Add Swig_obligatory_macros which must be called by each
target language to define SWIG_VERSION correctly
in the generated code, as well as the language specific
macro SWIGXXX where XXX is the target language name.
Drop the #ifdef SWIGXXX that was previously generated -
I can't see the point of this and if users are defining
this macro somehow, then users will need to change this
Closes#1050
Until now SWIG quietly ignored such errors unless -Wextra (or -Wall
which implies -Wextra) was passed, but this is unhelpful as it hides
problems. To illustrate this point, enabling this warning by
default revealled a typo in the preproc_defined.i testcase in
SWIG's own testsuite.
If you really don't want to see this warning, you can suppress it
with command line option -w202 or by using this in your interface
file:
%warnfilter(SWIGWARN_PP_EVALUATION);
Both will work with older versions of SWIG too.
Fixes#1465Fixes#2389
Unfortunately the changes of 26bf86322 (Use SWIG-specific for
non-overloaded synthesized functions too, 2021-11-09) did break some
existing code bases using SWIG as they hardcoded the old wrapper
function names.
So turn this off by default and add a global variable allowing to enable
this, which can be done for a specific language only. This is ugly but,
unfortunately, there is no way to use the Language object from the C
function Swig_MethodToFunction(), so the only alternative would be to
add another parameter to it, but it already has 6 of them, so it
wouldn't really be that much better.
See #2366, #2368, #2370.
This option was undocumented. If used it quietly did nothing
unless you'd installed the SWILL library before building SWIG, but
SWILL is no longer maintained and hasn't seen a release since
2008-04-10.
It's clear nobody has used this functionality for some time as the
code to support it in SWIG isn't actually valid ISO C++:
Modules/browser.cxx: In member function ‘void Browser::show_attributes(Node*)’:
Modules/browser.cxx:57:23: warning: ISO C++ forbids converting a string constant to ‘char*’ [-Wwrite-strings]
57 | char *trunc = "";
| ^~
Modules/browser.cxx:61:21: warning: ISO C++ forbids converting a string constant to ‘char*’ [-Wwrite-strings]
61 | trunc = "...";
| ^~~~~
Closes#2251
The debug command line options that display parse tree nodes
(-debug-module, -debug-top, -debug-symtabs) now display previously hidden
linked list pointers which are useful for debugging parse trees.
Added new command line option -debug-quiet. This suppresses the display
of most linked list pointers and symbol table pointers in the parse tree nodes.
The keys in the parse tree node are now shown in alphabetical order.
Use `#pragma GCC poison` (supported since GCC 3, maybe earlier) when
compiling with GCC to help prevent direct uses being introduced for
functions which DOH provides a wrapper for.
Exit() is a wrapper for exit() by default, but SetExitHandler() allows
specifying a function to call instead.
This means that failures within DOH (e.g. Malloc() failing due to lack
of memory) will now perform cleanup such as removing output files.
This commit also cleans up exit statuses so SWIG should now reliably
exit with status 0 if the run was successful and status 1 if there was
an error (or a warning and -Werror was in effect).
Previously in some situations SWIG would try to exit with the status set
to the number of errors encountered, but that's problematic - for
example if there were 256 errors this would result in exit status 0 on
most platforms. Also some error statuses have special meanings e.g.
those defined by <sysexits.h>.
Also SWIG/Javascript tried to exit with status -1 in a few places (which
typically results in exit status 255).
We were passing flags of DOH_REPLACE_ANY|DOH_REPLACE_FIRST in three
places, which doesn't make sense as those are mutually exclusive
concepts. In the current implementation DOH_REPLACE_FIRST wins in
this situation, so replace with that and clarify the docs.
Previously generated files were not removed,
potentially breaking Makefiles using file dependencies, especially when
-Werror (warnings as errors) was used.
Output is once again:
$ swig -versioon
Unable to find option or file '-versioon', Use 'swig -help' for more information.
instead of
$ swig -versioon
No target language specified
Three status: Disabled, Experimental and Supported.
Any target language classified as 'Experimental' will issue new warning
524 SWIGWARN_LANG_EXPERIMENTAL.
Any target language classified as 'Disabled' will error out.
Languages will be classified in forthcoming commits.
Issue #1437
SWIG now requires a target language to be specified instead of
defaulting to wrapping for Tcl. Specifying swig --help without
a target language now just shows the generic help. The -nolang
option has been removed.
