annotating_by_label : automatically make label groups.

annotating_by_label.py

@@ -102,16 +102,34 @@ def main():
 
     root_dir = args.input_path
     results = annotating.make_dictionary(root_dir)
-    labels_file = os.path.join(root_dir, "label_groups")
+    label_groups = os.path.join(root_dir, "label_groups")
+    all_labels = os.path.join(root_dir, "labels")
 
-    if not os.path.exists(labels_file):
-        print(f"Error: Labels file '{labels_file}' not found.")
-        sys.exit(1)
+    if not os.path.exists(label_groups):
+        print(f"Warning: Labels file '{label_groups}' not found, making it out of '{all_labels}' (do check it…)")
+        if not os.path.exists(all_labels):
+            print(f"Error: {all_labels} not found.")
+            sys.exit(1)
 
-    with open(labels_file, "r") as f:
+        with open(all_labels, 'r') as f:
+            lines = [l.strip() for l in f if l.strip()]
+
+        groups = {}
+        for line in lines:
+            # Key is the part before the colon, or the whole line if no colon
+            key = line.split(' : ')[0] if ' : ' in line else line
+            groups.setdefault(key, []).append(line)
+
+        with open(label_groups, 'w') as f:
+            for items in groups.values():
+                f.write(",".join(items) + "\n")
+
+    with open(label_groups, "r") as f:
         lines = [line.strip() for line in f if line.strip()]
 
     bgnot_dir = os.path.join(root_dir, "BGnot")
+    if args.overwrite and os.path.exists(bgnot_dir):
+        shutil.rmtree(bgnot_dir)
     os.makedirs(bgnot_dir, exist_ok=True)
 
     for line in lines: