WARNING! The roof generation is now quite significantly different. This
WILL change your geometry when you refresh roofs from existing models.
- The roof profile now always represents the top of eave. This better
represents how things are built rather than bottom of eave which was
previous.
- The roof therefore always grows down from the profile. The rafter
edge angle can only be acute. This means that the roof will never grow
larger than the footprint profile that you've drawn. (before, the
footprint was not guaranteed to match).
- The roof thickness is now the actual thickness of the roof, not the
"vertical dimension" of the roof. This means that the roof thickness can
now match intended layer thicknesses instead of you needing to do math
to work it out.
I've rewritten the internals of how roofs were generated to be hopefully
a lot simpler but as a tradeoff it's more restrictive. After the
skeleton is generated, non-uniform angles would be handled through
vertex splitting and moving. This works in simple scenarios but fails in
more complex ones. The new approach only handles non-uniform angles on
triangular faces. These faces are really easy to handle compared to
ngons, but are also more robust. The rafter edge angle is also now
handled using a clipping plane, which is a lot, lot simpler than vertex
sliding math.
Each face is now processed separately and then merged at the end. This
means that if a face has a different angle, it will now correctly
represent the different thickness at that portion of the roof.
In the process consistent roof thickness / rafter angle bugs were fixed.
E.g. two objects are selected - tesselation and profile and tesellation being active object. Using s-e with deselect tesselation object as it doens't have a usage but Bonsai would still try to run bim.enable_editing_extrusion_profile on it leading to the error.
Also fix possible issues for hotkey executed without active object (it seems all operators in hotkey do require active object)
Apparently (https://debbugs.gnu.org/cgi/bugreport.cgi?bug=62572) coreutils have changed `mv -n` to throw an error now instead of working silently, therefore our build was failing on Mac when Github Action switched from Ubuntu 22 to 24.
`--update=none` seems to restore the old behaviour.
The for/else block was too greedy, meaning that if nothing was selected,
it would still try to create slabs from walls. I find for/else confusing
so I restructured the section to do more early returns and unindent the
logic.
• Removed bpy.context.preferences.themes[0] from the input and hardcoded it in the function.
• color_path now only requires the attribute path suffix (e.g., "user_interface.wcol_regular.text").
• Updated error handling to default to "dm" instead of "lm".
If you move a wall, and that wall has features (e.g. openings), it's
desirable to also move those openings (because they are invisible). This
is a special exception to `should_transform_children` because the
definition of the feature (opening) is inherently tied to the parent
(wall).
What wasn't considered is that this would typically then also move
subchildren of the features (e.g. fills like doors). I'm surprised
nobody caught this earlier.
I did also consider another approach where if you move a wall, it moves
all unfilled openings, and if you move a door which fills a opening, it
moves the opening too. Intuitively it sounds nice, but it doesn't work
because:
- Openings can have multiple fillings. If you move all fillings, they
all fight to move the openings.
- All logic about children goes one way: a placement may have child
placements relative to it. This breaks the convention (if moving a door
instead moves its opening) which can make brains explode.
- It starts to conflate rules about relative / referenced placements
with spatial decomposition. We assume all IFCs are valid and follows the
convention of relative placement but we cannot guarantee this. This also
leads to brain explosion.
Previously the code looked for additive facets (entity, instance)
anywhere in the query, and then only added the default basket once at
the beginning of the query. This change makes it look for additive
facets in order (that way you could start with a default selection, and
then later use an additive facet) and does it per facet list (so you can
omit the additive facet at the beginning of each list).
The IDS documentation used to state that "optional" on requirement
facets was "for information only" and didn't affect the pass / fail
state. This was changed a while ago (which I missed) to a new behaviour:
both no values and passing values will pass.
Previously, to edit a single IfcMaterial pset, you'd do so in the
material manager. In contrast, to edit a IfcMaterialSet pset, you'd do
so only when assigned to an object. Not only is it bad to edit psets in
two locations, it also means you can't edit psets until you've first
assigned the material set to an object which is bad. Now it's less code,
and material / material set psets are in the same spot.
There still more consolidation work to do, to allow editing of sets in
the material manager and to figure out how to show item psets, but it's
a step in the right direction.
Went back to previous toolbar : icon + type name + dot grid icon which all launch the type manager.
Validating the type manager popup now only closes it. Functionally is the same thing as clicking Cancel. Don't really know what to do to fix it. Does cancel make sense in the type manager ? Do we want to be able to undo what is possible to be done from within the type manager ?
You can change the active type either with the dropdown in the popup and in the grid flow. I changed it so that changing the type in the dropdown jumps to the corresponding page. BTW first time using github copilot to figure out the maths and it worked magically :) . I'm not against removing the dropdown but I feel like it's nice to see a plain list to choose from.
You can input a specific page number or scrub the page number field. Might be usefull if user has dozens of pages to flip through.
Right now the search is very simple, it does not implement fuzzy search or anything fancy with *
Add back the type dropdown and looking glass icon in the toolbar
Clicking on a type in the manager no longer closes the popup anymore.
A visual indictor is displayed on the active type
Changed the parameters icon in the individual type interface
Fix bug in "Slab from Walls" Shift + A code
This highlights a potential bug that already existed due to the way the attribute system is designed. Where the user mass-edits attributes from different IFC types, resulting in discrepancies due to the same attributes having a different index. eg IfcWindow has the Tag attribute at index 3, IfcDoor at index 4. So using the ALT modifier feature from vanilla Blender to mass-modify similarly named attributes on the Tag attribute of the IfcWindow will modify the PredefinedType attribute on the IfcDoor (index 3). For this reason I added the ALT modifier to the buttons so mass-operations are a conscious effort for more knowledgeable users. Might reconsider if problems arise but this is IMO a good QOL feature.
Also fix floating point precision errors on Pset float values stored as strings (re last commit)
To do so Right Click on a material in the Asset Browser then `Set To External Style`. The style editor must be loaded in the properties editor.
Also run black on a file from previous commit and fix a bug soft-locking the style editor when removing a style that is being edited
It's a naive implementation and very inefficient. Moreover there may be false positives because it's testing overlapping the evaluated mesh element with all openings applied.
No functional change apart from in dissolve_triangulated_edges where it will try to load the attribute if it can. I believe it will be easier in the long term to use the builtin Attributes system. Unfortunately Curves don't support Attributes yet. Also, attributes are more lightweight compared to custom properties.
It uses a custom boolean attribute on the mesh to decide whether an edge should be displayed or not. I think later on it can be used to get rid of the dictionary accessors which make the blend file size skyrocket.
- Highlight new objects in red when in aggregate mode
- Introduce an operator to add new objects to the current aggregate, accessible via a button in the header menu
Now, when trying to move an object that is a part of an aggregate, the
`OverrideMoveAggregate` will be called, meaning that it will move the whole
aggregation. It works by creating constraints and moving the aggregate instead
of the selected object.
It also introduces the idea of editing an aggregate (no hotkey yet). When active, the user
will be able to move the parts of the aggregate instead of the whole aggregation: https://imgur.com/IcCGtPN
- Enables the vanilla Blender autosave feature back, since a running modal operator deactivates it
- ESC is not used by any other vanilla keymap in the 3D viewport (may be conflict with other addons)
- Hitting ESC over another type of editor (eg enum field in the properties editor) will not trigger disabling ifc item mode
- You can now hit ESC in IFC edit mode to cancel editing the geometry and go back to ifc object mode
Error occurred if IfcElementAssembly, that was currently refreshed, had another IfcElementAssembly as it's part. Then, it was removed twice - once during delete_objects and another time during delete_ifc_object, leading to the error.
If current set of rules (e.g. IfcOpenShell) doesn't provide a way to quantify some type of element, then user now have an option to fallback to other calculator (Blender).
Location - https://i.imgur.com/csIPOuo.png
This operator allows users to create walls from a slab.
