Update test_dataset.py

test_dataset.py

@@ -33,7 +33,10 @@ _DESCRIPTION = """\
 This is a test dataset used to demonstrate the process of creating a hugging face dataset
 """
 
-_URL = "https://340788.oss-cn-beijing.aliyuncs.com/animal%20classification.zip"
+_URLS = {
+    "cat": "https://340788.oss-cn-beijing.aliyuncs.com/cat.zip",
+    "dog": "https://340788.oss-cn-beijing.aliyuncs.com/dog.zip",
+}
 
 _NAMES = ["cat", "dog"]
 
@@ -65,30 +68,29 @@ class TestDataset(datasets.GeneratorBasedBuilder):
         )
 
     def _split_generators(self, dl_manager):
-        data_dir = dl_manager.download_and_extract(_URL)
-        sub_set = self.config.name
+        urls = _URLS[self.config.name]
+        data_dir = dl_manager.download_and_extract(urls)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "files": os.path.join(data_dir, sub_set, "train"),
+                    "files": dl_manager.iter_files(os.path.join(data_dir, "train")),
                 },
             ),
             datasets.SplitGenerator(
                 name=datasets.Split.TEST,
                 gen_kwargs={
-                    "files": os.path.join(data_dir, sub_set, "test"),
+                    "files": dl_manager.iter_files(os.path.join(data_dir, "test")),
                 },
             ),
         ]
 
-
     def _generate_examples(self, files):
-        for i, path in files:
+        for i, path in enumerate(files):
             file_name = os.path.basename(path)
             if file_name.endswith(".jpeg"):
                 yield i, {
                     "image_file_path": path,
                     "image": path,
                     "labels": os.path.basename(os.path.dirname(path)).lower(),
-                }
+                }