Update app.py

app.py

@@ -3,16 +3,17 @@ import os
 import random
 import time
 from datetime import datetime
-# import pandas as pd # 未在后续代码中使用
 from functools import partial
 import json
 import io
 from huggingface_hub import HfApi
 from huggingface_hub.hf_api import HfHubHTTPError
 import traceback
+from itertools import combinations
 
-# ==== 全局配置 ====
-BASE_IMAGE_DIR = "/data/images/images" 
+# ==== 全局配置 (部分保持不变) ====
+# ... (BASE_IMAGE_DIR, TARGET_DIR, METHOD_ROOTS, SUBJECTS, etc. 保持不变) ...
+BASE_IMAGE_DIR = "/data/images/images"
 TARGET_DIR_BASENAME = "gt"
 TARGET_DIR = os.path.join(BASE_IMAGE_DIR, TARGET_DIR_BASENAME)
 
@@ -31,19 +32,64 @@ if os.path.exists(BASE_IMAGE_DIR):
     except Exception as e: print(f"错误：在扫描 '{BASE_IMAGE_DIR}' 时发生错误: {e}"); METHOD_ROOTS = []
 else: print(f"警告：基础目录 '{BASE_IMAGE_DIR}' 不存在。将无法加载候选图片。")
 
-SUBJECTS = ["subj01", "subj02", "subj05", "subj07"] 
+SUBJECTS = ["subj01", "subj02", "subj05", "subj07"]
 SENTINEL_TRIAL_INTERVAL = 20
-NUM_TRIALS_PER_RUN = 100 
-LOG_BATCH_SIZE = 5# <--- 新增：每多少条选择日志进行一次批量上传
+NUM_TRIALS_PER_RUN = 100
+LOG_BATCH_SIZE = 5
 
 DATASET_REPO_ID = "YanmHa/image-aligned-experiment-data"
-INDIVIDUAL_LOGS_FOLDER = "individual_choice_logs" # 此文件夹可能不再主要使用，但变量保留
-BATCH_LOG_FOLDER = "run_logs_batch" 
+INDIVIDUAL_LOGS_FOLDER = "individual_choice_logs"
+BATCH_LOG_FOLDER = "run_logs_batch"
 CSS = ".gr-block {margin-top: 4px !important; margin-bottom: 4px !important;} .compact_button { padding: 4px 8px; min-width: auto; }"
 
-# ==== 加载所有可用的目标图片 ====
+
+# ==== 全局持久化历史记录 ====
+GLOBAL_HISTORY_FILE = "global_experiment_shown_pairs.json"
+global_shown_pairs_cache = {}
+global_history_has_unsaved_changes = False # <--- 新增：跟踪全局历史是否有未保存的更改
+
+def load_global_shown_pairs():
+    global global_shown_pairs_cache, global_history_has_unsaved_changes
+    if os.path.exists(GLOBAL_HISTORY_FILE):
+        try:
+            with open(GLOBAL_HISTORY_FILE, 'r', encoding='utf-8') as f:
+                data_from_file = json.load(f)
+                global_shown_pairs_cache = {
+                    target_img: {frozenset(pair) for pair in pairs_list}
+                    for target_img, pairs_list in data_from_file.items()
+                }
+            print(f"已成功从 '{GLOBAL_HISTORY_FILE}' 加载全局已展示图片对历史。")
+        except Exception as e:
+            print(f"错误：加载全局历史文件 '{GLOBAL_HISTORY_FILE}' 失败: {e}。将使用空历史记录。")
+            global_shown_pairs_cache = {}
+    else:
+        print(f"信息：全局历史文件 '{GLOBAL_HISTORY_FILE}' 未找到。将创建新的空历史记录。")
+        global_shown_pairs_cache = {}
+    global_history_has_unsaved_changes = False # 初始化或加载后，标记为无未保存更改
+
+def save_global_shown_pairs():
+    global global_shown_pairs_cache, global_history_has_unsaved_changes # 确保可以修改这个flag
+    # print("尝试保存全局图片对历史...") # 调试信息
+    try:
+        data_to_save = {
+            target_img: [sorted(list(pair_fset)) for pair_fset in pairs_set]
+            for target_img, pairs_set in global_shown_pairs_cache.items()
+        }
+        temp_file = GLOBAL_HISTORY_FILE + ".tmp"
+        with open(temp_file, 'w', encoding='utf-8') as f:
+            json.dump(data_to_save, f, ensure_ascii=False, indent=2)
+        os.replace(temp_file, GLOBAL_HISTORY_FILE)
+        print(f"已成功将全局已展示图片对历史保存到 '{GLOBAL_HISTORY_FILE}'。")
+        global_history_has_unsaved_changes = False # 保存成功后重置标志
+        return True
+    except Exception as e:
+        print(f"错误：保存全局历史文件 '{GLOBAL_HISTORY_FILE}' 失败: {e}")
+        return False
+
+load_global_shown_pairs()
+
+# ==== 加载所有可用的目标图片 (保持不变) ====
 master_image_list = []
-# ... (master_image_list 加载逻辑与您上一版代码相同) ...
 if os.path.exists(TARGET_DIR):
     try:
         master_image_list = sorted(
@@ -60,120 +106,169 @@ if not master_image_list: print(f"关键错误：由于 '{TARGET_DIR}' 问题，
 