When the wall tool is active, pressing `Shift + A` will generate walls based on the exterior polygon of the selected slab.
This operator allows users to create a slab from a selected closed loop of walls.
When the slab tool is active, pressing `Shift + A` will generate a slab based on the exterior polygon of the selected walls.
E.g. in #5890 pset was shared by 2 elements and they it worked is
1) it appended first element and it's rel to pset
2) it appended second element and skipped updating it's rel as it was appended before. Now it's going to update it to ensure it encludes second element too.
I believe, issue occurred after 8c28f52 when we stopped duplicating inverses and started to reuse them.
Mentioned in 618ba51, it probably will be more natural this way - if there are no textures, we hide UV attribute to indicate that it won't be saved and on save we either do nothing, if texture style wasn't saved before, or remove it.
It was suggesting that it's possible to create it with just shading style though it's not currently supported.
Now it also appears when user is about to enable texture style editing, not when they're already trying to save the changes.
Demo - https://imgchest.com/p/xny8qvox5yb
You can also now reuse 'bpy.data.texts['globalid-filter-3UvnD5Ml95zBDEcwAkAUIp']' kind of queries from Search in ExtractElements query.
1) Identify from previous pset value whether prop was an enum (by checking if previous value is a list), instead of searching through prop templates
2) Compare previous and new enum values - if they match, skip editing IFC.
1) If object had multiple representations for the same context, then switch simply wouldn't work as it would always pick up the first representation it could find. Now it prioritizes the one user is selecting in UI.
2) If some object is selected and it also has multiple representations for the same context, then switch representation would switch it to any representation in that context it first finds. As user doesn't provide representation explicitly in that case, now it will stick to the already active representation on selected object if it matches the context, to avoid accident switches.
Notcied by working with example from #5824
- The patch was failed when executed from Bonsai since self.src wasn't provided, fixed now.
- Added a temporary hack to diplsay file selector for this patch.
- small refactor.
It was only checking if it's existing path but wasn't checking whether it's a file, therefore directories with suffixes were slipping in and then failing during `IfcHeaderExtractor(filepath)`.
Mentioned in logs from #5820
Turned out, `area.spaces` also contain all other spaces this area used (e.g. it used be view3d before it became a nodegraph) leading to errors. Simplifying it with `area.spaces.active`.
the way it was intended to originally but absolute paths of ifc and ids got in the way
Replaced TemporaryDirectory with mkdtemp as sometimes .html report doesn't open fast enough and directory gets deleted. Also need to keep the directory to make sure .ods is not removed.
The titleblock is ifc relative, not .blend/cwd relative.
e.g. if you open Blender in C:\ and then would open .ifc from c:\path\to\project.ifc, it would try to create layouts folder in C:\layouts instead of c:\path\to\layouts
Error: Python: Traceback (most recent call last):
File "\bonsai\bim\ifc.py", line 443, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\profile\operator.py", line 172, in _execute
if not indices:
^^^^^^^
UnboundLocalError: cannot access local variable 'indices' where it is not associated with a value
example - https://imgchest.com/p/qb4zwj6ad4j
1) moved profile object selection to the same row as the profile type selection, so it will more noticable that they're related
2) added object requirements in the description
3) filtered only mesh objects to appear in the dropdown
There was an issue that it wouldn't load until you manually select it in the "Cost Item Rates" enum - otherwise items didn't appear the first time cost schedule opened and didn't not update when you makes changes to the active schedule of rates.
To avoid confusing situations when user would assign a class, move object around and save ifc file, then to realize that it was saved using just local transforms which can be completely different.
No idea what happened but in Blender <4.3 `type(bpy.types.SCENE_PT_rigid_body_world_settings.bl_rna)` is `<class 'bl_ui.properties_scene.SCENE_PT_rigid_body_world_settings'>` which is `bpy.types.Panel` but in Blender 4.3 it's now is `bpy.types.Struct`. Using `.base` seems to work for both versions.
Error: Python: Traceback (most recent call last):
File "\bonsai\bim\ifc.py", line 443, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\drawing\operator.py", line 2960, in _execute
core.edit_assigned_product(tool.Ifc, tool.Drawing, obj=context.active_object, product=product)
File "\bonsai\core\drawing.py", line 66, in edit_assigned_product
drawing.update_text_value(obj)
File "\bonsai\tool\drawing.py", line 983, in update_text_value
cls.import_text_attributes(obj)
File "\bonsai\tool\drawing.py", line 904, in import_text_attributes
props.font_size = str(text_data["FontSize"])
should_draw_decorations is updated as part of regular refresh_ui_data.
'removed_drawings' line was a dead code
Traceback (most recent call last):
File \bonsai\bim\module\drawing\prop.py", line 252, in update_should_draw_decorations
collection = context.scene.camera.BIMObjectProperties.collection
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'BIMObjectProperties'
File \bonsai\bim\module\drawing\prop.py", line 249, in update_should_draw_decorations
Items generators were exhausted at the first register leading to errors like 'TypeError: EnumProperty(..., default='-'): not found in enum members' preventing reregistration.
Traceback:
Traceback (most recent call last):
File "\bonsai\bim\module\patch\operator.py", line 187, in execute
core.run_migrate_patch(tool.Patch, infile=self.infile, outfile=self.outfile, schema=self.schema)
File "\bonsai\core\patch.py", line 30, in run_migrate_patch
patch.run_migrate_patch(infile, outfile, schema)
File "\bonsai\tool\patch.py", line 28, in run_migrate_patch
output = ifcpatch.execute(
^^^^^^^^^^^^^^^^^
File "\ifcpatch\__init__.py", line 85, in execute
patcher.patch()
File "\ifcpatch\recipes\Migrate.py", line 58, in patch
new_element = migrator.migrate(element, self.file_patched)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 363, in migrate
new_element = self.migrate_attributes(element, new_file, new_element, new_element_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 393, in migrate_attributes
self.migrate_attribute(attribute, element, new_file, new_element, new_element_schema)
File "\ifcopenshell\util\schema.py", line 494, in migrate_attribute
new_value.append(self.migrate(item, new_file))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 363, in migrate
new_element = self.migrate_attributes(element, new_file, new_element, new_element_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 393, in migrate_attributes
self.migrate_attribute(attribute, element, new_file, new_element, new_element_schema)
File "\ifcopenshell\util\schema.py", line 497, in migrate_attribute
setattr(new_element, attribute.name(), value)
File "\ifcopenshell\entity_instance.py", line 325, in __setattr__
self[index] = value
~~~~^^^^^^^
File "\ifcopenshell\entity_instance.py", line 363, in __setitem__
raise TypeError(
TypeError: attribute 'CountValue' for entity 'IFC4X3_ADD2.IfcQuantityCount' is expecting value of type 'INT', got 'float'.
Error: Python: Traceback (most recent call last):
File "\ifcopenshell\entity_instance.py", line 361, in __setitem__
self.method_list[idx](self.wrapped_data, idx, entity_instance.unwrap_value(value))
File "\ifcopenshell\ifcopenshell_wrapper.py", line 9271, in setArgumentAsInt
return _ifcopenshell_wrapper.entity_instance_setArgumentAsInt(self, i, v)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: in method 'entity_instance_setArgumentAsInt', argument 3 of type 'int'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "\ifcopenshell\entity_instance.py", line 325, in __setattr__
self[index] = value
~~~~^^^^^^^
File "\ifcopenshell\entity_instance.py", line 363, in __setitem__
raise TypeError(
TypeError: attribute 'CountValue' for entity 'IFC4X3_ADD2.IfcQuantityCount' is expecting value of type 'INT', got 'float'.
