gardel/Mofox-Core
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(exa_engine): 更新Exa搜索方法，优化搜索参数格式，增强结果处理逻辑

Browse Source 
feat(url_parser): 精简Exa客户端参数，优化结果摘要处理逻辑
This commit is contained in:
Windpicker-owo
2025-11-10 22:17:58 +08:00
parent 4b71cb4a88
commit e79494d3c2
2 changed files with 19 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

36
src/plugins/built_in/web_search_tool/engines/exa_engine.py
View File

@@ -39,7 +39,7 @@ class ExaSearchEngine(BaseSearchEngine):
return self.api_manager.is_available() return self.api_manager.is_available()
async def search(self, args: dict[str, Any]) -> list[dict[str, Any]]: async def search(self, args: dict[str, Any]) -> list[dict[str, Any]]:
"""执行优化的Exa搜索（使用answer模式）""" """执行优化的Exa搜索（使用新的search API）"""
if not self.is_available(): if not self.is_available():
return [] return []
@@ -47,12 +47,13 @@ class ExaSearchEngine(BaseSearchEngine):
num_results = min(args.get("num_results", 5), 5) # 默认5个结果，但限制最多5个 num_results = min(args.get("num_results", 5), 5) # 默认5个结果，但限制最多5个
time_range = args.get("time_range", "any") time_range = args.get("time_range", "any")
# 优化的搜索参数 - 更注重答案质量 # 使用新的搜索参数格式
exa_args = { exa_args = {
"num_results": num_results, "num_results": num_results,
"text": True, "contents": {
"highlights": True, "text": True,
"summary": True, # 启用自动摘要 "summary": True, # 启用自动摘要
},
} }
# 时间范围过滤 # 时间范围过滤
@@ -69,23 +70,20 @@ class ExaSearchEngine(BaseSearchEngine):
return [] return []
loop = asyncio.get_running_loop() loop = asyncio.get_running_loop()
# 使用search_and_contents获取完整内容，优化为answer模式 # 使用新的search方法
func = functools.partial(exa_client.search_and_contents, query, **exa_args) func = functools.partial(exa_client.search, query, **exa_args)
search_response = await loop.run_in_executor(None, func) search_response = await loop.run_in_executor(None, func)
# 优化结果处理 - 更注重答案质量 # 优化结果处理 - 更注重答案质量
results = [] results = []
for res in search_response.results: for res in search_response.results:
# 获取最佳内容片段 # 获取最佳内容片段
highlights = getattr(res, "highlights", [])
summary = getattr(res, "summary", "") summary = getattr(res, "summary", "")
text = getattr(res, "text", "") text = getattr(res, "text", "")
# 智能内容选择：摘要 > 高亮 > 文本开头 # 智能内容选择：摘要 > 文本开头
if summary and len(summary) > 50: if summary and len(summary) > 50:
snippet = summary.strip() snippet = summary.strip()
elif highlights:
snippet = " ".join(highlights).strip()
elif text: elif text:
snippet = text[:300] + "..." if len(text) > 300 else text snippet = text[:300] + "..." if len(text) > 300 else text
else: else:
@@ -105,7 +103,7 @@ class ExaSearchEngine(BaseSearchEngine):
return results return results
except Exception as e: except Exception as e:
logger.error(f"Exa answer模式搜索失败: {e}") logger.error(f"Exa搜索失败: {e}")
return [] return []
async def answer_search(self, args: dict[str, Any]) -> list[dict[str, Any]]: async def answer_search(self, args: dict[str, Any]) -> list[dict[str, Any]]:
@@ -119,9 +117,10 @@ class ExaSearchEngine(BaseSearchEngine):
# 精简的搜索参数 - 专注快速答案 # 精简的搜索参数 - 专注快速答案
exa_args = { exa_args = {
"num_results": num_results, "num_results": num_results,
"text": False, # 不需要全文 "contents": {
"highlights": True, # 只要关键高亮 "text": False, # 不需要全文
"summary": True, # 优先摘要 "summary": True, # 优先摘要
},
} }
try: try:
@@ -130,17 +129,16 @@ class ExaSearchEngine(BaseSearchEngine):
return [] return []
loop = asyncio.get_running_loop() loop = asyncio.get_running_loop()
func = functools.partial(exa_client.search_and_contents, query, **exa_args) func = functools.partial(exa_client.search, query, **exa_args)
search_response = await loop.run_in_executor(None, func) search_response = await loop.run_in_executor(None, func)
# 极简结果处理 - 只保留最核心信息 # 极简结果处理 - 只保留最核心信息
results = [] results = []
for res in search_response.results: for res in search_response.results:
summary = getattr(res, "summary", "") summary = getattr(res, "summary", "")
highlights = getattr(res, "highlights", [])
# 优先使用摘要，否则使用高亮 # 使用摘要作为答案
answer_text = summary.strip() if summary and len(summary) > 30 else " ".join(highlights).strip() answer_text = summary.strip() if summary else ""
if answer_text and len(answer_text) > 20: if answer_text and len(answer_text) > 20:
results.append({ results.append({

5
src/plugins/built_in/web_search_tool/tools/url_parser.py
View File

@@ -179,7 +179,7 @@ class URLParserTool(BaseTool):
logger.error("无法获取Exa客户端") logger.error("无法获取Exa客户端")
else: else:
loop = asyncio.get_running_loop() loop = asyncio.get_running_loop()
exa_params = {"text": True, "summary": True, "highlights": True} exa_params = {"text": True, "summary": True}
func = functools.partial(exa_client.get_contents, urls, **exa_params) func = functools.partial(exa_client.get_contents, urls, **exa_params)
contents_response = await loop.run_in_executor(None, func) contents_response = await loop.run_in_executor(None, func)
except Exception as e: except Exception as e:
@@ -197,9 +197,8 @@ class URLParserTool(BaseTool):
res = results_map.get(status.id) res = results_map.get(status.id)
if res: if res:
summary = getattr(res, "summary", "") summary = getattr(res, "summary", "")
highlights = " ".join(getattr(res, "highlights", []))
text_snippet = (getattr(res, "text", "")[:300] + "...") if getattr(res, "text", "") else "" text_snippet = (getattr(res, "text", "")[:300] + "...") if getattr(res, "text", "") else ""
snippet = summary or highlights or text_snippet or "无摘要" snippet = summary or text_snippet or "无摘要"
successful_results.append( successful_results.append(
{ {