1627 lines
57 KiB
Python
1627 lines
57 KiB
Python
# -*- coding: utf-8 -*-
|
||
"""
|
||
MediaWiki 最近变更同步工具 - 增强版
|
||
支持：
|
||
1. 正常全量同步（无参数）
|
||
2. 手动指定时间起点：--since 2025-11-28T00:00:00Z
|
||
3. 只同步单个页面：--title "页面名称"
|
||
4. 单个页面时可选更新全局时间戳：--update-timestamp
|
||
5. 获取历史版本并生成diff
|
||
6. 同步中文翻译版本
|
||
7. 生成双语对比网页
|
||
"""
|
||
|
||
import os
|
||
import argparse
|
||
from pathlib import Path
|
||
from datetime import datetime
|
||
import requests
|
||
from dotenv import load_dotenv
|
||
import difflib
|
||
import json
|
||
import re
|
||
from urllib.parse import quote
|
||
|
||
# ==================== 配置区 ====================
|
||
load_dotenv()
|
||
WIKI_API_URL_EN = os.getenv("WIKI_API_URL_EN", "https://wiki.projectdiablo2.com/w/api.php")
|
||
WIKI_API_URL_CN = os.getenv("WIKI_API_URL_CN", "https://wiki.projectdiablo2.cn/w/api.php")
|
||
OUTPUT_DIR = Path("wiki_sync_output")
|
||
OUTPUT_DIR.mkdir(exist_ok=True)
|
||
|
||
# 全局变量，存储本次执行的输出目录
|
||
CURRENT_OUTPUT_DIR = None
|
||
|
||
LAST_TIMESTAMP_FILE = "last_sync_timestamp.txt"
|
||
|
||
SESSION_EN = requests.Session()
|
||
SESSION_EN.headers.update({
|
||
"User-Agent": "WikiSyncTool/4.0 (your-email@example.com; MediaWiki Sync Bot)"
|
||
})
|
||
|
||
SESSION_CN = requests.Session()
|
||
SESSION_CN.headers.update({
|
||
"User-Agent": "WikiSyncTool/4.0 (your-email@example.com; MediaWiki Sync Bot)"
|
||
})
|
||
# ================================================
|
||
|
||
def load_last_timestamp():
|
||
if not os.path.exists(LAST_TIMESTAMP_FILE):
|
||
return None
|
||
with open(LAST_TIMESTAMP_FILE, encoding="utf-8") as f:
|
||
return f.read().strip()
|
||
|
||
def save_last_timestamp(ts):
|
||
with open(LAST_TIMESTAMP_FILE, "w", encoding="utf-8") as f:
|
||
f.write(ts)
|
||
|
||
def get_recent_changes(since):
|
||
"""获取自 since 时间后每个页面的最新 revid（自动去重）"""
|
||
params = {
|
||
"action": "query",
|
||
"list": "recentchanges",
|
||
"rcprop": "title|ids|timestamp",
|
||
"rctype": "edit|new",
|
||
"rcdir": "newer",
|
||
"rcstart": since,
|
||
"rclimit": 500,
|
||
"format": "json"
|
||
}
|
||
latest = {}
|
||
while True:
|
||
try:
|
||
r = SESSION_EN.get(WIKI_API_URL_EN, params=params)
|
||
r.raise_for_status()
|
||
response_data = r.json()
|
||
if "error" in response_data:
|
||
raise Exception(response_data["error"])
|
||
for rc in response_data.get("query", {}).get("recentchanges", []):
|
||
latest[rc["title"]] = (rc["revid"], rc["timestamp"])
|
||
if "continue" not in response_data:
|
||
break
|
||
params.update(response_data["continue"])
|
||
except Exception as e:
|
||
print(f"获取最近更改时出错: {e}")
|
||
break
|
||
return latest
|
||
|
||
def get_old_revid(title, end_time):
|
||
"""获取 ≤ end_time 的最后一次修订的 revid（用于 fromrev）"""
|
||
params = {
|
||
"action": "query",
|
||
"prop": "revisions",
|
||
"titles": title,
|
||
"rvprop": "ids|timestamp",
|
||
"rvlimit": 1,
|
||
"rvdir": "older",
|
||
"rvstart": end_time,
|
||
"format": "json"
|
||
}
|
||
try:
|
||
r = SESSION_EN.get(WIKI_API_URL_EN, params=params).json()
|
||
pages = r["query"]["pages"]
|
||
page = next(iter(pages.values()))
|
||
if "revisions" not in page:
|
||
print(f" 页面 '{title}' 在指定时间前没有找到修订版本")
|
||
return None
|
||
|
||
revisions = page["revisions"]
|
||
if len(revisions) >= 1:
|
||
return revisions[0]["revid"]
|
||
print(f" 页面 '{title}' 在指定时间前没有找到修订版本")
|
||
return None
|
||
except Exception as e:
|
||
print(f"获取旧版本ID时出错: {e}")
|
||
return None
|
||
|
||
def get_page_content(wiki_url, session, title, revid=None):
|
||
"""获取页面完整内容"""
|
||
params = {
|
||
"action": "query",
|
||
"prop": "revisions",
|
||
"titles": title,
|
||
"rvprop": "content|timestamp|ids",
|
||
"rvslots": "main",
|
||
"format": "json"
|
||
}
|
||
if revid:
|
||
params["rvstartid"] = revid
|
||
params["rvendid"] = revid
|
||
|
||
try:
|
||
r = session.get(wiki_url, params=params).json()
|
||
pages = r["query"]["pages"]
|
||
page = next(iter(pages.values()))
|
||
|
||
if "revisions" not in page:
|
||
return None, None, None
|
||
|
||
rev = page["revisions"][0]
|
||
content = rev["slots"]["main"]["*"]
|
||
timestamp = rev["timestamp"]
|
||
rev_id = rev["revid"]
|
||
|
||
return content, timestamp, rev_id
|
||
except Exception as e:
|
||
print(f"获取页面内容时出错: {e}")
|
||
return None, None, None
|
||
|
||
def generate_text_diff(old_text, new_text):
|
||
"""生成类似git diff的文本diff"""
|
||
if not old_text:
|
||
return "新创建页面"
|
||
|
||
old_lines = old_text.splitlines(keepends=True)
|
||
new_lines = new_text.splitlines(keepends=True)
|
||
|
||
differ = difflib.unified_diff(
|
||
old_lines,
|
||
new_lines,
|
||
lineterm='\n'
|
||
)
|
||
|
||
return ''.join(differ)
|
||
|
||
def parse_diff_with_line_numbers(diff_text):
|
||
"""解析diff文本，提取详细的行号信息"""
|
||
if not diff_text or diff_text.startswith("新创建页面"):
|
||
return []
|
||
|
||
parsed_lines = []
|
||
current_old_line = 0
|
||
current_new_line = 0
|
||
in_hunk = False
|
||
|
||
for line in diff_text.splitlines():
|
||
if line.startswith('@@'):
|
||
# 解析hunk头部，格式如: @@ -start,count +start,count @@
|
||
import re
|
||
match = re.match(r'@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@', line)
|
||
if match:
|
||
old_start = int(match.group(1))
|
||
old_count = int(match.group(2)) if match.group(2) else 1
|
||
new_start = int(match.group(3))
|
||
new_count = int(match.group(4)) if match.group(4) else 1
|
||
|
||
current_old_line = old_start
|
||
current_new_line = new_start
|
||
in_hunk = True
|
||
|
||
parsed_lines.append({
|
||
'type': 'hunk',
|
||
'content': line,
|
||
'old_start': old_start,
|
||
'old_count': old_count,
|
||
'new_start': new_start,
|
||
'new_count': new_count,
|
||
'old_line': None,
|
||
'new_line': None
|
||
})
|
||
else:
|
||
parsed_lines.append({
|
||
'type': 'other',
|
||
'content': line,
|
||
'old_line': None,
|
||
'new_line': None
|
||
})
|
||
elif line.startswith('---') or line.startswith('+++'):
|
||
# 文件头信息
|
||
continue
|
||
|
||
#parsed_lines.append({
|
||
# 'type': 'header',
|
||
# 'content': line,
|
||
# 'old_line': None,
|
||
# 'new_line': None
|
||
#})
|
||
elif in_hunk:
|
||
if line.startswith('-'):
|
||
# 删除的行
|
||
parsed_lines.append({
|
||
'type': 'removed',
|
||
'content': line[1:], # 去掉开头的 '-'
|
||
'old_line': current_old_line,
|
||
'new_line': None
|
||
})
|
||
current_old_line += 1
|
||
elif line.startswith('+'):
|
||
# 新增的行
|
||
parsed_lines.append({
|
||
'type': 'added',
|
||
'content': line[1:], # 去掉开头的 '+'
|
||
'old_line': None,
|
||
'new_line': current_new_line
|
||
})
|
||
current_new_line += 1
|
||
elif line.startswith(' '):
|
||
# 未变更的行
|
||
parsed_lines.append({
|
||
'type': 'context',
|
||
'content': line[1:], # 去掉开头的 ' '
|
||
'old_line': current_old_line,
|
||
'new_line': current_new_line
|
||
})
|
||
current_old_line += 1
|
||
current_new_line += 1
|
||
else:
|
||
# 其他行（如空行）
|
||
parsed_lines.append({
|
||
'type': 'other',
|
||
'content': line,
|
||
'old_line': None,
|
||
'new_line': None
|
||
})
|
||
else:
|
||
# 不在任何hunk中的行
|
||
parsed_lines.append({
|
||
'type': 'other',
|
||
'content': line,
|
||
'old_line': None,
|
||
'new_line': None
|
||
})
|
||
|
||
return parsed_lines
|
||
|
||
def search_chinese_page(title):
|
||
"""在中文wiki中搜索对应的页面"""
|
||
# 首先尝试精确匹配
|
||
params = {
|
||
"action": "query",
|
||
"list": "search",
|
||
"srsearch": f'"{title}"',
|
||
"srwhat": "title",
|
||
"srlimit": 5,
|
||
"format": "json"
|
||
}
|
||
|
||
try:
|
||
r = SESSION_CN.get(WIKI_API_URL_CN, params=params).json()
|
||
search_results = r.get("query", {}).get("search", [])
|
||
|
||
if search_results:
|
||
# 返回第一个匹配的结果
|
||
return search_results[0]["title"]
|
||
|
||
# 如果精确匹配没有结果，尝试模糊搜索
|
||
params["srsearch"] = title.replace(" ", "%20")
|
||
r = SESSION_CN.get(WIKI_API_URL_CN, params=params).json()
|
||
search_results = r.get("query", {}).get("search", [])
|
||
|
||
if search_results:
|
||
return search_results[0]["title"]
|
||
|
||
except Exception as e:
|
||
print(f"搜索中文页面时出错: {e}")
|
||
|
||
return None
|
||
|
||
def create_diff_html(title, en_diff, en_old_lines, en_new_lines, cn_content=None):
|
||
"""创建双语对比的HTML页面 - Word批注风格，英文变更直接显示在对应中文行右侧"""
|
||
# 准备中文内容行
|
||
cn_lines = []
|
||
if cn_content:
|
||
cn_lines = cn_content.splitlines()
|
||
|
||
# 解析diff并获取行号信息
|
||
parsed_diff = parse_diff_with_line_numbers(en_diff) if en_diff else []
|
||
|
||
# 构建行号到diff内容的映射 - 科学处理连续diff块
|
||
en_changes_by_line = {}
|
||
blank_lines_to_insert = {} # 记录需要在某行前插入的空白行及其对应的新增内容
|
||
|
||
if parsed_diff:
|
||
i = 0
|
||
while i < len(parsed_diff):
|
||
# 收集连续的diff块
|
||
diff_block = []
|
||
start_index = i
|
||
|
||
# 收集连续的添加/删除操作（跳过hunk和header）
|
||
while i < len(parsed_diff):
|
||
item = parsed_diff[i]
|
||
if item['type'] in ['added', 'removed']:
|
||
diff_block.append(item)
|
||
elif item['type'] in ['hunk', 'header'] or item['type'] == 'context':
|
||
if diff_block: # 如果已经有diff块，就停止
|
||
break
|
||
i += 1
|
||
|
||
# 处理连续的diff块 - 改进的连续匹配算法
|
||
if diff_block:
|
||
# 使用新的匹配算法：连续的减号和加号应该按顺序匹配
|
||
removed_items = []
|
||
added_items = []
|
||
|
||
# 分离删除和新增项目
|
||
for item in diff_block:
|
||
if item['type'] == 'removed':
|
||
removed_items.append(item)
|
||
elif item['type'] == 'added':
|
||
added_items.append(item)
|
||
|
||
# 进行匹配：连续块里的每一个减都应该和后续的加形成匹配替换
|
||
match_index = 0
|
||
for removed_item in removed_items:
|
||
if match_index < len(added_items):
|
||
# 匹配成功：形成替换
|
||
target_line = removed_item['old_line']
|
||
if target_line not in en_changes_by_line:
|
||
en_changes_by_line[target_line] = []
|
||
|
||
en_changes_by_line[target_line].append({
|
||
'type': 'replaced',
|
||
'old_content': removed_item['content'],
|
||
'new_content': added_items[match_index]['content']
|
||
})
|
||
match_index += 1
|
||
else:
|
||
# 没有匹配的加：这是删除
|
||
target_line = removed_item['old_line']
|
||
if target_line not in en_changes_by_line:
|
||
en_changes_by_line[target_line] = []
|
||
|
||
en_changes_by_line[target_line].append({
|
||
'type': 'removed',
|
||
'content': removed_item['content']
|
||
})
|
||
|
||
# 处理剩余的加（没有匹配的减）：这是新增，应该在左侧空行
|
||
if match_index < len(added_items):
|
||
# 找到基准行号（使用最后一个删除行的行号，如果没有则使用第一个新增的行号）
|
||
base_line = None
|
||
if removed_items:
|
||
base_line = removed_items[-1]['old_line']
|
||
elif added_items:
|
||
base_line = added_items[match_index]['new_line'] - len(added_items) + match_index
|
||
|
||
if base_line:
|
||
remaining_additions = added_items[match_index:]
|
||
blank_lines_to_insert[base_line + 1] = remaining_additions
|
||
|
||
# 继续处理剩余项
|
||
else:
|
||
i += 1
|
||
|
||
# HTML转义函数
|
||
def html_escape(text):
|
||
if not text:
|
||
return ""
|
||
return (str(text)
|
||
.replace("&", "&")
|
||
.replace("<", "<")
|
||
.replace(">", ">")
|
||
.replace('"', """)
|
||
.replace("'", "'"))
|
||
|
||
def generate_inline_diff(old_text, new_text):
|
||
"""生成GitHub风格的行内字符级diff"""
|
||
if not old_text or not new_text:
|
||
return html_escape(new_text or "")
|
||
|
||
escaped_old = html_escape(old_text)
|
||
escaped_new = html_escape(new_text)
|
||
|
||
# 使用difflib进行字符级别的比较
|
||
differ = difflib.SequenceMatcher(None, escaped_old, escaped_new)
|
||
|
||
result = []
|
||
for tag, i1, i2, j1, j2 in differ.get_opcodes():
|
||
if tag == 'equal':
|
||
# 相同的部分
|
||
result.append(escaped_new[j1:j2])
|
||
elif tag == 'replace':
|
||
# 替换的部分：删除的用红色背景，新增的用绿色背景
|
||
deleted = escaped_old[i1:i2]
|
||
added = escaped_new[j1:j2]
|
||
