互联网搜索支持多搜索引擎选择
- 新增搜索引擎选择下拉框 - 支持 Bing 中国（默认）、Bing 国际、Google、百度 - 后端 API 支持 engine 参数 - 添加百度搜索结果解析方法
This commit is contained in:
+3
-2
@@ -140,17 +140,18 @@ def internet_search():
|
||||
data = request.get_json()
|
||||
keyword = data.get('keyword', '')
|
||||
max_results = data.get('max_results', 10)
|
||||
save_to_library = data.get('save_to_library', False)
|
||||
engine = data.get('engine', 'bing_cn') # 默认 Bing 中国
|
||||
|
||||
if not keyword:
|
||||
return jsonify({'error': '请提供搜索关键词'}), 400
|
||||
|
||||
# 执行互联网搜索
|
||||
results = search_service.search_internet(keyword, max_results)
|
||||
results = search_service.search_internet(keyword, max_results, engine)
|
||||
|
||||
return jsonify({
|
||||
'success': True,
|
||||
'keyword': keyword,
|
||||
'engine': engine,
|
||||
'results': results,
|
||||
'count': len(results)
|
||||
})
|
||||
|
||||
@@ -33,18 +33,32 @@ class SearchService:
|
||||
)
|
||||
return result.stdout, result.stderr, result.returncode
|
||||
|
||||
def search_internet(self, keyword, max_results=None):
|
||||
def search_internet(self, keyword, max_results=None, engine='bing_cn'):
|
||||
"""
|
||||
从互联网搜索（使用 agent-browser 浏览器自动化）
|
||||
|
||||
支持的搜索引擎:
|
||||
- bing_cn: Bing 中国（默认）
|
||||
- bing_global: Bing 国际版
|
||||
- google: Google
|
||||
- baidu: 百度
|
||||
"""
|
||||
max_results = max_results or self.max_results
|
||||
results = []
|
||||
|
||||
try:
|
||||
# 1. 打开 Bing 搜索
|
||||
encoded_keyword = urllib.parse.quote(keyword)
|
||||
search_url = f"https://www.bing.com/search?q={encoded_keyword}"
|
||||
# 根据搜索引擎选择 URL
|
||||
encoded_keyword = urllib.parse.quote(keyword)
|
||||
search_urls = {
|
||||
'bing_cn': f"https://cn.bing.com/search?q={encoded_keyword}",
|
||||
'bing_global': f"https://www.bing.com/search?q={encoded_keyword}",
|
||||
'google': f"https://www.google.com/search?q={encoded_keyword}",
|
||||
'baidu': f"https://www.baidu.com/s?wd={encoded_keyword}"
|
||||
}
|
||||
|
||||
search_url = search_urls.get(engine, search_urls['bing_cn'])
|
||||
|
||||
try:
|
||||
# 1. 打开搜索引擎
|
||||
stdout, stderr, code = self._run_browser('open', search_url, '--timeout', '20000')
|
||||
if code != 0:
|
||||
print(f"打开搜索页面失败: {stderr}")
|
||||
@@ -66,9 +80,11 @@ class SearchService:
|
||||
print(f"解析 JSON 失败: {stdout[:500]}")
|
||||
return results
|
||||
|
||||
# 4. 从 accessibility tree 中提取搜索结果
|
||||
# Bing 搜索结果在 main[aria-label="搜索结果"] 区域内
|
||||
results = self._parse_bing_results(data, max_results)
|
||||
# 4. 根据搜索引擎选择解析方法
|
||||
if engine == 'baidu':
|
||||
results = self._parse_baidu_results(data, max_results)
|
||||
else:
|
||||
results = self._parse_bing_results(data, max_results)
|
||||
|
||||
# 5. 关闭浏览器
|
||||
self._run_browser('close')
|
||||
@@ -143,6 +159,44 @@ class SearchService:
|
||||
|
||||
return results
|
||||
|
||||
def _parse_baidu_results(self, snapshot_data, max_results=10):
|
||||
"""从百度搜索结果中解析标题和链接"""
|
||||
results = []
|
||||
|
||||
snapshot = snapshot_data.get('data', {}).get('snapshot', '')
|
||||
if not snapshot:
|
||||
return results
|
||||
|
||||
# 百度搜索结果解析
|
||||
refs = []
|
||||
lines = snapshot.split('\n')
|
||||
|
||||
for line in lines:
|
||||
line = line.strip()
|
||||
|
||||
# 百度结果通常在 link 标签中
|
||||
if 'link "' in line and '[ref=' in line:
