1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
|
Version 1.3.40 (in progress)
============================
2009-07-15: olly
[Perl] Don't specify Perl prototype "()" for a constructor with a
different name to the class, as such constructors can still take
parameters.
2009-07-05: olly
[PHP] Update the list of PHP keywords - "cfunction" is no longer a
keyword in PHP5 and PHP 5.3 added "goto", "namespace", "__DIR__",
and "__NAMESPACE__".
2009-07-03: olly
[Tcl] To complement USE_TCL_STUBS, add support for USE_TK_STUBS
and SWIG_TCL_STUBS_VERSION. Document all three in the Tcl chapter
of the manual. Based on patch from SF#2810380 by Christian
Gollwitzer.
2009-07-02: wsfulton
Fix -Wallkw option as reported by Solomon Gibbs.
2009-07-02: wsfulton
Fix syntax error when a nested struct contains a comment containing a * followed eventually by a /.
Regression from 1.3.37, reported by Solomon Gibbs.
2009-06-30: olly
[Ruby] Undefine close and connect macros defined by Ruby API
headers as we don't need them and they can clash with C++ methods
being wrapped. Patch from Vit Ondruch in SF#2814430.
2009-06-26: olly
[Ruby] Fix to handle FIXNUM values greater than MAXINT passed for a
double parameter.
2009-06-24: wsfulton
Fix wrapping methods with default arguments and the compactdefaultargs feature
where a class is passed by value and is assigned a default value. The SwigValueWrapper
template workaround for a missing default constructor is no longer used as the code
generated does not call the default constructor.
2009-06-16: wsfulton
[Java,C#] Fix enum marshalling when %ignore is used on one of the enum items.
Incorrect enum values were being passed to the C++ layer or compilation errors resulted.
2009-06-02: talby
[Perl] Resolved reference.i overload support problem
identified by John Potowsky.
2009-05-26: wsfulton
[C#] Improved std::map wrappers based on patch from Yuval Baror. The C# proxy
now implements System.Collections.Generic.IDictionary<>.
These std:map wrappers have a non-backwards compatible overhaul to make them
like a .NET IDictionary. Some method names have changed as following:
set -> setitem (use this[] property now)
get -> getitem (use this[] property now)
has_key -> ContainsKey
del -> Remove
clear -> Clear
The following macros used for std::map wrappers are deprecated and will no longer work:
specialize_std_map_on_key
specialize_std_map_on_value
specialize_std_map_on_both
*** POTENTIAL INCOMPATIBILITY ***
2009-05-14: bhy
[Python] Fix the wrong pointer value returned by SwigPyObject_repr().
2009-05-13: mutandiz (Mikel Bancroft)
[allegrocl] Minor tweak when wrapping in -nocwrap mode.
2009-05-11: wsfulton
[C#] Improved std::vector wrappers on the C# proxy side from Yuval Baror. These
implement IList<> instead of IEnumerable<> where possible.
2009-04-29: wsfulton
[Java, C#] Add the 'notderived' attribute to the javabase and csbase typemaps.
When this attribute is set, the typemap will not apply to classes that are derived
from a C++ base class, eg
%typemap(csbase, notderived="1") SWIGTYPE "CommonBase"
2009-04-29: olly
[Python] Don't attempt to acquire the GIL in situations where we
know that it will already be locked. This avoids some dead-locks
with mod_python (due to mod_python bugs which are apparently
unlikely to ever be fixed), and results in smaller wrappers which
run a little faster (in tests with Xapian on x86-64 Ubuntu 9.04,
the stripped wrapper library was 11% smaller and ran 2.7% faster).
2009-04-21: wsfulton
[C#] Fix #2753469 - bool &OUTPUT and bool *OUTPUT typemaps initialisation.
2009-04-09: wsfulton
Fix #2746858 - C macro expression using floating point numbers
2009-03-30: olly
[PHP] The default out typemap for char[ANY] now returns the string up to a
zero byte, or the end of the array if there is no zero byte. This
is the same as Python does, and seems more generally useful than
the previous behaviour of returning the whole contents of the array
including any zero bytes. If you want the old behaviour, you can provide
your own typemap to do this:
%typemap(out) char [ANY]
%{
RETVAL_STRINGL($1, $1_dim0, 1);
%}
|
