Remove alpha layer from PNG files for img2pdf

src/paperless_tesseract/parsers.py

@@ -66,6 +66,10 @@ def is_image(self, mime_type):
             "image/gif",
         ]
 
+    def has_alpha(self, image):
+        with Image.open(image) as im:
+            return im.mode in ('RGBA', 'LA')
+
     def get_dpi(self, image):
         try:
             with Image.open(image) as im:
@@ -182,6 +186,19 @@ def construct_ocrmypdf_parameters(self,
         if self.is_image(mime_type):
             dpi = self.get_dpi(input_file)
             a4_dpi = self.calculate_a4_dpi(input_file)
+
+            if self.has_alpha(input_file):
+                self.log(
+                    "info",
+                    f"Removing alpha layer from {input_file} "
+                    "for compatibility with img2pdf"
+                )
+                with Image.open(input_file) as im:
+                    background = Image.new('RGBA', im.size, (255, 255, 255))
+                    background.alpha_composite(im)
+                    background = background.convert('RGB')
+                    background.save(input_file, format=im.format)
+
             if dpi:
                 self.log(
                     "debug",

src/paperless_tesseract/tests/test_parser.py

@@ -181,13 +181,14 @@ def test_image_simple(self):
 
         self.assertContainsStrings(parser.get_text(), ["This is a test document."])
 
-    def test_image_simple_alpha_fail(self):
+    def test_image_simple_alpha(self):
         parser = RasterisedDocumentParser(None)
 
-        def f():
-            parser.parse(os.path.join(self.SAMPLE_FILES, "simple-alpha.png"), "image/png")
+        parser.parse(os.path.join(self.SAMPLE_FILES, "simple-alpha.png"), "image/png")
 
-        self.assertRaises(ParseError, f)
+        self.assertTrue(os.path.isfile(parser.archive_path))
+
+        self.assertContainsStrings(parser.get_text(), ["This is a test document."])
 
     def test_image_calc_a4_dpi(self):
         parser = RasterisedDocumentParser(None)