summaryrefslogtreecommitdiffstats
path: root/Tools/bgen
Commit message (Collapse)AuthorAgeFilesLines
* Oops, old-style types don't have a tp_free slot. Call PyObject_FreeJack Jansen2002-12-231-2/+6
| | | | directly in that case.
* Always use self->ob_type->tp_free when freeing an object.Jack Jansen2002-12-191-1/+1
|
* Always output an inheritance-aware version of the xxxx_Check() macro. AlsoJack Jansen2002-12-191-5/+1
| | | | fixed that macro to actually work:-)
* Added an optional longname argument to Module, which gives the full,Jack Jansen2002-12-171-2/+7
| | | | | | externally visible name of the module. This is so that type names can be shown as "Carbon.File.FSSpec" even though the real name of the module is "_File".
* Fixed typo.Jack Jansen2002-12-051-1/+1
|
* Added PEP253 support.Jack Jansen2002-12-031-11/+114
|
* Fixed two silly bugs in the PEP252 support code, added an assertJack Jansen2002-11-281-3/+10
| | | | that basechain isn't set, and made the output a bit prettier.
* Added a class PEP252Mixin. By adding this to your ObjectDefinition youJack Jansen2002-11-251-2/+127
| | | | | | | | | | | get PEP-252 style objects in stead of old-fashioned objects. In stead of defining a GetattrHook you declare a class variable getsetlist, which contains tuples (name, getcode, setcode, docstring). Only lightly tested: the code still works if you don't inherit PEP252Mixin and the code works if you inherit it but don't define any getters or setters. Also, this will not work together with the "poor mans inheritance" offered by method chains, so the CF module will remain with old-style objects until PEP253 is supported too.
* Apply diff2.txt from SF patch http://www.python.org/sf/572113Walter Dörwald2002-09-113-23/+19
| | | | | | | | (with one small bugfix in bgen/bgen/scantools.py) This replaces string module functions with string methods for the stuff in the Tools directory. Several uses of string.letters etc. are still remaining.
* Moved CoreFoundation type support to bgen/macsupport.Jack Jansen2002-08-221-0/+11
|
* iUse PyDoc_STR() around docstrings.Jack Jansen2002-08-161-1/+1
|
* Be a lot less verbose by default.Jack Jansen2002-08-052-6/+7
|
* Fixed to run better in unix-Python, and to cater for bgenlocationsJack Jansen2002-08-051-3/+7
| | | | possibly being missing.
* Moved bgenlocations to the Mac/Lib directory. Not perfect, but better thanJack Jansen2002-08-051-13/+0
| | | | | | | where it was: it is really a configuration file, not a normal module. By moving it into Mac/Lib we can now also store the location of bgen itself in there, which is needed because bgen isn't installed.
* Specify pathnames in a way that works on both OS9 and OSX.Jack Jansen2002-08-041-3/+3
| | | | You'll still have to manually edit it, though...
* Use universal newline input when scanning header files.Jack Jansen2002-08-041-3/+3
|
* staticforward bites the dust.Jeremy Hylton2002-07-171-2/+2
| | | | | | | | | | | | | | | The staticforward define was needed to support certain broken C compilers (notably SCO ODT 3.0, perhaps early AIX as well) botched the static keyword when it was used with a forward declaration of a static initialized structure. Standard C allows the forward declaration with static, and we've decided to stop catering to broken C compilers. (In fact, we expect that the compilers are all fixed eight years later.) I'm leaving staticforward and statichere defined in object.h as static. This is only for backwards compatibility with C extensions that might still use it. XXX I haven't updated the documentation.
* escape 8-bit chars when generating .py files. fixes bug #566302Just van Rossum2002-06-091-1/+16
|
* Converted to use re in stead of regex and regsub (finally:-).Jack Jansen2002-04-231-54/+65
|
* Oops: we used PyMem_DEL() to clean up objects, and that's a problem sinceJack Jansen2002-04-191-1/+1
| | | | | | | | pymalloc, apparently. Fixed, but this means all bgen-generated modules will have to be re-generated. I hope (and expect) that the pymalloc fixes aren't bugfix candidates, because if they are this is one too.
* - Added support for inherent pointer types (typedefs of arrays)Jack Jansen2002-04-121-0/+34
| | | | - Added a debug class variable to enable parser debugging.
* Give type name when complaining about using input-only type for outputJack Jansen2002-04-121-2/+2
| | | | or v.v.