Traceback:
Traceback (most recent call last):
File "\bonsai\bim\module\patch\operator.py", line 187, in execute
core.run_migrate_patch(tool.Patch, infile=self.infile, outfile=self.outfile, schema=self.schema)
File "\bonsai\core\patch.py", line 30, in run_migrate_patch
patch.run_migrate_patch(infile, outfile, schema)
File "\bonsai\tool\patch.py", line 28, in run_migrate_patch
output = ifcpatch.execute(
^^^^^^^^^^^^^^^^^
File "\ifcpatch\__init__.py", line 85, in execute
patcher.patch()
File "\ifcpatch\recipes\Migrate.py", line 58, in patch
new_element = migrator.migrate(element, self.file_patched)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 358, in migrate
new_element = self.migrate_attributes(element, new_file, new_element, new_element_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 388, in migrate_attributes
self.migrate_attribute(attribute, element, new_file, new_element, new_element_schema)
File "\ifcopenshell\util\schema.py", line 489, in migrate_attribute
new_value.append(self.migrate(item, new_file))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 358, in migrate
new_element = self.migrate_attributes(element, new_file, new_element, new_element_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 388, in migrate_attributes
self.migrate_attribute(attribute, element, new_file, new_element, new_element_schema)
File "\ifcopenshell\util\schema.py", line 484, in migrate_attribute
value = self.migrate(value, new_file)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\util\schema.py", line 346, in migrate
return new_file.create_entity(element.is_a(), element.wrappedValue)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\file.py", line 400, in create_entity
e[idx] = arg
~^^^^^
File "\ifcopenshell\entity_instance.py", line 363, in __setitem__
raise TypeError(
TypeError: attribute 'wrappedValue' for entity 'IFC4X3_ADD2.IfcCountMeasure' is expecting value of type 'INT', got 'float'.
Error: Python: Traceback (most recent call last):
File "\ifcopenshell\entity_instance.py", line 361, in __setitem__
self.method_list[idx](self.wrapped_data, idx, entity_instance.unwrap_value(value))
File "\ifcopenshell\ifcopenshell_wrapper.py", line 9271, in setArgumentAsInt
return _ifcopenshell_wrapper.entity_instance_setArgumentAsInt(self, i, v)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: in method 'entity_instance_setArgumentAsInt', argument 3 of type 'int'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "\ifcopenshell\file.py", line 400, in create_entity
e[idx] = arg
~^^^^^
File "\ifcopenshell\entity_instance.py", line 363, in __setitem__
raise TypeError(
TypeError: attribute 'wrappedValue' for entity 'IFC4X3_ADD2.IfcCountMeasure' is expecting value of type 'INT', got 'float'.
Traceback
2024-11-18:11:09:15,441 ERROR [log.py:69] Uncaught exception
Traceback (most recent call last):
File "\bonsai\bim\module\pset_template\ui.py", line 66, in draw
self.draw_pset_template()
File "\bonsai\bim\module\pset_template\ui.py", line 110, in draw_pset_template
value_name = self.props.active_prop_template.get_value_name()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\pset_template\prop.py", line 175, in get_value_name
ifc_data_type = IfcStore.get_schema().declaration_by_name(self.primary_measure_type)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\ifcopenshell_wrapper.py", line 10119, in declaration_by_name
return _ifcopenshell_wrapper.schema_definition_declaration_by_name(self, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Entity with name '-' not found in schema 'IFC4'
Full error traceback:
Error: Python: Traceback (most recent call last):
File "\ifcopenshell\entity_instance.py", line 361, in __setitem__
self.method_list[idx](self.wrapped_data, idx, entity_instance.unwrap_value(value))
File "\ifcopenshell\ifcopenshell_wrapper.py", line 9271, in setArgumentAsInt
return _ifcopenshell_wrapper.entity_instance_setArgumentAsInt(self, i, v)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: in method 'entity_instance_setArgumentAsInt', argument 3 of type 'int'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "\bonsai\bim\ifc.py", line 443, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\pset\operator.py", line 142, in _execute
ifcopenshell.api.run(
File "\ifcopenshell\api\__init__.py", line 92, in run
return usecase_function(ifc_file, should_run_listeners=should_run_listeners, **settings)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\__init__.py", line 258, in wrapper
raise e
File "\ifcopenshell\api\__init__.py", line 252, in wrapper
result = usecase(*args, **settings)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\pset\edit_qto.py", line 135, in edit_qto
return usecase.execute()
^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\pset\edit_qto.py", line 149, in execute
self.update_existing_properties()
File "\ifcopenshell\api\pset\edit_qto.py", line 166, in update_existing_properties
self.update_existing_property(prop)
File "\ifcopenshell\api\pset\edit_qto.py", line 180, in update_existing_property
prop[3] = float(value)
~~~~^^^
File "\ifcopenshell\entity_instance.py", line 363, in __setitem__
raise TypeError(
TypeError: attribute 'CountValue' for entity 'IFC4X3_ADD2.IfcQuantityCount' is expecting value of type 'INT', got 'float'.
Example - https://imgchest.com/p/bp452xxbx45
cc @steverugi @DimitriosThe maybe you'll find useful too if you use custom quantity sets for products in 4d/5d
tuple(resource) was resulting in (GlobalId, OwnerHistory, Name, ...) instead of (resource,), so the check never passed and orphaned IfcRelAssignsToResource was never removed.
* Fix Shift-clicking SVG-DXF conversion
* Fix Schedules and References not incrementing the ID number properly
* Remove the superfluous Create and Open buttons from the Active Drawing pane
* Add active camera indicator to the Drawing list
* Improved enable/disabled icon behaviour
As many operators as possible now use polling method to set enabled/disabled state.
Icons set their enabled/disabled state intelligently depending on what is selected.
When an operator is disabled it says why it is disabled in the tooltip.
* Rename core open_sheet, operator OpenSheet/bim.open_sheet to better reflect what it actually does, i.e. open_layout
* Added selection checkmarks to Sheets, and operator to select all/none, to do batch ops on sheets
* Missed adding the renamed operator to __init__ in commit 4cfc0bd9f3
* Decouple creating and viewing drawings/sheets to avoid focus stealing popups polluting display on long batch runs
* Improve drawing selections in OpenDrawing operator
* Add the Shift+click power move to Create Sheet
* Add the Open Sheet + Shift+click power move
* Improve drawing selections in RemoveDrawing operator
* All operators are now documented with standardised language and terms.
* Add status bar Info messages for feedback on create tasks that now don't have visual feedback
* Fix enable/disable for ActivateDrawing operator icons.
This was a strange one. They have common invoke and execute methods,
but the polling has different conditions.
I tried subclassing the original ActivateDrawing for the Sheets version,
but something about Blenders operators breaks the classmethod decorated
poll. This was the cleanest solution I found, but maybe there's a better
one.
* Placeholder operator for DuplicateSheet function.
Does nothing but report Not Implemented. Could be omitted if this
kind of thing is frowned upon.
* Change icons to be more consistent
* Final changes to toolbars
Change button order and spacing for better usability.
Add a checkmark all/none button, to do batch ops on sheets.
Moved Bonsai Web UI button up to sheets toolbar.
Some general cleanup of unnecessary clutter in the toolbar code.
* Sheet/Drawing checkmarks remember their state properly
* Adjust sheet/drawing operators to view after singular create, but not when batch creating more than one.
* By special request, add back the opening windows when creating, but requiring Ctrl to be pressed.
* black the files
* Second attempt at blacking ui.py because I used an old black
* Change method of getting active drawing to use one in tool import as requested
* Fix for incorrect d1fc5158d - actually checked it this time
On large projects selecting cost items is very slow due to all data that needs to be loaded. On the project I was testing, it was taking about 0.8sec, now this time is reduced to 0.17sec.
That way all assigned resources will be added as cost values to cost items and it will be more apparent what resources were used for the last calculation and what costs/quantities they had at the time.
cc @Andrej730 we need this for validate.buildingsmart.org because for express rule evaluation we depend on include_id=False and entity_instance.__eq__ has been changed to get_info_2.
Previously Bonsai would show a wall of errors, now it's show a simple error message and redirecting to system console for the details.
