test commitscheduler

app.py

@@ -2,8 +2,24 @@ import streamlit as st
 from datasets import load_dataset
 import datetime as dt
 import random
+import json
 import os
-from huggingface_hub import HfApi
+from huggingface_hub import HfApi, CommitScheduler
+import uuid
+
+
+
+def save_feedback(input_text:str, output_1: str, output_2:str, user_choice: int) -> None:
+    """
+    Append input/outputs and user feedback to a JSON Lines file 
+    using a thread lock to avoid concurrent writes from different users.
+    """
+    with scheduler.lock:
+        with submissions_file.open("a") as f:
+            f.write(json.dumps({"input": input_text, "output_1": output_1, "output_2": output_2, "user_choice": user_choice}))
+            f.write("\n")
+
+
 
 
 HF_API_KEY = os.environ.get("HF_TOKEN", None)
@@ -13,6 +29,16 @@ api = HfApi(token=HF_API_KEY)
 REPO_ID = "imomayiz/darija-english"
 DATASET_REPO_URL = f"https://huggingface.co/datasets/{REPO_ID}"
 
+submissions_folder = "submissions"
+submissions_file = os.path.join(submissions_folder, f"submissions_{uuid.uuid4()}.json")
+
+scheduler = CommitScheduler(
+    repo_id=REPO_ID,
+    repo_type="dataset",
+    folder_path=submissions_folder,
+    path_in_repo="submissions",
+    every=10,
+)
 
 def load_data(repo_id):
     dataset = load_dataset(f'{repo_id}', name='sentences', split='sentences')
@@ -30,39 +56,37 @@ def fetch_sentence(dataset, column_name="darija_ar"):
 
     return random_sentence
 
-def store_submission(api: HfApi, sentence: str, translation: str, translation_fr: str):
-
+def store_submission(api: HfApi, sentence: str, translation: str, translation_fr: str): 
+    """
+    Append input/outputs and user feedback to a JSON Lines file 
+    using a thread lock to avoid concurrent writes from different users.
+    """
     ts = dt.datetime.now().strftime("%Y-%m-%d_%H-%M-%S-%f")
-    folder_path = "submissions"
-    os.makedirs(folder_path, exist_ok=True)
-    filename = os.path.join(folder_path, f"submissions_{ts}.txt")
-
-    with open(filename, "w", encoding="utf-8") as f:
-        f.write(f"darija,eng,darija_ar\n{sentence},{translation},{translation_fr}")
-
-    print(filename)
-    with open(filename, 'r') as f:
-        print(f.read())
+    # folder_path = "submissions"
+    # os.makedirs(folder_path, exist_ok=True)
+    # filename = os.path.join(folder_path, f"submissions_{ts}.txt")
 
-
-    # api.upload_folder(
-    #         folder_path=folder_path,
-    #         path_in_repo=folder_path,
-    #         repo_id=REPO_ID,
-    #         repo_type="dataset",
-    #         commit_message="New submission",
-    #     )
+    # with open(filename, "w", encoding="utf-8") as f:
+    #     f.write(f"darija,eng,darija_ar\n{sentence},{translation},{translation_fr}")
         
-    api.upload_file(
-                path_or_fileobj=filename,
-                path_in_repo=filename,
-                repo_id=REPO_ID,
-                repo_type="dataset",
-            )
-    
+    # api.upload_file(
+    #             path_or_fileobj=filename,
+    #             path_in_repo=filename,
+    #             repo_id=REPO_ID,
+    #             repo_type="dataset",
+    #         )
+
+    with scheduler.lock:
+        with submissions_file.open("a") as f:
+            f.write(json.dumps({
+                "darija": translation_fr, 
+                "eng": translation, 
+                "darija_ar": sentence}))
+            f.write("\n")
+
     st.success(
         f"""Translation submitted successfully to 
-        {DATASET_REPO_URL}/tree/main/{folder_path}"""
+        {DATASET_REPO_URL}/tree/main/{submissions_folder}"""
         )
 
 
@@ -89,7 +113,7 @@ The translated sentence will be submitted to the dataset
 [here](https://huggingface.co/datasets/imomayiz/darija-english)."""
 )
 
-st.text("")
+st.divider()
 
 st.write(f"""
     <div style="
@@ -126,7 +150,7 @@ with translation_input_placeholder_fr.container():
     st.session_state.translation_input_fr = translation_input_fr
 
 if st.button("Submit Translation"):
-    if not st.session_state.translation_input_fr and st.session_state.translation_input:
+    if not st.session_state.translation_input_fr or st.session_state.translation_input:
         st.warning("Please enter a translation before submitting.")
     else:
         store_submission(api,