fix: 将平台业务ID统一改为String并自动建表

- 抖音、B站、快手、微博的帖子/视频/评论ID从BigInteger改为String，
  避免PostgreSQL下字符串写入BIGINT报错及未来ID溢出风险
- B站dynamic_id改为String，修复API返回id_str被强转int导致的精度丢失
- 知乎提取器对content_id/question_id显式str()转换
- main.py启动数据库保存模式时自动建表，无需手动--init_db
- 同步更新相关老化测试

tests/test_douyin_no_user_info.py

@@ -282,7 +282,7 @@ def test_douyin_store_end_to_end_sqlite(monkeypatch):
     # dict 多了已删列会 TypeError,少了非空必填列 SQLAlchemy 也会报错;
     # 此处证明 captured 的 key 与删列后 ORM 列完全对得上,不抛异常。
     obj = DouyinAweme(**captured)
-    assert int(obj.aweme_id) == int(aweme["aweme_id"])
+    assert obj.aweme_id == aweme["aweme_id"]
     assert obj.creator_hash == anonymize_user_id(raw_uid)
     assert obj.creator_hash != raw_uid
     assert obj.nickname == mask_nickname(raw_nick)
@@ -310,7 +310,7 @@ def test_douyin_store_end_to_end_sqlite(monkeypatch):
         # 查询回读
         async with db_session.get_session() as session:
             res = await session.execute(
-                select(DouyinAweme).where(DouyinAweme.aweme_id == int(aweme["aweme_id"]))
+                select(DouyinAweme).where(DouyinAweme.aweme_id == aweme["aweme_id"])
             )
             row = res.scalar_one_or_none()
         await engine.dispose()
@@ -320,7 +320,7 @@ def test_douyin_store_end_to_end_sqlite(monkeypatch):
 
     # ---- 4. 回读断言 ----
     assert row is not None, "作品未写入 SQLite"
-    assert int(row.aweme_id) == int(aweme["aweme_id"])
+    assert row.aweme_id == aweme["aweme_id"]
     assert row.creator_hash == anonymize_user_id(raw_uid)
     assert row.creator_hash != raw_uid
     assert row.nickname == mask_nickname(raw_nick)
@@ -348,7 +348,7 @@ def test_douyin_store_end_to_end_sqlite(monkeypatch):
         ds.DouyinStoreFactory.create_store = orig_c
     captured_comment = fake_c.comments[0]
     comment_obj = DouyinAwemeComment(**captured_comment)  # 不抛异常即对得上
-    assert int(comment_obj.comment_id) == int(comment["cid"])
+    assert comment_obj.comment_id == comment["cid"]
     assert comment_obj.creator_hash == anonymize_user_id(comment["user"]["uid"])
     assert comment_obj.nickname == mask_nickname(comment["user"]["nickname"])
     _assert_no_forbidden(captured_comment, "douyin_comment_orm_construct")

tests/test_store_factory.py

@@ -33,6 +33,7 @@ from store.xhs._store_impl import (
     XhsMongoStoreImplement,
     XhsExcelStoreImplement
 )
+from store.excel_store_base import ExcelStoreBase
 
 
 class TestXhsStoreFactory:
@@ -73,7 +74,8 @@ class TestXhsStoreFactory:
         """Test creating Excel store"""
         # ContextVar cannot be mocked, so we test with actual value
         store = XhsStoreFactory.create_store()
-        assert isinstance(store, XhsExcelStoreImplement)
+        # XhsExcelStoreImplement 返回 ExcelStoreBase 单例，实际类型为 ExcelStoreBase
+        assert isinstance(store, ExcelStoreBase)
 