Example error - https://i.imgur.com/uJJfjUW.png
Example validation error details in console - https://i.imgur.com/3mcVBUh.png
Same details but in text:
Validation error details:
failed validating {'dataType': 'IFCBOOLEAN', 'ursi': 'https://google.com', 'cardinality': 'required', 'instructions': "Make sure it's true"} with XsdAttributeGroup(['dataType', 'uri', 'cardinality', 'instructions']):
Reason: 'ursi' attribute not allowed for element
Schema component:
<xs:extension xmlns:xs="http://www.w3.org/2001/XMLSchema" base="ids:propertyType">
<xs:attribute name="uri" type="xs:anyURI" use="optional" />
<xs:attribute name="cardinality" type="ids:conditionalCardinality" use="optional" default="required" />
<xs:attribute name="instructions" type="xs:string" use="optional">
<xs:annotation>
<xs:documentation>Author of the IDS can leave instructions for the authors of the IFC. This text could/should be displayed in the BIM/IFC authoring tool.</xs:documentation>
</xs:annotation>
</xs:attribute>
</xs:extension>
Instance type: <class 'xml.etree.ElementTree.Element'>
Instance:
<ids:property xmlns:ids="http://standards.buildingsmart.org/IDS" dataType="IFCBOOLEAN" ursi="https://google.com" cardinality="required" instructions="Make sure it's true">
<ids:propertySet>
<ids:simpleValue>Pset_WallCommon</ids:simpleValue>
</ids:propertySet>
<ids:baseName>
<ids:simpleValue>Combustible</ids:simpleValue>
</ids:baseName>
<ids:value>
<ids:simpleValue>false</ids:simpleValue>
</ids:value>
</ids:property>
Path: /ids:ids/ids:specifications/ids:specification/ids:requirements/ids:property[1]
Got report that there was an issue with get_parts not returning all parts if there are multiple IfcRelAggregates involved though it is valid ifc. Fixed the same possible issue for ContainsElements, IsNestedBy
* fix: Rename the "MergeProject" recipe to "MergeProjects"
This rename, alongside renaming "filepath" to "filepaths" and adjusting the docstring has been done to clarify, that this patch can be used to merge multiple IFC files into one (and not just two)
* fix: Fix `test_Merge_Project.py`, which was broken due to renaming the patch recipe
Traceback
```
Error: Python: Traceback (most recent call last):
File "\bonsai\bim\ifc.py", line 443, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\pset\operator.py", line 126, in _execute
ifcopenshell.api.run(
File "\ifcopenshell\api\__init__.py", line 92, in run
return usecase_function(ifc_file, should_run_listeners=should_run_listeners, **settings)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\__init__.py", line 252, in wrapper
result = usecase(*args, **settings)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\pset\edit_pset.py", line 166, in edit_pset
return usecase.execute()
^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\pset\edit_pset.py", line 176, in execute
existing_props = self.update_existing_properties()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\pset\edit_pset.py", line 225, in update_existing_properties
prop = self.update_existing_prop_enum(prop)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\ifcopenshell\api\pset\edit_pset.py", line 251, in update_existing_prop_enum
primary_measure_type = prop.EnumerationReference.EnumerationValues[
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'EnumerationValues'
```
Noticed working on #5596 and #5708 a few issue when pset template is changing the prop types.
E.g. prop had string data type and now template suggests that it should be a float.
Previously, it might error trying to cast a string to a float. Now it will show a float until user will set it to null, then next time it will be loaded from the template data type.
Another example - prop was a string and now it's a boolean. Previously, it would be cast to boolean implicitly. E.g. "test" -> bool("test") -> True. And if user wouldn't be paying attention then enabling pset editing and saving it would make them lose "test" value.
Now - https://imgchest.com/p/na7ke99wb78
Before - https://imgchest.com/p/5xy23rr3z4l
Basically it won't allow setting values below 0 by accident (especially important for representation item props as it may lead to a crash).
1) IfcPositiveLengthMeasure will affect bunch of props like Depth in IfcExtrudedAreaSolid.
2) IfcNonNegativeLengthMeasure will affect MaterialLayer and some profiles attributes.
As it's the quicker way to select it than searching it in outliner - https://i.imgur.com/p2NVcdR.png
Couldn't find a simpler a way to do it in UI and had to create bunch of columns. Tried to use splits but they end up readjusting the buttons.
1) Fixed non-existent tool.Blender.get_screen
2) ui.py - remove try/except for the case when tabs are not loaded yet as it's covered by BIMTabProperties
3) removed previous_tab, alt_tab props never used in ui
4) general method for getting area properties that includes BIMTabProperties automatically
I guess by accident in d443c5e I've made this method too strict and it broke get_references using it for IfcPropertySets. Maybe it's a good idea to make it more strict in the future, for now just restoring the previous behaviour.
Basically now all length ifc attributes should be displayed in UI as Blender length properties - they display the units and allow math, typing different units, etc.
Example - https://imgchest.com/p/dl7p93gz64o
Now it's stored as a collection property in the project properties, so in the most cases we delegate to Blender to ensure it's handled properly on undo and redo.
Button location - https://i.imgur.com/6zSv8vV.png
Now it's possible to "bake" openings to the geometry, The openings representation will be marked as "Reference".
Example - https://i.imgur.com/v8p4D57.png (it's ignoring walls).
Since we now filter valid elements in poll_container_obj, then there is no need to check it twice in update_container_obj.
Previously for both cardinality optional/required ifctester would show same results.
E.g.
<ids:property dataType="IFCBOOLEAN" cardinality="required">
<ids:property dataType="IFCBOOLEAN" cardinality="optional">
Were resulting in something like
LoadBearing data shall be true and in the dataset Pset_WallCommon
LoadBearing data shall be true and in the dataset Pset_WallCommon
Now it will result in
LoadBearing data shall be true and in the dataset Pset_WallCommon
LoadBearing data may be true and in the dataset Pset_WallCommon
`transition_curve.py` is now obsolete and has been removed.
A python port of IfcAlignmentHelper.cpp has also been started.
It needs further development to accommodate vertical alignment.
We try to apply scales using bim.update_representation and if didn't worked (presumably because object has openings) we reset the scales.
It was confusing for users that they scaled an object, saved IFC and they would know that scaling didn't worked only after they reopen the IFC project.
Currently there is an UX problem that it's not clear for users when it's possible to use scale or not. We definitely won't be able to safely apply scales for objects with openings, so will try to lock them from the start.
* Implement rounding of elevation
Implement rounding of elevation based on precision stored in IfcGeometricRepresentationContext
* Move formating function to tool
The problem was that shared folders names are starting with `\\`, e.g. `\\SERVERNAME\shared_folder`. When we pass path strings we usually sanitize them and save them as posix, so it's saved as `//SERVERNAME/shared_folder` and Blender has a convention that paths starting with `//` it interprets as filepaths relative to the current .blend file, therefore it was giving some weird error that `C:\Users\xxxx\SERVERNAME\shared_folder\test.blend` is missing instead of `\\SERVERNAME\shared_folder`.
So need to be careful passing strings to Blender operators and use str instead of as_posix for those operators, so \\` wouldn't be misinterpreted.
Turn out storing bpy.types.PropetyGroups was not very reliable idea as they get invalidated, just as Objects.
It's still breaking if you'd try to rename the style and then undo.
Previously transaction was never finished and therefore is_top_level_operator would never be triggered (without restarting project) making any IFC undo useless.
Previously it wsan't finishing transaction therefore:
- it was possible to undo ifc changes in the last step that had an error
- since transaction was never finished, IfcStore.current_transaction was never cleared and therefore is_top_level_operator will be never be triggered for the next operators, so new transactions are never created breaking undo for any further ifc changes.
Now user can try to undo the last action to save the file.
OperationData was always including guid key, so guid_map[None] was pointing to some object (typically a style). Also covered by rebuild_element_maps in the most cases.