result.append(f'<span class="diff-char-removed">{deleted}</span>')
|
||
result.append(f'<span class="diff-char-added">{added}</span>')
|
||
elif tag == 'delete':
|
||
# 删除的部分用红色背景
|
||
deleted = escaped_old[i1:i2]
|
||
result.append(f'<span class="diff-char-removed">{deleted}</span>')
|
||
elif tag == 'insert':
|
||
# 新增的部分用绿色背景
|
||
added = escaped_new[j1:j2]
|
||
result.append(f'<span class="diff-char-added">{added}</span>')
|
||
|
||
return ''.join(result)
|
||
|
||
def generate_clean_new_content(old_text, new_text):
|
||
"""生成干净的新内容，只显示新增部分的高亮，不包含删除部分"""
|
||
if not old_text or not new_text:
|
||
return html_escape(new_text or "")
|
||
|
||
escaped_old = html_escape(old_text)
|
||
escaped_new = html_escape(new_text)
|
||
|
||
# 使用difflib进行字符级别的比较
|
||
differ = difflib.SequenceMatcher(None, escaped_old, escaped_new)
|
||
|
||
result = []
|
||
for tag, i1, i2, j1, j2 in differ.get_opcodes():
|
||
if tag == 'equal':
|
||
# 相同的部分
|
||
result.append(escaped_new[j1:j2])
|
||
elif tag == 'replace':
|
||
# 替换的部分：只显示新增的内容（绿色高亮），跳过删除的内容
|
||
added = escaped_new[j1:j2]
|
||
result.append(f'<span class="diff-char-added">{added}</span>')
|
||
elif tag == 'delete':
|
||
# 删除的部分：跳过，不显示
|
||
continue
|
||
elif tag == 'insert':
|
||
# 新增的部分用绿色背景
|
||
added = escaped_new[j1:j2]
|
||
result.append(f'<span class="diff-char-added">{added}</span>')
|
||
|
||
return ''.join(result)
|
||
|
||
# 收集变更块信息用于导航
|
||
change_blocks = []
|
||
change_block_id = 0
|
||
|
||
# 生成HTML
|
||
html = f'''<!DOCTYPE html>
|
||
<html lang="zh-CN">
|
||
<head>
|
||
<meta charset="UTF-8">
|
||
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
||
<title>Wiki Diff: {title}</title>
|
||
<style>
|
||
* {{
|
||
margin: 0;
|
||
padding: 0;
|
||
box-sizing: border-box;
|
||
}}
|
||
|
||
body {{
|
||
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, 'Helvetica Neue', Arial, sans-serif;
|
||
background-color: #f5f5f5;
|
||
line-height: 1.6;
|
||
padding: 20px;
|
||
}}
|
||
|
||
.header {{
|
||
background-color: #fff;
|
||
padding: 20px;
|
||
box-shadow: 0 2px 4px rgba(0,0,0,0.1);
|
||
margin-bottom: 20px;
|
||
border-radius: 8px;
|
||
}}
|
||
|
||
.header h1 {{
|
||
color: #333;
|
||
font-size: 24px;
|
||
margin-bottom: 10px;
|
||
}}
|
||
|
||
.header .meta {{
|
||
color: #666;
|
||
font-size: 14px;
|
||
}}
|
||
|
||
.content-container {{
|
||
max-width: 1200px;
|
||
margin: 0 auto;
|
||
background-color: #fff;
|
||
border-radius: 8px;
|
||
box-shadow: 0 2px 8px rgba(0,0,0,0.1);
|
||
overflow: hidden;
|
||
}}
|
||
|
||
.content-header {{
|
||
background-color: #e9ecef;
|
||
padding: 15px 20px;
|
||
font-weight: bold;
|
||
color: #495057;
|
||
border-bottom: 1px solid #dee2e6;
|
||
}}
|
||
|
||
.diff-content {{
|
||
padding: 0;
|
||
}}
|
||
|
||
.line-wrapper {{
|
||
display: flex;
|
||
border-bottom: 1px solid #f0f0f0;
|
||
position: relative;
|
||
}}
|
||
|
||
.line-wrapper:hover {{
|
||
background-color: rgba(0, 123, 255, 0.02);
|
||
}}
|
||
|
||
.line-wrapper.has-changes {{
|
||
background-color: rgba(255, 193, 7, 0.05);
|
||
}}
|
||
|
||
.main-line {{
|
||
display: flex;
|
||
flex: 1;
|
||
min-height: 24px;
|
||
align-items: center;
|
||
}}
|
||
|
||
.line-number {{
|
||
width: 60px;
|
||
text-align: right;
|
||
padding: 8px 12px;
|
||
background-color: #f8f9fa;
|
||
color: #6c757d;
|
||
font-size: 12px;
|
||
user-select: none;
|
||
flex-shrink: 0;
|
||
border-right: 1px solid #e9ecef;
|
||
}}
|
||
|
||
.line-content {{
|
||
flex: 1;
|
||
padding: 8px 12px;
|
||
font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;
|
||
font-size: 13px;
|
||
line-height: 1.5;
|
||
white-space: pre-wrap;
|
||
word-break: break-word;
|
||
color: #333;
|
||
}}
|
||
|
||
/* 批注样式 */
|
||
.annotation {{
|
||
width: 400px;
|
||
background-color: #f8f9fa;
|
||
border-left: 1px solid #dee2e6;
|
||
padding: 10px 14px;
|
||
font-size: 14px;
|
||
display: none;
|
||
}}
|
||
|
||
.line-wrapper.has-changes .annotation {{
|
||
display: block;
|
||
}}
|
||
|
||
.annotation-item {{
|
||
margin-bottom: 8px;
|
||
padding: 8px 10px;
|
||
border-radius: 4px;
|
||
font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;
|
||
font-size: 13px;
|
||
line-height: 1.5;
|
||
}}
|
||
|
||
.annotation-item:last-child {{
|
||
margin-bottom: 0;
|
||
}}
|
||
|
||
.annotation-item.added {{
|
||
background-color: #e6ffec;
|
||
border-left: 3px solid #28a745;
|
||
color: #155724;
|
||
}}
|
||
|
||
.annotation-item.removed {{
|
||
background-color: #ffeef0;
|
||
border-left: 3px solid #dc3545;
|
||
color: #721c24;
|
||
text-decoration: line-through;
|
||
}}
|
||
|
||
.annotation-item.replaced {{
|
||
margin-bottom: 8px;
|
||
}}
|
||
|
||
.annotation-item.replaced .old-content {{
|
||
background-color: #ffeef0;
|
||
border-left: 3px solid #dc3545;
|
||
color: #721c24;
|
||
text-decoration: line-through;
|
||
padding: 4px 6px;
|
||
border-radius: 3px;
|
||
margin-bottom: 4px;
|
||
}}
|
||
|
||
.annotation-item.replaced .new-content {{
|
||
background-color: #e6ffec;
|
||
border-left: 3px solid #28a745;
|
||
color: #155724;
|
||
padding: 4px 6px;
|
||
border-radius: 3px;
|
||
}}
|
||
|
||
.annotation-header {{
|
||
font-size: 11px;
|
||
color: #6c757d;
|
||
margin-bottom: 6px;
|
||
font-weight: bold;
|
||
text-transform: uppercase;
|
||
letter-spacing: 0.5px;
|
||
}}
|
||
|
||
/* GitHub风格的字符级diff样式 */
|
||
.diff-char-added {{
|
||
background-color: #acf2bd;
|
||
color: #24292f;
|
||
border-radius: 2px;
|
||
padding: 1px 2px;
|
||
}}
|
||
|
||
.diff-char-removed {{
|
||
background-color: #ffd5d5;
|
||
color: #24292f;
|
||
border-radius: 2px;
|
||
padding: 1px 2px;
|
||
}}
|
||
|
||
/* 批注项内的字符级diff样式调整 */
|
||
.annotation-item.removed {{
