Merge pull request #2044 from microbiomedata/metagnome-input-id

add metagnome input id pattern constraint
microbiomedata · Jun 5, 2024 · 46750c0 · 46750c0
2 parents b2dc818 + fbeb467
commit 46750c0
Show file tree

Hide file tree

Showing 8 changed files with 29 additions and 7 deletions.
diff --git a/nmdc_schema/nmdc.py b/nmdc_schema/nmdc.py
@@ -1,5 +1,5 @@
 # Auto generated from nmdc.yaml by pythongen.py version: 0.0.1
-# Generation date: 2024-06-04T21:02:10
+# Generation date: 2024-06-05T15:38:19
 # Schema: NMDC
 #
 # id: https://w3id.org/nmdc/nmdc
@@ -5707,17 +5707,23 @@ class MetagenomeSequencingActivity(WorkflowExecutionActivity):
     id: Union[str, MetagenomeSequencingActivityId] = None
     execution_resource: str = None
     git_url: str = None
-    has_input: Union[Union[str, NamedThingId], List[Union[str, NamedThingId]]] = None
     type: str = None
     started_at_time: str = None
     ended_at_time: str = None
+    has_input: Union[Union[str, NamedThingId], List[Union[str, NamedThingId]]] = None
 
     def __post_init__(self, *_: List[str], **kwargs: Dict[str, Any]):
         if self._is_empty(self.id):
             self.MissingRequiredField("id")
         if not isinstance(self.id, MetagenomeSequencingActivityId):
             self.id = MetagenomeSequencingActivityId(self.id)
 
+        if self._is_empty(self.has_input):
+            self.MissingRequiredField("has_input")
+        if not isinstance(self.has_input, list):
+            self.has_input = [self.has_input] if self.has_input is not None else []
+        self.has_input = [v if isinstance(v, NamedThingId) else NamedThingId(v) for v in self.has_input]
+
         super().__post_init__(**kwargs)
 
 
@@ -11737,6 +11743,9 @@ class slots:
                    model_uri=NMDC.MetagenomeSequencingActivity_id, domain=MetagenomeSequencingActivity, range=Union[str, MetagenomeSequencingActivityId],
                    pattern=re.compile(r'^[a-zA-Z0-9][a-zA-Z0-9_\.]+:[a-zA-Z0-9_][a-zA-Z0-9_\-\/\.,]*$'))
 
+slots.MetagenomeSequencingActivity_has_input = Slot(uri=NMDC.has_input, name="MetagenomeSequencingActivity_has_input", curie=NMDC.curie('has_input'),
+                   model_uri=NMDC.MetagenomeSequencingActivity_has_input, domain=MetagenomeSequencingActivity, range=Union[Union[str, NamedThingId], List[Union[str, NamedThingId]]])
+
 slots.ReadQcAnalysisActivity_id = Slot(uri=NMDC.id, name="ReadQcAnalysisActivity_id", curie=NMDC.curie('id'),
                    model_uri=NMDC.ReadQcAnalysisActivity_id, domain=ReadQcAnalysisActivity, range=Union[str, ReadQcAnalysisActivityId],
                    pattern=re.compile(r'^[a-zA-Z0-9][a-zA-Z0-9_\.]+:[a-zA-Z0-9_][a-zA-Z0-9_\-\/\.,]*$'))

diff --git a/nmdc_schema/nmdc_materialized_patterns.schema.json b/nmdc_schema/nmdc_materialized_patterns.schema.json
@@ -6033,7 +6033,7 @@
                 "has_input": {
                     "description": "An input to a process.",
                     "items": {
-                        "pattern": "^(nmdc):dobj-([0-9][a-z]{0,6}[0-9])-([A-Za-z0-9]{1,})$",
+                        "pattern": "^(nmdc):(bsm|procsm)-([0-9][a-z]{0,6}[0-9])-([A-Za-z0-9]{1,})$",
                         "type": "string"
                     },
                     "type": "array"

diff --git a/nmdc_schema/nmdc_materialized_patterns.yaml b/nmdc_schema/nmdc_materialized_patterns.yaml
@@ -20809,6 +20809,12 @@ classes:
         structured_pattern:
           syntax: '{id_nmdc_prefix}:wfmsa-{id_shoulder}-{id_blade}{id_version}$'
           interpolated: true
+      has_input:
+        name: has_input
+        pattern: ^(nmdc):(bsm|procsm)-([0-9][a-z]{0,6}[0-9])-([A-Za-z0-9]{1,})$
+        structured_pattern:
+          syntax: '{id_nmdc_prefix}:(bsm|procsm)-{id_shoulder}-{id_blade}$'
+          interpolated: true
   ReadQcAnalysisActivity:
     name: ReadQcAnalysisActivity
     description: A workflow execution activity that performs quality control on raw

diff --git a/nmdc_schema/nmdc_schema_merged.yaml b/nmdc_schema/nmdc_schema_merged.yaml
@@ -20755,6 +20755,11 @@ classes:
         structured_pattern:
           syntax: '{id_nmdc_prefix}:wfmsa-{id_shoulder}-{id_blade}{id_version}$'
           interpolated: true
+      has_input:
+        name: has_input
+        structured_pattern:
+          syntax: '{id_nmdc_prefix}:(bsm|procsm)-{id_shoulder}-{id_blade}$'
+          interpolated: true
   ReadQcAnalysisActivity:
     name: ReadQcAnalysisActivity
     description: A workflow execution activity that performs quality control on raw

diff --git a/project.Makefile b/project.Makefile
@@ -213,6 +213,8 @@ local/mongo_as_unvalidated_nmdc_database.yaml:
 		--selected-collections metagenome_annotation_activity_set \
 		--selected-collections metagenome_annotation_set \
 		--selected-collections metagenome_assembly_set \
+		--selected-collections metagenome_sequencing_activity_set \
+		--selected-collections metagenome_sequencing_set \
 		--selected-collections metap_gene_function_aggregation \
 		--selected-collections metatranscriptome_activity_set \
 		--selected-collections metatranscriptome_analysis_set \

diff --git a/...se-with-MetagenomeSequencingActivity.yaml → ...se-with-MetagenomeSequencingActivity.yaml b/...se-with-MetagenomeSequencingActivity.yaml → ...se-with-MetagenomeSequencingActivity.yaml
diff --git a/...metagenome_seequencing_activity_json.yaml → ...metagenome_seequencing_activity_json.yaml b/...metagenome_seequencing_activity_json.yaml → ...metagenome_seequencing_activity_json.yaml
diff --git a/src/schema/workflow_execution_activity.yaml b/src/schema/workflow_execution_activity.yaml
@@ -265,10 +265,10 @@ classes:
         structured_pattern:
           syntax: "{id_nmdc_prefix}:wfmsa-{id_shoulder}-{id_blade}{id_version}$"
           interpolated: true
-  #      has_input:
-  #        structured_pattern:
-  #          syntax: ".*" # todo this accounts for an extreme edge case and should nto be repeated anywhere
-  #          interpolated: true
+      has_input:
+        structured_pattern:
+          syntax: "{id_nmdc_prefix}:(bsm|procsm)-{id_shoulder}-{id_blade}$"
+          interpolated: true
 
 
   ReadQcAnalysisActivity: