added file type checks to the parsers to prevent temporary files from being consumed. Also: parsers announce file types they wish to use as default for each mime type.

src/paperless_tesseract/signals.py

@@ -5,9 +5,9 @@ def tesseract_consumer_declaration(sender, **kwargs):
     return {
         "parser": RasterisedDocumentParser,
         "weight": 0,
-        "mime_types": [
-            "application/pdf",
-            "image/jpeg",
-            "image/png"
-        ]
+        "mime_types": {
+            "application/pdf": ".pdf",
+            "image/jpeg": ".jpg",
+            "image/png": ".png"
+        }
     }