|
||
text-decoration: none; /* 移除删除线 */
|
||
}}
|
||
|
||
.annotation-item.replaced .old-content {{
|
||
text-decoration: none; /* 移除删除线 */
|
||
}}
|
||
|
||
/* 新页面提示 */
|
||
.new-page-notice {{
|
||
background-color: #d4edda;
|
||
color: #155724;
|
||
padding: 15px 20px;
|
||
margin: 15px;
|
||
border-radius: 4px;
|
||
border-left: 4px solid #28a745;
|
||
}}
|
||
|
||
.no-translation {{
|
||
background-color: #fff3cd;
|
||
color: #856404;
|
||
padding: 15px 20px;
|
||
margin: 15px;
|
||
border-radius: 4px;
|
||
border-left: 4px solid #ffc107;
|
||
}}
|
||
|
||
/* 响应式设计 */
|
||
@media (max-width: 1024px) {{
|
||
.annotation {{
|
||
width: 300px;
|
||
}}
|
||
}}
|
||
|
||
@media (max-width: 768px) {{
|
||
body {{
|
||
padding: 10px;
|
||
}}
|
||
|
||
.annotation {{
|
||
width: 100%;
|
||
display: block !important;
|
||
border-left: none;
|
||
border-top: 1px solid #dee2e6;
|
||
}}
|
||
|
||
.line-wrapper {{
|
||
flex-direction: column;
|
||
}}
|
||
|
||
.main-line {{
|
||
border-bottom: none;
|
||
}}
|
||
}}
|
||
|
||
/* 高亮效果 */
|
||
.line-wrapper.highlight {{
|
||
background-color: rgba(255, 235, 59, 0.3) !important;
|
||
animation: highlight 2s ease-in-out;
|
||
}}
|
||
|
||
@keyframes highlight {{
|
||
0% {{ background-color: rgba(255, 235, 59, 0.6); }}
|
||
100% {{ background-color: rgba(255, 235, 59, 0.3); }}
|
||
}}
|
||
|
||
/* 空行样式 */
|
||
.line-wrapper.empty-line .line-content {{
|
||
min-height: 24px;
|
||
color: #999;
|
||
font-style: italic;
|
||
}}
|
||
|
||
/* 空白占位行样式 */
|
||
.line-wrapper.blank-placeholder {{
|
||
background-color: #fafafa;
|
||
border-bottom: 1px solid #e9ecef;
|
||
display: flex;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .main-line {{
|
||
min-height: 24px;
|
||
flex: 1;
|
||
display: flex;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .line-number {{
|
||
color: #dee2e6;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .line-content {{
|
||
color: #dee2e6;
|
||
font-style: italic;
|
||
min-height: 24px;
|
||
display: flex;
|
||
align-items: center;
|
||
}}
|
||
|
||
/* 空白占位行的批注样式 */
|
||
.line-wrapper.blank-placeholder .annotation {{
|
||
width: 400px;
|
||
background-color: #f8f9fa;
|
||
border-left: 1px solid #dee2e6;
|
||
padding: 10px 14px;
|
||
font-size: 14px;
|
||
display: block;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .annotation .annotation-item {{
|
||
margin-bottom: 8px;
|
||
padding: 8px 10px;
|
||
border-radius: 4px;
|
||
font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;
|
||
font-size: 13px;
|
||
line-height: 1.5;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .annotation .annotation-item:last-child {{
|
||
margin-bottom: 0;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .annotation .annotation-item.added {{
|
||
background-color: #e6ffec;
|
||
border-left: 3px solid #28a745;
|
||
color: #155724;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder .annotation .annotation-header {{
|
||
font-size: 10px;
|
||
color: #6c757d;
|
||
margin-bottom: 4px;
|
||
font-weight: bold;
|
||
text-transform: uppercase;
|
||
letter-spacing: 0.5px;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder:hover {{
|
||
background-color: #f8f9fa;
|
||
}}
|
||
|
||
.line-wrapper.blank-placeholder:hover .main-line {{
|
||
background-color: rgba(0, 123, 255, 0.02);
|
||
}}
|
||
|
||
/* 变更块高亮样式 */
|
||
.line-wrapper.change-block {{
|
||
border-left: 3px solid #007bff;
|
||
}}
|
||
|
||
/* 导航浮窗样式 */
|
||
.navigation-float {{
|
||
position: fixed;
|
||
left: 20px;
|
||
top: 50%;
|
||
transform: translateY(-50%);
|
||
width: 280px;
|
||
max-height: 70vh;
|
||
background-color: #fff;
|
||
border: 1px solid #dee2e6;
|
||
border-radius: 8px;
|
||
box-shadow: 0 4px 12px rgba(0, 0, 0, 0.15);
|
||
z-index: 1000;
|
||
overflow: hidden;
|
||
}}
|
||
|
||
.navigation-header {{
|
||
background-color: #007bff;
|
||
color: white;
|
||
padding: 12px 16px;
|
||
font-weight: bold;
|
||
font-size: 14px;
|
||
display: flex;
|
||
justify-content: space-between;
|
||
align-items: center;
|
||
}}
|
||
|
||
.navigation-toggle {{
|
||
background: none;
|
||
border: none;
|
||
color: white;
|
||
cursor: pointer;
|
||
font-size: 16px;
|
||
padding: 4px 8px;
|
||
border-radius: 4px;
|
||
transition: background-color 0.2s;
|
||
}}
|
||
|
||
.navigation-toggle:hover {{
|
||
background-color: rgba(255, 255, 255, 0.2);
|
||
}}
|
||
|
||
.navigation-content {{
|
||
max-height: calc(70vh - 50px);
|
||
overflow-y: auto;
|
||
padding: 8px;
|
||
}}
|
||
|
||
.navigation-item {{
|
||
padding: 10px 12px;
|
||
margin-bottom: 6px;
|
||
background-color: #f8f9fa;
|
||
border-radius: 4px;
|
||
cursor: pointer;
|
||
transition: all 0.2s;
|
||
border-left: 3px solid transparent;
|
||
}}
|
||
|
||
.navigation-item:hover {{
|
||
background-color: #e9ecef;
|
||
border-left-color: #007bff;
|
||
}}
|
||
|
||
.navigation-item.active {{
|
||
background-color: #e3f2fd;
|
||
border-left-color: #007bff;
|
||
}}
|
||
|
||
.navigation-item-header {{
|
||
display: flex;
|
||
justify-content: space-between;
|
||
align-items: center;
|
||
margin-bottom: 4px;
|
||
}}
|
||
|
||
.navigation-item-number {{
|
||
background-color: #007bff;
|
||
color: white;
|
||
font-size: 11px;
|
||
padding: 2px 6px;
|
||
border-radius: 10px;
|
||
font-weight: bold;
|
||
}}
|
||
|
||
.navigation-item-type {{
|
||
font-size: 11px;
|
||
padding: 2px 6px;
|
||
border-radius: 3px;
|
||
font-weight: bold;
|
||
text-transform: uppercase;
|
||
}}
|
||
|
||
.navigation-item-type.added {{
|
||
background-color: #28a745;
|
||
color: white;
|
||
}}
|
||
|
||
.navigation-item-type.replaced {{
|
||
background-color: #ffc107;
|
||
color: #212529;
|
||
}}
|
