# METADATA # { # "name": "google_search", # # "display_name": { # "zh": "Google 搜索", # "en": "Google Search" # }, # "description": { # "zh": "提供 Google 普通搜索与 Google Scholar 学术搜索能力，支持设置语言与返回条数。", # "en": "Google web search and Google Scholar search, with configurable language and result count." # }, # "enabledByDefault": false, # "category": "Search", # "tools": [ # { # "name": "search_web", # "description": { "zh": "执行 Google 普通搜索，返回网页搜索结果。", "en": "Perform a regular Google search and return web results." }, # "parameters": [ # { "name": "query", "description": { "zh": "搜索关键词", "en": "Search query" }, "type": "string", "required": true }, # { "name": "max_results", "description": { "zh": "返回结果数量，默认 10，最大 20", "en": "Number of results to return (default: 10, max: 20)" }, "type": "number", "required": false }, # { "name": "language", "description": { "zh": "界面语言参数，默认 en", "en": "Interface language (default: en)" }, "type": "string", "required": false }, # { "name": "region", "description": { "zh": "地区参数，例如 us、cn。默认 us", "en": "Region parameter, e.g. us, cn (default: us)" }, "type": "string", "required": false }, # { "name": "includeLinks", "description": { "zh": "是否在结果中包含可点击的链接列表，默认为false。", "en": "Whether to include a clickable link list in results (default: false)" }, "type": "boolean", "required": false } # ] # }, # { # "name": "search_scholar", # "description": { "zh": "执行 Google Scholar 学术搜索，返回学术文献结果。", "en": "Perform a Google Scholar search and return academic results." }, # "parameters": [ # { "name": "query", "description": { "zh": "搜索关键词", "en": "Search query" }, "type": "string", "required": true }, # { "name": "max_results", "description": { "zh": "返回结果数量，默认 10，最大 20", "en": "Number of results to return (default: 10, max: 20)" }, "type": "number", "required": false }, # { "name": "language", "description": { "zh": "界面语言参数，默认 en", "en": "Interface language (default: en)" }, "type": "string", "required": false }, # { "name": "includeLinks", "description": { "zh": "是否在结果中包含可点击的链接列表，默认为false。", "en": "Whether to include a clickable link list in results (default: false)" }, "type": "boolean", "required": false } # ] # }, # { # "name": "search_scholar_mirror", # "description": { "zh": "通过镜像站执行 Google Scholar 学术搜索，以绕过人机验证。", "en": "Use a mirror site to perform Google Scholar searches to bypass CAPTCHA checks." }, # "parameters": [ # { "name": "query", "description": { "zh": "搜索关键词", "en": "Search query" }, "type": "string", "required": true }, # { "name": "max_results", "description": { "zh": "返回结果数量，默认 10，最大 20", "en": "Number of results to return (default: 10, max: 20)" }, "type": "number", "required": false }, # { "name": "language", "description": { "zh": "界面语言参数，默认 en", "en": "Interface language (default: en)" }, "type": "string", "required": false }, # { "name": "includeLinks", "description": { "zh": "是否在结果中包含可点击的链接列表，默认为false。", "en": "Whether to include a clickable link list in results (default: false)" }, "type": "boolean", "required": false } # ] # } # ] # } import json import re from urllib.parse import quote GOOGLE_SEARCH_URL = "https://www.google.com/search" GOOGLE_SCHOLAR_URL = "https://scholar.google.com/scholar" GOOGLE_SCHOLAR_MIRRORS = [ "https://xs.cntpj.com/scholar", ] MAX_RESULTS = 20 def buildUrl(base, params): queryString = "&".join(f"{quote(k)}={quote(v)}" for k, v in params.items()) return f"{base}?{queryString}" def getHostname(rawUrl): match = re.match(r'^[a-zA-Z][a-zA-Z\d+\-.]*:\/\/([^/?