Fix: invalid COMPONENT_EXEC_TIMEOUT (#10278)

### What problem does this PR solve? Fix invalid COMPONENT_EXEC_TIMEOUT. #10273 ### Type of change - [x] Bug Fix (non-breaking change which fixes an issue)
2025-12-08 20:42:30 +08:00 · 2025-09-25 14:11:09 +08:00
parent 4aa1abd8e5
commit daea357940
18 changed files with 56 additions and 61 deletions
--- a/agent/component/categorize.py
+++ b/agent/component/categorize.py
@ -80,7 +80,7 @@ Here's description of each category:
 - Prioritize the most specific applicable category
 - Return only the category name without explanations
 - Use "Other" only when no other category fits
- 
+
 """.format(
            "\n - ".join(list(self.category_description.keys())),
            "\n".join(descriptions)
@ -96,7 +96,7 @@ Here's description of each category:
 class Categorize(LLM, ABC):
    component_name = "Categorize"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
    def _invoke(self, **kwargs):
        msg = self._canvas.get_history(self._param.message_history_window_size)
        if not msg:
@ -112,7 +112,7 @@ class Categorize(LLM, ABC):

        user_prompt = """
 ---- Real Data ----
-{} → 
+{} →
 """.format(" | ".join(["{}: \"{}\"".format(c["role"].upper(), re.sub(r"\n", "", c["content"], flags=re.DOTALL)) for c in msg]))
        ans = chat_mdl.chat(self._param.sys_prompt, [{"role": "user", "content": user_prompt}], self._param.gen_conf())
        logging.info(f"input: {user_prompt}, answer: {str(ans)}")
@ -134,4 +134,4 @@ class Categorize(LLM, ABC):
        self.set_output("_next", cpn_ids)

    def thoughts(self) -> str:
-        return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))
+        return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))
--- a/agent/component/string_transform.py
+++ b/agent/component/string_transform.py
@ -56,7 +56,7 @@ class StringTransform(Message, ABC):
            "type": "line"
        } for k, o in self.get_input_elements_from_text(self._param.script).items()}

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
    def _invoke(self, **kwargs):
        if self._param.method == "split":
            self._split(kwargs.get("line"))
--- a/agent/tools/arxiv.py
+++ b/agent/tools/arxiv.py
@ -61,7 +61,7 @@ class ArXivParam(ToolParamBase):
 class ArXiv(ToolBase, ABC):
    component_name = "ArXiv"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -97,6 +97,6 @@ class ArXiv(ToolBase, ABC):

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-                """.format(self.get_input().get("query", "-_-!"))
+                """.format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/code_exec.py
+++ b/agent/tools/code_exec.py
@ -129,7 +129,7 @@ module.exports = { main };
 class CodeExec(ToolBase, ABC):
    component_name = "CodeExec"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
    def _invoke(self, **kwargs):
        lang = kwargs.get("lang", self._param.lang)
        script = kwargs.get("script", self._param.script)
--- a/agent/tools/duckduckgo.py
+++ b/agent/tools/duckduckgo.py
@ -73,7 +73,7 @@ class DuckDuckGoParam(ToolParamBase):
 class DuckDuckGo(ToolBase, ABC):
    component_name = "DuckDuckGo"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -115,6 +115,6 @@ class DuckDuckGo(ToolBase, ABC):

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-                """.format(self.get_input().get("query", "-_-!"))
+                """.format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/email.py
+++ b/agent/tools/email.py
@ -98,8 +98,8 @@ class EmailParam(ToolParamBase):

 class Email(ToolBase, ABC):
    component_name = "Email"
-    
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
    def _invoke(self, **kwargs):
        if not kwargs.get("to_email"):
            self.set_output("success", False)
@ -212,4 +212,4 @@ class Email(ToolBase, ABC):
 To: {}
 Subject: {}
 Your email is on its way—sit tight!
-""".format(inputs.get("to_email", "-_-!"), inputs.get("subject", "-_-!"))
+""".format(inputs.get("to_email", "-_-!"), inputs.get("subject", "-_-!"))
--- a/agent/tools/exesql.py
+++ b/agent/tools/exesql.py
@ -78,7 +78,7 @@ class ExeSQLParam(ToolParamBase):
 class ExeSQL(ToolBase, ABC):
    component_name = "ExeSQL"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
    def _invoke(self, **kwargs):

