Update app.py

app.py

@@ -6,6 +6,7 @@ import random
 import shutil
 import time
 import collections
+from functools iport wraps
 from filelock import FileLock
 from datasets import load_dataset, Audio
 from huggingface_hub import HfApi
@@ -362,7 +363,94 @@ def navigate_dimensions(direction, q_idx, d_idx, selections, *slider_values):
             next_btn_update,
         ) + tuple(slider_updates)
 
-def submit_question_and_advance(q_idx, d_idx, selections, final_choice, all_results, user_data):
+# ==============================================================================
+# 重连函数定义 (Retry Function Definitions)
+# ==============================================================================
+
+# Function for handling connection error
+def retry_with_timeout(max_retries=3, timeout=10, backoff=1):
+    def decorator(func):
+        @wraps(func)
+        def wrapper(*args, **kwargs):
+            last_exception = None
+            for attempt in range(max_retries):
+                try:
+                    import signal
+                    class TimeoutError(Exception):
+                        update_test_dimension_view
+
+                    def handle_timeout(signum, frame):
+                        raise TimeoutError("Operation timed out")
+
+                    signal.signal(signal.SIGALRM, handle_timeout)
+                    signal.alarm(timeout)
+
+                    try:
+                        result = func(*args, **kwargs)
+                        signal.alarm(0) # Cancel the clocker
+                        return result
+                    except TimeoutError:
+                        signal.alarm(0)
+                        raise
+                except Exception as e:
+                    last_exception = e
+                    print(f"Attempt {attempt + 1} failed: {str(e)}")
+                    if attempt < max_retries - 1:
+                        time.sleep(backoff * (attempt + 1))
+                finally:
+                    signal.alarm(0) # Guarantee that the clocker will be canceled
+
+            print(f"All {max_retries} attempts failed")
+            if last_exception:
+                raise last_exception
+            raise Exception("Unkown error occured")
+        return wrapper
+    return decorator
+
+def save_with_retry(all_results, user_data, count_data):
+    try:
+        save_all_results_to_file(all_results, user_data, count_data)
+        return True
+    except Exception as e:
+        print(f"Fail to upload file to HuggingFace Dataset: {e}")
+        return False
+
+def save_locally_with_retry(data, filename, max_retries=3):
+    for attempt in range(max_retries):
+        try:
+            with open(filename, 'w', encoding='utf-8') as f:
+                json.dump(data, f, indent=4, ensure_ascii=False)
+            return True
+        except Exception as e:
+            print(f"Fail to save file to HugginigFace workspace: {e} for {attempt + 1} time")
+            if attempt < max_retries - 1:
+                time.sleep(1)
+    return False
+
+def update_count_with_retry(count_data, question_set, max_retries=3):
+    for attempt in range(max_retries):
+        try:
+            lock_path = COUNT_JSON_PATH + ".lock"
+            with FileLock(lock_path, timeout=10):
+                # Remove unfinished question(s) from count.json
+                for question in question_set:
+                    filename = os.path.basename(question['audio'])
+                    if filename in count_data and count_data[filename] < 1:
+                        count_data[filename] = 0 # Mark unfinished data as 0
+
+                with open(COUNT_JSON_PATH, 'w', encoding='utf-8') as f:
+                    json.dump(count_data, f, indent=4, ensure_ascii=False)
+            return True
+        except Exception as e:
+            print(f"Fail to update count.json {e} for {attempt + 1} time")
+            if attempt < max_retries - 1:
+                time.sleep(1)
+    return False
+
+# ==============================================================================
+
+# Previous version of submit_question_and_advance
+"""def submit_question_and_advance(q_idx, d_idx, selections, final_choice, all_results, user_data):
     # selections["final_choice"] = final_choice
     
     """final_question_result = {
@@ -395,11 +483,12 @@ def submit_question_and_advance(q_idx, d_idx, selections, final_choice, all_resu
     all_results.append(final_question_result)
     
     q_idx += 1
-    
-    # if q_idx < len(QUESTION_SET):
+
+    # If q_idx hasn't reached the last one
     if q_idx < len(user_data["question_set"]):
-        init_q_updates = init_test_question(user_data, q_idx)
+        init_q_updates = init_test_question(user_data, q_idx) # Case 1: jam happens when initialize next question
         return init_q_updates + (all_results, gr.update(value=""))
+    # If q_idx has reached the last one
     else:
         result_str = "### 测试全部完成！\n\n你的提交结果概览：\n"
         for res in all_results:
@@ -420,9 +509,92 @@ def submit_question_and_advance(q_idx, d_idx, selections, final_choice, all_resu
             q_idx, d_idx, {},
             gr.update(), gr.update(), gr.update(), gr.update(), gr.update(),
             gr.update(), gr.update(),
-        ) + (gr.update(),) * MAX_SUB_DIMS + (all_results, result_str)
+        ) + (gr.update(),) * MAX_SUB_DIMS + (all_results, result_str)"""
 
