Update hebrew_news.py

hebrew_news.py

@@ -6,6 +6,8 @@ import os.path
 
 import datasets
 from urllib.parse import urlparse
+import io
+
 
 _CITATION = """
 """
@@ -66,17 +68,15 @@ class HebrewNews(datasets.GeneratorBasedBuilder):
 
     def _generate_examples(self, archive_iter):
         for filepath, f in archive_iter:
-            print(type(f))
-            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"],
-                    }
-    
+            csv_file = io.TextIOWrapper(f, encoding="utf-8")
+            fieldnames = ["articleBody", "description", "headline", "title"]
+            reader = csv.DictReader(csv_file, 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"],
+                }