resolving card bugs

py_legislation.py

@@ -275,40 +275,52 @@ class PY_legislation(datasets.GeneratorBasedBuilder):
         # TODO: labeled subset has two splits
         urls_to_download = _metadata["urls"]
 
-        directory = os.getcwd()
+        root = os.getcwd()
 
         downloaded_files = dl_manager.download_and_extract(urls_to_download)
 
+        rt_train = dl_manager.download_and_extract(
+            os.path.join(root, _metadata["urls"]["raw_text"])
+        )
+
+        us_train = dl_manager.download_and_extract(
+            os.path.join(root, _metadata["urls"]["unlabeled_sentences"])
+        )
+
+        ls_train = dl_manager.download_and_extract(
+            os.path.join(root, _metadata["urls"]["labeled_sentences_train"])
+        )
+
+        ls_test = dl_manager.download_and_extract(
+            os.path.join(root, _metadata["urls"]["labeled_sentences_test"])
+        )
+
         generators = [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "filepath": downloaded_files["raw_text"],
-                    "split":"train"
+                    "filepath": rt_train # downloaded_files["raw_text"],
                 },
             ),
 
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "filepath": downloaded_files["unlabeled_sentences"],
-                    "split":"train"
+                    "filepath": us_train # downloaded_files["unlabeled_sentences"],
                 }
             ),
 
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "filepath": downloaded_files["labeled_sentences_train"],
-                    "split":"train"
+                    "filepath": ls_train # downloaded_files["labeled_sentences_train"],
                 },
             ),
 
             datasets.SplitGenerator(
                 name=datasets.Split.TEST,
                 gen_kwargs={
-                    "filepath": downloaded_files["labeled_sentences_test"],
-                    "split":"test"
+                    "filepath": ls_test # downloaded_files["labeled_sentences_test"],
                 },
             )
         ]