#]+)', rawUrl.strip()) if not match or not match.group(1): return rawUrl authority = match.group(1) host = authority.split("@")[-1] or authority if "@" in authority else authority return re.sub(r':\d+$', '', host) async def fetchHtmlViaWebVisit(url): result = await Tools.Net.visit(url) # The result can be a string if the underlying tool returns a simple string. # We'll normalize it to a dict object. if isinstance(result, str): return { "url": url, "title": "", "content": result, "links": [], } return result # 解析相关逻辑已移除，直接返回 visit 的纯文本结果 async def performSearch(url, includeLinks=False, sourceName=""): try: result = await fetchHtmlViaWebVisit(url) content = result.get("content", "") if isinstance(result, dict) else "" parts = [] if isinstance(result, dict) and result.get("visitKey"): parts.append(f"visit_key: {result['visitKey']}") if includeLinks and isinstance(result, dict) and result.get("links") and len(result["links"]) > 0: linksLines = [f"[{index + 1}] {link.get('text')}" for index, link in enumerate(result["links"])] parts.append("\n".join(linksLines)) parts.append(content) return { "success": True, "message": f"{sourceName} 搜索成功", "data": "\n\n".join(parts) } except Exception as error: msg = str(getattr(error, "message", error)) return { "success": False, "message": f"{sourceName} 搜索失败: {msg}" } async def searchWeb(params): if not params.get("query") or params.get("query", "").strip() == "": raise Exception("请提供有效的 query 参数。") maxResults = min(max(params.get("max_results") or 10, 1), MAX_RESULTS) language = params.get("language") or "en" region = params.get("region") or "us" url = buildUrl(GOOGLE_SEARCH_URL, { "q": params["query"], "hl": language, "gl": region, "num": str(maxResults), "pws": "0", }) return await performSearch(url, params.get("includeLinks"), "Google") async def searchScholar(params): if not params.get("query") or params.get("query", "").strip() == "": raise Exception("请提供有效的 query 参数。") maxResults = min(max(params.get("max_results") or 10, 1), MAX_RESULTS) language = params.get("language") or "en" url = buildUrl(GOOGLE_SCHOLAR_URL, { "q": params["query"], "hl": language, "as_sdt": "0,5", "num": str(maxResults) }) return await performSearch(url, params.get("includeLinks"), "Google Scholar") async def searchScholarMirror(params): if not params.get("query") or params.get("query", "").strip() == "": raise Exception("请提供有效的 query 参数。") maxResults = min(max(params.get("max_results") or 10, 1), MAX_RESULTS) language = params.get("language") or "en" mirrorUrls = [ buildUrl(mirror, { "q": params["query"], "hl": language, "as_sdt": "0,5", "num": str(maxResults) }) for mirror in GOOGLE_SCHOLAR_MIRRORS ] if len(mirrorUrls) == 0: return { "success": False, "message": "没有可用的 Google Scholar 镜像地址。" } lastError = None for currentUrl in mirrorUrls: try: searchResult = await performSearch(currentUrl, params.get("includeLinks"), f"Google Scholar 镜像 ({getHostname(currentUrl)})") if searchResult.get("success") and searchResult.get("data"): # Check for CAPTCHA in the content if "recaptcha" in searchResult["data"] or "人机身份验证" in searchResult["data"]: raise Exception("CAPTCHA required") return searchResult # If performSearch itself fails, it will throw and be caught below. except Exception as error: lastError = error msg = str(getattr(error, "message", error)) console.log(f"Attempt with {currentUrl} failed: {msg}") errMsg = str(getattr(lastError, "message", lastError)) if lastError else "Unknown error" return { "success": False, "message": f"Google Scholar 镜像搜索在尝试所有镜像后失败: {errMsg}" } async def main(): console.log("--- Testing Web Search ---") webResult = await searchWeb({"query": "TypeScript"}) console.log(json.dumps(webResult, indent=2, ensure_ascii=False)) console.log("\n--- Testing Scholar Search ---") scholarResult = await searchScholar({"query": "Large Language Models"}) console.log(json.dumps(scholarResult, indent=2, ensure_ascii=False)) console.log("\n--- Testing Scholar Mirror Search ---") scholarMirrorResult = await searchScholarMirror({"query": "Quantum Computing"}) console.log(json.dumps(scholarMirrorResult, indent=2, ensure_ascii=False)) def wrap(coreFunction): async def wrapped(params): # The core function expects the params object directly. return await coreFunction(params) return wrapped exports.search_web = wrap(searchWeb) exports.search_scholar = wrap(searchScholar) exports.search_scholar_mirror = wrap(searchScholarMirror) exports.main = main