Update app.py: Fix missing environment variables and improve dataframe filtering

app.py

@@ -26,9 +26,12 @@ REMOVE_ORGS = {
 HF_TOKEN = os.getenv("HF_TOKEN")
 USER_AGENT = os.getenv("USER_AGENT")
 
+if not HF_TOKEN or not USER_AGENT:
+    raise ValueError(
+        "Missing required environment variables. Please ensure both HF_TOKEN and USER_AGENT are set."
+    )
 
-headers = {"authorization": f"Bearer ${HF_TOKEN}", "user-agent": USER_AGENT}
-
+headers = {"authorization": f"Bearer {HF_TOKEN}", "user-agent": USER_AGENT}
 
 client = Client(
     headers=headers,
@@ -99,9 +102,7 @@ def render_model_hub_link(hub_id):
 def get_datasets():
     return list(
         tqdm(
-            iter(
-                list_datasets(limit=LIMIT, full=True, sort="createdAt", direction=-1)
-            )
+            iter(list_datasets(limit=LIMIT, full=True, sort="createdAt", direction=-1))
         )
     )
 
@@ -130,7 +131,7 @@ columns_to_drop = [
     "_id",
     "private",
     "author",
-  #  "citation",
+    #  "citation",
     "lastModified",
 ]
 
@@ -165,15 +166,22 @@ def filter_by_readme_len(df, min_len=None):
 
 
 def filter_df(max_age_days=None, min_len=None, needs_server_preview: bool = False):
-    df = prep_dataframe()
-    if needs_server_preview:
-        df = df[df["server_preview"] == True]
-    if max_age_days is not None:
-        df = filter_df_by_max_age(df, max_age_days=max_age_days)
-    if min_len is not None:
-        df = filter_by_readme_len(df, min_len=min_len)
-    df = df.sort_values(by=["likes", "downloads", "len"], ascending=False)
-    return df
+    try:
+        df = prep_dataframe()
+        if needs_server_preview:
+            df = df[df["server_preview"] == True]
+        if max_age_days is not None:
+            df = filter_df_by_max_age(df, max_age_days=max_age_days)
+        if min_len is not None:
+            df = filter_by_readme_len(df, min_len=min_len)
+        df = df.sort_values(by=["likes", "downloads", "len"], ascending=False)
+        return df
+    except Exception as e:
+        print(f"Error filtering dataframe: {str(e)}")
+        # Return empty dataframe with same columns if there's an error
+        return pd.DataFrame(
+            columns=["id", "likes", "downloads", "len", "createdAt", "server_preview"]
+        )
 
 
 with gr.Blocks() as demo:
@@ -204,21 +212,34 @@ with gr.Blocks() as demo:
             interactive=True,
         )
 
-    output = gr.DataFrame(filter_df, datatype="markdown", min_width=160 * 2.5, height=1000)
-    max_age_days.input(
-        filter_df,
-        inputs=[max_age_days, min_len, needs_server_preview],
-        outputs=[output],
-    )
-    min_len.input(
-        filter_df,
-        inputs=[max_age_days, min_len, needs_server_preview],
-        outputs=[output],
-    )
-    needs_server_preview.change(
-        filter_df,
-        inputs=[max_age_days, min_len, needs_server_preview],
-        outputs=[output],
+    # gr.Markdown(
+    #     """
+    #     <style>
+    #     #dataset_table {
+    #         height: 1000px;
+    #         overflow: auto;
+    #     }
+    #     </style>
+    #     """
+    # )
+
+    output = gr.DataFrame(
+        value=filter_df(7, 300, False),  # Set initial values explicitly
+        interactive=False,
+        datatype="markdown",
+        min_width=160 * 2.5,
+        elem_id="dataset_table",
     )
 
+    def update_df(age, length, preview):
+        return filter_df(age, length, preview)
+
+    # Use a single update function for all inputs
+    for component in [max_age_days, min_len, needs_server_preview]:
+        component.change(
+            fn=update_df,
+            inputs=[max_age_days, min_len, needs_server_preview],
+            outputs=[output],
+        )
+
 demo.launch()