||
|
||
.navigation-item-type.removed {{
|
||
background-color: #dc3545;
|
||
color: white;
|
||
}}
|
||
|
||
.navigation-item-preview {{
|
||
font-size: 12px;
|
||
color: #6c757d;
|
||
font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;
|
||
line-height: 1.3;
|
||
word-break: break-word;
|
||
}}
|
||
|
||
.navigation-item-line {{
|
||
font-size: 10px;
|
||
color: #adb5bd;
|
||
margin-top: 2px;
|
||
}}
|
||
|
||
/* 导航浮窗收起状态 */
|
||
.navigation-float.collapsed .navigation-content {{
|
||
display: none;
|
||
}}
|
||
|
||
.navigation-float.collapsed {{
|
||
width: auto;
|
||
min-width: 160px;
|
||
}}
|
||
|
||
/* 响应式设计 */
|
||
@media (max-width: 768px) {{
|
||
.navigation-float {{
|
||
left: 10px;
|
||
width: 240px;
|
||
}}
|
||
|
||
.navigation-float.collapsed {{
|
||
min-width: 120px;
|
||
}}
|
||
}}
|
||
|
||
@media (max-width: 480px) {{
|
||
.navigation-float {{
|
||
position: relative;
|
||
left: auto;
|
||
top: auto;
|
||
transform: none;
|
||
width: 100%;
|
||
max-height: none;
|
||
margin-bottom: 20px;
|
||
}}
|
||
}}
|
||
</style>
|
||
</head>
|
||
<body>
|
||
<div class="header">
|
||
<h1>{title}</h1>
|
||
<div class="meta">
|
||
<span>英文Wiki: wiki.projectdiablo2.com</span>
|
||
{f' | 中文Wiki: wiki.projectdiablo2.cn' if cn_content else ''}
|
||
</div>
|
||
</div>
|
||
|
||
<div class="navigation-float" id="navigation-float">
|
||
<div class="navigation-header">
|
||
<span>变更导航</span>
|
||
<button class="navigation-toggle" id="navigation-toggle">−</button>
|
||
</div>
|
||
<div class="navigation-content" id="navigation-content">
|
||
<!-- Navigation items will be generated here -->
|
||
</div>
|
||
</div>
|
||
|
||
<div class="content-container">
|
||
<div class="content-header">中文翻译（含英文变更批注）</div>
|
||
<div class="diff-content">
|
||
'''
|
||
|
||
# 添加中文内容和英文变更批注
|
||
if cn_content:
|
||
for i, line in enumerate(cn_lines, 1):
|
||
# 检查是否需要在此行之前插入空白行
|
||
if i in blank_lines_to_insert:
|
||
additions_list = blank_lines_to_insert[i]
|
||
change_block_id += 1
|
||
|
||
# 添加变更块到导航列表
|
||
preview_text = additions_list[0][:50] + "..." if len(additions_list[0]) > 50 else additions_list[0]
|
||
change_blocks.append({
|
||
'id': change_block_id,
|
||
'line': i,
|
||
'type': '新增',
|
||
'preview': preview_text,
|
||
'count': len(additions_list)
|
||
})
|
||
|
||
for idx, addition_content in enumerate(additions_list):
|
||
html += f'<div class="line-wrapper blank-placeholder change-block" data-change-id="{change_block_id}">'
|
||
html += '<div class="main-line">'
|
||
html += '<span class="line-number"> </span>' # 不显示行号
|
||
html += f'<span class="line-content">（新增英文内容占位）</span>'
|
||
html += '</div>'
|
||
|
||
# 为空白行添加对应的新增批注
|
||
escaped_addition = html_escape(addition_content)
|
||
html += '<div class="annotation">'
|
||
html += f'<div class="annotation-item added">'
|
||
html += f'<div class="annotation-header">新增</div>'
|
||
html += f'<div>{escaped_addition}</div>'
|
||
html += '</div>'
|
||
html += '</div>'
|
||
|
||
html += '</div>'
|
||
|
||
escaped_line = html_escape(line)
|
||
has_changes = i in en_changes_by_line
|
||
changes = en_changes_by_line.get(i, [])
|
||
|
||
# 判断是否为空行
|
||
is_empty = not line.strip()
|
||
|
||
# 如果有变更（除了新增），添加到导航列表
|
||
if has_changes and any(change['type'] in ['replaced', 'removed'] for change in changes):
|
||
change_block_id += 1
|
||
preview_text = line[:50] + "..." if len(line) > 50 else line
|
||
change_type = "替换" if any(change['type'] == 'replaced' for change in changes) else "删除"
|
||
|
||
change_blocks.append({
|
||
'id': change_block_id,
|
||
'line': i,
|
||
'type': change_type,
|
||
'preview': preview_text,
|
||
'count': 1
|
||
})
|
||
|
||
html += f'<div class="line-wrapper {"has-changes" if has_changes else ""} {"empty-line" if is_empty else ""} {"change-block" if has_changes else ""}" data-change-id="{change_block_id if has_changes else ""}">'
|
||
html += f'<div class="main-line">'
|
||
html += f'<span class="line-number">{i}</span>'
|
||
html += f'<span class="line-content">{escaped_line if not is_empty else "（空行）"}</span>'
|
||
html += '</div>'
|
||
|
||
# 添加英文变更批注（只显示替换和删除操作，新增操作已经在空白行中显示）
|
||
if has_changes:
|
||
html += '<div class="annotation">'
|
||
for change in changes:
|
||
if change['type'] == 'added':
|
||
# 新增内容已经在空白行中显示，这里跳过
|
||
continue
|
||
elif change['type'] == 'removed':
|
||
escaped_change = html_escape(change['content'])
|
||
html += f'<div class="annotation-item removed">'
|
||
html += f'<div class="annotation-header">删除</div>'
|
||
html += f'<div>{escaped_change}</div>'
|
||
html += '</div>'
|
||
elif change['type'] == 'replaced':
|
||
# 生成干净的新内容（只显示新增部分的高亮，不包含删除部分）
|
||
clean_new_content = generate_clean_new_content(change['old_content'], change['new_content'])
|
||
html += f'<div class="annotation-item replaced">'
|
||
html += f'<div class="annotation-header">替换</div>'
|
||
html += f'<div class="old-content">{html_escape(change["old_content"])}</div>'
|
||
html += f'<div class="new-content">{clean_new_content}</div>'
|
||
html += '</div>'
|
||
html += '</div>'
|
||
|
||
html += '</div>'
|
||
else:
|
||
html += '<div class="no-translation">未找到对应的中文翻译页面</div>'
|
||
|
||
# 调试日志：打印change_blocks信息
|
||
print(f"DEBUG: Final change_blocks length = {len(change_blocks)}")
|
||
for i, block in enumerate(change_blocks):
|
||
print(f"DEBUG: Final block {i}: {block}")
|
||
|
||
html += '''
|
||
</div>
|
||
</div>
|
||
|
||
<script>
|
||
// 动态生成导航项
|
||
function generateNavigationItems() {
|
||
const navContent = document.getElementById('navigation-content');
|
||
const changeBlocks = document.querySelectorAll('.change-block');
|
||
|
||
// 清空现有内容
|
||
navContent.innerHTML = '';
|
||
|
||
// 变更块信息已经在HTML中通过data属性标记
|
||
const changes = [];
|
||
|
||
// 收集所有变更块信息
|
||
changeBlocks.forEach((block, index) => {
|
||
const changeId = block.getAttribute('data-change-id');
|
||
if (changeId) {
|
||
// 获取变更类型和预览文本
|
||
const annotation = block.querySelector('.annotation');
|
||
if (annotation) {
|
||
const addedItem = annotation.querySelector('.annotation-item.added .annotation-header');
|
||
const replacedItem = annotation.querySelector('.annotation-item.replaced .annotation-header');
|
||
const removedItem = annotation.querySelector('.annotation-item.removed .annotation-header');
|
||
|
||
let type = '变更';
|
||
let preview = '';
|
||
|
||
if (addedItem) {
|
||
type = '新增';
|
||
const content = annotation.querySelector('.annotation-item.added > div:last-child');
|
||
preview = content ? content.textContent.substring(0, 50) : '';
|
||
} else if (replacedItem) {
|
||
type = '替换';
|
||
const content = annotation.querySelector('.annotation-item.replaced .old-content');
|
||
preview = content ? content.textContent.substring(0, 50) : '';
|
||
} else if (removedItem) {
|
||
type = '删除';
|
||
const content = annotation.querySelector('.annotation-item.removed > div:last-child');
|
||
preview = content ? content.textContent.substring(0, 50) : '';
|
||
}
|
||
|
||
if (preview) {
|
||
preview += '...';
|
||
}
|
||
|
||
// 获取行号
|
||
const lineNumber = block.querySelector('.line-number');
|
||
const line = lineNumber ? lineNumber.textContent : '?';
|
||
|
||
changes.push({
|
||
id: changeId,
|
||
type: type,
|
||
preview: preview || '变更内容',
|
||
line: line
|
||
});
|
||
}
|
||
}
|
||
});
|
||
|
||
// 如果没有变更，显示未发现变更
|
||
if (changes.length === 0) {
|
||
navContent.innerHTML = '<div class="navigation-item"><div class="navigation-item-preview">未发现变更</div></div>';
|
||
return;
|
||
}
|
||
|
||
// 生成导航项
|
||
changes.forEach(change => {
|
||
const navItem = document.createElement('div');
|
||
navItem.className = 'navigation-item';
|
||
navItem.setAttribute('data-change-id', change.id);
|
||
|
||
navItem.innerHTML = `
|
||
<div class="navigation-item-header">
|
||
<span class="navigation-item-number">${change.id}</span>
|
||
<span class="navigation-item-type ${change.type.toLowerCase()}">${change.type}</span>
|
||
</div>
|
||
<div class="navigation-item-preview">${change.preview}</div>
|
||
<div class="navigation-item-line">行 ${change.line}</div>
|
||
`;
|
||
|
||
// 添加点击事件
|
||
navItem.addEventListener('click', () => {
|
||
const targetBlock = document.querySelector(`[data-change-id="${change.id}"].change-block`);
|
||
if (targetBlock) {
|
||
targetBlock.scrollIntoView({
|
||
behavior: 'smooth',
|
||
block: 'center'
|
||
});
|
||
updateActiveNavItem();
|
||
highlightBlock(targetBlock);
|
||
}
|
||
});
|
||
|
||
navContent.appendChild(navItem);
|
||
});
|
||
}
|
||
|
||
// 导航功能
|
||
class DiffNavigation {
|
||
constructor() {
|
||
this.navFloat = document.getElementById('navigation-float');
|
||
this.navToggle = document.getElementById('navigation-toggle');
|
||
// 延迟获取导航项，确保它们已经生成
|
||
this.updateNavigationElements();
|
||
this.isCollapsed = false;
|
||
|
||
// 调试日志
|
||
console.log('DiffNavigation Constructor:');
|
||
console.log(' navFloat:', this.navFloat);
|
||
console.log(' navToggle:', this.navToggle);
|
||
console.log(' navItems length:', this.navItems.length);
|
||
console.log(' changeBlocks length:', this.changeBlocks.length);
|
||
|
||
// 打印所有导航项的详细信息
|
||
this.navItems.forEach((item, index) => {
|
||
console.log(` Navigation item ${index}:`, {
|
||
id: item.getAttribute('data-change-id'),
|
||
hasType: !!item.querySelector('.navigation-item-type'),
|
||
hasPreview: !!item.querySelector('.navigation-item-preview')
|
||
});
|
||
});
|
||
|
||
this.init();
|
||
}
|
||
|
||
updateNavigationElements() {
|
||
this.navItems = document.querySelectorAll('.navigation-item');
|
||
this.changeBlocks = document.querySelectorAll('.change-block');
|
||
}
|
||
|
||
init() {
|
||
// 绑定收起/展开按钮
|
||
this.navToggle.addEventListener('click', () => {
|
||
this.toggleCollapse();
|
||
});
|
||
|
||
// 绑定导航项点击事件
|
||
this.navItems.forEach(item => {
|
||
item.addEventListener('click', () => {
|
||
this.navigateToChange(item);
|
||
});
|
||
});
|
||
|
||
// 监听滚动，更新当前激活的导航项
|
||
window.addEventListener('scroll', () => {
|
||
this.updateActiveNavItem();
|
||
});
|
||
|
||
// 初始化当前激活项
|
||
this.updateActiveNavItem();
|
||
}
|
||
|
||
toggleCollapse() {
|
||
this.isCollapsed = !this.isCollapsed;
|
||
if (this.isCollapsed) {
|
||
this.navFloat.classList.add('collapsed');
|
||
this.navToggle.textContent = '+';
|
||
} else {
|
||
this.navFloat.classList.remove('collapsed');
|
||
this.navToggle.textContent = '−';
|
||
}
|
||
}
|
||
|
||
navigateToChange(navItem) {
|
||
const changeId = navItem.getAttribute('data-change-id');
|
||
const targetBlock = document.querySelector(`[data-change-id="${changeId}"].change-block`);
|
||
|
||
if (targetBlock) {
|
||
// 滚动到目标位置
|
||
targetBlock.scrollIntoView({
|
||
behavior: 'smooth',
|
||
block: 'center'
|
||
});
|
||
|
||
// 更新激活状态
|
||
this.updateActiveNavItem();
|
||
|
||
// 添加临时高亮效果
|
||
this.highlightBlock(targetBlock);
|
||
}
|
||
}
|
||
|
||
updateActiveNavItem() {
|
||
const scrollPosition = window.scrollY + window.innerHeight / 2;
|
||
let activeItem = null;
|
||
|
||
this.changeBlocks.forEach(block => {
|
||
const blockTop = block.offsetTop;
|
||
const blockBottom = blockTop + block.offsetHeight;
|
||
|
||
if (scrollPosition >= blockTop && scrollPosition <= blockBottom) {
|
||
const changeId = block.getAttribute('data-change-id');
|
||
activeItem = document.querySelector(`[data-change-id="${changeId}"].navigation-item`);