Probably was harmless, it was just flooding the console with errors such as:
object address : 000002485A776380
object refcount : 2
object type : 00007FFBD1249050
object type name: RecursionError
object repr :
lost sys.stderr
File "\bonsai\bim\module\model\prop.py", line 707, in <lambda>
object address : 000002485A774520
object refcount : 2
object type : 00007FFBD1249050
object type name: RecursionError
object repr :
lost sys.stderr
File "\bonsai\bim\module\model\prop.py", line 713, in <lambda>
Not sure when this occurred but get_bmesh_from_polygon was expecting Polygon in project units but it was SI. Since it's probably better to move everything to SI, added an option to specify whether polygon is in SI or not.
Since 5677768 recalculate_dumb_wall_origin is never used and bim.align_wall is either just changing the location (align_centerline) or might change object's representation but it's already saved to ifc (exterior/interior).
Noticed that after 189bcd5 some representations were missing styles - the ones that are using IfcSurfaceStyle with just IfcExternallyDefinedSurfaceStyle.
Ping @aothms just in case
Mentioned in #5604. Since IfcSurfaceStyle.Side is used during representation generation, we need to regenerate them if the value was changed. Previously it would require manual update / project reload.
Fixes:
$ IfcConvert
IfcConvert: error while loading shared libraries: libcityjson_converter.so: cannot open shared object file: No such file or directory
Mentioned in #5596
Example - https://imgchest.com/p/5xy235xvj4l
Adds 2 things:
1) UI indicator whether pset is based on template or not
2) When pset is not based on a template, indicator becomes clickable and works as a shortcut for creating pset template based on the current pset.
- as it was only used inside bim.add_pset_template_file operator
- also remove second unnecessary ":" in the prop label (":" is automatically added by Blender)
Previously it would show obscure error like:
RuntimeError: Type held at index 0 is class Blank and not class std::vector<class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> >,class std::allocator<class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > > >
The problem is index.json is using urls pointing to zip files from the releases and it's possible that installation from unstable repo will be broken (user will get error message like '"Archive size mismatch "bonsai", expected 87417726, was 87418259"') in 2 cases:
1) during 5 minutes after commit while builds are still uploaded to the releases but index.json is not yet updated
2) if some builds were uploaded and some builds failed to upload, then for platforms that were uploaded successfully unstable installation will be broken until we fix the builds
When we add hours and minutes to the tag, that should be enough to make urls unique for each build avoiding this problem. If needed we can came up with some mechanism to clean up all builds besides the last one.
Underlay filename wasn't sanitized when saved to 'cache' and when added to .svg but later when it's moved from 'cache' to 'drawings' folder the filename was regenerated using sanitized main .svg filename leading to issues.
By mistake it was always using currently edited external style, even if external style was not edited anymore and even if it was different style currently edited.
While drawing the polyline the user can press:
- F to flip the wall. Changes the direction sense.
- O to change the offset type between Exterior, Center and Interior. This will change the offset from reference line.
Those attributes can be verified in the Object Materials panels of each wall after they are created.
Basically those two properties (Relating Whole and Related Part) are now showing only objects that are valid for the assigning an aggregate - https://i.imgur.com/bqQsXhG.png
otherwise it seems it was never closed (even if you open some other ifc project in current blender session) and .sqlite file is locked from being deleted (at least on Windows)
Was resulting in error below for meshes with > 1000 face indices (>~333 tris)
last_error: Traceback (most recent call last):
File "C:\Users\Ryan Schultz\AppData\Roaming\Blender Foundation\Blender\4.2\extensions\.local\lib\python3.11\site-packages\bonsai\bim\ifc.py", line 408, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\Ryan Schultz\AppData\Roaming\Blender Foundation\Blender\4.2\extensions\.local\lib\python3.11\site-packages\bonsai\bim\module\project\operator.py", line 1858, in _execute
ifc_file = ifcopenshell.open(queried_obj["ifc_filepath"])
~~~~~~~~~~~^^^^^^^^^^^^^^^^
KeyError: 'bpy_struct[key]: key "ifc_filepath" not found'
Ping @Gorgious56 as perhaps you might be interested in this weird case - it seems passing Operator with context_pointer_set makes it OperatorProperties which doesn't have original __annotations__.
- Corrected the alignment of the Type Manager preview in tool header
- Improved click-ability of the thumbnail icon in the Type Manger preview in the side bar
The first implementation was based directly on Blender increment snap.
The new one works just by rounding distance and angle to a better number.
The rounding is still contextual to the window zoom.
The function was only passing the closest option, but this is not always
the best option. It still defaults to the shortest, but now the user can
select the other option by cycling through snapping points with the "M"
key.
- Press "L" while the tool is active to lock axis to the current angle.
- Press "Shift" + "Mouse Wheel" up or down to adjust the angle.
If using with the measure tool it will only work if you have a plane method selected.
* Remove length from wall tool
No longer relevant with the new draw method
* Selected Type Preview and Launch Type Manager
Redesign of thumbnail preview
* Update to Type Manager
Moved Create new type to end of paginated loop
Added Predefined type label under thumbnail
* No Types Found
Modified to match Types Manager tiles
Quick Default icon added
* No Types Found
Modified to match Types Manager tiles
Quick Default icon added
* Description in thumbnail linked to value
---------
Co-authored-by: tim <tjrhyder@gmail.com>
- Single: draws a single measurement that shows the lines and dimensions for x, y and z.
- Polyline: it's how it was already working
- Area: show the area value in the input panel and creates the polygon shape that represents the area. Only works for coplanar points.
The mode can be chosen by clicking the option icon in the workspace menu.
This is a very common complaint: missing elements on large models. It seems as better default is to warn the user that the model is large, but give them the option to filter prior to loading the model.
KeyError: 'bpy_struct[key]: key "ios_edges_item_ids" not found'
It's still not possible to edit curve representations, just fixing an error trying to import representation items.
Apparently Blender is strict that items should be tuples, not lists. Interestingly enough, though it did failed to register, operator still was working fine.
Example error:
TypeError: EnumProperty(...): expected a tuple containing (identifier, name, description) and optionally an icon name and unique number
ValueError: bpy_struct "BIM_OT_select_container" registration error: 'selection_mode' EnumProperty could not register (see previous error)
Apparently after recent changes with how types are added, it's now required to add a representation for them explicitly instead of just marking them as edited
It was reflecting active material but it was only updated when ifc UI data was refreshed (usually after some IFC operators), now it should be always up to date.
Previously it would fail e.g. if IfcFillAreaStyle was already assigned to the material but UI was showing errors because enum items contained only surface styles.
* Conditions added to show Edit vs Create for each Tool
* Void conditions revised
Apply Voids now only show when number of selected objects ==2, and one of them is a void
* Conditions added to show create vs edit for each tool
example: to edit a wall -> the wall tool must be the active tool and a wall must be selected, to add a wall -> the wall tool must be the active tool and a wall must not be selected
* Extend wall to slab icons added
dark and light mode versions added
* Reset vertex icons added
dark and light mode version added
* Type Manager Preview added
---------
Co-authored-by: tim <tjrhyder@gmail.com>
Co-authored-by: Dion Moult <dion@thinkmoult.com>
Even if user decide not to delete style during unlinking, we will create a material copy and relink style to it, so it will be safe to assume that each surface style is always linked to blender material.
Added a debug and run script to the documentation section. Added a tip and clarifies how to roll back to a previous version, when running unstable release.
In my case it was IfcIndexedPolyCurve with just 1 vertex. Though kernel->convert does return a boolean value to indicate if conversion was successful, we never used it - so it failed silently leading to segfault later on.
Now we also stop processing those polycurves during mapping stage but checking kernel->convert result still might be useful in some other cases.
Noticed in the file attached in #5474 (note that it's not the issue that's causing the segfault), it was failing due IFCSWEPTDISKSOLID's IFCINDEXEDPOLYCURVE having just 1 point, resulting in something like '[Error] [ 14:07:00] No segment successfully converted: 80=IfcIndexedPolyCurve(79,$)'.