 # ==== 辅助函数 ====
 def get_next_trial_info(current_trial_idx_in_run, current_run_image_list_for_trial, num_trials_in_this_run_for_trial):
-    # ... (此函数与您上一版代码完全一致) ...
-    global TARGET_DIR, METHOD_ROOTS, SUBJECTS, SENTINEL_TRIAL_INTERVAL
-    if not current_run_image_list_for_trial or current_trial_idx_in_run >= num_trials_in_this_run_for_trial: return None, current_trial_idx_in_run
+    global TARGET_DIR, METHOD_ROOTS, SUBJECTS, SENTINEL_TRIAL_INTERVAL, global_shown_pairs_cache, global_history_has_unsaved_changes
+
+    # ... (函数开始部分与上一版相同，直到哨兵试验逻辑结束) ...
+    if not current_run_image_list_for_trial or current_trial_idx_in_run >= num_trials_in_this_run_for_trial:
+        return None, current_trial_idx_in_run
+
     img_filename_original = current_run_image_list_for_trial[current_trial_idx_in_run]
     target_full_path = os.path.join(TARGET_DIR, img_filename_original)
     trial_number_for_display = current_trial_idx_in_run + 1
+
     pool = []
     for m_root_path in METHOD_ROOTS:
         method_name = os.path.basename(m_root_path)
         for s_id in SUBJECTS:
-            base, ext = os.path.splitext(img_filename_original); reconstructed_filename = f"{base}_0{ext}"
+            base, ext = os.path.splitext(img_filename_original)
+            reconstructed_filename = f"{base}_0{ext}"
             candidate_path = os.path.join(m_root_path, s_id, reconstructed_filename)
-            if os.path.exists(candidate_path): pool.append((f"{method_name}/{s_id}/{reconstructed_filename}", candidate_path))
+            if os.path.exists(candidate_path):
+                internal_label = f"{method_name}/{s_id}/{reconstructed_filename}"
+                pool.append((internal_label, candidate_path))
+
     trial_info = {"image_id": img_filename_original, "target_path": target_full_path, "cur_no": trial_number_for_display, "is_sentinel": False,
                   "left_display_label": "N/A", "left_internal_label": "N/A", "left_path": None,
                   "right_display_label": "N/A", "right_internal_label": "N/A", "right_path": None}
+
     is_potential_sentinel_trial = (trial_number_for_display > 0 and trial_number_for_display % SENTINEL_TRIAL_INTERVAL == 0)
+
     if is_potential_sentinel_trial:
-        if not pool: print(f"警告：哨兵图 '{img_filename_original}' (trial {trial_number_for_display}) 无候选。")
+        # 哨兵试验逻辑不变
+        if not pool:
+            print(f"警告：哨兵图 '{img_filename_original}' (trial {trial_number_for_display}) 无候选。")
         else:
             print(f"生成哨兵试验 for '{img_filename_original}' (trial {trial_number_for_display})")
             trial_info["is_sentinel"] = True
             sentinel_candidate_target_tuple = ("目标图像", target_full_path)
             random_reconstruction_candidate_tuple = random.choice(pool)
             candidates_for_sentinel = [
-                (("目标图像", target_full_path), sentinel_candidate_target_tuple[0]), 
+                (("目标图像", target_full_path), sentinel_candidate_target_tuple[0]),
                 (("重建图", random_reconstruction_candidate_tuple[1]), random_reconstruction_candidate_tuple[0])
             ]
             random.shuffle(candidates_for_sentinel)
             trial_info.update({
-                "left_display_label": candidates_for_sentinel[0][0][0], "left_path": candidates_for_sentinel[0][0][1], "left_internal_label": candidates_for_sentinel[0][1],
-                "right_display_label": candidates_for_sentinel[1][0][0], "right_path": candidates_for_sentinel[1][0][1], "right_internal_label": candidates_for_sentinel[1][1],
+                "left_display_label": candidates_for_sentinel[0][0][0],
+                "left_path": candidates_for_sentinel[0][0][1],
+                "left_internal_label": candidates_for_sentinel[0][1],
+                "right_display_label": candidates_for_sentinel[1][0][0],
+                "right_path": candidates_for_sentinel[1][0][1],
+                "right_internal_label": candidates_for_sentinel[1][1],
             })
-    if not trial_info["is_sentinel"]:
-        if len(pool) < 2: print(f"警告：常规图 '{img_filename_original}' (trial {trial_number_for_display}) 候选少于2 (找到 {len(pool)})。"); return None, current_trial_idx_in_run
-        selected_candidates_tuples = random.sample(pool, 2)
-        if random.random() > 0.5: selected_candidates_tuples = selected_candidates_tuples[::-1]
+    else: # 常规试验
+        if len(pool) < 2:
+            print(f"警告：常规图 '{img_filename_original}' (trial {trial_number_for_display}) 候选少于2 (找到 {len(pool)})。此试验无法进行。")
+            return None, current_trial_idx_in_run
+
+        target_global_history_set = global_shown_pairs_cache.setdefault(img_filename_original, set())
+        all_possible_pairs_in_pool = []
+        for c1, c2 in combinations(pool, 2):
+            pair_labels_fset = frozenset({c1[0], c2[0]})
+            all_possible_pairs_in_pool.append( ((c1, c2), pair_labels_fset) )
+
+        unseen_globally_pairs_with_data = [
+            item for item in all_possible_pairs_in_pool
+            if item[1] not in target_global_history_set
+        ]
+        selected_candidates_tuples = None
+
+        if unseen_globally_pairs_with_data:
+            chosen_pair_data_and_labels = random.choice(unseen_globally_pairs_with_data)
+            selected_candidates_tuples = chosen_pair_data_and_labels[0]
+            chosen_pair_frozenset = chosen_pair_data_and_labels[1]
+            target_global_history_set.add(chosen_pair_frozenset)
+            global_history_has_unsaved_changes = True # <--- 标记全局历史已更新
+            # 不再在此处调用 save_global_shown_pairs()
+            # print(f"调试：目标 '{img_filename_original}': 新全局唯一对 {chosen_pair_frozenset} 已添加至缓存。未保存更改标志: {global_history_has_unsaved_changes}")
+        else:
+            # ... (处理所有对都已展示过的情况，与上一版相同) ...
+            print(f"警告：目标图 '{img_filename_original}': 来自当前池的所有 ({len(all_possible_pairs_in_pool)}) 个候选对均已在全局展示过。")
+            if all_possible_pairs_in_pool:
+                print("将随机选择一个重复的对（全局重复）。")
+                chosen_pair_data_and_labels = random.choice(all_possible_pairs_in_pool)
+                selected_candidates_tuples = chosen_pair_data_and_labels[0]
+            else:
+                print(f"错误：即使允许全局重复，也无法从池中选择图片对（池大小 {len(pool)}）。")
+                return None, current_trial_idx_in_run
+
+        display_order_candidates = list(selected_candidates_tuples)
+        if random.random() > 0.5:
+            display_order_candidates = display_order_candidates[::-1]
+
         trial_info.update({
-            "left_display_label": "候选图 1", "left_path": selected_candidates_tuples[0][1], "left_internal_label": selected_candidates_tuples[0][0],
-            "right_display_label": "候选图 2", "right_path": selected_candidates_tuples[1][1], "right_internal_label": selected_candidates_tuples[1][0],
+            "left_display_label": "候选图 1", "left_path": display_order_candidates[0][1], "left_internal_label": display_order_candidates[0][0],
+            "right_display_label": "候选图 2", "right_path": display_order_candidates[1][1], "right_internal_label": display_order_candidates[1][0],
         })
     return trial_info, current_trial_idx_in_run + 1
 
-# def save_single_log_to_hf_dataset(log_entry, user_identifier_str): # <--- 不再需要此函数，注释或删除
-#     pass 
-
-# --- 批量保存累积日志的函数 ---
-def save_collected_logs_batch(list_of_log_entries, user_identifier_str, batch_identifier): # batch_identifier 可以是轮次号或时间戳
+# ==== 批量保存用户选择日志函数 (保持不变) ====
+def save_collected_logs_batch(list_of_log_entries, user_identifier_str, batch_identifier):
     global DATASET_REPO_ID, BATCH_LOG_FOLDER
     if not list_of_log_entries:
-        print("批量保存：没有累积的日志。")
-        return True 
+        print("批量保存用户日志：没有累积的日志。")
+        return True # 认为无日志即成功
 
     identifier_safe = str(user_identifier_str if user_identifier_str else "unknown_user_session").replace('.', '_').replace(':', '_').replace('/', '_').replace(' ', '_')
-    print(f"用户 {identifier_safe} - 准备批量保存 {len(list_of_log_entries)} 条日志 (批次标识: {batch_identifier})...")
-    
+    print(f"用户 {identifier_safe} - 准备批量保存 {len(list_of_log_entries)} 条选择日志 (批次标识: {batch_identifier})...")
     try:
         token = os.getenv("HF_TOKEN")
-        if not token: print("错误：HF_TOKEN 未设置。无法批量保存日志。"); return False
-        if not DATASET_REPO_ID: print("错误：DATASET_REPO_ID 未配置。无法批量保存日志。"); return False
+        if not token: print("错误：HF_TOKEN 未设置。无法批量保存选择日志。"); return False
+        if not DATASET_REPO_ID: print("错误：DATASET_REPO_ID 未配置。无法批量保存选择日志。"); return False
 
         api = HfApi(token=token)
         timestamp_str = datetime.now().strftime('%Y%m%d_%H%M%S_%f')
         batch_filename = f"batch_user-{identifier_safe}_id-{batch_identifier}_{timestamp_str}_logs-{len(list_of_log_entries)}.jsonl"
-        path_in_repo = f"{BATCH_LOG_FOLDER}/{identifier_safe}/{batch_filename}" 
-
+        path_in_repo = f"{BATCH_LOG_FOLDER}/{identifier_safe}/{batch_filename}"
         jsonl_content = ""
         for log_entry in list_of_log_entries:
             try:
                 if isinstance(log_entry, dict): jsonl_content += json.dumps(log_entry, ensure_ascii=False) + "\n"
-                else: print(f"警告：批量保存时，日志条目非字典：{log_entry}")
+                else: print(f"警告：批量保存选择日志时，条目非字典：{log_entry}")
             except Exception as json_err:
-                print(f"错误：批量保存序列化单条日志时出错: {log_entry}. 错误: {json_err}")
-                jsonl_content += json.dumps({"error": "serialization_failed_in_batch", "original_data_preview": str(log_entry)[:100],"timestamp": datetime.now().isoformat()}, ensure_ascii=False) + "\n"
-        
-        if not jsonl_content.strip(): print(f"用户 {identifier_safe} (批次 {batch_identifier}) 无可序列化日志。"); return True
+                print(f"错误：批量保存选择日志序列化单条时出错: {log_entry}. 错误: {json_err}")
+                jsonl_content += json.dumps({"error": "serialization_failed_in_batch_user_log", "original_data_preview": str(log_entry)[:100],"timestamp": datetime.now().isoformat()}, ensure_ascii=False) + "\n"
+        if not jsonl_content.strip(): print(f"用户 {identifier_safe} (批次 {batch_identifier}) 无可序列化选择日志。"); return True
 
