Update app.py

app.py

@@ -5,10 +5,38 @@ import pandas as pd
 import json
 
 async def get_splits(dataset_name: str) -> Dict[str, List[Dict]]:
-    URL = f"https://datasets-server.huggingface.co/splits?dataset={dataset_name}"
+    URL = f"https://huggingface.co/api/datasets/{dataset_name}"
     async with httpx.AsyncClient() as session:
         response = await session.get(URL)
-        return response.json()
+        dataset_info = response.json()
+        return {
+            "splits": [
+                {"split": split_name, "config": config_name}
+                for config_name, config_info in dataset_info.get("config", {}).items()
+                for split_name in config_info.get("splits", [])
+            ]
+        }
+
+async def get_first_rows(dataset: str, config: str, split: str) -> Dict[str, Dict[str, List[Dict]]]:
+    URL = f"https://huggingface.co/datasets/{dataset}/resolve/main/dataset_info.json"
+    async with httpx.AsyncClient() as session:
+        response = await session.get(URL)
+        dataset_info = response.json()
+        split_info = dataset_info["splits"][split]
+        first_rows = {
+            "rows": [
+                {"row": row} for row in split_info["examples"][:10]
+            ]
+        }
+        return first_rows
+
+# Guido von Roissum: https://www.youtube.com/watch?v=-DVyjdw4t9I
+async def update_URL(dataset: str, config: str, split: str) -> str:
+    URL = f"https://huggingface.co/datasets/{dataset}/tree/main/{config}/{split}"
+    return URL
+
+    
+
 
 async def get_valid_datasets() -> List[str]:
     URL = f"https://huggingface.co/api/datasets"
@@ -20,13 +48,6 @@ async def get_valid_datasets() -> List[str]:
             datasets = []  # Set a default value if the response is not in the expected format
         return datasets
 
-async def get_first_rows(dataset: str, config: str, split: str) -> Dict[str, Dict[str, List[Dict]]]:
-    URL = f"https://datasets-server.huggingface.co/first-rows?dataset={dataset}&config={config}&split={split}"
-    async with httpx.AsyncClient() as session:
-        response = await session.get(URL)
-        print(URL)
-        gr.Markdown(URL)
-        return response.json()
 
 def get_df_from_rows(api_output):
     dfFromSort = pd.DataFrame([row["row"] for row in api_output["rows"]])
@@ -53,11 +74,6 @@ async def update_dataset(split_name: str, config_name: str, dataset_name: str):
     df = get_df_from_rows(rows)
     return df
 
-# Guido von Roissum: https://www.youtube.com/watch?v=-DVyjdw4t9I
-async def update_URL(dataset: str, config: str, split: str) -> str:
-    URL = f"https://datasets-server.huggingface.co/first-rows?dataset={dataset}&config={config}&split={split}"
-    URL = f"https://huggingface.co/datasets/{split}"
-    return (URL)
    
 async def openurl(URL: str) -> str:
     html = f"<a href={URL} target=_blank>{URL}</a>"