     @patch('config.SAVE_DATA_OPTION', 'jsonl')
     def test_create_jsonl_store(self):

tests/test_tieba_extractor.py

@@ -20,7 +20,7 @@ def test_extract_search_note_list_from_keyword_page():
     assert notes[0].note_id == "9117888152"
     assert notes[0].title.startswith("武汉交互空间科技")
     assert notes[0].tieba_name == "武汉交互空间"
-    assert notes[0].user_nickname == "VR虚拟达人"
+    assert notes[0].user_nickname == "V***人"
 
 
 def test_extract_search_note_list_from_current_pc_card_page():
@@ -56,7 +56,7 @@ def test_extract_search_note_list_from_current_pc_card_page():
     assert notes[0].desc == "培训班需求，数学，英语，编程老师，专职兼职都可"
     assert notes[0].tieba_name == "诸城吧"
     assert notes[0].tieba_link.endswith("kw=%E8%AF%B8%E5%9F%8E")
-    assert notes[0].user_nickname == "754023117"
+    assert notes[0].user_nickname == "7***7"
     assert notes[0].publish_time == "2026-3-15"
     assert notes[0].total_replay_num == 19
 
@@ -147,17 +147,17 @@ def test_extract_note_detail_and_comments_from_current_pc_api():
     assert note.note_id == "10451142633"
     assert note.title == "这X尔斯对比巴尔斯，我只能说ID正确，允许居功自傲"
     assert note.desc == "皮队败决处刑德国编程钢琴师兼职数学家"
-    assert note.user_nickname == "泰高祖蒙斯克"
+    assert note.user_nickname == "泰***克"
     assert note.tieba_name == "dota2吧"
     assert note.total_replay_num == 15
     assert note.total_replay_page == 1
-    assert note.ip_location == "广东"
+    # 教学版已移除 ip_location 等可定位真人字段
     assert len(comments) == 1
     assert comments[0].comment_id == "153154097267"
     assert comments[0].content == "xg现在大树阵容另一个辅助不选控制"
-    assert comments[0].user_nickname == "期胡希3"
+    assert comments[0].user_nickname == "期***3"
     assert comments[0].sub_comment_count == 4
-    assert comments[0].ip_location == "河北"
+    # 教学版已移除 ip_location 等可定位真人字段
 
 
 def test_extract_creator_info_and_threads_from_current_pc_api():
@@ -191,12 +191,10 @@ def test_extract_creator_info_and_threads_from_current_pc_api():
     creator = extractor.extract_creator_info_from_api(creator_api)
     thread_ids = extractor.extract_creator_thread_id_list_from_api(feed_api)
 
-    assert creator.user_id == "3546493137"
-    assert creator.user_name == "拜月教Alice"
-    assert creator.nickname == "米米世界大手子"
+    assert creator.user_nickname == "米***子"
     assert creator.fans == 58
     assert creator.follows == 1
-    assert creator.ip_location == "广东"
+    # 教学版已移除 user_id、user_name、ip_location 等可定位真人字段
     assert creator.registration_duration == "7.8"
     assert thread_ids == ["10208192951", "9835114923"]
 
@@ -225,7 +223,7 @@ def test_extract_tieba_note_list_from_bigpipe_thread_page():
     assert len(notes) == 48
     assert notes[0].note_id == "9079949995"
     assert notes[0].title == "盗墓笔记全集+txt小说，已整理"
-    assert notes[0].user_nickname == "公子伯仲"
+    assert notes[0].user_nickname == "公***仲"
     assert notes[0].tieba_name == "盗墓笔记吧"
     assert notes[0].tieba_link.endswith("kw=%E7%9B%97%E5%A2%93%E7%AC%94%E8%AE%B0&ie=utf-8")
 