         log_bytes = jsonl_content.encode('utf-8')
         file_like_object = io.BytesIO(log_bytes)
-
-        print(f"准备批量上传日志文件: {path_in_repo} ({len(log_bytes)} bytes)")
+        print(f"准备批量上传选择日志文件: {path_in_repo} ({len(log_bytes)} bytes)")
         api.upload_file(
             path_or_fileobj=file_like_object, path_in_repo=path_in_repo, repo_id=DATASET_REPO_ID, repo_type="dataset",
-            commit_message=f"Batch logs for user {identifier_safe}, batch_id {batch_identifier} ({len(list_of_log_entries)} entries)"
+            commit_message=f"Batch user choice logs for {identifier_safe}, batch_id {batch_identifier} ({len(list_of_log_entries)} entries)"
         )
-        print(f"批量日志已成功保存到 HF Dataset: {DATASET_REPO_ID}/{path_in_repo}")
+        print(f"批量选择日志已成功保存到 HF Dataset: {DATASET_REPO_ID}/{path_in_repo}")
         return True
     except Exception as e:
-        print(f"批量保存日志 (user {identifier_safe}, batch_id {batch_identifier}) 失败: {e}"); traceback.print_exc()
+        print(f"批量保存选择日志 (user {identifier_safe}, batch_id {batch_identifier}) 失败: {e}"); traceback.print_exc()
         return False
 
+
 # ==== 主要的 Gradio 事件处理函数 ====
 def process_experiment_step(
     s_trial_idx_val, s_run_no_val, s_user_logs_val, s_current_trial_data_val, s_user_session_id_val,
     s_current_run_image_list_val, s_num_trials_this_run_val,
     action_type=None, choice_value=None, request: gr.Request = None
 ):
-    global master_image_list, NUM_TRIALS_PER_RUN, outputs_ui_components_definition, LOG_BATCH_SIZE
+    global master_image_list, NUM_TRIALS_PER_RUN, outputs_ui_components_definition, LOG_BATCH_SIZE, global_history_has_unsaved_changes
 
+    # ... (函数开始部分与上一版类似) ...
     output_s_trial_idx = s_trial_idx_val; output_s_run_no = s_run_no_val
     output_s_user_logs = list(s_user_logs_val); output_s_current_trial_data = dict(s_current_trial_data_val) if s_current_trial_data_val else {}
     output_s_user_session_id = s_user_session_id_val; output_s_current_run_image_list = list(s_current_run_image_list_val)
     output_s_num_trials_this_run = s_num_trials_this_run_val
     user_ip_fallback = request.client.host if request else "unknown_ip"
     user_identifier_for_logging = output_s_user_session_id if output_s_user_session_id else user_ip_fallback
+
     len_ui_outputs = len(outputs_ui_components_definition)
     def create_ui_error_tuple(message, progress_msg_text): return (gr.update(visible=False),) * 3 + ("", "", message, progress_msg_text) + (gr.update(interactive=True), gr.update(interactive=False), gr.update(interactive=False)) + (gr.update(visible=False),)
     def create_no_change_tuple(): return (gr.update(),) * len_ui_outputs
     user_id_display_text = output_s_user_session_id if output_s_user_session_id else "用户ID待分配"
 
+
     if action_type == "record_choice":
+        # ... (日志记录逻辑与上一版相同) ...
         if output_s_current_trial_data.get("data") and output_s_current_trial_data["data"].get("left_internal_label"):
             chosen_internal_label = (output_s_current_trial_data["data"]["left_internal_label"] if choice_value == "left" else output_s_current_trial_data["data"]["right_internal_label"])
-            # ... (log_entry 创建逻辑，与上一版一致，包含 chosen_method, chosen_subject, chosen_filename) ...
             parsed_chosen_method, parsed_chosen_subject, parsed_chosen_filename = "N/A", "N/A", "N/A"
             if chosen_internal_label == "目标图像": parsed_chosen_method, parsed_chosen_subject, parsed_chosen_filename = "TARGET", "GT", output_s_current_trial_data["data"]["image_id"]
             else:
@@ -182,84 +277,102 @@ def process_experiment_step(
                 elif len(parts) == 2: parsed_chosen_method, parsed_chosen_subject = parts[0].strip(), parts[1].strip()
                 elif len(parts) == 1: parsed_chosen_method = parts[0].strip()
             log_entry = {
-                "timestamp": datetime.now().isoformat(), "user_identifier": user_identifier_for_logging, "run_no": output_s_run_no, 
+                "timestamp": datetime.now().isoformat(), "user_identifier": user_identifier_for_logging, "run_no": output_s_run_no,
                 "image_id": output_s_current_trial_data["data"]["image_id"],
-                "left_internal_label": output_s_current_trial_data["data"]["left_internal_label"], 
+                "left_internal_label": output_s_current_trial_data["data"]["left_internal_label"],
                 "right_internal_label": output_s_current_trial_data["data"]["right_internal_label"],
-                "chosen_side": choice_value, "chosen_internal_label": chosen_internal_label, 
+                "chosen_side": choice_value, "chosen_internal_label": chosen_internal_label,
                 "chosen_method": parsed_chosen_method, "chosen_subject": parsed_chosen_subject, "chosen_filename": parsed_chosen_filename,
-                "trial_sequence_in_run": output_s_current_trial_data["data"]["cur_no"], 
+                "trial_sequence_in_run": output_s_current_trial_data["data"]["cur_no"],
                 "is_sentinel": output_s_current_trial_data["data"]["is_sentinel"]
             }
             output_s_user_logs.append(log_entry)
             print(f"用户 {user_identifier_for_logging} 记录选择 (img: {log_entry['image_id']})。当前批次日志数: {len(output_s_user_logs)}")
 
-            # !!! 修改：每 LOG_BATCH_SIZE 条选择后尝试批量保存 !!!
+
+            # !!! 修改：当用户日志达到批量大小时，同时尝试保存全局历史（如果它有更改）!!!
             if len(output_s_user_logs) >= LOG_BATCH_SIZE:
-                print(f"累积日志达到 {LOG_BATCH_SIZE} 条，准备批量保存...")
-                # 使用当前轮次号和时间戳组合作为批次标识符，或者仅用时间戳
-                batch_id_for_filename = f"run{output_s_run_no}_trial{output_s_trial_idx}_count{len(output_s_user_logs)}"
-                save_success = save_collected_logs_batch(list(output_s_user_logs), user_identifier_for_logging, batch_id_for_filename)
-                if save_success:
-                    print("批量日志已成功（或尝试）保存，将清空累积日志列表。")
-                    output_s_user_logs = [] # 清空已保存的日志
+                print(f"累积用户选择日志达到 {LOG_BATCH_SIZE} 条，准备批量保存...")
+                batch_id_for_filename = f"run{output_s_run_no}_trialidx{output_s_trial_idx}_logcount{len(output_s_user_logs)}"
+                
+                # 1. 保存用户选择日志
+                user_logs_save_success = save_collected_logs_batch(list(output_s_user_logs), user_identifier_for_logging, batch_id_for_filename)
+                if user_logs_save_success:
+                    print("批量用户选择日志已成功（或尝试）保存，将清空累积的用户选择日志列表。")
+                    output_s_user_logs = [] # 清空已保存的用户日志
                 else:
-                    print("警告：批量日志保存失败。日志将继续累积，下次达到阈值时重试。")
-        else:
-            # ... (错误处理) ...
-            print(f"用户 {user_identifier_for_logging} 错误：记录选择时数据为空！") # ... (rest of error handling) ...
+                    print("警告：批量用户选择日志保存失败。选择日志将继续累积，下次达到阈值时重试。")
+
+                # 2. 检查并保存全局图片对历史（如果自上次保存后有更改）
+                if global_history_has_unsaved_changes:
+                    print("检测到全局图片对历史自上次保存后有更新，将一并保存...")
+                    save_global_shown_pairs() # 此函数内部会在成功保存后将 global_history_has_unsaved_changes 置为 False
+                else:
+                    print("全局图片对历史自上次保存后无更新，无需保存。")
+        else: # 处理记录选择时数据为空的错误
+            print(f"用户 {user_identifier_for_logging} 错误：记录选择时数据为空！")
             error_ui_updates = create_ui_error_tuple("记录选择时内部错误。", f"用户ID: {user_id_display_text} | 进度：{output_s_trial_idx}/{output_s_num_trials_this_run}")
+            # 返回所有状态变量的当前值以及错误UI更新
             return output_s_trial_idx, output_s_run_no, output_s_user_logs, output_s_current_trial_data, output_s_user_session_id, output_s_current_run_image_list, output_s_num_trials_this_run, *error_ui_updates
 
-    # --- "start_experiment" 逻辑 ---
+    # ... (start_experiment 逻辑与上一版相同) ...
     if action_type == "start_experiment":
         is_first = (output_s_num_trials_this_run == 0 and output_s_trial_idx == 0 and output_s_run_no == 1)
         is_completed_for_restart = (output_s_num_trials_this_run > 0 and output_s_trial_idx >= output_s_num_trials_this_run)
         if is_first or is_completed_for_restart:
             if not master_image_list: error_ui = create_ui_error_tuple("错误: 无可用目标图片!", f"用户ID: {user_id_display_text} | 进度: 0/0"); return 0, output_s_run_no, output_s_user_logs, {}, output_s_user_session_id, [], 0, *error_ui
-            
-            # 如果是完成一轮后重启，并且此时累积的日志未达到 LOG_BATCH_SIZE，它们会保留到下一批
-            # 轮次号的增加在这里是正确的，用于显示
-            if is_completed_for_restart: 
+
+            if is_completed_for_restart:
                 output_s_run_no += 1
-            
+
             num_avail = len(master_image_list); run_size = min(num_avail, NUM_TRIALS_PER_RUN)
             if run_size == 0: error_ui = create_ui_error_tuple("错误: 采样图片数为0!", f"用户ID: {user_id_display_text} | 进度: 0/0"); return 0, output_s_run_no, output_s_user_logs, {}, output_s_user_session_id, [], 0, *error_ui
-            
+
             output_s_current_run_image_list = random.sample(master_image_list, run_size)
             output_s_num_trials_this_run = run_size
             output_s_trial_idx = 0
-            # output_s_user_logs 不在这里重置，它会持续累积直到达到 LOG_BATCH_SIZE
             output_s_current_trial_data = {}
-            
-            # 用户会话ID在欢迎页已设置，这里不再修改，除非有特殊需求
             print(f"开始/继续轮次 {output_s_run_no} (用户ID: {output_s_user_session_id}). 随机选择 {output_s_num_trials_this_run} 张图片.")
-        else: # 在轮次中途点击开始
+        else:
             print(f"用户 {user_identifier_for_logging} 在第 {output_s_run_no} 轮，试验 {output_s_trial_idx} 点击开始，但轮次未完成。忽略。")
             no_change_ui = create_no_change_tuple()
             return output_s_trial_idx, output_s_run_no, output_s_user_logs, output_s_current_trial_data, output_s_user_session_id, output_s_current_run_image_list, output_s_num_trials_this_run, *no_change_ui
-            
-    # --- 轮次结束的UI处理 (不再包含批量保存逻辑) ---
+
+    # ... (轮次结束处理与上一版相同) ...
     if output_s_trial_idx >= output_s_num_trials_this_run and output_s_num_trials_this_run > 0:
-        # 仅打印完成信息，实际的批量保存由 record_choice 中的计数器触发
         print(f"用户 {output_s_user_session_id} 已完成第 {output_s_run_no} 轮。等待下一批或下一轮开始。")
+        # 检查是否有未保存的全局历史，即使日志批次未满，也可能在轮次结束时考虑保存
+        # 但当前逻辑是仅在日志批次满时保存全局历史，这里可以保持一致或添加额外逻辑
+        if global_history_has_unsaved_changes:
+             print(f"提示：轮次 {output_s_run_no} 结束，仍有未保存的全局图片对历史更改。将在下次日志批量保存时一并处理。")
+
         prog_text = f"用户ID: {output_s_user_session_id} | 进度：{output_s_num_trials_this_run}/{output_s_num_trials_this_run} | 第 {output_s_run_no} 轮 🎉"
         ui_updates = list(create_ui_error_tuple(f"🎉 第 {output_s_run_no} 轮完成！请点击“开始试验 / 下一轮”继续或开始新批次。", prog_text))
-        # ... (UI 更新与之前一致) ...
         ui_updates[7]=gr.update(interactive=True); ui_updates[8]=gr.update(interactive=False); ui_updates[9]=gr.update(interactive=False)
         ui_updates[0]=gr.update(value=None,visible=False); ui_updates[1]=gr.update(value=None,visible=False); ui_updates[2]=gr.update(value=None,visible=False)
         yield output_s_trial_idx, output_s_run_no, output_s_user_logs, output_s_current_trial_data, output_s_user_session_id, output_s_current_run_image_list, output_s_num_trials_this_run, *ui_updates; return
 
-    # --- 获取并显示下一个试验 (与之前一致) ---
-    # ... (省略这部分代码，与您上一版相同，确保 prog_text 使用了更新后的用户ID格式) ...
+
+    # ... (获取并显示下一个试验的逻辑，与上一版相同，调用 get_next_trial_info 不需要 user_id for history) ...
     if not output_s_current_run_image_list or output_s_num_trials_this_run == 0:
         error_ui = create_ui_error_tuple("错误: 无法加载试验图片 (列表为空)", f"用户ID: {user_id_display_text} | 进度: N/A")
         return output_s_trial_idx, output_s_run_no, output_s_user_logs, {"data": None}, output_s_user_session_id, [], 0, *error_ui
-    trial_info, next_s_trial_idx_for_state = get_next_trial_info(output_s_trial_idx, output_s_current_run_image_list, output_s_num_trials_this_run)
+
+    trial_info, next_s_trial_idx_for_state = get_next_trial_info(
+        output_s_trial_idx,
+        output_s_current_run_image_list,
+        output_s_num_trials_this_run
+    )
+
     if trial_info is None:
-        error_ui = create_ui_error_tuple("错误：无法加载试验图片。", f"用户ID: {user_id_display_text} | 进度：{output_s_trial_idx}/{output_s_num_trials_this_run}")
+        print(f"错误：用户 {user_identifier_for_logging}，轮次 {output_s_run_no}，试验 {output_s_trial_idx}: get_next_trial_info 返回 None。")
+        error_msg_display = "无法加载下一个试验，可能是因为候选图片不足或所有唯一组合已用尽。"
+        if len(METHOD_ROOTS) * len(SUBJECTS) < 2 :
+            error_msg_display = "候选图片来源不足，无法形成对比试验。"
+        error_ui_updates = create_ui_error_tuple(error_msg_display, f"用户ID: {user_id_display_text} | 进度：{output_s_trial_idx}/{output_s_num_trials_this_run}")
         output_s_current_trial_data = {"data": None}
-        return output_s_trial_idx, output_s_run_no, output_s_user_logs, output_s_current_trial_data, output_s_user_session_id, output_s_current_run_image_list, output_s_num_trials_this_run, *error_ui
+        return output_s_trial_idx, output_s_run_no, output_s_user_logs, output_s_current_trial_data, output_s_user_session_id, output_s_current_run_image_list, output_s_num_trials_this_run, *error_ui_updates
+
     output_s_current_trial_data = {"data": trial_info}
     prog_text = f"用户ID: {output_s_user_session_id} | 进度：{trial_info['cur_no']}/{output_s_num_trials_this_run} | 第 {output_s_run_no} 轮"
     ui_show_target_updates = list(create_no_change_tuple())
@@ -275,7 +388,9 @@ def process_experiment_step(
     ui_show_candidates_updates[7]=gr.update(interactive=False); ui_show_candidates_updates[8]=gr.update(interactive=True); ui_show_candidates_updates[9]=gr.update(interactive=True)
     yield next_s_trial_idx_for_state, output_s_run_no, output_s_user_logs, output_s_current_trial_data, output_s_user_session_id, output_s_current_run_image_list, output_s_num_trials_this_run, *ui_show_candidates_updates
 
-# ==== Gradio UI 定义 ====
+
+# ==== Gradio UI 定义 和 程序入口 (保持不变) ====
+# ... (welcome_page_markdown, handle_agree_and_start, gr.Blocks, if __name__ == "__main__": etc.) ...
 welcome_page_markdown = """
 ## 欢迎加入实验！
 您好！非常感谢您抽出宝贵时间参与我们的视觉偏好评估实验。您的选择将帮助我们改进重建算法，让机器生成的图像更贴近人类视觉体验!
@@ -303,7 +418,6 @@ welcome_page_markdown = """
 再次感谢您的参与与支持！您每一次认真选择都对我们的研究意义重大。祝您一切顺利，实验愉快！
 """
 def handle_agree_and_start(name, gender, age, education, request: gr.Request):
-    # ... (此函数与您上一版代码完全一致) ...
     error_messages_list = []
     if not name or str(name).strip() == "": error_messages_list.append("姓名 不能为空。")
     if gender is None or str(gender).strip() == "": error_messages_list.append("性别 必须选择。")
@@ -311,32 +425,30 @@ def handle_agree_and_start(name, gender, age, education, request: gr.Request):
     elif not (isinstance(age, (int, float)) and 1 <= age <= 120):
         try: num_age = float(age);
         except (ValueError, TypeError): error_messages_list.append("年龄必须是一个有效的数字。")
-        else: 
+        else:
             if not (1 <= num_age <= 120): error_messages_list.append("年龄必须在 1 到 120 之间。")
-    if education is None or str(education).strip() == "其他": error_messages_list.append("学历 必须选择。") # 修正：如果选“其他”但希望是有效输入，这里的逻辑可能需要调整
+    if education is None or str(education).strip() == "其他": error_messages_list.append("学历 必须选择。")
     if error_messages_list:
         full_error_message = "请修正以下错误：\n" + "\n".join([f"- {msg}" for msg in error_messages_list])
         print(f"用户输入验证失败: {full_error_message}")
         return gr.update(), False, gr.update(visible=True), gr.update(visible=False), full_error_message
     s_name = str(name).strip().replace(" ","_").replace("/","_").replace("\\","_")
     s_gender = str(gender).strip().replace(" ","_").replace("/","_").replace("\\","_")
-    s_age = str(int(float(age))) 
+    s_age = str(int(float(age)))
     s_education = str(education).strip().replace(" ","_").replace("/","_").replace("\\","_")
     user_id_str = f"N-{s_name}_G-{s_gender}_A-{s_age}_E-{s_education}"
     print(f"用户信息收集完毕，生成用户ID: {user_id_str}")
     return user_id_str, True, gr.update(visible=False), gr.update(visible=True), ""
 
 with gr.Blocks(css=CSS, title="图像重建主观评估") as demo:
-    # ... (所有 State 变量定义，与您上一版代码相同) ...
     s_show_experiment_ui = gr.State(False); s_trial_index = gr.State(0); s_run_no = gr.State(1)
     s_user_logs = gr.State([]); s_current_trial_data = gr.State({}); s_user_session_id = gr.State(None)
     s_current_run_image_list = gr.State([]); s_num_trials_this_run = gr.State(0)
-    
-    welcome_container = gr.Column(visible=True) 
+
+    welcome_container = gr.Column(visible=True)
     experiment_container = gr.Column(visible=False)
 
     with welcome_container:
-        # ... (欢迎页UI，与您上一版代码相同) ...
         gr.Markdown(welcome_page_markdown)
         with gr.Row(): user_name_input = gr.Textbox(label="请输入您的姓名或代号 (例如 张三 或 User001)", placeholder="例如：张三 -> ZS"); user_gender_input = gr.Radio(label="性别", choices=["男", "女"])
         with gr.Row(): user_age_input = gr.Number(label="年龄 (请输入1-120的整数)", minimum=1, maximum=120, step=1); user_education_input = gr.Dropdown(label="学历", choices=["其他","初中及以下","高中（含中专）", "大专（含在读）", "本科（含在读）", "硕士（含在读）", "博士（含在读）"])
@@ -344,7 +456,6 @@ with gr.Blocks(css=CSS, title="图像重建主观评估") as demo:
         btn_agree_and_start = gr.Button("我已阅读上述说明并同意参与实验")
 
     with experiment_container:
-        # ... (实验主界面UI，与您上一版代码相同，包括新的按钮布局和隐藏的标签) ...
         gr.Markdown("## 🧠 图像重建主观评估实验"); gr.Markdown(f"每轮实验大约有 {NUM_TRIALS_PER_RUN} 次比较。")
         with gr.Row():
             with gr.Column(scale=1, min_width=300): left_img = gr.Image(label="左候选图", visible=False, height=400, interactive=False); left_lbl = gr.Textbox(label="左图信息", visible=False, interactive=False, max_lines=1); btn_left = gr.Button("选择左图 (更相似)", interactive=False, elem_classes="compact_button")
@@ -355,17 +466,17 @@ with gr.Blocks(css=CSS, title="图像重建主观评估") as demo:
         with gr.Row(): btn_start = gr.Button("开始试验 / 下一轮")
         file_out_placeholder = gr.File(label=" ", visible=False, interactive=False)
 
-    outputs_ui_components_definition = [ 
+    outputs_ui_components_definition = [
         target_img, left_img, right_img, left_lbl, right_lbl, status_text, progress_text,
-        btn_start, btn_left, btn_right, file_out_placeholder 
+        btn_start, btn_left, btn_right, file_out_placeholder
     ]
-    click_inputs_base = [ 
+    click_inputs_base = [
         s_trial_index, s_run_no, s_user_logs, s_current_trial_data, s_user_session_id,
         s_current_run_image_list, s_num_trials_this_run
     ]
-    event_outputs = [ 
+    event_outputs = [
         s_trial_index, s_run_no, s_user_logs, s_current_trial_data, s_user_session_id,
-        s_current_run_image_list, s_num_trials_this_run, *outputs_ui_components_definition 
+        s_current_run_image_list, s_num_trials_this_run, *outputs_ui_components_definition
     ]
 
     btn_agree_and_start.click(fn=handle_agree_and_start, inputs=[user_name_input, user_gender_input, user_age_input, user_education_input], outputs=[s_user_session_id, s_show_experiment_ui, welcome_container, experiment_container, welcome_error_msg])
@@ -373,9 +484,7 @@ with gr.Blocks(css=CSS, title="图像重建主观评估") as demo:
     btn_left.click(fn=partial(process_experiment_step, action_type="record_choice", choice_value="left"), inputs=click_inputs_base, outputs=event_outputs, queue=True)
     btn_right.click(fn=partial(process_experiment_step, action_type="record_choice", choice_value="right"), inputs=click_inputs_base, outputs=event_outputs, queue=True)
 
-# ==== 程序入口 ====
 if __name__ == "__main__":
-    # ... (与您上一版相同的启动检查和打印逻辑, 确保 allowed_paths 正确) ...
     if not master_image_list: print("\n关键错误：程序无法启动，因无目标图片。"); exit()
     else:
         print(f"从 '{TARGET_DIR}' 加载 {len(master_image_list)} 张目标图片。每轮选 {NUM_TRIALS_PER_RUN} 张。")
@@ -383,18 +492,18 @@ if __name__ == "__main__":
         else: print(f"方法根目录: {METHOD_ROOTS}")
         if not SUBJECTS: print("警告: SUBJECTS 列表为空。")
         else: print(f"Subjects: {SUBJECTS}")
-        print(f"日志保存到 Dataset: '{DATASET_REPO_ID}'")
-        if BATCH_LOG_FOLDER: print(f"  - 批量日志文件夹: '{BATCH_LOG_FOLDER}/'")
-        # INDIVIDUAL_LOGS_FOLDER 的打印可以保留或移除，因为现在主要用批量
-        # if INDIVIDUAL_LOGS_FOLDER: print(f"  - 单个选择日志文件夹 (可能未使用): '{INDIVIDUAL_LOGS_FOLDER}/'") 
-        if not os.getenv("HF_TOKEN"): print("警告: HF_TOKEN 未设置。日志无法保存。\n      请在 Space Secrets 中设置 HF_TOKEN。")
+        print(f"用户选择日志保存到 Dataset: '{DATASET_REPO_ID}' 的 '{BATCH_LOG_FOLDER}/' 文件夹")
+        if not os.getenv("HF_TOKEN"): print("警告: HF_TOKEN 未设置。日志无法保存到Hugging Face Dataset。\n      请在 Space Secrets 中设置 HF_TOKEN。")
         else: print("HF_TOKEN 已找到。")
+        print(f"全局图片对历史将从 '{GLOBAL_HISTORY_FILE}' 加载/保存到此文件。")
+
         path_to_allow_serving_from = BASE_IMAGE_DIR
         allowed_paths_list = []
         if os.path.exists(path_to_allow_serving_from) and os.path.isdir(path_to_allow_serving_from):
             allowed_paths_list.append(os.path.abspath(path_to_allow_serving_from))
             print(f"Gradio `demo.launch()` 配置 allowed_paths: {allowed_paths_list}")
         else: print(f"关键警告：图片基础目录 '{path_to_allow_serving_from}' ({os.path.abspath(path_to_allow_serving_from) if path_to_allow_serving_from else 'N/A'}) 不存在或非目录。")
+
         print("启动 Gradio 应用...")
         if allowed_paths_list: demo.launch(allowed_paths=allowed_paths_list)
         else: demo.launch()