Not sure if this kind of polyline is considered valid in IFC but it doesn't trigger any validation errors, so I guess it shouldn't stop IFC project from loading in BBIM.
Recently released Python 3.12.6 dropped support for Mac OS 10.9-10.12 - https://www.python.org/downloads/release/python-3126/
And Python 3.12.6 is the one that's used by compiled builds from now - in our case it was fonttools (used by ezdxf) that compiled binaries and it's using cibuildwheel which is also started using Python 3.12.6 as it's released.
Replaced "product=wall" with "products=[wall]" in:
# Place our wall in the ground floor
ifcopenshell.api.spatial.assign_container(model, relating_structure=storey, products=[wall])
As code example fails otherwise
As hinted in the TODO note for linking collections safely, try-catch is actually much faster. This brings it to 10% faster than 0.7 speeds. Previously the regression made it 10x slower.
- add "Linked rate" column,
- assign a cost rate to multiple cost items by selecting linked rate cells & clicking the "assign" button on the chosen cost rate
#5369
Mentioined in #5457
Errors such as:
Exception ignored in: <function file.__del__ at 0x7f4cfcc9f920>
Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/ifcopenshell/file.py", line 284, in __del__
del file_dict[self.file_pointer()]
^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/ifcopenshell/file.py", line 444, in __getattr__
return getattr(self.wrapped_data, attr)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'file_pointer'
Example error messages now:
RuntimeError: Failed to process shape. Product: #3855=IfcSite('3ZGD7y6S5209$mGLi_sPll',#20,'Surface:411452',$,$,#3854,#3850,$,.ELEMENT.,(42,12,46,799999),(-71,-2,0,-599999),0.,$,$), representation: #3849=IfcShapeRepresentation(#93,'FootPrint','Curve2D',(#3841,#31427,#31471,#45600,#45628))
RuntimeError: Failed to process shape. Instance: #3841=IfcPolyline((#3800,#3801,#3802,#3803,#3804,#3805,#3806,#3807,#3808,#3809,#3810,#3811,#3812,#3813,#3814,#3815,#3816,#3817,#3818,#3819,#3820,#3821,#3822,#3823,#3824,#3825,#3826,#3827,#3828,#3829,#3830,#3831,#3832,#3833,#3834,#3835,#3836,#3837,#3838,#3839,#3840,#3800))
Previously, the way to add an IFC element would be to 1) create a blender element then assign or 2) activate bim tool, launch type manager, click create type, fill out field, or 3) using various menus for parametric stuff like stairs. This is now the new proposed way to do things and everything will consolidate to here - and it allows you to directly create non-mesh geometry immediately too!
- add setting options to hide/show columns
- Cost classification column
- New Cost Classification Form to display active classification library, add/delete classification references
* Updates to some icons, new icons added for CAD tools, and new layout for CAD tools
* updates to icons
* WIP changes to BIM Tool UI
* Separated BimToolUI, added Draw wall icon
Separated BimToolUI into AddElementUI and ModifySelectedUI classes, added Draw wall icon
* Voids back in UI
* WIP UI tools and icons
partial support for draw wall tool added
* 40d95c294 WIP UI tools and icons
Minor tweaks
* WIP on profile-axis-ui
* WIP on profile-axis-ui
* Added Merge, Split, Rotate, and Flip back into wall edit tools
* Separators added
* Restore commented out code in CAD workspace.py
---------
Co-authored-by: tim <tjrhyder@gmail.com>
Co-authored-by: Dion Moult <dion@thinkmoult.com>
Now we're going to use bar ids instead their detalization ids (e.g. expanded task ids, task ids, milestone ids) because it is the way how it works in the software pp file originates from.
When loading an ifc project it will search it for bsdd ifcclassifications and will set one as active if finds it, so now it will be a bit more convenient working with bsdd classification assignment as it active bsdd previously was always getting reset when you reload ifc file.
Example - https://imgur.com/a/stlJqwm
The input value can be a string or number, but 1.0 schema only accepts strings. The processing of int and float was missing. The casting to string might no longer be needed in IDS 1.1. https://github.com/buildingSMART/IDS/issues/342
Viewing changes necessarily sets object colours and switches the
viewport to object colouring. Previously most other operations switched
the viewport 'back' to material colouring. This was an annoyance for
users whose workflow makes use of manual object colouring.
Now only operators that reload the model (revert, switch and merge), the
refresh button, and attempting to view changes with the current revision
switch to material colouring.
Couldn't find if we actually use it anywhere in the code but better to keep it safe, especially because it's exposed to Python and by accident user might provide some too bit integer instead of a string.
Noticed that prop_value could be other types besides str (e.g. int when using context menu on UIList item) and we can skip them to avoid redundant api calls.
By accident noticed that `declaration_by_name(-1)` crashes python. If we use size_t it will just throw an error like below
NotImplementedError: Wrong number or type of arguments for overloaded function 'schema_definition_declaration_by_name'.
Possible C/C++ prototypes are:
IfcParse::schema_definition::declaration_by_name(std::string const &) const
IfcParse::schema_definition::declaration_by_name(size_t) const
`repo.path` is always setup using current platform type of slashes since it's normalized by get_path_dir but when we pass a path to index.add it's not normalized leading to errors
As on first iteration it may remove only unused material sets and on the next iteration it will be able to remove IfcMaterials that were not used anywhere else besides those material sets.
It wasn't working anymore after 27a2c94. Importing IFC file as just geometry should be a separate option from 'should_start_fresh_session' and I've added it as 'import_without_ifc_data' property during project load (now both options are also available for users in project load dialog - https://i.imgur.com/Fvbv4Dy.png)
Previously it was using taxonomy::cast that was also trying to upgrade current item - due to the strictness of dispatch_conversion (it works only if type matches exactly because of `N == item_kind`) upgrades will never succeed so we may just skip them.
Polyline Wall and Measure tool now share most of the logic by inheriting from a base class PolylineOperator.
This should facilitate the implementation of other tools that use polyline.
Apparently it's part of #4832, previously creating a shape for IfcTypeProduct with literals would error, we ignored the error and type product end up being an empty. After 0.8.0 create_shape for literals creates an empty shape (0 verts) and we need to skip it, otherwise there would be an empty mesh representaiton.
- assign cost item products from Blender selection
- display list of selected products ( with available quantities, subtotals, and highlighting already assigned products)
- display and refresh assigned products
- highlight loaded cost schedule
to avoid issues importing csvs that used it to extract correct predefined type
Example warning:
WARNING. Assigning 'predefined_type' is not yet supported. Skipping value 'ELECTRICACTUATOR' for element: '#89=IfcActuator('1RjCen6XL6WAdhz7HumLFU',$,'Actuator.002',$,$,#179,$,$,.ELECTRICACTUATOR.)'.
Traceback
Traceback (most recent call last):
File "\bonsai\bim\module\type\operator.py", line 63, in execute
return IfcStore.execute_ifc_operator(self, context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\ifc.py", line 408, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\type\operator.py", line 78, in _execute
active_context = active_representation.ContextOfItems
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'ContextOfItems'
Error: Python: Traceback (most recent call last):
File "\bonsai\bim\ifc.py", line 408, in execute_ifc_operator
result = getattr(operator, "_execute")(context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\bonsai\bim\module\type\operator.py", line 78, in _execute
active_context = active_representation.ContextOfItems
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'ContextOfItems'
It was leaving an orphaned pset if 'products' was a list of all elements pset is assigned to. Then it would create a new pset for each element leaving original pset orphaned.
In IFC it's possible for a property set to be assigned to multiple elements and which may lead to confusing behaviour when you edit a pset on one element and other element seems to get edited too.
Which makes it worse is that that it is possible that some software is might be doing this unintentionally when exporting IFC (as some sort of optimization as storing 1 is more optimal than n copies of it).
So now there are some tools in Bonsai and in IfcOpenShell to handle the shared psest:
1) Indication that property is shared - https://imgur.com/a/9dd3jST (similar to how Blender indicates ID data-block users). You can click on it to "unshare" the pset - a new copy for the pset will be created and it's going to be linked only to the active object.
2) api pset.unshare_pset method that does the same. And util.element.get_elements_using_pset method that encapsulates schema differences and different approaches for occurrences/types.
3) ifcpatch recipe 'UnsharePsets' that's making all property sets in the IFC file to have just 1 element that's using them. You can limit the affected elements by providing query.
ifcpatch recipe is also available in Bonsai - https://i.imgur.com/aOCx7HI.png
Kind of experimental. The idea is that it will reassign class e.g. from IfcWindow to IfcWindowType (or vice versa) and will keep the IFC representations and property sets.
Example - https://imgur.com/a/X7MHR0s
To avoid creating invalid material profile sets, representation items. Currently the only inverse that is allowed before removal is profile properties.
Example - https://i.imgur.com/UobI4pL.png
1) For psets there was an issue - it was instantiating abstract IfcProfileProperties
2) For materials it was only able to instantiate IfcExtendedMaterialProperties though there are other pset types too in ifc2x3.
It was failing on Windows with something like "ValueError: path is on mount 'L:', start on mount 'C:'".
I guess it doesn't really occur in bonsai as sheets are typically stored next to the titleblocks (we used to use absolute paths awhile ago and now everything is relative to the ifc project)
This functions was built to work with the input calculations for the polytool.
Part of its code was being duplicated in another part of the code. The refactor
keeps most of the calculation in the same place.
As currently we do not support multilayer/profile geometry and therefore we can't correctly represent them when they have multiple styles. As a workaround we'll use the material/style from the first layer and apply it to the entire object.
This function uses another method to calculate the closest points in two edges that are not parallel.
The `intersect_line_line` from mathutils were not getting good results when using orthogonal view.
Added more control on how the snapping system selects the snapping point, and "M" key now updates the input panel.
Added `intersect_edge_2` to Cad tool
Revert "Improved snap "stickiness"."
This reverts commit 85e67d3996ab8002b4cef7c0b859e9da251d39bf.
A few snap and raycast changes to improve the snap feel
Restore old behaviour (new behaviour was introduced by accident in 786b796) where it would unhide all objects instead of keeping only objects in the current drawing.
Mentioned in #5233
Also some refactor:
- checking IfcProduct instead of IfcProject. Not sure when this issue occur in general but if it occurs then it might fail for other non-IfcProducts too, not just IfcProject
- small performance optimization
The `intersect_line_line` were returning points that were outside
the object edges, so it was added a function to check if the point
return is on the object edge.
Previously it would fail with:
PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'bonsai\\bim\\data\\cache\\988c190483ae6bab7cf6b00cee78694d.h5'
Those two panels were requiring drawing's camera to be active object which was a bit confusing since "Active Drawing" section above worked fine without selecting camera explicitly.
https://i.imgur.com/wcmA73Z.png
You can now lock into an axis while snapping to an object.
The result will be to closer point. In the future, there should
be a way for the user to select between different points.
Approach in 358c834 didn't worked out because annotations are assigned to the drawing after assign_class and we can't check in assign_class whether it's going to be part of the drawing.
This function was hard coded to round length angle, now it has an option.
I assume it was originally built like this because it was supposed to
work only with Grease Pencil.
By pressing "M", the user can now cycle through active snapping points when
there are many options close to each other. Still lacks a text to give
information about the active point.
removed absolute positioning as it made spreadsheets on top of each other
and made the spreadsheet's width and height as full screen as possible. also, add a separating margin between two spreadsheets.
Previously it wasn't handled and would always result in errors. It seems the only way to load point cloud representation to BBIM was by loading IFC file and only if this is the only representation of the element.
See #5214 for details, can't rely on 'bpy.context.preferences.addons' as it may contain a missing addons too (if addon was manually uninstalled and it won't be able to load but it will still appear in 'addons').
Noticed that added drawing camera was linked to the scene collection and after collector.assign it was linked to the 2 collections instead of just 1 - it's own drawing collection and the scene collection.
The remove *.ifc Association removes any stale Linux BlenderBIM
associations as well as Bonsai associations, this should now give a
clean upgrade path. Closes#5212
Apparently we can't trust `obj.data == None` to identify empties since in Blender there are empties that can use `bpy.types.Image` as their `obj.data`.
There was a problem with assigning uv workflow if object wasn't originally from IFC - saving uv requires object to have IfcSurfaceStyle but assigning IfcSurfaceStyle would reload representation discarding UV.
So, the current workflow for saving non-ifc objects uv to ifc:
1) Assign class to ifc object
2) Create IfcSurfaceStyle with a texture (currently also requires a rendering style)
3) Use "Save UV To Style"
Example - https://imgur.com/a/GZqo38M
extracting files in general is also improved - now it's possible to extract a file that was just added to bcf in memory and not yet saved to the disk.
Update model.py
bcfxml doesn't actually have "author" field - this field is not stored in bcf, it's just needed as an UI for author for newly added comments and topics.
Previously you had to type in exactly manually.
Besides the dropdown it should also show now a topic description in the dropdown items tooltips.
Removed "bim.add_bcf_related_topic" poll as it should be now covered by the enum.
Example - https://imgur.com/a/xhgcGWI
this solves the issue when a user has not killed a previous server so the port of that server wasn't removed from the pid file, which would make has_started return true even if server hasn't started yet.
previously it was just checking if there was a server running but not if blender was connected to a server, this was being handled in the operator itself which isn't the best.
Since 06c8a9e there's no longer issue with loaded binaries preventing to reload the addon on Windows, which works for 4.2.0 and in 4.2.1 we'll have https://projects.blender.org/blender/blender/issues/125049 which essentially does the same, moves the binaries out of the way.
Upgrade also works fine and we have a warning from 3d4c85c to indicate that user should restart Blender after upgrade.
It wasn't considering that self._topics could be an empty dict because there are no topics and it would also reload it. Now we have None value to distinguish when it actually wasn't loaded before.
Same for viewpoints, reference files and document references.
I've also imade mplementations identical/more similar between v2 and v3.
1) bpy.ops.object.hide_view_set doesn't actually work with context provided selected objects - it's checking whether they're actually selected
2) bpy.ops.object.hide_view_clear wasn't using select=False, selecting revealed objects which might get in the way later.
3) used new api for getting visibility settings
1) added more high level api for getting/setting selected/visible elements in bcf
2) in bbim when you add a new viewpoint it will remember currently selected and visible objects:
https://imgur.com/a/2pZ79Z2
Previously they weren't editable if they were not defined in .bcf already, so previously it was kind of requiring them to be defined by some other software, not BBIM.
Also changed couple prop names to make them more clear to users.
disabling it as it currently won't allows users to create a different viewpoint for the same IFC element (in the same or other bcf file) and also entity position can change between `build_viewpoint` calls
According to docs bcf is using meters as their length unit - https://github.com/BuildingSMART/BCF-XML/tree/release_3_0/Documentation
We should create some kind of common code folder to keep this sane...
Example warning:
WARNING. Snapshot with viewpoint guid '6aff8c6e-216c-48c9-8120-3101816def3e' won't be saved to bcf. Only snapshot data (VisualizationInfoHandler.snapshot) is provided and snapshot filename (ViewPoint.snapshot) is missing.
1) it was throwing errors if object was linked
2) bim.override_mode_set_edit wasn't considering that toggle_edit_mode might not work out and was changing BIMGeometryProperties.mode though it wasn't needed.
Previous method uses bounding box which could be wrong if the app (Revit) doesn't extrude in +Z and has a slope (e.g. most pipes). Also Revit tends to use rectangle profiles not for the cross section, but instead for the footprint which is crazy.
