Merge pull request #15 from linkml/issue-14

issue 14

Author: cmungall

schemasheets/conf/configschema.py

@@ -1,5 +1,5 @@
 # Auto generated from configschema.yaml by pythongen.py version: 0.9.0
-# Generation date: 2022-01-03T11:25:54
+# Generation date: 2022-02-02T11:11:10
 # Schema: configschema
 #
 # id: https://w3id.org/linkml/configschema
@@ -70,6 +70,7 @@ class ColumnSettings(YAMLRoot):
     suffix: Optional[str] = None
     template: Optional[str] = None
     vmap: Optional[Union[Dict[Union[str, ValueMapMapKey], Union[dict, "ValueMap"]], List[Union[dict, "ValueMap"]]]] = empty_dict()
+    inner_key: Optional[str] = None
     applies_to_class: Optional[str] = None
     applies_to_slot: Optional[str] = None
     tag: Optional[str] = None
@@ -89,6 +90,9 @@ def __post_init__(self, *_: List[str], **kwargs: Dict[str, Any]):
 
         self._normalize_inlined_as_dict(slot_name="vmap", slot_type=ValueMap, key_name="map_key", keyed=True)
 
+        if self.inner_key is not None and not isinstance(self.inner_key, str):
+            self.inner_key = str(self.inner_key)
+
         if self.applies_to_class is not None and not isinstance(self.applies_to_class, str):
             self.applies_to_class = str(self.applies_to_class)
 
@@ -225,6 +229,9 @@ class slots:
 slots.columnSettings__vmap = Slot(uri=THIS.vmap, name="columnSettings__vmap", curie=THIS.curie('vmap'),
                    model_uri=THIS.columnSettings__vmap, domain=None, range=Optional[Union[Dict[Union[str, ValueMapMapKey], Union[dict, ValueMap]], List[Union[dict, ValueMap]]]])
 
+slots.columnSettings__inner_key = Slot(uri=THIS.inner_key, name="columnSettings__inner_key", curie=THIS.curie('inner_key'),
+                   model_uri=THIS.columnSettings__inner_key, domain=None, range=Optional[str])
+
 slots.columnSettings__applies_to_class = Slot(uri=THIS.applies_to_class, name="columnSettings__applies_to_class", curie=THIS.curie('applies_to_class'),
                    model_uri=THIS.columnSettings__applies_to_class, domain=None, range=Optional[str])
 

schemasheets/conf/configschema.yaml

@@ -54,6 +54,10 @@ classes:
         range: ValueMap
         multivalued: true
         inlined: true
+      inner_key:
+        range: string
+        description: |-
+          for complex fields such as annotations that are modeled as dicts, this is used to specify which key in the dict the value should be applied to
       applies_to_class:
         description: |-
           if a value C is specified, then this column in the relevant row is interpreted as

schemasheets/schemamaker.py

@@ -16,7 +16,7 @@
     SlotDefinition, EnumDefinition, PermissibleValue, SubsetDefinition, TypeDefinition, Element
 from linkml_runtime.utils.schemaview import SchemaView
 
-from schemasheets.utils.configschema import ColumnSettings, Shortcuts, Cardinality
+from schemasheets.conf.configschema import ColumnSettings, Shortcuts, Cardinality
 from schemasheets.utils.prefixtool import guess_prefix_expansion
 
 
@@ -245,9 +245,7 @@ def merge_sheet(self, file_name: str, delimiter='\t') -> None:
                         if v is not None and v.startswith('>'):
                             v = v.replace('>', '')
                         if v:
-                            #print(f'PARSING: {v}')
                             meta_obj = yaml.safe_load(v)
-                            #print(f'PARSING: {v} => {meta_obj}')
                             table_config.add_info(k, meta_obj)
                         else:
                             logging.debug(f'Empty val {v} for {k} in {row}')
@@ -286,16 +284,31 @@ def add_row(self, row: Dict[str, Any], table_config: TableConfig):
                     if cc.maps_to == 'cardinality':
                         self.set_cardinality(actual_element, v)
                     elif cc.metaslot:
-                        if isinstance(v, list):
+                        if cc.maps_to == 'annotations' and not cc.settings.inner_key:
+                            anns = yaml.load(v[0])
+                            for ann_key, ann_val in anns.items():
+                                actual_element.annotations[ann_key] = ann_val
+                        elif isinstance(v, list):
+                            #print(f'SETTING {k} to {v}')
                             setattr(actual_element, cc.maps_to, getattr(actual_element, cc.maps_to, []) + v)
+                        elif isinstance(v, dict):
+                            for v_k, v_v in v.items():
+                                curr_dict = getattr(actual_element, cc.maps_to)
+                                curr_dict[v_k] = v_v
                         else:
-                            curr_val = getattr(actual_element, cc.maps_to)
+                            if cc.settings.inner_key:
+                                curr_val = getattr(actual_element, cc.maps_to).get(cc.settings.inner_key, None)
+                            else:
+                                curr_val = getattr(actual_element, cc.maps_to)
                             if curr_val and curr_val != 'TEMP' and curr_val != v and \
                                     not isinstance(actual_element, SchemaDefinition) and \
                                     not isinstance(actual_element, Prefix):
                                 logging.warning(f'Overwriting value for {k}, was {curr_val}, now {v}')
                                 raise ValueError(f'Cannot reset value for {k}, was {curr_val}, now {v}')
-                            setattr(actual_element, cc.maps_to, v)
+                            if cc.settings.inner_key:
+                                getattr(actual_element, cc.maps_to)[cc.settings.inner_key] = v
+                            else:
+                                setattr(actual_element, cc.maps_to, v)
                     elif cc.is_element_type:
                         logging.debug(f'Already accounted for {k}')
                     elif cc.maps_to == 'metatype':
@@ -493,7 +506,7 @@ def normalize_value(self, v: str, column_config: ColumnConfig = None) -> Any:
                     v = bmap[v.lower()]
                 else:
                     v = bool(v)
-        if metaslot and metaslot.multivalued:
+        if metaslot and metaslot.multivalued and not column_config.settings.inner_key:
             if not isinstance(v, list):
                 if v is None:
                     v = []