|
||||
match = re.search(r'link "([^"]+)" \[ref=(e\d+)\]', line)
|
||||
if match:
|
||||
title = match.group(1)
|
||||
ref = match.group(2)
|
||||
# 过滤百度内部链接和广告
|
||||
if len(title) > 10 and '百度' not in title[:6]:
|
||||
refs.append((title, ref))
|
||||
|
||||
# 获取每个结果的 URL
|
||||
for title, ref in refs[:max_results]:
|
||||
url = self._get_link_url(ref)
|
||||
if url and 'baidu.com' not in url:
|
||||
results.append({
|
||||
'title': title,
|
||||
'url': url,
|
||||
'snippet': '',
|
||||
'source': 'baidu'
|
||||
})
|
||||
|
||||
return results
|
||||
|
||||
def _get_link_url(self, ref):
|
||||
"""通过 agent-browser 获取链接的 URL"""
|
||||
try:
|
||||
|
||||
@@ -87,6 +87,21 @@
|
||||
font-size: 16px;
|
||||
}
|
||||
|
||||
.engine-select {
|
||||
padding: 15px;
|
||||
border: 2px solid #e9ecef;
|
||||
border-radius: 8px;
|
||||
font-size: 16px;
|
||||
background: white;
|
||||
cursor: pointer;
|
||||
min-width: 120px;
|
||||
}
|
||||
|
||||
.engine-select:focus {
|
||||
outline: none;
|
||||
border-color: #667eea;
|
||||
}
|
||||
|
||||
.search-options {
|
||||
display: flex;
|
||||
gap: 20px;
|
||||
|
||||
+9
-2
@@ -19,6 +19,7 @@ document.addEventListener('DOMContentLoaded', () => {
|
||||
async function doSearch() {
|
||||
const keyword = document.getElementById('search-keyword').value.trim();
|
||||
const maxResults = parseInt(document.getElementById('search-count').value) || 10;
|
||||
const engine = document.getElementById('search-engine').value;
|
||||
const autoFetch = document.getElementById('auto-fetch').checked;
|
||||
const autoSave = document.getElementById('auto-save').checked;
|
||||
|
||||
@@ -28,7 +29,13 @@ async function doSearch() {
|
||||
}
|
||||
|
||||
// 更新状态
|
||||
updateStatus('searching', '搜索中...');
|
||||
const engineNames = {
|
||||
'bing_cn': 'Bing 中国',
|
||||
'bing_global': 'Bing 国际',
|
||||
'google': 'Google',
|
||||
'baidu': '百度'
|
||||
};
|
||||
updateStatus('searching', `正在通过 ${engineNames[engine]} 搜索...`);
|
||||
document.getElementById('search-btn').disabled = true;
|
||||
|
||||
// 显示进度
|
||||
@@ -41,7 +48,7 @@ async function doSearch() {
|
||||
const response = await fetch(`${API_BASE}/api/articles/internet-search`, {
|
||||
method: 'POST',
|
||||
headers: { 'Content-Type': 'application/json' },
|
||||
body: JSON.stringify({ keyword, max_results: maxResults })
|
||||
body: JSON.stringify({ keyword, max_results: maxResults, engine })
|
||||
});
|
||||
|
||||
const data = await response.json();
|
||||
|
||||
@@ -29,6 +29,12 @@
|
||||
<!-- 搜索区域 -->
|
||||
<div class="search-box">
|
||||
<div class="search-input-group">
|
||||
<select id="search-engine" class="engine-select" title="选择搜索引擎">
|
||||
<option value="bing_cn" selected>Bing 中国</option>
|
||||
<option value="bing_global">Bing 国际</option>
|
||||
<option value="google">Google</option>
|
||||
<option value="baidu">百度</option>
|
||||
</select>
|
||||
<input type="text" id="search-keyword" placeholder="输入关键词搜索..." class="search-input-large">
|
||||
<input type="number" id="search-count" value="10" min="1" max="20" class="count-input" title="结果数量">
|
||||
<button onclick="doSearch()" class="btn btn-primary btn-large" id="search-btn">
|
||||
|
||||
Reference in New Issue
Block a user