From d2f62abe56a4dfbb07bbef42836c5eb0c7713c58 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Marek=20Po=C5=82om?= <marek.polom@deepsense.ai>
Date: Wed, 8 May 2024 17:43:24 +0200
Subject: [PATCH] Fix

---
 unstructured/partition/pdf_image/pdfminer_processing.py | 4 ++--
 unstructured/partition/utils/constants.py               | 1 +
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/unstructured/partition/pdf_image/pdfminer_processing.py b/unstructured/partition/pdf_image/pdfminer_processing.py
index e82bb4292..00d46d06f 100644
--- a/unstructured/partition/pdf_image/pdfminer_processing.py
+++ b/unstructured/partition/pdf_image/pdfminer_processing.py
@@ -87,7 +87,7 @@ def process_data_with_pdfminer(
                 x2 * coef,
                 y2 * coef,
                 text=_text,
-                source="pdftext",
+                source=Source.PDFTEXT,
             )
 
             if text_region.bbox is not None and text_region.bbox.area > 0:
@@ -104,7 +104,7 @@ def process_data_with_pdfminer(
                     x2 * coef,
                     y2 * coef,
                     text=None,
-                    source="pdftext",
+                    source=Source.PDFTEXT,
                 )
                 if image_region.bbox is not None and image_region.bbox.area > 0:
                     layout.append(image_region)
diff --git a/unstructured/partition/utils/constants.py b/unstructured/partition/utils/constants.py
index c1864a9e5..374a5dda7 100644
--- a/unstructured/partition/utils/constants.py
+++ b/unstructured/partition/utils/constants.py
@@ -3,6 +3,7 @@ from enum import Enum
 
 
 class Source(Enum):
+    PDFTEXT = "pdftext"
     PDFMINER = "pdfminer"
     OCR_TESSERACT = "ocr_tesseract"
     OCR_PADDLE = "ocr_paddle"