Replace plain-text type information with the type syntax used
for Python's type annotations as it's more concise, especially for
callbacks which often didn't include useful type information.
Note that this change only applies to inline doc-strings,
generated doc-strings from RNA need to be updated separately.
Details:
- Many minor corrections were made when "list" was incorrectly used
instead of "sequence".
- Some type information wasn't defined in the doc-strings and has been
added.
- Verbose type info would benefit from support for type aliases.
For C/C++ doc-strings should be located in headers,
move function comments into the headers, in some cases merging
with existing doc-strings, in other cases, moving implementation
notes into the function body.
Refactor 1dbe94c8ac restored the 'old' proper way to directly call the
python type (i.e. use the python's implementation to create objects),
instead of using the 'specialized' type creation code.
However, the handling of the ugly `rna_disallow_writes` global was only
added later to the 'workaround' part of the code, but not to the original
'canonical' one.
This commit copies the handling of `rna_disallow_writes` back into the
now active part of the code.
This solves the `is_readonly_init` unused variable build warning.
NOTE: At some point the BPY code needs a good cleanup pass, there are
way to many pieces of codes #ifdef'ed there.
For now, PointerRNA is made non-trivial by giving explicit default
values to its members.
Besides of BPY python binding code, the change is relatively trivial.
The main change (besides the creation/deletion part) is the replacement
of `memset` by zero-initialized assignment (using `{}`).
makesrna required changes are quite small too.
The big piece of this PR is the refactor of the BPY RNA code.
It essentially brings back allocation and deletion of the BPy_StructRNA,
BPy_Pointer etc. python objects into 'cannonical process', using `__new__`,
and `__init__` callbacks (and there matching CAPI functions).
Existing code was doing very low-level manipulations to create these
data, which is not really easy to understand, and AFAICT incompatible
with handling C++ data that needs to be constructed and destructed.
Unfortunately, similar change in destruction code (using `__del__` and
matching `tp_finalize` CAPI callback) is not possible, because of technical
low-level implementation details in CPython (see [1] for details).
`std::optional` pointer management is used to encapsulate PointerRNA
data. This allows to keep control on _when_ actual RNA creation is done,
and to have a safe destruction in `tp_dealloc` callbacks.
Note that a critical change in Blender's Python API will be that classes
inherinting from `bpy_struct` etc. will now have to properly call the
base class `__new__` and/or `__init__`if they define them.
Implements #122431.
[1] https://discuss.python.org/t/cpython-usage-of-tp-finalize-in-c-defined-static-types-with-no-custom-tp-dealloc/64100
While copying this used to be very cheap, this will soon be less
trivial.
Note that there are many more places where this type of cleanup will be
needed.
Remove a workaround for #23871 which manipulated the module
to prevent classes defined in the text editor from having their
name-space cleared.
This caused the "multiprocessing" module to store the "__main__" module
as "__mp_main__" for later use.
Accessing attributes from this module would then attempt to read from
with a null "mp_dict" which crashed. This happened when showing the
extensions preferences but would have occurred if "__mp_main__" was
accessed from elsewhere too.
Resolve by removing the workaround since it has not been needed
since Python 3.2.
- `_PySet_NextEntry` has been removed, use generic iterator access
which will has some additional overhead as it needs to create
an iterator to access the values.
- Add v3.13 compatibility defines to account for renaming:
_PyObject_LookupAttr -> PyObject_GetOptionalAttr
_PyLong_AsInt -> PyLong_AsInt
- Unfortunately use of Python's internal API needs to be used to
inspect op-codes in `bpy_driver.cc`.
Testing GLTF/FBX IO there isn't any significant performance impact
from these changes.
Resolves#123871.
Essentially, our current code would not properly remove (dereference)
its python objects matching various RNA data created during execution.
Some cases are fairly trivial to understand (like the lack of handling
of unregstering for our 'startup' operators and UI), other were more
subtle (like unregistered PropertyGroups who would remove/free their RNA
struct definition, without releasing first the potential matching python
object).
Co-authored-by: Campbell Barton <campbell@blender.org>
Pull Request: https://projects.blender.org/blender/blender/pulls/128899
Move most of the string preprocessing used for MSL
compatibility to `glsl_preprocess`.
Enforce some changes like matrix constructor and
array constructor to the GLSL codebase. This is
for C++ compatibility.
Additionally reduce the amount of code duplication
inside the compatibility code.
Pull Request: https://projects.blender.org/blender/blender/pulls/128634
The `pre` handler is called after blender internal code is done populating
the link/append context with data to be processed, and before this data
starts being linked from library files.
The `post` handler is called after blender is done linking, and
potentailly appending and/or instantiating, the requested data and all
of their dependencies.
Both handlers are called with a single argument, the link/append
context.
An new RNA sets of wrappers have been added to expose relevant info from
these internal C++ structures.
NOTE: !113658 is very similar (but tied to asset drag & drop), whereas
this PR is more general (these could probably live hand-in-hand / side-
by-side).
Implements #122357
Pull Request: https://projects.blender.org/blender/blender/pulls/128279
-----------------
Some quick py example code:
```python
import bpy
def my_handler_pre(lapp_context):
print("About to {}:\n\t".format("link" if "LINK" in lapp_context.options else "append"),
"\n\t".join("{} '{}', from libs ['{}']".format(item.id_type, item.name,
"', '".join([l.filepath for l in item.source_libraries]))
for item in lapp_context.import_items))
def my_handler_post(lapp_context):
print("{}:\n\t".format("Linked" if "LINK" in lapp_context.options else "Appended"),
"\n\t".join("{} '{}', from lib '{}'".format(item.id.id_type, item.id.name, item.source_library.filepath)
for item in lapp_context.import_items))
bpy.app.handlers.link_append_pre.append(my_handler_pre)
bpy.app.handlers.link_append_post.append(my_handler_post)
```
This replaces the older dynamic c array macros with blender::Vector in
bmesh_polygon_edgenet. This is the only remaining use of the old array
machinery and removal of `BLI_array.h / .c` can happen immediately
afterwards.
See #103343
Pull Request: https://projects.blender.org/blender/blender/pulls/119975
