chore: import upstream snapshot with attribution

application/parser/schema/base.py

@@ -0,0 +1,34 @@
+"""Base schema for readers."""
+from dataclasses import dataclass
+
+from langchain_core.documents import Document as LCDocument
+from application.parser.schema.schema import BaseDocument
+
+
+@dataclass
+class Document(BaseDocument):
+    """Generic interface for a data document.
+
+    This document connects to data sources.
+
+    """
+
+    def __post_init__(self) -> None:
+        """Post init."""
+        if self.text is None:
+            raise ValueError("text field not set.")
+
+    @classmethod
+    def get_type(cls) -> str:
+        """Get Document type."""
+        return "Document"
+
+    def to_langchain_format(self) -> LCDocument:
+        """Convert struct to LangChain document format."""
+        metadata = self.extra_info or {}
+        return LCDocument(page_content=self.text, metadata=metadata)
+
+    @classmethod
+    def from_langchain_format(cls, doc: LCDocument) -> "Document":
+        """Convert struct from LangChain document format."""
+        return cls(text=doc.page_content, extra_info=doc.metadata)