-ing it on dispell.
238 + Fix : Undefining MGf_DUP caused the wizard's magic to be wrongly
239 initialized. MGf_DUP now appears as true from the user point of
240 view, but the dup callback isn't actually set.
241 + Fix : Warnings with blead due to copy callbacks now taking an I32 (API
243 + Fix : vmg_svt_val() and vmg_uvar_del() aren't required if no uvar
245 + Tst : Tests now output the version of optional modules loaded.
247 0.13 2008-03-19 14:35 UTC
248 + Doc : Link to coverage report.
249 + Fix : Correct dependencies listing in META.yml.
250 + Tst : Improved test coverage.
251 + Tst : Print the patchlevel as a comment.
252 + Tst : Use is() where it's relevant.
253 + Tst : t/16-huf.t now really tests interaction with H::U::FH.
255 0.12 2008-02-07 18:15 UTC
256 + Fix : POD error. Thanks to Chris Williams (BinGOs) for the quick
259 0.11 2008-02-07 17:55 UTC
260 + Add : Copy callbacks now receive the current key/index in $_[2].
261 The current value/element is now in $_[3].
262 + Chg : The trigger of copy magic on list assignment with perls greater
263 than 5.10.0 was caused by uvar magic. Hence,
264 VMG_COMPAT_HASH_LISTASSIGN_COPY was removed. Use VMG_UVAR
266 + Fix : Build failures on patched perls.
267 + Tst : Added missing exported symbols to 01-import.t.
269 0.10 2008-02-04 11:30 UTC
270 + Add : New script : samples/vm_vs_tie.pl, that benchmarks our uvar
271 magic versus tied hashes.
272 + Add : The VMG_COMPAT_* constants can be used from userspace to check
273 perl magic abilities.
274 + Fix : Callbacks that returned undef made us croak, breaking the
275 variable behaviour (regression test in 14-callbacks.t).
276 + Fix : uvar callbacks weren't tested for non-NULL-ity before being
277 called (regression test in 28-uvar.t).
278 + Tst : Fix typo in 25-copy.t that prevented Tie::Hash tests to be ran.
280 0.09 2008-02-02 11:30 UTC
281 + Doc : Explicitely say that uvar callbacks are safely ignored for
283 + Doc : Document caveats and fix the usual set of typos.
284 + Fix : vmg_dispell() didn't check if the ext magic were ours when
285 counting wizards that have uvar callbacks, resulting in a
286 possible memory misread.
287 + Fix : getdata() now returns directly the data object, and no longer a
288 copy. This caused a leak.
289 + Tst : Prefix author tests by 9*-.
290 + Tst : New optional author test : 95-portability-files.t, that uses
291 Test::Portability::Files when it's present.
292 + Tst : New test : 14-self.t, that tests application of magic on the
294 + Tst : Move Hash::Util::FieldHash tests out of 11-multiple.t to
297 0.08 2008-02-01 16:55 UTC
298 + Add : copy magic for tied arrays/hashes.
300 + Add : uvar magics : fetch, store, exists, delete for hashes.
303 Internal development versions.
305 0.06 2007-11-20 10:10 UTC
306 + Chg : 5.7.3 is now officially required.
307 + Fix : "data" test failures on 5.8.{0,2}.
308 + Fix : Drand01() vs rand().
310 0.05 2007-11-19 09:10 UTC
311 + Fix : 5.10.0_RC1 compatibility fix.
313 0.04 2007-08-28 12:25 UTC
314 + Chg : s/require (XSLoader)/use $1/.
315 + Fix : Tests are now strict.
316 + Fix : Complete dependencies.
318 0.03 2007-08-01 17:20 UTC
319 + Add : Passing the signature of an already defined magic to wizard()
320 now returns the corresponding magic object.
321 + Add : You can pass the numeric signature as the wizard argument of
322 cast(), dispell() and getdata().
323 + Add : Any argument specified after the wizard (or the signature) in a
324 call to cast() is now passed to the private data constructor in
326 + Chg : $_[0] is now always a reference to the magic variable in all
327 callbacks. The reason for changing from the previous behaviour
328 is that one may want to apply the same magic to a plain scalar
329 and to a scalar reference, and needs a way to distinguish
330 between them in the callback (say, ref()).
331 + Fix : Wizard object destruction used not to free the signature.
333 0.02 2007-07-27 13:50 UTC
334 + Fix : In response to test report 548152 :
335 Newx() and SvMAGIC_set() not present on older perls.
336 + Fix : In response to test report 548275 :
337 Since perl 5.9.5, 'clear' magic is invoked when an array is
338 undefined (bug #43357). Moreover, 'len' magic is no longer
339 called by pushing an element since perl 5.9.3.
340 + Fix : Missing glob test in MANIFEST.
342 0.01 2007-07-25 16:15 UTC
343 First version, released on an unsuspecting world.