Update: Skip processing substitution_definition nodes

[Andrej730]

And leaving them as-is is producing errors later in format_validator transformer.
Resolved blocking issue from #399

[nutti]

Please do this in the transformer.
https://github.com/nutti/fake-bpy-module/blob/main/src/fake_bpy_module/transformer/rst_specific_node_cleaner.py

This should be CodeDocumentNode to remain its contents as a comment.

[Andrej730]

@nutti moved handling to the transformer. But couldn't get past the line endings restrictions to commit my changes without applying line endings normalization first, see ##409

[nutti]

Can you share the final html document which has substitution_definition node?
I want to check if is OK to skip or not.

[Andrej730]

@nutti

Can you share the final html document which has substitution_definition node?
I want to check if is OK to skip or not.

https://docs.blender.org/api/current/bmesh.types.html

Used replacements:

SUBSTITUTION REPLACEMENT 'UV_STICKY_SELECT_MODE_REF':
(bpy.types.ToolSettings.uv_sticky_select_mode which may be passed in directly).

SUBSTITUTION REPLACEMENT 'UV_STICKY_SELECT_MODE_TYPE':
Literal['SHARED_LOCATION', 'DISABLED', 'SHARED_VERTEX']

SUBSTITUTION REPLACEMENT 'UV_SELECT_FLUSH_MODE_NEEDED':
This function selection-mode independent, typically bmesh.types.BMesh.uv_select_flush_mode should be called afterwards.

SUBSTITUTION REPLACEMENT 'UV_SELECT_SYNC_TO_MESH_NEEDED':
This function doesn't flush the selection to the mesh, typically bmesh.types.BMesh.uv_select_sync_to_mesh should be called afterwards.

[Andrej730]

Rebased to resolve conflicts, no functional changes.

[Andrej730]

Ahhh, errors due to conflicting line endings... rebasing. Should be good now.

[nutti]

@nutti

Can you share the final html document which has substitution_definition node?
I want to check if is OK to skip or not.

https://docs.blender.org/api/current/bmesh.types.html

Used replacements:

SUBSTITUTION REPLACEMENT 'UV_STICKY_SELECT_MODE_REF':
(bpy.types.ToolSettings.uv_sticky_select_mode which may be passed in directly).

SUBSTITUTION REPLACEMENT 'UV_STICKY_SELECT_MODE_TYPE':
Literal['SHARED_LOCATION', 'DISABLED', 'SHARED_VERTEX']

SUBSTITUTION REPLACEMENT 'UV_SELECT_FLUSH_MODE_NEEDED':
This function selection-mode independent, typically bmesh.types.BMesh.uv_select_flush_mode should be called afterwards.

SUBSTITUTION REPLACEMENT 'UV_SELECT_SYNC_TO_MESH_NEEDED':
This function doesn't flush the selection to the mesh, typically bmesh.types.BMesh.uv_select_sync_to_mesh should be called afterwards.

Still I could not understand the position.
Can you share a snapshot which indicates the substitution_definition part?

[nutti]

Some files include the changes which is not related to this.
I hope this comes from the line break format differentiation.
Can you make a separate PR?

[Andrej730]

Some files include the changes which is not related to this.
I hope this comes from the line break format differentiation.
Can you make a separate PR?

See

But couldn't get past the line endings restrictions to commit my changes without applying line endings normalization first, see ##409

The problem is tests\python\fake_bpy_module_test\fake_bpy_module_test\generator_test_data\code_document_node_translator_test\expect\basic\basic.py affected by both this PR and #409.

So because of the line-endings issues, it was hard to commit the change separately, because git wants files to be renormalized according to .gitattributes... So I included commits from #409 to this PR, so it will be possible to merge without conflicts if #409 if accepted.

[nutti]

Thank you @Andrej730
Now #409 is merged. Please rebase again.