Stub for implementing uspto backend meta-data extraction

Signed-off-by: Viktor Kuropiatnyk <vku@zurich.ibm.com>

docling/backend/abstract_backend.py

@@ -38,6 +38,10 @@ class AbstractDocumentBackend(ABC):
     def supported_formats(cls) -> Set["InputFormat"]:
         pass
 
+    @abstractmethod
+    def extract_metadata(self) -> Dict[str, Any]:
+        return {}
+
 
 class PaginatedDocumentBackend(AbstractDocumentBackend):
     """DeclarativeDocumentBackend.

docling/backend/xml/uspto_backend.py

@@ -148,6 +148,10 @@ class PatentUsptoDocumentBackend(DeclarativeDocumentBackend):
                 f"name={self.file.name}) because the backend failed to init."
             )
     
+    @override
+    def extract_metadata(self) -> Dict[str, Any]:
+        return {}
+
 
 class PatentUspto(ABC):
     """Parser of patent documents from the US Patent Office."""

docling/datamodel/document.py

@@ -207,6 +207,7 @@ class ConversionResult(BaseModel):
     confidence: ConfidenceReport = Field(default_factory=ConfidenceReport)
 
     document: DoclingDocument = _EMPTY_DOCLING_DOC
+    metadata: Dict[str, Any] = {}
 
     @property
     @deprecated("Use document instead.")

docling/pipeline/simple_pipeline.py

@@ -38,6 +38,7 @@ class SimplePipeline(ConvertPipeline):
         # a DoclingDocument straight.
         with TimeRecorder(conv_res, "doc_build", scope=ProfilingScope.DOCUMENT):
             conv_res.document = conv_res.input._backend.convert()
+            conv_res.metadata = conv_res.input._backend.extract_metadata()
         return conv_res
 
     def _determine_status(self, conv_res: ConversionResult) -> ConversionStatus: