Rename missing_df to filtered_df

Author: EmersonFras

src/cautiousrobot/utils.py

@@ -104,7 +104,7 @@ def check_existing_images(csv_path, img_dir, source_df, filename_col, subfolders
 
     Returns:
         updated_df (pd.DataFrame): DataFrame with new column 'in_img_dir' indicating presence in img_dir.
-        missing_df (pd.DataFrame): DataFrame filtered to only files not present in img_dir.
+        filtered_df (pd.DataFrame): DataFrame filtered to only files not present in img_dir.
     """
     # Create a copy to avoid modifying the original DataFrame
     df = source_df.copy()
@@ -148,15 +148,15 @@ def check_existing_images(csv_path, img_dir, source_df, filename_col, subfolders
     df = df.drop(columns=["expected_path"])
 
     # Create filtered DataFrame
-    missing_df = df[~df["in_img_dir"]].copy()
+    filtered_df = df[~df["in_img_dir"]].copy()
 
     # Exit if all images are already there
-    if missing_df.empty:
+    if filtered_df.empty:
         sys.exit(f"'{img_dir}' already contains all images. Exited without executing.")
     else:
         # Print directory status message - pre-download
         num_existing = len(existing_files)
         expected_num = df.shape[0] - (starting_idx)
         print(f"There are {num_existing} files in {img_dir}. Based on {csv_path}, there should be {expected_num} images.")
 
-    return df, missing_df
+    return df, filtered_df

tests/test_existing_images.py

@@ -16,27 +16,27 @@ def setUp(self):
     @patch("cautiousrobot.utils.os.path.exists", return_value=False)
     def test_directory_does_not_exist(self, mock_exists):
         """If image directory doesn't exist, all images marked as not in directory."""
-        updated_df, missing_df = check_existing_images(
+        updated_df, filtered_df = check_existing_images(
             self.csv_path, self.img_dir, self.sample_df.copy(), self.filename_col
         )
 
         self.assertFalse(any(updated_df["in_img_dir"]))
-        self.assertEqual(len(missing_df), len(self.sample_df))
+        self.assertEqual(len(filtered_df), len(self.sample_df))
         mock_exists.assert_called_once_with(self.img_dir)
 
     @patch("cautiousrobot.utils.os.path.exists", return_value=True)
     @patch("cautiousrobot.utils.gather_file_paths", return_value=["test_images/a.jpg"])
     @patch("cautiousrobot.utils.print")
     def test_some_files_exist(self, mock_print, mock_gather, mock_exists):
         """Should mark existing files correctly and print status."""
-        updated_df, missing_df = check_existing_images(
+        updated_df, filtered_df = check_existing_images(
             self.csv_path, self.img_dir, self.sample_df.copy(), self.filename_col
         )
 
         self.assertTrue(updated_df.loc[0, "in_img_dir"])   # a.jpg exists
         self.assertFalse(updated_df.loc[1, "in_img_dir"])  # b.jpg missing
         self.assertFalse(updated_df.loc[2, "in_img_dir"])  # c.jpg missing
-        self.assertEqual(len(missing_df), 2)
+        self.assertEqual(len(filtered_df), 2)
         mock_print.assert_called_once()
         self.assertIn("There are 1 files", mock_print.call_args[0][0])
 
@@ -57,12 +57,12 @@ def test_all_files_exist_exits(self, mock_gather, mock_exists):
     @patch("cautiousrobot.utils.print")
     def test_no_files_exist(self, mock_print, mock_gather, mock_exists):
         """If no files exist, should mark all as missing and print message."""
-        updated_df, missing_df = check_existing_images(
+        updated_df, filtered_df = check_existing_images(
             self.csv_path, self.img_dir, self.sample_df.copy(), self.filename_col
         )
 
         self.assertFalse(any(updated_df["in_img_dir"]))
-        self.assertEqual(len(missing_df), len(self.sample_df))
+        self.assertEqual(len(filtered_df), len(self.sample_df))
         mock_print.assert_called_once()
         self.assertIn("There are 0 files", mock_print.call_args[0][0])
 
@@ -75,14 +75,14 @@ def test_subfolders_handling(self, mock_gather, mock_exists):
             self.filename_col: ["a.jpg", "b.jpg"]
         })
 
-        updated_df, missing_df = check_existing_images(
+        updated_df, filtered_df = check_existing_images(
             self.csv_path, self.img_dir, sub_df.copy(), self.filename_col, subfolders="subfolder"
         )
 
         # species1/a.jpg should be marked present, species2/b.jpg missing
         self.assertTrue(updated_df.loc[0, "in_img_dir"])
         self.assertFalse(updated_df.loc[1, "in_img_dir"])
-        self.assertEqual(len(missing_df), 1)
+        self.assertEqual(len(filtered_df), 1)
 
 
 if __name__ == "__main__":