Implement v3.0 of the metamodel specification #174
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The problem was, that with V3.0RC02, an Identifer is a string. Therefore the distinction between File-Identifier and AAS-Identifier could not be made anymore. As a solution, remove the option to input the already hashed File-Identifier.
implement global and model references rename AASReference to ModelReference add Reference/referredSemanticId remove redundant definitions from XML schema (AAS and IEC61360)
rename SpecificAssetId/key to SpecificAssetId/name change type of SpecificAssetId/externalSubjectId from Reference to GlobalReference
also make it an Iterable
make CONCEPT_DICTIONARY a private member since it has been removed from the spec
reflect this change in the json/xml schemata add missing comments to protected KeyTypes enum members
In V3.0RC02 DataTypeDef has been split into DataTypeDefRdf and DataTypeDefXsd. Since DataTypeDefRdf only consists of rdf::langString and currently isn't used anywhere in the DotAAS specification, just rename DataTypeDef to DataTypeDefXsd.
In V3.0RC02 the attributes `allowDuplicates` and `ordered` were removed from `SubmodelElementCollection`. Additionally the `semanticId` of contained elements is no longer unique. Because of that, the extra classes `SubmodelElementCollectionOrdered`, `SubmodelElementCollectionUnordered`, `SubmodelElementCollectionOrderedUniqueSemanticId` and `SubmodelElementCollectionUnorderedUniqueSemanticId` aren't needed anymore. Also ignore two tests which will be modified and re-enabled later.
The `item_add_hook` is a function that is called whenever items are added to the `NamespaceSet`, even on initialization. It is supplied to the `NamespaceSet` on initialization and allows the model class to check constraints whenever a new item is added.
In turn re-enable the unittests that were disabled in 639b4b9.
Improve/coverage/model.datatypes
This commit makes `Operation` inherit from `UniqueIdShortNamespace`, to implement Constraint AASd-134: For an Operation, the idShort of all inputVariable/value, outputVariable/value, and inoutputVariable/value shall be unique. In the DotAAS spec, the attributes `inputVariable`, `outputVariable` and `inoutputVariable` of `Operation` are defined to be a collection of `OperationVariable` instances, which themselves just contain a single `SubmodelElement`. Thus, the `OperationVariable` isn't really required for `Operation`, as the `Operation` can just contain the `SubmodelElements` directly, without an unnecessary wrapper. This makes `Operation` less tedious to use and also allows us to use normal `NamespaceSets` for the 3 attributes, which together with the `UniqueIdShortNamespace` ensure, that the `idShort` of all contained `SubmodelElements` is unique across all 3 attributes. Aside this, the examples are updated since `SubmodelElements` as children of an `Operation` are now linked to the parent. This prevents us from reusing other `SubmodelElements` as `OperationVariables` as it was done previously, since each `SubmodelElement` can only have one parent. Fix #146 #148
…e of a conflict Currently a string constraints decorator silently overwrites existing attributes in case of a naming conflict. This behavior is changed such that the decorator checks for existing attributes beforehand and raises an exception in case of a conflict. Futhermore, tests for this behavior are added.
- Refactor `NamespaceSet.add()` as too big - Extract some methods, in particular `_execute_item_del_hook` to avoid code duplication - As we check different constraints for uniqueness in the namespace, I defined `ATTRIBUTES_CONSTRAINT_IDS`. The dict will be used when throwing exception. The solution with the dict is temporary, we need another solution here. - Use correct constraint ids for each NamespaceSet in tests, use 000 if not constraint id is suitable
These went unnoticed previously since the JSON schema doesn't prohibit additional attributes. Now that the `modelType` is defined as a constant string per definition, a `Property` object only matches against the corresponding definition, allowing more strict validation, that catches errors such as these null values.
This removes the two XSD datatypes - `xs:dateTimeStamp` - `xs:dayTimeDuration` from the `DataTypeDefXsd` in the JSON schema, as they were removed from the specification in v3.0
This commit remanes `valueDataType_t` from the XSD to `valueDataType`. Furthermore, it adds a missing `<reference>` tag around the `refersTo` References of `Extension`.
The current order of the elements in `SubmodelElementList` was wrong. This updates the order.
Version 3.0 of the spec removes the attribute `kind` from `SubmodelElement`s. While we already implemented this, it was still missing in the XSD Schema, as well as the examples. This commit fixes that.
Currently, the XML deserialization missed the `<aas:refersTo>` wrapper around the single references inside `Extension.refers_to`. This commit fixes that.
Currently, `Extension.refers_to` is declared as a `Iterable[Reference]`. This implies, that we can not necessarily check, whether or not the attribute is empty or not. This creates a problem with the XML serialization, since the `<aas.refersTo>` element should only appear if there is at least one `Reference` inside. This commit changes the `Extension.refers_to` to be a set of `Reference`s, as well as adapting a more clear check whether or not the attribute is empty in `adapter.xml.xml_serialization`.
remove constraint, that if the second part of lang tag exists, it must be a region See #157
Previously, the compliance tool would ignore `Extensions` of objects. However, during the update to V3.0 and the changes to the JSON/XML formats, we noticed that the adapters aren't properly tested this way. The validation of `Extensions` can be disabled via the flag `--dont-check-extensions` or the keyword argument `check_extensions` to `AASDataChecker`, since many users are probably only interested in comparing the main model, without any extensions.
These comments were mistakenly re-added in a previous merge of main into improve/V30.
Frosty2500
pushed a commit
to rwth-iat/basyx-python-sdk
that referenced
this pull request
Aug 22, 2024
Implement v3.0 of the metamodel specification For the detailed documentation of the changes, please refer to the [prior PRs](https://github.com/eclipse-basyx/basyx-python-sdk/pulls?page=4&q=is%3Apr+base%3Aimprove%2FV30)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR merges the v3.0 implementation of the metamodel specification into
main.For the detailed documentation of the changes, please refer to the prior PRs