2 files changed, 34 insertions, 0 deletions
diff --git a/rag/retriever/parser/__init__.py b/rag/retriever/parser/__init__.py
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/rag/retriever/parser/__init__.py
diff --git a/rag/retriever/parser/pdf.py b/rag/retriever/parser/pdf.py
new file mode 100644
index 0000000..410f027
--- /dev/null
+++ b/rag/retriever/parser/pdf.py
@@ -0,0 +1,34 @@
+import os
+from pathlib import Path
+from typing import List, Optional
+
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_core.documents import Document
+from langchain_community.document_loaders.parsers.pdf import (
+    PyPDFParser,
+)
+from langchain_community.document_loaders.blob_loaders import Blob
+
+
+class PDFParser:
+    def __init__(self) -> None:
+        self.parser = PyPDFParser(password=None, extract_images=False)
+
+    def from_data(self, blob: Blob) -> List[Document]:
+        return self.parser.parse(blob)
+
+    def from_path(self, path: Path) -> Blob:
+        return Blob.from_path(path)
+
+    def chunk(
+        self, document: List[Document], source: Optional[str] = None
+    ) -> List[Document]:
+        splitter = RecursiveCharacterTextSplitter(
+            chunk_size=int(os.environ["CHUNK_SIZE"]),
+            chunk_overlap=int(os.environ["CHUNK_OVERLAP"]),
+        )
+        chunks = splitter.split_documents(document)
+        if source is not None:
+            for c in chunks:
+                c.metadata["source"] = source
+        return chunks