enhancement: preserve image format (#107)

PIL records the filetype of the image when it's loaded, but the way we were handling the image destroyed the format metadata. This change keeps the image format metadata so that it can be recovered downstream.

Author: qued

CHANGELOG.md

@@ -1,5 +1,6 @@
-## 0.4.5-dev0
+## 0.4.5-dev1
 
+* Preserve image format in PIL.Image.Image when loading
 * Added ONNX version of Detectron2
 
 ## 0.4.4

unstructured_inference/__version__.py

@@ -1 +1 @@
-__version__ = "0.4.5-dev0"  # pragma: no cover
+__version__ = "0.4.5-dev1"  # pragma: no cover

unstructured_inference/inference/layout.py

@@ -103,7 +103,10 @@ def from_image_file(
         """Creates a DocumentLayout from an image file."""
         logger.info(f"Reading image file: {filename} ...")
         try:
-            image = Image.open(filename).convert("RGB")
+            image = Image.open(filename)
+            format = image.format
+            image = image.convert("RGB")
+            image.format = format
         except Exception as e:
             if os.path.isdir(filename) or os.path.isfile(filename):
                 raise e