mirror of
https://github.com/NanmiCoder/MediaCrawler.git
synced 2026-03-05 13:40:46 +08:00
i18n: translate all Chinese comments, docstrings, and logger messages to English
Comprehensive translation of Chinese text to English across the entire codebase: - api/: FastAPI server documentation and logger messages - cache/: Cache abstraction layer comments and docstrings - database/: Database models and MongoDB store documentation - media_platform/: All platform crawlers (Bilibili, Douyin, Kuaishou, Tieba, Weibo, Xiaohongshu, Zhihu) - model/: Data model documentation - proxy/: Proxy pool and provider documentation - store/: Data storage layer comments - tools/: Utility functions and browser automation - test/: Test file documentation Preserved: Chinese disclaimer header (lines 10-18) for legal compliance 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
@@ -17,9 +17,9 @@
|
||||
# 使用本代码即表示您同意遵守上述原则和LICENSE中的所有条款。
|
||||
|
||||
# persist-1<persist1@126.com>
|
||||
# 原因:将 db.py 改造为模块,移除直接执行入口,修复相对导入问题。
|
||||
# 副作用:无
|
||||
# 回滚策略:还原此文件。
|
||||
# Reason: Refactored db.py into a module, removed direct execution entry point, fixed relative import issues.
|
||||
# Side effects: None
|
||||
# Rollback strategy: Restore this file.
|
||||
import asyncio
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
@@ -406,9 +406,9 @@ class ZhihuContent(Base):
|
||||
last_modify_ts = Column(BigInteger)
|
||||
|
||||
# persist-1<persist1@126.com>
|
||||
# 原因:修复 ORM 模型定义错误,确保与数据库表结构一致。
|
||||
# 副作用:无
|
||||
# 回滚策略:还原此行
|
||||
# Reason: Fixed ORM model definition error, ensuring consistency with database table structure.
|
||||
# Side effects: None
|
||||
# Rollback strategy: Restore this line
|
||||
|
||||
class ZhihuComment(Base):
|
||||
__tablename__ = 'zhihu_comment'
|
||||
|
||||
@@ -16,7 +16,7 @@
|
||||
# 详细许可条款请参阅项目根目录下的LICENSE文件。
|
||||
# 使用本代码即表示您同意遵守上述原则和LICENSE中的所有条款。
|
||||
|
||||
"""MongoDB存储基类:提供连接管理和通用存储方法"""
|
||||
"""MongoDB storage base class: Provides connection management and common storage methods"""
|
||||
import asyncio
|
||||
from typing import Dict, List, Optional
|
||||
from motor.motor_asyncio import AsyncIOMotorClient, AsyncIOMotorDatabase, AsyncIOMotorCollection
|
||||
@@ -25,7 +25,7 @@ from tools import utils
|
||||
|
||||
|
||||
class MongoDBConnection:
|
||||
"""MongoDB连接管理(单例模式)"""
|
||||
"""MongoDB connection management (singleton pattern)"""
|
||||
_instance = None
|
||||
_client: Optional[AsyncIOMotorClient] = None
|
||||
_db: Optional[AsyncIOMotorDatabase] = None
|
||||
@@ -37,7 +37,7 @@ class MongoDBConnection:
|
||||
return cls._instance
|
||||
|
||||
async def get_client(self) -> AsyncIOMotorClient:
|
||||
"""获取客户端"""
|
||||
"""Get client"""
|
||||
if self._client is None:
|
||||
async with self._lock:
|
||||
if self._client is None:
|
||||
@@ -45,7 +45,7 @@ class MongoDBConnection:
|
||||
return self._client
|
||||
|
||||
async def get_db(self) -> AsyncIOMotorDatabase:
|
||||
"""获取数据库"""
|
||||
"""Get database"""
|
||||
if self._db is None:
|
||||
async with self._lock:
|
||||
if self._db is None:
|
||||
@@ -53,7 +53,7 @@ class MongoDBConnection:
|
||||
return self._db
|
||||
|
||||
async def _connect(self):
|
||||
"""建立连接"""
|
||||
"""Establish connection"""
|
||||
try:
|
||||
mongo_config = db_config.mongodb_config
|
||||
host = mongo_config["host"]
|
||||
@@ -62,14 +62,14 @@ class MongoDBConnection:
|
||||
password = mongo_config["password"]
|
||||
db_name = mongo_config["db_name"]
|
||||
|
||||
# 构建连接URL(有认证/无认证)
|
||||
# Build connection URL (with/without authentication)
|
||||
if user and password:
|
||||
connection_url = f"mongodb://{user}:{password}@{host}:{port}/"
|
||||
else:
|
||||
connection_url = f"mongodb://{host}:{port}/"
|
||||
|
||||
self._client = AsyncIOMotorClient(connection_url, serverSelectionTimeoutMS=5000)
|
||||
await self._client.server_info() # 测试连接
|
||||
await self._client.server_info() # Test connection
|
||||
self._db = self._client[db_name]
|
||||
utils.logger.info(f"[MongoDBConnection] Connected to {host}:{port}/{db_name}")
|
||||
except Exception as e:
|
||||
@@ -77,7 +77,7 @@ class MongoDBConnection:
|
||||
raise
|
||||
|
||||
async def close(self):
|
||||
"""关闭连接"""
|
||||
"""Close connection"""
|
||||
if self._client is not None:
|
||||
self._client.close()
|
||||
self._client = None
|
||||
@@ -86,24 +86,24 @@ class MongoDBConnection:
|
||||
|
||||
|
||||
class MongoDBStoreBase:
|
||||
"""MongoDB存储基类:提供通用的CRUD操作"""
|
||||
"""MongoDB storage base class: Provides common CRUD operations"""
|
||||
|
||||
def __init__(self, collection_prefix: str):
|
||||
"""初始化存储基类
|
||||
"""Initialize storage base class
|
||||
Args:
|
||||
collection_prefix: 平台前缀(xhs/douyin/bilibili等)
|
||||
collection_prefix: Platform prefix (xhs/douyin/bilibili, etc.)
|
||||
"""
|
||||
self.collection_prefix = collection_prefix
|
||||
self._connection = MongoDBConnection()
|
||||
|
||||
async def get_collection(self, collection_suffix: str) -> AsyncIOMotorCollection:
|
||||
"""获取集合:{prefix}_{suffix}"""
|
||||
"""Get collection: {prefix}_{suffix}"""
|
||||
db = await self._connection.get_db()
|
||||
collection_name = f"{self.collection_prefix}_{collection_suffix}"
|
||||
return db[collection_name]
|
||||
|
||||
async def save_or_update(self, collection_suffix: str, query: Dict, data: Dict) -> bool:
|
||||
"""保存或更新数据(upsert)"""
|
||||
"""Save or update data (upsert)"""
|
||||
try:
|
||||
collection = await self.get_collection(collection_suffix)
|
||||
await collection.update_one(query, {"$set": data}, upsert=True)
|
||||
@@ -113,7 +113,7 @@ class MongoDBStoreBase:
|
||||
return False
|
||||
|
||||
async def find_one(self, collection_suffix: str, query: Dict) -> Optional[Dict]:
|
||||
"""查询单条数据"""
|
||||
"""Query a single record"""
|
||||
try:
|
||||
collection = await self.get_collection(collection_suffix)
|
||||
return await collection.find_one(query)
|
||||
@@ -122,7 +122,7 @@ class MongoDBStoreBase:
|
||||
return None
|
||||
|
||||
async def find_many(self, collection_suffix: str, query: Dict, limit: int = 0) -> List[Dict]:
|
||||
"""查询多条数据(limit=0表示不限制)"""
|
||||
"""Query multiple records (limit=0 means no limit)"""
|
||||
try:
|
||||
collection = await self.get_collection(collection_suffix)
|
||||
cursor = collection.find(query)
|
||||
@@ -134,7 +134,7 @@ class MongoDBStoreBase:
|
||||
return []
|
||||
|
||||
async def create_index(self, collection_suffix: str, keys: List[tuple], unique: bool = False):
|
||||
"""创建索引:keys=[("field", 1)]"""
|
||||
"""Create index: keys=[("field", 1)]"""
|
||||
try:
|
||||
collection = await self.get_collection(collection_suffix)
|
||||
await collection.create_index(keys, unique=unique)
|
||||
|
||||
Reference in New Issue
Block a user