1 Revision history for Variable-Magic
3 0.24 2008-10-12 14:55 UTC
6 0.23 2008-10-11 17:25 UTC
7 + Fix : Building on perls with multiplicity but not ithreads.
9 0.22 2008-09-29 19:40 UTC
10 + Doc : Explain how magic differs from overloading.
11 + Fix : Disable thread safety for perl 5.8 on Windows. If you want it
12 enabled, reliable feedback is welcome.
14 0.21_02 2008-09-28 23:10 UTC
15 + Fix : Hide previously deleted magic tokens in vmg_svt_free (Yuval
18 0.21_01 2008-09-22 13:25 UTC
19 + Chg : Shrink the context data.
20 + Fix : Enable thread safety features for perls that just lack
22 + Tst : Output threads and threads::shared versions in threads tests.
24 0.20 2008-09-20 23:35 UTC
25 + Fix : The module is now thread safe. You can test if it had been
26 built with thread safety features enabled through the
27 VMG_THREADSAFE constant.
28 + Fix : Correct backtrace when wizard() croaks.
29 + Fix : 'cast $a, undef' should throw an 'invalid wizard' error instead
30 of an 'invalid signature'.
32 0.19 2008-07-05 15:05 UTC
33 + Add : The patchlevel with which the module has been built is now
34 exported with the VMG_PERL_PATCHLEVEL constant.
35 + Chg : Version checking macros overhaul.
36 + Fix : 'uvar' magic should only be enabled in maint-5.10 and higher
37 since it's never going to be backported to 5.8.
38 + Fix : Likewise, the copy callback API change should only be enabled in
40 + Fix : Instead of those ackward fake refcount increments in the wizard
41 destructor, flag the wizard with SVf_BREAK and force temps
42 cleanup. This should fix the wizard destruction for all perls,
43 and solves the leak introduced by the previous solution.
45 0.18 2008-07-04 12:20 UTC
46 + Fix : Enable uvar magic when patchlevel >= 28419.
47 + Fix : Set VMG_COMPAT_ARRAY_PUSH_NOLEN for Windows ActiveState Perl
50 0.17 2008-05-11 09:05 UTC
51 + Fix : 5.8.4 and lower don't have mPUSHi. Gah. Yeah, I know, I should
54 0.16 2008-05-10 22:05 UTC
55 + Add : The samples/copy.pl script.
56 + Chg : The sv_magical() + vmg_mg_magical() combo was simplified into
58 + Tst : t/33-code.t was testing scalars, not code.
59 + Tst : is() and like() are better than ok().
60 + Tst : 100% coverage reached.
62 0.15 2008-04-11 18:25 UTC
63 + Chg : Factor vmg_cb_call{1,2,3}() into one function.
64 + Fix : len magic is no longer called when taking the length() of a
65 magical scalar since p32969. The VMG_COMPAT_SCALAR_LENGTH_NOLEN
66 constant was added to cover this.
67 + Tst : More tests for t/22-len.t.
69 0.14 2008-03-24 12:35 UTC
70 + Fix : t/16-huf.t failures on Solaris and FreeBSD caused by not
71 updating mg->mg_ptr after Renew-ing it on dispell.
72 + Fix : Undefining MGf_DUP caused the wizard's magic to be wrongly
73 initialized. MGf_DUP now appears as true from the user point of
74 view, but the dup callback isn't actually set.
75 + Fix : Warnings with blead due to copy callbacks now taking an I32 (API
77 + Fix : vmg_svt_val() and vmg_uvar_del() aren't required if no uvar
79 + Tst : Tests now output the version of optional modules loaded.
81 0.13 2008-03-19 14:35 UTC
82 + Doc : Link to coverage report.
83 + Fix : Correct dependencies listing in META.yml.
84 + Tst : Improved test coverage.
85 + Tst : Print the patchlevel as a comment.
86 + Tst : Use is() where it's relevant.
87 + Tst : t/16-huf.t now really tests interaction with H::U::FH.
