good

Author: JinZhou5042

generate_data.py

@@ -3,16 +3,27 @@
 from pathlib import Path
 from src.data_parse import DataParser
 
+def remove_duplicates_preserve_order(seq):
+    seen = set()
+    result = []
+    for item in seq:
+        name = Path(item).name
+        if name not in seen:
+            seen.add(name)
+            result.append(name)
+    return result
+
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()
     parser.add_argument('runtime_templates', type=str, nargs='+', help='list of log directories (e.g., log1 log2 log3)')
     parser.add_argument('--subgraphs-only', action='store_true', help='had previously parsed the debug file, only generate subgraphs')
     args = parser.parse_args()
 
+    deduped_names = remove_duplicates_preserve_order(args.runtime_templates)
+
     root_dir = os.path.join(os.getcwd(), 'logs')
-    full_paths = [os.path.join(root_dir, Path(rt).name) for rt in args.runtime_templates]
+    full_paths = [os.path.join(root_dir, name) for name in deduped_names]
 
-    # verify all paths exist
     missing = [p for p in full_paths if not os.path.exists(p)]
     if missing:
         print("❌ The following log directories do not exist:")