2024-08-08 15:13:34 +10:00
1615 changed files with 178890 additions and 167124 deletions
| bsdd | Library to query the bSDD API | LGPL-3.0-or-later | [](https://pypi.org/project/bsdd/) |
| ifc2ca | Utility to convert IFC structural analysis models to Code_Aster | LGPL-3.0-or-later |
| ifc4d | Convert to and from IFC and project management software | LGPL-3.0-or-later | [](https://pypi.org/project/ifc4d/) |
BlenderBIM is an open-source project, and its development is driven by the contributions of a dedicated community of developers,
architects, engineers, and enthusiasts. If you're interested in contributing to the project, whether by submitting bug reports,
suggesting new features, or contributing code, your involvement is highly encouraged and appreciated.
This part of the documentation covers various aspects of the BlenderBIM development process, including:
-:doc:`Writing User Documentation </contribute/writing_docs>`
-:doc:`Translations and Internationalisation </contribute/translations>`
- Contributing Code
- :doc:`Installation and Setting up a Development Environment </devs/installation>`
- Understanding the Project Structure and Codebase
- :doc:`Hello, World! </devs/hello_world>`
- :doc:`Undo System </devs/undo_system>`
- :doc:`Code Style Guidelines and Best Practices </devs/code_style>`
..
- :doc:`Submitting Pull Requests and Contributing Code </devs/contributing_code>`
-:doc:`Testing and Quality Assurance </devs/running_tests>`
- :doc:`Running Tests </devs/running_tests>`
-:doc:`System Support and Multiplatform Compatibility </devs/system_support>`
-:doc:`User Experience and User Interface Guidelines </devs/ux_guidelines>`
The BlenderBIM Developer Documentation is a living resource maintained by the core development team and the open-source community. It serves as a central hub for developers who want to get involved in the project, ensuring a consistent and efficient development process.
By contributing to BlenderBIM, you'll not only be helping to improve and expand the capabilities of this powerful open-source BIM authoring platform but also be part of a vibrant community driving innovation in the AEC industry.
After you modify your code in the Git repository, you will need to restart
Blender for the changes to take effect.
The downside with this approach is that if a new dependency is added, or a
compiled dependency version requirement has changed, or the build system
changes, you'll need to fix your setup manually. But this is relatively rare.
Reviewing the Makefile history, `here <https://github.com/IfcOpenShell/IfcOpenShell/commits/v0.8.0/src/blenderbim/Makefile>`__, is one quick way to see if a dependency has changed.
The BlenderBIM interface extends Blender's powerful 3D environment with specialized tools for Building Information Modeling (BIM). This guide will help you navigate the BlenderBIM interface and understand its key components.
BlenderBIM Workspace
====================
When you start Blender with BlenderBIM installed, you can switch to the BlenderBIM workspace
by clicking on the BIM workspace tab at the top of the Blender window.
This workspace is preconfigured with the most commonly used panels and tools for BIM workflows.
Key Interface Elements
======================
Properties Extended
-------------------
The Properties Editor is extended to provide IFC-specific properties and settings.
- Scene
- Project Information
- Tool
- Object
BIM Toolbar
-----------
The BIM toolbar, typically located at the left of the 3D Viewport, contains shortcuts to frequently used BIM tools:
- Explore tool
- Create various building elements
- Create annotations, measure distance and angles, calculate volumes, etc
IFC Tree View
-------------
Found in the Outliner, the IFC Tree View displays the hierarchical structure of your BIM model:
- Spatial Structure (Site, Building, Storey)
- Building Elements
- Types
Customizing the Interface
=========================
BlenderBIM respects Blender's highly customizable interface. You can:
- Rearrange editors and panels
- Save custom workspace layouts
- Create custom shortcuts for BIM operations
Tips for Efficient Use
======================
1. Familiarize yourself with IFC classes and their properties.
2. Utilize Blender's search function (F3) to quickly access BlenderBIM tools.
Next Steps
==========
Now that you're familiar with the BlenderBIM interface, you have two main paths to continue your learning:
1. Proceed with IFC Modeling:
If you're ready to start creating and working with BIM models, you can move on to:
results, and construction types. The data may be relevant to multiple
disciplines, such as architecture, engineering, and construction.
..note::
BIM data is very different from a regular 3D model. In fact, geometry is
optional, and most data is non-geometric. This means that it is not simply a
3D format that you can import or export from and expect meaningful results.
**Industry Foundation Classes**, or **IFC**, is an international standard for
**BIM**. **IFC** is the most well-established open digital language for our
built environment. Most software will be able to describe their **BIM** data
using **IFC**. Most commonly, **IFC** models will be shared as a ``.ifc`` file.
For example, **IFC** will define a wall as an object that can have a name,
construction type, and quantities. **IFC** will also describe that a wall that
be associated with a location, like a building storey, or have an associated
cost item in a schedule.
When you use the BlenderBIM Add-on, you will be able to view and create **BIM**
objects and relationships using the **IFC** standard.
Things you can do
-----------------
The BlenderBIM Add-on is designed to be a comprehensive and truly native
BIM authoring platform. Its capabilities include a wide range of tasks
and workflows typically found across various BIM and CAD software, costing
programs, scheduling tools, and simulation applications. While not
an exhaustive list, some of the key things you can do with BlenderBIM include:
- View and explore IFC models, including spaces, properties, and relationships
- Edit and extract attributes, properties, and metadata directly from IFC data
- Move, rotate, and modify the geometry of objects while preserving IFC semantics
- Create new objects using predefined library elements or custom parametric types
- Manage classification systems, document references, and link to external libraries
- Generate 2D drawing views like plans, sections, and elevations with customizable annotations
- Investigate and edit structural analysis models with support for various steel profiles
- Model and manage complex distribution systems like HVAC, plumbing, and electrical (MEP)
- Create construction schedules, perform critical path analysis, and generate sequence animations
- Derive quantities from model elements and create cost schedules with formulas
- Perform clash detection and coordinate models, managing issues across disciplines
- Integrate non-geometric data like costing, scheduling, and asset management
Roadmap and Upcoming Features
-----------------------------
While BlenderBIM already offers a comprehensive set of BIM authoring capabilities,
the development team and community are continuously working to expand its functionality and improve existing workflows.
Some of the most ambitious features and enhancements on the roadmap include:
Usability and Workflow Enhancements
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- Improve user-friendliness to equip average "Joe" with capabilities to model simple projects like single family homes, making BlenderBIM a viable alternative to SketchUp.
- Make BlenderBIM more approachable for users across different skill levels, from advanced BIM experts to architects working on smaller residential projects.
Improved Drawing and Documentation Workflows
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- Continued enhancements to the drawing generation capabilities, including better dimensioning tools and support for associative dimensions linked to model geometry.
- Advanced annotation tools with customizable tags, callouts, and the ability to define reusable annotation styles and templates.
- Streamlined sheet layout management and improved integration with external tools like Inkscape for creating title blocks and sheet compositions.
- Continued improvement of BlenderBIM's ability to generate comprehensive documentation sets, including drawing annotations, door schedules, wall schedules, and material tagging.
Expanded Parametric Modeling Capabilities
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- Native modeling tools for complex building elements like curtain wall systems and facade panels.
- Expanded libraries with support for country-specific and manufacturer-provided object types.
- Enhanced support for multi-story modeling, enabling efficient duplication and coordination of building elements across various levels.
- Expanded parametric relationships and automating more common BIM tasks to reduce manual effort and enhance productivity.
Enhanced Coordination and Collaboration Features
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- Robust model merge and coordination workflows, with detailed clash detection capabilities across different disciplines and data sources.
- Built-in support for version control and model sharing using Git repositories, enabling better team collaboration and change tracking.
The development roadmap is continuously updated based on user feedback, industry requirements,
and contributions from the open-source community. By embracing an open and collaborative approach,
BlenderBIM aims to push the boundaries of what's possible in BIM authoring,
ensuring that it remains at the forefront of innovation in the AEC industry.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Dion Moult
