From b25dbcb5b3494cc32476115a75e21f0560036ec4 Mon Sep 17 00:00:00 2001
From: Tim Asp <707699+timothyasp@users.noreply.github.com>
Date: Tue, 28 Mar 2023 13:22:05 -0700
Subject: [PATCH] add missing `source` field to pymupdf output (#2110)

To be consistent with other loaders for use with the `Sources` vector
workflows.
---
 langchain/document_loaders/pdf.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/langchain/document_loaders/pdf.py b/langchain/document_loaders/pdf.py
index b7e9cd1c..0ff912b8 100644
--- a/langchain/document_loaders/pdf.py
+++ b/langchain/document_loaders/pdf.py
@@ -156,6 +156,7 @@ class PyMuPDFLoader(BasePDFLoader):
                 page_content=page.get_text(**kwargs).encode("utf-8"),
                 metadata=dict(
                     {
+                        "source": file_path,
                         "file_path": file_path,
                         "page_number": page.number + 1,
                         "total_pages": len(doc),