without this printing a unicode string may raise an error which is a real pain especially since script authors often forget this and print the path of a file for example on export which can make a script fail outright when writing to paths with certain encodings.
Previous state:
Right now, there are "memories" of the "old" (less than a month!) translation way:
* A few remaining calls to BLF_gettext() (only UI_translate_do_iface and UI_translate_do_tooltip should be used).
* The _() macro still also calls BLF_gettext()!
New state:
Here are the changes made by the patch:
* Removing the no more needed _() macro.
* Removing most N_() and _() calls, only keeping the few needed ones (i.e. strings that are in no other way findable by xgettext and/or update_msg script).
* Defining in UI_interface.h IFACE_() and TIP_() macros (resp. for UI_translate_do_iface and UI_translate_do_tooltip).
* Replacing all calls to BLF_gettext by relevant IFACE_ or TIP_ one.
* Replacing all calls to UI_translate_do_iface by IFACE_.
* Replacing all calls to UI_translate_do_tooltip by TIP_.
All this somewhat clarifies and simplifies the code.
On the bf-translations scripts side, this only implies adding IFACE_ and TIP_ as detection markers for xgettext.
It also allows to reduce POTFILES.in quite notably (only 20 files remaining in it).
Please also have a look at those pages:
* Coder POV: http://wiki.blender.org/index.php/Dev:2.5/Source/Interface/Internationalization
* Translator POV: http://wiki.blender.org/index.php/Dev:2.5/Doc/How_to/Translate_Blender
was infact a very old bug where an empty title took the following word for the title, eg: "%t|First Item|Second Item"
the first item would be evaluated as a title.
- use Py_ssize_t when dealing with python sequence sizes
- dont call PySequence_Size(py_b) in a loop (its slow).
- use faster sequence/float parsing in aud.Factory.filter
Merging the node changes required a lot of conflict resolution, fixed the
issues I could find but if you want stability you might want to wait a bit
before updating.
Use this to raise errors when assigning a string property fails even though the value to assign *is* a string.
Before:
TypeError: bpy_struct: item.attr= val: Object.name expected a string type, not str
After:
TypeError: bpy_struct: item.attr= val: Object.name error assigning string, UnicodeEncodeError('utf-8' codec can't encode character '\udce9' in position 23: surrogates not allowed)
- rearrange structs to work for 64bit
- define all vars before goto's
- ifdefs for qsort_r/qsort_s
- dont cast pointers to int only for NULL checks
- dont printf STR_String directly, get the char pointer from it
also minor change to gpu py module, no need to pass empty tuple to PyObject_CallObject, can just be NULL
