Update hebrew_news.py

hebrew_news.py

@@ -55,25 +55,26 @@ class HebrewNews(datasets.GeneratorBasedBuilder):
     def _split_generators(self, dl_manager):
         """Returns SplitGenerators."""
         urls_to_download = "data/news.tar.gz"
-        downloaded_files = dl_manager.download_and_extract(urls_to_download)
-        filepath = os.path.join(downloaded_files, "news.csv")
+        archive_iter = dl_manager.iter_archive(urls_to_download)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
-                gen_kwargs={"filepath": filepath},
+                gen_kwargs={"archive_iter":archive_iter },
             ),
         ]
 
-    def _generate_examples(self, filepath):
-        with open(filepath, encoding="utf-8") as f:
-            fieldnames = ["articleBody", "description", "headline", "title"]
-            reader = csv.DictReader(f, delimiter=",", fieldnames=fieldnames)
-            next(reader)
-            for idx, row in enumerate(reader):
-                yield idx, {
-                    "id": idx,
-                    "articleBody": row["articleBody"],
-                    "description": row["description"],
-                    "headline": row["headline"],
-                    "title": row["title"],
-                }
+    def _generate_examples(self, archive_iter):
+        for path, f in archive_iter:
+            ##with open(filepath, encoding="utf-8") as f:
+                fieldnames = ["articleBody", "description", "headline", "title"]
+                reader = csv.DictReader(f, delimiter=",", fieldnames=fieldnames)
+                next(reader)
+                for idx, row in enumerate(reader):
+                    yield idx, {
+                        "id": idx,
+                        "articleBody": row["articleBody"],
+                        "description": row["description"],
+                        "headline": row["headline"],
+                        "title": row["title"],
+                    }
+