Update app.py

app.py

@@ -2,6 +2,7 @@ import gradio as gr
 import pandas as pd
 import io
 import os
+from pathlib import Path
 from huggingface_hub import HfApi, Repository
 
 # Global DataFrame
@@ -11,12 +12,12 @@ def upload_csv(file):
     global df
     df = pd.read_csv(file.name)
     if "text" not in df.columns or "label" not in df.columns:
-        return gr.update(visible=False), "❌ CSV must have 'text' and 'label' columns."
+        return gr.update(visible=False), "❌ CSV must contain ‘text’ and ‘label’ columns."
     df["label"] = df["label"].fillna("")
-    # Show the dataframe and status
+    # Show the table and set status
     return (
         gr.update(value=df[["text","label"]], visible=True),
-        "✅ File uploaded — you can now edit the labels below."
+        "✅ File uploaded — you can now edit labels."
     )
 
 def save_changes(edited_table):
@@ -26,20 +27,41 @@ def save_changes(edited_table):
 
 def download_csv():
     global df
-    csv_bytes = df.to_csv(index=False).encode("utf-8")
-    return gr.File.update(value=io.BytesIO(csv_bytes), filename="annotated_data.csv")
+    out_path = "annotated_data.csv"
+    df.to_csv(out_path, index=False)
+    return out_path
 
-def push_to_hub(repo_name, hf_token):
+def push_to_hub(repo_name: str, hf_token: str) -> str:
     global df
-    repo_url = f"https://huggingface.co/datasets/{repo_name}"
-    local_path = f"./{repo_name.replace('/', '_')}"
-    if os.path.exists(local_path):
-        os.system(f"rm -rf {local_path}")
-    HfApi().create_repo(repo_id=repo_name, token=hf_token, repo_type="dataset", exist_ok=True)
-    repo = Repository(local_dir=local_path, clone_from=repo_url, token=hf_token)
-    df.to_csv(f"{local_path}/data.csv", index=False)
-    repo.push_to_hub()
-    return f"🚀 Data pushed to {repo_url}"
+    try:
+        api = HfApi()
+        api.create_repo(
+            repo_id=repo_name,
+            token=hf_token,
+            repo_type="dataset",
+            exist_ok=True
+        )
+
+        local_dir = Path(f"./{repo_name.replace('/', '_')}")
+        if local_dir.exists():
+            for child in local_dir.iterdir():
+                child.unlink()
+            local_dir.rmdir()
+
+        repo = Repository(
+            local_dir=str(local_dir),
+            clone_from=repo_name,
+            use_auth_token=hf_token
+        )
+
+        csv_path = local_dir / "data.csv"
+        df.to_csv(csv_path, index=False)
+
+        repo.push_to_hub(commit_message="📑 Update annotated data")
+        return f"🚀 Pushed to https://huggingface.co/datasets/{repo_name}"
+
+    except Exception as e:
+        return f"❌ Push failed: {e}"
 
 with gr.Blocks(theme=gr.themes.Default()) as app:
     gr.Markdown("## 🏷️ Label It! Text Annotation Tool")
@@ -51,16 +73,16 @@ with gr.Blocks(theme=gr.themes.Default()) as app:
 
     df_table = gr.Dataframe(
         headers=["text","label"],
+        label="📝 Editable Table",
         interactive=True,
-        visible=False,
-        label="📝 Editable Table"
+        visible=False
     )
     status = gr.Textbox(label="Status", interactive=False)
 
     with gr.Row():
-        save_btn    = gr.Button("💾 Save")
-        download_btn= gr.Button("⬇️ Download CSV")
-        download_out= gr.File(label="Download")
+        save_btn     = gr.Button("💾 Save")
+        download_btn = gr.Button("⬇️ Download CSV")
+        download_out = gr.File(label="📥 Downloaded File")
 
     with gr.Accordion("📦 Push to Hugging Face Hub", open=False):
         repo_input  = gr.Textbox(label="Repo (username/dataset-name)")
@@ -68,12 +90,11 @@ with gr.Blocks(theme=gr.themes.Default()) as app:
         push_btn    = gr.Button("🚀 Push")
         push_status = gr.Textbox(label="Push Status", interactive=False)
 
-    # ------------------------
-    upload_btn.click(upload_csv,     inputs=file_input,        outputs=[df_table, status])
-    save_btn.click(  save_changes,   inputs=df_table,          outputs=status)
+    # Event bindings
+    upload_btn.click(upload_csv,     inputs=file_input,              outputs=[df_table, status])
+    save_btn.click(  save_changes,   inputs=df_table,                outputs=status)
     download_btn.click(download_csv, outputs=download_out)
-    push_btn.click(  push_to_hub,    inputs=[repo_input,token_input], outputs=push_status)
-    # ------------------------
+    push_btn.click(  push_to_hub,    inputs=[repo_input, token_input], outputs=push_status)
 
-    # Launch the app
-    app.launch()
+# Launch the app
+app.launch()