89 0.12 2008-02-07 18:15 UTC
90 + Fix : POD error. Thanks to Chris Williams (BinGOs) for the quick
93 0.11 2008-02-07 17:55 UTC
94 + Add : Copy callbacks now receive the current key/index in $_[2].
95 The current value/element is now in $_[3].
96 + Chg : The trigger of copy magic on list assignment with perls greater
97 than 5.10.0 was caused by uvar magic. Hence,
98 VMG_COMPAT_HASH_LISTASSIGN_COPY was removed. Use VMG_UVAR
100 + Fix : Build failures on patched perls.
101 + Tst : Added missing exported symbols to 01-import.t.
103 0.10 2008-02-04 11:30 UTC
104 + Add : New script : samples/vm_vs_tie.pl, that benchmarks our uvar
105 magic versus tied hashes.
106 + Add : The VMG_COMPAT_* constants can be used from userspace to check
107 perl magic abilities.
108 + Fix : Callbacks that returned undef made us croak, breaking the
109 variable behaviour (regression test in 14-callbacks.t).
110 + Fix : uvar callbacks weren't tested for non-NULL-ity before being
111 called (regression test in 28-uvar.t).
112 + Tst : Fix typo in 25-copy.t that prevented Tie::Hash tests to be ran.
114 0.09 2008-02-02 11:30 UTC
115 + Doc : Explicitely say that uvar callbacks are safely ignored for
117 + Doc : Document caveats and fix the usual set of typos.
118 + Fix : vmg_dispell() didn't check if the ext magic were ours when
119 counting wizards that have uvar callbacks, resulting in a
120 possible memory misread.
121 + Fix : getdata() now returns directly the data object, and no longer a
122 copy. This caused a leak.
123 + Tst : Prefix author tests by 9*-.
124 + Tst : New optional author test : 95-portability-files.t, that uses
125 Test::Portability::Files when it's present.
126 + Tst : New test : 14-self.t, that tests application of magic on the
128 + Tst : Move Hash::Util::FieldHash tests out of 11-multiple.t to
131 0.08 2008-02-01 16:55 UTC
132 + Add : copy magic for tied arrays/hashes.
134 + Add : uvar magics : fetch, store, exists, delete for hashes.
137 Internal development versions.
139 0.06 2007-11-20 10:10 UTC
140 + Chg : 5.7.3 is now officially required.
141 + Fix : "data" test failures on 5.8.{0,2}.
142 + Fix : Drand01() vs rand().
144 0.05 2007-11-19 09:10 UTC
145 + Fix : 5.10.0_RC1 compatibility fix.
147 0.04 2007-08-28 12:25 UTC
148 + Chg : s/require (XSLoader)/use $1/.
149 + Fix : Tests are now strict.
150 + Fix : Complete dependencies.
152 0.03 2007-08-01 17:20 UTC
153 + Add : Passing the signature of an already defined magic to wizard()
154 now returns the corresponding magic object.
155 + Add : You can pass the numeric signature as the wizard argument of
156 cast(), dispell() and getdata().
157 + Add : Any argument specified after the wizard (or the signature) in a
158 call to cast() is now passed to the private data constructor in
160 + Chg : $_[0] is now always a reference to the magic variable in all
161 callbacks. The reason for changing from the previous behaviour
162 is that one may want to apply the same magic to a plain scalar
163 and to a scalar reference, and needs a way to distinguish
164 between them in the callback (say, ref()).
165 + Fix : Wizard object destruction used not to free the signature.
167 0.02 2007-07-27 13:50 UTC
168 + Fix : In response to test report 548152 :
169 Newx() and SvMAGIC_set() not present on older perls.
170 + Fix : In response to test report 548275 :
171 Since perl 5.9.5, 'clear' magic is invoked when an array is
172 undefined (bug #43357). Moreover, 'len' magic is no longer
173 called by pushing an element since perl 5.9.3.
174 + Fix : Missing glob test in MANIFEST.
176 0.01 2007-07-25 16:15 UTC
177 First version, released on an unsuspecting world.