        def convert_decimals(obj):
--- a/agent/tools/github.py
+++ b/agent/tools/github.py
@ -57,7 +57,7 @@ class GitHubParam(ToolParamBase):
 class GitHub(ToolBase, ABC):
    component_name = "GitHub"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -88,4 +88,4 @@ class GitHub(ToolBase, ABC):
        assert False, self.output()

    def thoughts(self) -> str:
-        return "Scanning GitHub repos related to `{}`.".format(self.get_input().get("query", "-_-!"))
+        return "Scanning GitHub repos related to `{}`.".format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/google.py
+++ b/agent/tools/google.py
@ -116,7 +116,7 @@ class GoogleParam(ToolParamBase):
 class Google(ToolBase, ABC):
    component_name = "Google"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("q"):
            self.set_output("formalized_content", "")
@ -154,6 +154,6 @@ class Google(ToolBase, ABC):

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/googlescholar.py
+++ b/agent/tools/googlescholar.py
@ -63,7 +63,7 @@ class GoogleScholarParam(ToolParamBase):
 class GoogleScholar(ToolBase, ABC):
    component_name = "GoogleScholar"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -93,4 +93,4 @@ class GoogleScholar(ToolBase, ABC):
        assert False, self.output()

    def thoughts(self) -> str:
-        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
+        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/pubmed.py
+++ b/agent/tools/pubmed.py
@ -33,7 +33,7 @@ class PubMedParam(ToolParamBase):
        self.meta:ToolMeta = {
            "name": "pubmed_search",
            "description": """
-PubMed is an openly accessible, free database which includes primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics. 
+PubMed is an openly accessible, free database which includes primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics.
 In addition to MEDLINE, PubMed provides access to:
 - older references from the print version of Index Medicus, back to 1951 and earlier
 - references to some journals before they were indexed in Index Medicus and MEDLINE, for instance Science, BMJ, and Annals of Surgery
@ -69,7 +69,7 @@ In addition to MEDLINE, PubMed provides access to:
 class PubMed(ToolBase, ABC):
    component_name = "PubMed"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -105,4 +105,4 @@ class PubMed(ToolBase, ABC):
        assert False, self.output()

    def thoughts(self) -> str:
-        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
+        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/retrieval.py
+++ b/agent/tools/retrieval.py
@ -74,7 +74,7 @@ class RetrievalParam(ToolParamBase):
 class Retrieval(ToolBase, ABC):
    component_name = "Retrieval"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", self._param.empty_response)
@ -164,18 +164,18 @@ class Retrieval(ToolBase, ABC):

        # Format the chunks for JSON output (similar to how other tools do it)
        json_output = kbinfos["chunks"].copy()
-        
+
        self._canvas.add_reference(kbinfos["chunks"], kbinfos["doc_aggs"])
        form_cnt = "\n".join(kb_prompt(kbinfos, 200000, True))
-        
+
        # Set both formalized content and JSON output
        self.set_output("formalized_content", form_cnt)
        self.set_output("json", json_output)
-        
+
        return form_cnt

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/searxng.py
+++ b/agent/tools/searxng.py
@ -77,7 +77,7 @@ class SearXNGParam(ToolParamBase):
 class SearXNG(ToolBase, ABC):
    component_name = "SearXNG"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        # Gracefully handle try-run without inputs
        query = kwargs.get("query")
@ -94,7 +94,6 @@ class SearXNG(ToolBase, ABC):
        last_e = ""
        for _ in range(self._param.max_retries+1):
            try:
-                # 构建搜索参数
                search_params = {
                    'q': query,
                    'format': 'json',
@ -104,33 +103,29 @@ class SearXNG(ToolBase, ABC):
                    'pageno': 1
                }