@@ -235,11 +233,11 @@ def test_extract_note_detail_from_post_page():
 
     assert note.note_id == "9117905169"
     assert note.title == "对于一个父亲来说，这个女儿14岁就死了"
-    assert note.user_nickname == "章景轩"
+    assert note.user_nickname == "章***轩"
     assert note.tieba_name == "以太比特吧"
     assert note.total_replay_num == 786
     assert note.total_replay_page == 13
-    assert note.ip_location == "广东"
+    # 教学版已移除 ip_location 等可定位真人字段
 
 
 def test_extract_parent_comments_from_post_page():
@@ -251,9 +249,9 @@ def test_extract_parent_comments_from_post_page():
     assert len(comments) == 30
     assert comments[0].comment_id == "150726491368"
     assert comments[0].content == "中国队第22金！无悬念！"
-    assert comments[0].user_nickname == "heinzfrentzen"
+    assert comments[0].user_nickname == "h***n"
     assert comments[0].tieba_name == "网球风云吧"
-    assert comments[0].ip_location == "福建"
+    # 教学版已移除 ip_location 等可定位真人字段
 
 
 def test_extract_sub_comments_with_class_token_matching():
@@ -275,4 +273,4 @@ def test_extract_sub_comments_with_class_token_matching():
     assert len(comments) >= 10
     assert comments[0].comment_id
     assert comments[0].parent_comment_id == parent.comment_id
-    assert comments[0].user_link.startswith("https://tieba.baidu.com/home/main")
+    # 教学版已移除 user_link 等可定位真人字段的采集

tests/test_weibo_no_user_info.py

@@ -27,8 +27,8 @@ RAW_USER_ID = 7654321
 RAW_NICKNAME = "微博达人"
 RAW_COMMENT_USER_ID = 111222
 RAW_COMMENT_NICKNAME = "评论员小张"
-NOTE_ID = 5123456789
-COMMENT_ID = 998877
+NOTE_ID = "5123456789"
+COMMENT_ID = "998877"
 # 合法 RFC2822 时间串(weekday 与日期已对齐:2025-06-14 是周六)
 RFC2822_TIME = "Sat Jun 14 12:00:00 +0800 2025"
 
@@ -168,7 +168,7 @@ def test_weibo_comment_masks_user_info():
     fake = _FakeStore()
     wb_, orig = _patch_factory(fake)
     try:
-        asyncio.run(wb.update_weibo_note_comment(str(NOTE_ID), make_mock_comment()))
+        asyncio.run(wb.update_weibo_note_comment(NOTE_ID, make_mock_comment()))
     finally:
         _restore(wb_, orig)
 
@@ -194,7 +194,7 @@ def test_weibo_comment_masks_user_info():
 
     # 4. 内容字段正确
     assert "说得好" in captured["content"]
-    assert captured["comment_id"] == str(COMMENT_ID)
+    assert captured["comment_id"] == COMMENT_ID
     assert captured["comment_like_count"] == "5"
     assert captured["sub_comment_count"] == "3"
     assert captured["parent_comment_id"] == "parent_abc"
@@ -218,7 +218,7 @@ def test_weibo_store_end_to_end_sqlite():
     wb_, orig = _patch_factory(fake)
     try:
         asyncio.run(wb.update_weibo_note(make_mock_note()))
-        asyncio.run(wb.update_weibo_note_comment(str(NOTE_ID), make_mock_comment()))
+        asyncio.run(wb.update_weibo_note_comment(NOTE_ID, make_mock_comment()))
     finally:
         _restore(wb_, orig)
 
@@ -250,11 +250,8 @@ def test_weibo_store_end_to_end_sqlite():
         # 确认没有 desc 列存任何用户描述
         assert "desc" not in note_cols
 
-        # ---- 4. comment:同上。注意 store_comment 实现会把 comment_id/note_id/create_time
-        #         转成 int 后再构造 ORM,此处忠实复刻该转换以走通 BigInteger 列写入 ----
+        # ---- 4. comment:同上。ID 已统一为字符串类型,create_time 保持 int ----
         cc = dict(captured_comment)
-        cc["comment_id"] = int(cc["comment_id"])
-        cc["note_id"] = int(cc.get("note_id", 0) or 0)
         cc["create_time"] = int(cc.get("create_time", 0) or 0)
         comment_obj = WeiboNoteComment(**cc)  # 不抛异常 => key 与 ORM 列对得上
         session.add(comment_obj)