-def save_all_results_to_file(all_results, user_data, count_data=None):
+def submit_question_and_advance(q_idx, d_idx, selections, final_choice, all_results, user_data):
+    try:
+        # Data preperation
+        cleaned_selections = {}
+        for dim_title, sub_scores in selections.items():
+            cleaned_selections["final_choice"] = final_choice
+            cleaned_sub_scores = {}
+            for sub_dim, score in sub_scores.items():
+                cleaned_sub_scores[sub_dim] = None if score == 0 else score
+            cleaned_selections[dim_title] = cleaned_sub_scores
+
+        final_question_result = {
+            "question_id": q_idx,
+            "audio_file": user_data["question_set"][q_idx]['audio'],
+            "selections": cleaned_selections
+        }
+        
+        all_results.append(final_question_result)
+        q_idx += 1
+
+        if q_idx < len(user_data["question_set"]):
+            init_q_updates = init_test_question(user_data, q_idx)
+            return init_q_updates + (all_results, gr.update(value=""))
+        else:
+            result_str = "### 测试全部完成！\n\n你的提交结果概览：\n"
+            for res in all_results:
+                result_str += f"##### 最终判断: **{res['selections'].get('final_choice', '未选择')}**\n"
+                for dim_title, dim_data in res['selections'].items():
+                    if dim_title == 'final_choice': continue
+                    result_str += f"- **{dim_title}**:\n"
+                    for sub_dim, score in dim_data.items():
+                        result_str += f"  - *{sub_dim[:20]}...*: {score}/5\n"
+            
+            # 尝试上传（带重试）
+            success = save_with_retry(all_results, user_data, user_data.get("updated_count_data"))
+            
+            if not success:
+                username = user_data.get("username", "anonymous")
+                timestamp = pd.Timestamp.now().strftime('%Y%m%d_%H%M%S')
+                local_filename = f"submission_{username}_{timestamp}.json"
+ 
+                user_info_clean = {
+                    k: v for k, v in user_data.items() if k not in ["question_set", "updated_count_data"]
+                }
+                final_data_package = {
+                    "user_info": user_info_clean,
+                    "results": all_results
+                }
+                
+                # 尝试保存到本地
+                local_success = save_locally_with_retry(final_data_package, local_filename)
+                
+                if local_success:
+                    result_str += f"\n\n⚠️ 上传失败，结果已保存到本地文件: {local_filename}"
+                else:
+                    result_str += "\n\n❌ 上传失败且无法保存到本地文件，请联系管理员"
+                
+                # 更新count.json（剔除未完成的题目）
+                count_update_success = update_count_with_retry(
+                    user_data.get("updated_count_data", {}),
+                    user_data["question_set"]
+                )
+                
+                if not count_update_success:
+                    result_str += "\n\n⚠️ 无法更新题目计数，请联系管理员"
+            
+            return (
+                gr.update(visible=False), gr.update(visible=False), gr.update(visible=False), gr.update(visible=True),
+                q_idx, d_idx, {},
+                gr.update(), gr.update(), gr.update(), gr.update(), gr.update(),
+                gr.update(), gr.update(),
+            ) + (gr.update(),) * MAX_SUB_DIMS + (all_results, result_str)
+    except Exception as e:
+        print(f"提交过程中发生错误: {e}")
+        # 返回错误信息
+        error_msg = f"提交过程中发生错误: {str(e)}"
+        return (
+            gr.update(), gr.update(), gr.update(), gr.update(),
+            q_idx, d_idx, selections,
+            gr.update(), gr.update(), gr.update(), gr.update(), gr.update(),
+            gr.update(), gr.update(),
+        ) + (gr.update(),) * MAX_SUB_DIMS + (all_results, error_msg)
+
+"""def save_all_results_to_file(all_results, user_data, count_data=None):
     repo_id = "intersteller2887/Turing-test-dataset"
     username = user_data.get("username", "user")
     timestamp = pd.Timestamp.now().strftime('%Y%m%d_%H%M%S')
@@ -472,8 +644,56 @@ def save_all_results_to_file(all_results, user_data, count_data=None):
             )
 
     except Exception as e:
-        print(f"上传出错: {e}")
+        print(f"上传出错: {e}")"""
 
+@retry_with_timeout(max_retries=3, timeout=10)
+def save_all_results_to_file(all_results, user_data, count_data=None):
+    repo_id = "intersteller2887/Turing-test-dataset"
+    username = user_data.get("username", "user")
+    timestamp = pd.Timestamp.now().strftime('%Y%m%d_%H%M%S')
+    submission_filename = f"submissions_{username}_{timestamp}.json"
+
+    user_info_clean = {
+        k: v for k, v in user_data.items() if k not in ["question_set", "updated_count_data"]
+    }
+    
+    final_data_package = {
+        "user_info": user_info_clean,
+        "results": all_results
+    }
+    json_string = json.dumps(final_data_package, ensure_ascii=False, indent=4)
+    hf_token = os.getenv("HF_TOKEN")
+
+    if not hf_token:
+        raise Exception("HF_TOKEN not found. Cannot upload to the Hub.")
+
+    api = HfApi()
+
+    # 上传提交文件
+    api.upload_file(
+        path_or_fileobj=bytes(json_string, "utf-8"),
+        path_in_repo=f"submissions/{submission_filename}",
+        repo_id=repo_id,
+        repo_type="dataset",
+        token=hf_token,
+        commit_message=f"Add new submission from {username}",
+        timeout=30
+    )
+
+    if count_data:
+        with FileLock(COUNT_JSON_PATH + ".lock", timeout=5):
+            with open(COUNT_JSON_PATH, "w", encoding="utf-8") as f:
+                json.dump(count_data, f, indent=4, ensure_ascii=False)
+    
+        api.upload_file(
+            path_or_fileobj=COUNT_JSON_PATH,
+            path_in_repo=COUNT_JSON_REPO_PATH,
+            repo_id=repo_id,
+            repo_type="dataset",
+            token=hf_token,
+            commit_message=f"Update count.json after submission by {username}",
+            timeout=30
+        )
 
 def toggle_reference_view(current):
     if current == "参考":