Merge pull request #282 from dkpro/bugfix/280-JSON-CAS-parsing-does-n…

…ot-handle-DocumentAnnotation-properly #280 - JSON CAS parsing does not handle DocumentAnnotation properly
dkpro · Mar 21, 2023 · 3b4a75b · 3b4a75b
2 parents 97f4fe0 + 4721cf5
commit 3b4a75b
Show file tree

Hide file tree

Showing 4 changed files with 64 additions and 4 deletions.
diff --git a/cassis/json.py b/cassis/json.py
@@ -74,8 +74,10 @@ def deserialize(self, source: Union[IO, str], typesystem: Optional[TypeSystem] =
         self._max_sofa_num = 0
         self._post_processors = []
 
-        embedded_typesystem = TypeSystem()
         json_typesystem = data.get(TYPES_FIELD)
+        embedded_typesystem = TypeSystem(
+            add_document_annotation_type=not (json_typesystem.get(FLAG_DOCUMENT_ANNOTATION))
+        )
 
         # First, build a dependency graph to support cases where a child type is defined before its super type
         type_dependencies = defaultdict(set)
@@ -84,7 +86,7 @@ def deserialize(self, source: Union[IO, str], typesystem: Optional[TypeSystem] =
 
         # Second, load all the types but no features since features of a type X might be of a later loaded type Y
         for type_name in toposort_flatten(type_dependencies):
-            if is_predefined(type_name):
+            if is_predefined(type_name) or embedded_typesystem.contains_type(type_name):
                 continue
 
             self._parse_type(embedded_typesystem, type_name, json_typesystem[type_name])

diff --git a/cassis/typesystem.py b/cassis/typesystem.py
@@ -1102,8 +1102,8 @@ def _defines_predefined_type(self, type_name):
         self._predefined_types.add(type_name)
 
     def _add_document_annotation_type(self):
-        t = self.create_type(name=_DOCUMENT_ANNOTATION_TYPE, supertypeName="uima.tcas.Annotation")
-        self.create_feature(t, name="language", rangeType="uima.cas.String")
+        t = self.create_type(name=_DOCUMENT_ANNOTATION_TYPE, supertypeName=TYPE_NAME_ANNOTATION)
+        self.create_feature(t, name="language", rangeType=TYPE_NAME_STRING)
 
     def transitive_closure(self, seed_types: Set[Type], built_in: bool = False) -> Set[Type]:
         # Build transitive closure of used types by following parents, features, etc.

diff --git a/tests/test_files/json/fs_as_array/ser-ref/casExtendingDocumentAnnotation/data.json b/tests/test_files/json/fs_as_array/ser-ref/casExtendingDocumentAnnotation/data.json
@@ -0,0 +1,54 @@
+{
+  "%TYPES" : {
+    "de.tudarmstadt.ukp.dkpro.core.api.metadata.type.DocumentMetaData" : {
+      "%NAME" : "de.tudarmstadt.ukp.dkpro.core.api.metadata.type.DocumentMetaData",
+      "%SUPER_TYPE" : "uima.tcas.DocumentAnnotation",
+      "documentTitle" : {
+        "%NAME" : "documentTitle",
+        "%RANGE" : "uima.cas.String"
+      },
+      "documentId" : {
+        "%NAME" : "documentId",
+        "%RANGE" : "uima.cas.String"
+      },
+      "documentUri" : {
+        "%NAME" : "documentUri",
+        "%RANGE" : "uima.cas.String"
+      },
+      "collectionId" : {
+        "%NAME" : "collectionId",
+        "%RANGE" : "uima.cas.String"
+      },
+      "documentBaseUri" : {
+        "%NAME" : "documentBaseUri",
+        "%RANGE" : "uima.cas.String"
+      },
+      "isLastSegment" : {
+        "%NAME" : "isLastSegment",
+        "%RANGE" : "uima.cas.Boolean"
+      }
+    }
+  },
+  "%FEATURE_STRUCTURES" : [ {
+    "%ID" : 2,
+    "%TYPE" : "uima.cas.Sofa",
+    "sofaNum" : 1,
+    "sofaID" : "_InitialView",
+    "mimeType" : "text",
+    "sofaString" : "This is a test ."
+  }, {
+    "%ID" : 1,
+    "%TYPE" : "de.tudarmstadt.ukp.dkpro.core.api.metadata.type.DocumentMetaData",
+    "@sofa" : 2,
+    "begin" : 0,
+    "end" : 16,
+    "documentId" : "doc",
+    "isLastSegment" : false
+  } ],
+  "%VIEWS" : {
+    "_InitialView" : {
+      "%SOFA" : 2,
+      "%MEMBERS" : [ 1 ]
+    }
+  }
+}
diff --git a/tests/test_json.py b/tests/test_json.py
@@ -77,6 +77,10 @@
             ["uima.tcas.DocumentAnnotation", 0, 6, "這是一個測試"],
         ],
     ),
+    (
+        os.path.join(SER_REF_DIR, "casExtendingDocumentAnnotation"),
+        [["de.tudarmstadt.ukp.dkpro.core.api.metadata.type.DocumentMetaData", 0, 16, "This is a test ."]],
+    ),
 ]
-Original file line number
+Diff line change
@@ Expand Up / @@ -77,6 +77,10 @@ @@
                 ["uima.tcas.DocumentAnnotation", 0, 6, "這是一個測試"],
             ],
         ),
+        (
+            os.path.join(SER_REF_DIR, "casExtendingDocumentAnnotation"),
+            [["de.tudarmstadt.ukp.dkpro.core.api.metadata.type.DocumentMetaData", 0, 16, "This is a test ."]],
+        ),
     ]
@@ Expand Down @@