Read rather than readlines

code_clippy_github.py

@@ -176,7 +176,7 @@ class CodeClippyGithub(datasets.GeneratorBasedBuilder):
         key = 0
         for file_idx, file in enumerate(files):
             with xopen(file, "rb") as f:
-                uncompressed_data = gzip.decompress(f.readlines())
+                uncompressed_data = gzip.decompress(f.read())
                 
                 for batch_idx, code_base in enumerate(uncompressed_data):
                     j_dict = json.loads(code_base.decode('utf-8'))