tests/input/personinfo.tsv

@@ -1,15 +1,15 @@
-record	field	key	multiplicity	range	parents	desc	schema.org	wikidata	belongs	status	notes
-> class	slot	identifier	cardinality	range	is_a	description	exact_mappings: {curie_prefix: sdo}	exact_mappings: {curie_prefix: wikidata}	in_subset	status	ignore
->								curie_prefix: wikidata		vmap: {T: testing, R: release}	
-	id	yes	1	string		any identifier	identifier				
-	description	no	0..1	string		a textual description	description				
-Person		n/a	n/a	n/a		a person,living or dead	Person	Q215627		R	
-Person	id	yes	1	string		identifier for a person	identifier				
-Person|Organization	name	no	1	string		full name	name				
-Person	age	no	0..1	decimal		age in years					
-Person	gender	no	0..1	decimal		age in years					
-Person	has medical history	no	0..*	MedicalEvent		medical history				T	
-Event						grouping class for events		Q1656682	a	R	
-MedicalEvent		n/a	n/a	n/a	Event	a medical encounter			b	T	
-ForProfit					Organization						
-NonProfit					Organization			Q163740			foo
+record	field	key	multiplicity	range	parents	desc	schema.org	wikidata	belongs	status	special	special2	notes
+> class	slot	identifier	cardinality	range	is_a	description	exact_mappings: {curie_prefix: sdo}	exact_mappings: {curie_prefix: wikidata}	in_subset	status	annotations	annotations	ignore
+>								curie_prefix: wikidata		vmap: {T: testing, R: release}	inner_key: special	inner_key: special2	
+	id	yes	1	string		any identifier	identifier						
+	description	no	0..1	string		a textual description	description				my_val	my_val2	
+Person		n/a	n/a	n/a		a person,living or dead	Person	Q215627		R			
+Person	id	yes	1	string		identifier for a person	identifier						
+Person|Organization	name	no	1	string		full name	name				my_val		
+Person	age	no	0..1	decimal		age in years							
+Person	gender	no	0..1	decimal		age in years							
+Person	has medical history	no	0..*	MedicalEvent		medical history				T			
+Event						grouping class for events		Q1656682	a	R			
+MedicalEvent		n/a	n/a	n/a	Event	a medical encounter			b	T			
+ForProfit					Organization								
+NonProfit					Organization			Q163740					foo

tests/input/personinfo_with_types.tsv

@@ -9,7 +9,7 @@ F	age	Person	no	0..1	decimal		age in years
 F	gender	Person	no	0..1	decimal		age in years					
 F	has medical history	Person	no	0..*	MedicalEvent		medical history				T	
 C	Person						a person,living or dead	Person	Q215627		R	
-C	Event						grouping class for events		Q1656682	a	R
+C	Event						grouping class for events		Q1656682	a	R	
 C	MedicalEvent					Event	a medical encounter			b	T	
 C	ForProfit					Organization						
 C	NonProfit					Organization			Q163740			foo

tests/test_schemamaker.py

@@ -37,6 +37,11 @@ def test_classes_slots():
     assert person_cls.slot_usage['id'].identifier
     assert person_cls.slot_usage['has medical history'].multivalued
     assert person_cls.status == 'release'
+    anns = schema.slots['description'].annotations
+    assert anns
+    assert anns['special']
+    assert anns['special'] == 'my_val'
+    assert anns['special2'] == 'my_val2'
     assert not person_cls.slot_usage['has medical history'].required
     assert person_cls.slot_usage['has medical history'].status == 'testing'
     assert 'name' in organization_cls.slots