PHP5 is no longer actively supported by the PHP developers and security
support for it ends completely at the end of 2018, so it doesn't make
sense to include support for it in the upcoming SWIG 4.0.0 release.
See #701.
Fixes#1059
Methods with rvalue ref-qualifiers are ignored by default as it is not
possible to have an rvalue temporary from the target language (which is
needed to call the rvalue ref-qualified method).
A warning 405 is shown mentioning the ignored rvalue ref-qualifier method
which can be seen with the -Wextra option.
cpp_refqualifier.i:15: Warning 405: Method with rvalue ref-qualifier ignored h() const &&.
Usually rvalue and lvalue ref-qualifier overloaded methods are written - the
lvalue method will then be wrapped.
This is SWIG's default for every language except PHP, and now is a
good time to make this change so that once we drop PHP5 support, we'll
have the same default everywhere. It's easy to override the default
with -cppext cxx to get the PHP5 behaviour.
Provide -cppext as a general command line option for setting the
extension used for generated C++ files (previously it was specific
to the PHP backend). Deprecate the equivalent -suffix option
provided by the Ocaml backend, but continue to support that for
now.
I believe the aim is to keep these below 2KB to avoid string literal
length limits on some compiler, but they slowly creep up in size, so
split into more chunks and rebalance the contents.
This is just a simple code refactor, moving and function renaming to
remove the %extend code out of the parser into its own file now
that it isn't just used in the parser.
Error checking for this combination implemented as well as correcting
Octave
Also refactor, replacing CPlusPlusOut variable with cparse_cplusplusout
for an implementation which more closely resembles cparse_cplusplus which
is also required in both .c and .cxx files.
Closes#89
Squash merge branch 'master' of https://github.com/wkalinin/swig into wkalinin-nested
By Vladimir Kalinin
* 'master' of https://github.com/wkalinin/swig:
CPlusPlusOut mode for Octave
nested class illustration
fixed "Abstract" flag for nested classes added an example enabled anonymous nested structs runtime test
porting
warnings disabled
porting fixes
java runtime tests ported
nested class closing bracket offset fixed
removed double nested template (not supported by %template parsing)
template_nested test extended
parent field made public
property access fixed
replaced tabs with spaces
warning W-reorder
deprecated warnings removed, derived_nested runtime test added
optimized string indenting
Nested classes indenting
nested classes docs
fixed the order in which flattened inner classes are added after the outer
Private nested classes were getting into the type table.
Java getProxyName() fix for nested classes fixes the case when nested classes is forward declared
Fix for a case when a nested class inherits from the same base as the outer. (Base class constructor declaration is found first in this case)
merge fix
nested C struct first immediate declaration incorrectly renamed sample fixed
tests updated to reflect nested classes support
Java nested classes support (1)
flattening should remove the link to the outer class
access mode correctly set/restored for nested classes
nested templates should be skipped while flattening (template nodes themselves, not expanded versions) also non-public nested classes should be ignored
If nested classes are not supported, default behaviour is flattening, not ignoring flag "nested" is preserved, so, the nested classes can be ignored by user
nested workaround test updated
template instantiated within a class is marked as nested for ignoring purposes
%ignore not applied to the nested classed, because "nested" flag is set too late
typedef name takes precedence over the real name (reason?)
unnamed structs should be processed for all the languages
nested C struct instances are wrapped as "immutable"
tree building
typedef declaration for unnamed C structures fixed
nested classes "flattening"
fixed %ignoring nested classes
renamed "nested" attribute to "nested:outer" added "nested" flag, to be used with $ignore (it is not removed while flattening) added nestedClassesSupported() function to the Language interface
renamed "nested" attribute to "nested:outer" added "nested" flag, to be used with $ignore (it is not removed while flattening) added nestedClassesSupported() function to the Language interface
tree iteration fix
dirclassname variable names unified memory issue fixed
merge error
ignore unnamed structs for C++
unnamed nested C structs naming & unnesting
class added to classes hash under typedef name
private nested classes skipped
test updated due to nested templates support
anonymous structs with inheritance fixed nested_class test to allow anonymous structs w/o declarator
tests updated: nested workaround removed from namespace_class.i propagated nested template declaration to the C++ file
injected members scope
nested tempplates fixes, nested structures in "C" mode parsing added utility function "appendSibling" (like "appendChild")
nested unnamed structures parsing fixes, access mode restored on nested class end, tdname is properly patched with outer class name prefix
memory management fixes
nested templates (1)
Nested unnamed structs
Nested class support (1)
Nested class support (1)