-                # 发送搜索请求
                response = requests.get(
                    f"{searxng_url}/search",
                    params=search_params,
                    timeout=10
                )
                response.raise_for_status()
-                
+
                data = response.json()
-                
-                # 验证响应数据
+
                if not data or not isinstance(data, dict):
                    raise ValueError("Invalid response from SearXNG")
-                
+
                results = data.get("results", [])
                if not isinstance(results, list):
                    raise ValueError("Invalid results format from SearXNG")
-                
-                # 限制结果数量
+
                results = results[:self._param.top_n]
-                
-                # 处理搜索结果
+
                self._retrieve_chunks(results,
                                      get_title=lambda r: r.get("title", ""),
                                      get_url=lambda r: r.get("url", ""),
                                      get_content=lambda r: r.get("content", ""))
-                
+
                self.set_output("json", results)
                return self.output("formalized_content")

@ -151,6 +146,6 @@ class SearXNG(ToolBase, ABC):

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Searching with SearXNG for relevant results...
                """.format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/tavily.py
+++ b/agent/tools/tavily.py
@ -31,7 +31,7 @@ class TavilySearchParam(ToolParamBase):
        self.meta:ToolMeta = {
            "name": "tavily_search",
            "description": """
-Tavily is a search engine optimized for LLMs, aimed at efficient, quick and persistent search results. 
+Tavily is a search engine optimized for LLMs, aimed at efficient, quick and persistent search results.
 When searching:
   - Start with specific query which should focus on just a single aspect.
   - Number of keywords in query should be less than 5.
@ -101,7 +101,7 @@ When searching:
 class TavilySearch(ToolBase, ABC):
    component_name = "TavilySearch"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -136,7 +136,7 @@ class TavilySearch(ToolBase, ABC):

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
                """.format(self.get_input().get("query", "-_-!"))

@ -199,7 +199,7 @@ class TavilyExtractParam(ToolParamBase):
 class TavilyExtract(ToolBase, ABC):
    component_name = "TavilyExtract"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
    def _invoke(self, **kwargs):
        self.tavily_client = TavilyClient(api_key=self._param.api_key)
        last_e = None
@ -224,4 +224,4 @@ class TavilyExtract(ToolBase, ABC):
        assert False, self.output()

    def thoughts(self) -> str:
-        return "Opened {}—pulling out the main text…".format(self.get_input().get("urls", "-_-!"))
+        return "Opened {}—pulling out the main text…".format(self.get_input().get("urls", "-_-!"))
--- a/agent/tools/wencai.py
+++ b/agent/tools/wencai.py
@ -68,7 +68,7 @@ fund selection platform: through AI technology, is committed to providing excell
 class WenCai(ToolBase, ABC):
    component_name = "WenCai"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("report", "")
@ -111,4 +111,4 @@ class WenCai(ToolBase, ABC):
        assert False, self.output()

    def thoughts(self) -> str:
-        return "Pulling live financial data for `{}`.".format(self.get_input().get("query", "-_-!"))
+        return "Pulling live financial data for `{}`.".format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/wikipedia.py
+++ b/agent/tools/wikipedia.py
@ -64,7 +64,7 @@ class WikipediaParam(ToolParamBase):
 class Wikipedia(ToolBase, ABC):
    component_name = "Wikipedia"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
    def _invoke(self, **kwargs):
        if not kwargs.get("query"):
            self.set_output("formalized_content", "")
@ -99,6 +99,6 @@ class Wikipedia(ToolBase, ABC):

    def thoughts(self) -> str:
        return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))
--- a/agent/tools/yahoofinance.py
+++ b/agent/tools/yahoofinance.py
@ -72,7 +72,7 @@ class YahooFinanceParam(ToolParamBase):
 class YahooFinance(ToolBase, ABC):
    component_name = "YahooFinance"

-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
    def _invoke(self, **kwargs):
        if not kwargs.get("stock_code"):
            self.set_output("report", "")
@ -111,4 +111,4 @@ class YahooFinance(ToolBase, ABC):
        assert False, self.output()

    def thoughts(self) -> str:
-        return "Pulling live financial data for `{}`.".format(self.get_input().get("stock_code", "-_-!"))
+        return "Pulling live financial data for `{}`.".format(self.get_input().get("stock_code", "-_-!"))