|
||
}
|
||
});
|
||
|
||
// 更新激活状态
|
||
this.navItems.forEach(item => {
|
||
item.classList.remove('active');
|
||
});
|
||
|
||
if (activeItem) {
|
||
activeItem.classList.add('active');
|
||
// 确保激活项在视口内
|
||
this.ensureNavItemVisible(activeItem);
|
||
}
|
||
}
|
||
|
||
ensureNavItemVisible(navItem) {
|
||
const navContent = document.getElementById('navigation-content');
|
||
const itemTop = navItem.offsetTop;
|
||
const itemBottom = itemTop + navItem.offsetHeight;
|
||
const scrollTop = navContent.scrollTop;
|
||
const contentHeight = navContent.clientHeight;
|
||
|
||
if (itemTop < scrollTop) {
|
||
navContent.scrollTop = itemTop - 10;
|
||
} else if (itemBottom > scrollTop + contentHeight) {
|
||
navContent.scrollTop = itemBottom - contentHeight + 10;
|
||
}
|
||
}
|
||
|
||
highlightBlock(block) {
|
||
// 移除所有高亮
|
||
document.querySelectorAll('.line-wrapper.highlight').forEach(line => {
|
||
line.classList.remove('highlight');
|
||
});
|
||
|
||
// 添加高亮效果
|
||
block.classList.add('highlight');
|
||
|
||
// 2秒后移除高亮
|
||
setTimeout(() => {
|
||
block.classList.remove('highlight');
|
||
}, 2000);
|
||
}
|
||
|
||
// 公共方法：跳转到指定变更
|
||
goToChange(changeId) {
|
||
const navItem = document.querySelector(`[data-change-id="${changeId}"].navigation-item`);
|
||
if (navItem) {
|
||
this.navigateToChange(navItem);
|
||
}
|
||
}
|
||
|
||
// 公共方法：获取所有变更列表
|
||
getChanges() {
|
||
const changes = [];
|
||
this.navItems.forEach(item => {
|
||
const changeId = item.getAttribute('data-change-id');
|
||
const typeElement = item.querySelector('.navigation-item-type');
|
||
const previewElement = item.querySelector('.navigation-item-preview');
|
||
const lineElement = item.querySelector('.navigation-item-line');
|
||
|
||
if (changeId && typeElement && previewElement && lineElement) {
|
||
changes.push({
|
||
id: parseInt(changeId),
|
||
type: typeElement.textContent,
|
||
preview: previewElement.textContent,
|
||
line: lineElement.textContent
|
||
});
|
||
}
|
||
});
|
||
return changes;
|
||
}
|
||
}
|
||
|
||
// 初始化导航
|
||
let diffNavigation;
|
||
document.addEventListener('DOMContentLoaded', () => {
|
||
// 首先生成导航项
|
||
generateNavigationItems();
|
||
|
||
// 调试日志
|
||
console.log('=== Diff Navigation Debug ===');
|
||
console.log('Navigation float:', document.getElementById('navigation-float'));
|
||
console.log('Navigation content:', document.getElementById('navigation-content'));
|
||
console.log('Navigation items:', document.querySelectorAll('.navigation-item'));
|
||
console.log('Change blocks:', document.querySelectorAll('.change-block'));
|
||
console.log('Has changes lines:', document.querySelectorAll('.line-wrapper.has-changes'));
|
||
|
||
diffNavigation = new DiffNavigation();
|
||
|
||
// 添加键盘快捷键支持
|
||
document.addEventListener('keydown', (e) => {
|
||
if (e.ctrlKey || e.metaKey) {
|
||
if (e.key >= '1' && e.key <= '9') {
|
||
e.preventDefault();
|
||
const changeId = parseInt(e.key);
|
||
diffNavigation.goToChange(changeId);
|
||
}
|
||
}
|
||
});
|
||
});
|
||
|
||
// 原有的点击高亮功能
|
||
document.querySelectorAll('.line-wrapper.has-changes').forEach(lineWrapper => {
|
||
lineWrapper.addEventListener('click', () => {
|
||
// 移除所有高亮
|
||
document.querySelectorAll('.line-wrapper.highlight').forEach(line => {
|
||
line.classList.remove('highlight');
|
||
});
|
||
|
||
// 高亮当前行
|
||
lineWrapper.classList.add('highlight');
|
||
});
|
||
});
|
||
</script>
|
||
</body>
|
||
</html>'''
|
||
|
||
return html
|
||
|
||
def save_files(title, diff_html, diff_text, full_text, timestamp, note="", revid=None, cn_content=None, old_full_text=None):
|
||
global CURRENT_OUTPUT_DIR
|
||
|
||
# 确保本次执行的输出目录已经创建
|
||
if CURRENT_OUTPUT_DIR is None:
|
||
current_time_str = datetime.now().strftime("%Y%m%d_%H%M%S")
|
||
CURRENT_OUTPUT_DIR = OUTPUT_DIR / current_time_str
|
||
CURRENT_OUTPUT_DIR.mkdir(exist_ok=True)
|
||
# 创建子目录
|
||
(CURRENT_OUTPUT_DIR / "new_pages").mkdir(exist_ok=True)
|
||
(CURRENT_OUTPUT_DIR / "changed_pages").mkdir(exist_ok=True)
|
||
print(f"创建本次执行的输出目录: {CURRENT_OUTPUT_DIR}")
|
||
|
||
safe_title = "".join(c if c.isalnum() or c in " -_." else "_" for c in title)
|
||
time_str = timestamp[:19].replace("-", "").replace(":", "").replace("T", "_")
|
||
base_filename = f"{safe_title}-{time_str}-{revid}" if revid else f"{safe_title}-{time_str}"
|
||
|
||
# 判断是否为新页面
|
||
is_new_page = diff_text.startswith("新创建页面")
|
||
|
||
if is_new_page:
|
||
# 新页面：只保存完整内容到 new_pages 目录
|
||
target_dir = CURRENT_OUTPUT_DIR / "new_pages"
|
||
print(f" 检测到新页面，只保存完整内容")
|
||
|
||
# 保存最新完整内容
|
||
full_file = target_dir / f"{base_filename}.full.txt"
|
||
if full_text:
|
||
try:
|
||
with open(full_file, "w", encoding="utf-8") as f:
|
||
f.write(full_text)
|
||
print(f" → 已保存: {full_file.relative_to(OUTPUT_DIR)} (新页面完整内容)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {full_file} 时出错: {e}")
|
||
|
||
else:
|
||
# 变更页面：保存主要文件到 changed_pages 目录，其他文件到 files 目录
|
||
target_dir = CURRENT_OUTPUT_DIR / "changed_pages"
|
||
files_dir = target_dir / "files"
|
||
files_dir.mkdir(exist_ok=True)
|
||
|
||
# 1. 保存最新完整内容（主文件）
|
||
full_file = target_dir / f"{base_filename}.full.txt"
|
||
if full_text:
|
||
try:
|
||
with open(full_file, "w", encoding="utf-8") as f:
|
||
f.write(full_text)
|
||
print(f" → 已保存: {full_file.relative_to(OUTPUT_DIR)} (完整内容)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {full_file} 时出错: {e}")
|
||
|
||
# 2. 保存中文翻译内容（主文件）
|
||
if cn_content:
|
||
cn_file = target_dir / f"{base_filename}.cn.txt"
|
||
try:
|
||
with open(cn_file, "w", encoding="utf-8") as f:
|
||
f.write(cn_content)
|
||
print(f" → 已保存: {cn_file.relative_to(OUTPUT_DIR)} (中文翻译)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {cn_file} 时出错: {e}")
|
||
|
||
# 3. 创建双语对比HTML页面（主文件）
|
||
en_new_lines = full_text.splitlines() if full_text else []
|
||
en_old_lines = old_full_text.splitlines() if old_full_text else []
|
||
|
||
comparison_html = create_diff_html(title, diff_text, en_old_lines, en_new_lines, cn_content)
|
||
comparison_file = target_dir / f"{base_filename}.comparison.html"
|
||
try:
|
||
with open(comparison_file, "w", encoding="utf-8") as f:
|
||
f.write(comparison_html)
|
||
print(f" → 已保存: {comparison_file.relative_to(OUTPUT_DIR)} (双语对比页面)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {comparison_file} 时出错: {e}")
|
||
|
||
# 保存参考文件到 files 目录
|
||
if diff_html:
|
||
diff_file = files_dir / f"{base_filename}.diff.html"
|
||
try:
|
||
with open(diff_file, "w", encoding="utf-8") as f:
|
||
f.write(diff_html)
|
||
print(f" → 已保存: {diff_file.relative_to(OUTPUT_DIR)} (官方diff - 参考)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {diff_file} 时出错: {e}")
|
||
|
||
if diff_text:
|
||
text_diff_file = files_dir / f"{base_filename}.diff.txt"
|
||
try:
|
||
with open(text_diff_file, "w", encoding="utf-8") as f:
|
||
f.write(diff_text)
|
||
print(f" → 已保存: {text_diff_file.relative_to(OUTPUT_DIR)} (文本diff - 参考)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {text_diff_file} 时出错: {e}")
|
||
|
||
if old_full_text:
|
||
old_full_file = files_dir / f"{base_filename}.old.txt"
|
||
try:
|
||
with open(old_full_file, "w", encoding="utf-8") as f:
|
||
f.write(old_full_text)
|
||
print(f" → 已保存: {old_full_file.relative_to(OUTPUT_DIR)} (历史版本 - 参考)")
|
||
except Exception as e:
|
||
print(f" → 保存文件 {old_full_file} 时出错: {e}")
|
||
|
||
def process_single_page(title, since_time, update_timestamp=False):
|
||
"""只处理单个页面"""
|
||
print(f"正在单独处理页面：{title}")
|
||
|
||
# 获取当前最新 revid
|
||
try:
|
||
latest_content, latest_ts, latest_revid = get_page_content(WIKI_API_URL_EN, SESSION_EN, title)
|
||
if latest_content is None:
|
||
print("页面不存在或被删除")
|
||
return None
|
||
|
||
# 获取旧 revid
|
||
old_revid = get_old_revid(title, since_time)
|
||
|
||
# 初始化变量
|
||
diff_html = None
|
||
diff_text = None
|
||
old_content = None
|
||
cn_content = None
|
||
|
||
if old_revid:
|
||
# 获取历史版本内容
|
||
old_content, old_ts, _ = get_page_content(WIKI_API_URL_EN, SESSION_EN, title, old_revid)
|
||
|
||
if old_content is not None:
|
||
# 生成文本diff
|
||
diff_text = generate_text_diff(old_content, latest_content)
|
||
print(f" 生成了文本diff ({len(diff_text)} 字符)")
|
||
else:
|
||
print(f" 无法获取历史版本内容")
|
||
else:
|
||
# 新页面
|
||
print(" 这是新创建的页面")
|
||
|
||
# 搜索对应的中文页面
|
||
print(" 搜索中文翻译...")
|
||
cn_title = search_chinese_page(title)
|
||
if cn_title:
|
||
print(f" 找到中文页面: {cn_title}")
|
||
cn_content, cn_ts, cn_revid = get_page_content(WIKI_API_URL_CN, SESSION_CN, cn_title)
|
||
if cn_content:
|
||
print(f" 获取中文内容成功 ({len(cn_content)} 字符)")
|
||
else:
|
||
print(" 无法获取中文页面内容")
|
||
else:
|
||
print(" 未找到对应的中文翻译页面")
|
||
|
||
# 获取官方diff（可选）
|
||
if old_revid:
|
||
diff_params = {
|
||
"action": "compare",
|
||
"fromrev": old_revid,
|
||
"torev": latest_revid,
|
||
"format": "json"
|
||
}
|
||
try:
|
||
diff_resp = SESSION_EN.get(WIKI_API_URL_EN, params=diff_params).json()
|
||
diff_html = diff_resp.get("compare", {}).get("*", "")
|
||
except Exception as e:
|
||
print(f" 获取官方HTML diff时出错: {e}")
|
||
|
||
# 保存所有文件
|
||
save_files(title, diff_html, diff_text, latest_content, latest_ts, "", latest_revid, cn_content, old_content)
|
||
|
||
if update_timestamp:
|
||
save_last_timestamp(latest_ts)
|
||
print(f"已更新全局时间戳 → {latest_ts}")
|
||
|
||
return latest_ts
|
||
except Exception as e:
|
||
print(f"处理页面 '{title}' 时出错: {e}")
|
||
return None
|
||
|
||
def process_all_pages_since(since_time):
|
||
"""处理自指定时间以来的所有页面变更"""
|
||
print("正在获取最近变更列表...")
|
||
changes = get_recent_changes(since_time)
|
||
if not changes:
|
||
print("没有发现任何变更")
|
||
return
|
||
|
||
latest_global_ts = since_time
|
||
for title, (latest_revid, ts) in changes.items():
|
||
print(f"\n处理：{title}")
|
||
# 复用单页处理逻辑
|
||
page_latest_ts = process_single_page(title, since_time)
|
||
|
||
if page_latest_ts and page_latest_ts > latest_global_ts:
|
||
latest_global_ts = page_latest_ts
|
||
|
||
save_last_timestamp(latest_global_ts)
|
||
print(f"\n全量同步完成！本次最新时间戳已更新为：{latest_global_ts}")
|
||
print(f"文件保存在：{CURRENT_OUTPUT_DIR.resolve() if CURRENT_OUTPUT_DIR else OUTPUT_DIR.resolve()}")
|
||
|
||
def main():
|
||
parser = argparse.ArgumentParser(description="MediaWiki 同步工具 - 增强版支持双语对比")
|
||
parser.add_argument("--since", type=str, help="强制从指定时间开始同步，格式如 2025-11-28T00:00:00Z")
|
||
parser.add_argument("--title", type=str, help="只同步指定的单个页面标题")
|
||
parser.add_argument("--update-timestamp", action="store_true",
|
||
help="在单页模式下，完成后仍然更新全局 last_sync_timestamp.txt")
|
||
parser.add_argument("--run", action="store_true",
|
||
help="执行同步操作（必须提供此参数才能真正执行同步）")
|
||
|
||
args = parser.parse_args()
|
||
|
||
# 如果没有提供 --run 参数，则显示帮助信息并退出
|
||
if not args.run:
|
||
parser.print_help()
|
||
return
|
||
|
||
# 确定实际使用的 since 时间
|
||
if args.since:
|
||
since_time = args.since
|
||
print(f"使用命令行指定的时间起点：{since_time}")
|
||
else:
|
||
since_time = load_last_timestamp()
|
||
if not since_time:
|
||
from datetime import timedelta
|
||
since_time = (datetime.utcnow() - timedelta(days=1)).isoformat(timespec='seconds') + "Z"
|
||
print(f"使用上次记录的时间起点：{since_time}")
|
||
|
||
# 单页面模式
|
||
if args.title:
|
||
process_single_page(args.title.strip(), since_time, args.update_timestamp)
|
||
return
|
||
|
||
# 全量模式 - 使用复用的单页处理逻辑
|
||
process_all_pages_since(since_time)
|
||
|
||
if __name__ == "__main__":
|
||
main() |