chg: [reprocess tool] add OcrExtractor module + filter image gif

bin/lib/objects/Images.py

@@ -2,6 +2,7 @@
 # -*-coding:UTF-8 -*
 
 import base64
+import magic
 import os
 import sys
 
@@ -64,6 +65,14 @@ class Image(AbstractDaterangeObject):
         filename = os.path.join(IMAGE_FOLDER, self.get_rel_path())
         return os.path.realpath(filename)
 
+    def is_gif(self, filepath=None):
+        if not filepath:
+            filepath = self.get_filepath()
+        mime = magic.from_file(filepath, mime=True)
+        if mime == 'image/gif':
+            return True
+        return False
+
     def get_file_content(self):
         filepath = self.get_filepath()
         with open(filepath, 'rb') as f:

bin/modules/OcrExtractor.py

@@ -89,6 +89,10 @@ class OcrExtractor(AbstractModule):
         if self.is_cached():
             return None
 
+        if self.obj.is_gif():
+            self.logger.warning(f'Ignoring GIF: {self.obj.id}')
+            return None
+
         if not ocr.exists():
             path = image.get_filepath()
             languages = get_model_languages(image)

tools/reprocess_objects.py

@@ -30,15 +30,21 @@ from lib.objects import ail_objects
 # from modules.Telegram import Telegram
 
 from modules.Languages import Languages
+from modules.OcrExtractor import OcrExtractor
 
 MODULES = {
-    'Languages': Languages
+    'Languages': Languages,
+    'OcrExtractor': OcrExtractor
+
 }
 
 def reprocess_message_objects(object_type, module_name=None):
     if module_name:
         module = MODULES[module_name]()
         for obj in ail_objects.obj_iterator(object_type, filters={}):
+            if not obj.exists():
+                print(f'ERROR: object does not exist, {obj.id}')
+                continue
             module.obj = obj
             module.compute(None)
     else: