lazy loading for parsers

src/documents/tests/test_parsers.py

@@ -114,8 +114,8 @@ class TestParserAvailability(TestCase):
         self.assertEqual(get_default_file_extension('application/zip'), ".zip")
         self.assertEqual(get_default_file_extension('aasdasd/dgfgf'), "")
 
-        self.assertEqual(get_parser_class_for_mime_type('application/pdf'), RasterisedDocumentParser)
+        self.assertIsInstance(get_parser_class_for_mime_type('application/pdf')(logging_group=None), RasterisedDocumentParser)
-        self.assertEqual(get_parser_class_for_mime_type('text/plain'), TextDocumentParser)
+        self.assertIsInstance(get_parser_class_for_mime_type('text/plain')(logging_group=None), TextDocumentParser)
         self.assertEqual(get_parser_class_for_mime_type('text/sdgsdf'), None)
 
         self.assertTrue(is_file_ext_supported('.pdf'))

src/paperless_tesseract/signals.py

@@ -1,9 +1,13 @@
-from .parsers import RasterisedDocumentParser
+
+def get_parser(*args, **kwargs):
+    from .parsers import RasterisedDocumentParser
+
+    return RasterisedDocumentParser(*args, **kwargs)
 
 
 def tesseract_consumer_declaration(sender, **kwargs):
     return {
-        "parser": RasterisedDocumentParser,
+        "parser": get_parser,
         "weight": 0,
         "mime_types": {
             "application/pdf": ".pdf",

src/paperless_text/signals.py

@@ -1,9 +1,13 @@
-from .parsers import TextDocumentParser
+
+def get_parser(*args, **kwargs):
+    from .parsers import TextDocumentParser
+
+    return TextDocumentParser(*args, **kwargs)
 
 
 def text_consumer_declaration(sender, **kwargs):
     return {
-        "parser": TextDocumentParser,
+        "parser": get_parser,
         "weight": 10,
         "mime_types": {
             "text/plain": ".txt",

src/paperless_tika/signals.py

@@ -1,9 +1,13 @@
-from .parsers import TikaDocumentParser
+
+def get_parser(*args, **kwargs):
+    from .parsers import TikaDocumentParser
+
+    return TikaDocumentParser(*args, **kwargs)
 
 
 def tika_consumer_declaration(sender, **kwargs):
     return {
-        "parser": TikaDocumentParser,
+        "parser": get_parser,
         "weight": 10,
         "mime_types": {
             "application/msword": ".doc",