* Some structures should be passed to Py_BuildValue by reference, not by value,Jack Jansen2002-03-182-2/+10
| | | | | | notably FSSpec and FSRef objects. First half of fix for #531291. 2.2.1 candidate.
* Added some support for Uicode arguments. Still leaves something to be ↵Jack Jansen2002-01-013-1/+23
| | | | desired, especially because of poor unicode support in PyArg_ParseTuple.
* Updated to current pathnames.Jack Jansen2002-01-011-3/+3
|
* Small change to allow for generation of QuickTime module for Windows.Jack Jansen2001-12-181-0/+4
|
* OSX tweak: recognize both EXTERN_API_C and EXTERN_API declarations.Just van Rossum2001-12-121-2/+2
| | | | Jack: I hope I didn't break anything for you!
* The module generator now tells its object generators about the module name ↵Jack Jansen2001-12-092-1/+9
| | | | (through the new setmodulename() method). If the module name has been set the object generators output it as part of the tp_name field for the object type.
* Merged changes made on r22b2-branch between r22b2 and r22b2-mac (theJack Jansen2001-11-301-0/+2
| | | | changes from start of branch upto r22b2 were already merged, of course).
* Don't initialize tp_type statically, it won't work on Windows. SpottedJack Jansen2001-11-141-1/+1
| | | | by Thomas Heller (patch 459442).
* Don't override asplit_pat in Scanner_UH3. No reason to do so, and it brokeJack Jansen2001-10-301-1/+0
| | | | generating Waste.
* Only output the buffer size error label if it is used.Jack Jansen2001-09-052-3/+7
| | | | Shuts up another couple of gcc warnings.
* Don't use a default "int" return type, gcc gives a warning about it.Jack Jansen2001-09-041-1/+1
|
* Fixes by Thomas Heller:Jack Jansen2001-08-275-12/+13
| | | | | | - make the selftests work again (they were apparently not used since very early in bgen's development), with some minor cleanup by me - make emacs python mode happier
* Oops, left in a debug statement. It's gone now.Jack Jansen2001-07-171-1/+0
|
* Allow [] after a parameter name. We currently take this to be the same as * ↵Jack Jansen2001-07-131-13/+7
| | | | in front, which isn't 100% correct but good enough.
* Added WeakLink...Generator classes (should have done that ages ago). These ↵Jack Jansen2001-07-012-0/+22
| | | | | | check the c-function pointer for being NULL before calling it and raise UnimplementedError if it is. This allows system libs to be weak-linked, thereby allowing us to generate functions that are only available on some OS versions without getting a NULL dereference if the function isn't available.
* Make basechain a class variable in stead of initializing it in __init__. ↵Jack Jansen2001-06-281-1/+1
| | | | That way it's more easily overridden.
* Added a couple more types that Apple has taken a fancy to.Jack Jansen2001-06-271-1/+3
|
* Added a Parser_OSX class that can parse new-style (C-only) headers,Jack Jansen2001-06-261-4/+19
| | | | such as the Core Foundation ones.
* Generate prototype-style function headers in stead of K&R style. Makes life ↵Jack Jansen2001-05-193-33/+12
| | | | easier with gcc -Wstrict-function-prototypes.
* Dialog and Window objects are (finally) different beasts.Jack Jansen2001-02-271-2/+1
|
* Use the filename, not the pathname, in the definitions fileJack Jansen2001-02-271-1/+1
| | | | | comment. This way the generated files are identical when generated on different machines.
* Optionally weed out duplicate prototypes for the same function (whichJack Jansen2000-12-123-3/+11
| | | | happens because the scanner ignores preprocessor #ifs).
* Adapted to new standard for initmodule() routine: don't callJack Jansen2000-12-121-2/+1
| | | | Py_FatalError on errors.
* Various new standard types.Jack Jansen2000-12-121-1/+16
|
* Added support for generating a single module from multiple .h files.Jack Jansen2000-12-121-3/+20
| | | | Allow /* */ comments within function prototypes.
* Modified the standard mac preamble so we include pymactoolbox.h inJack Jansen2000-07-151-35/+1
| | | | stead of defining lots of function prototypes in each module.
* ScannerPREUH3.initpatterns() should call Scanner.initpatterns() so theJack Jansen2000-07-151-0/+1
| | | | new comment patterns get included.
* Make a distinction between shorts and unsigned shorts.Jack Jansen2000-07-062-1/+2
|