datawhalechina/hello-agents · error · HTTPException

执行失败: {str(e)}

Error message

执行失败: {str(e)}

What it means

Catch-all HTTPException 500 '执行失败: {str(e)}' at api/routes/workflow.py:288 for the search-and-analyze route. Raised when the inline call to analyze_paper (imported from api.routes.analysis and invoked directly as a coroutine, bypassing HTTP) throws after search succeeded — e.g. the paper URL is unreachable, the PDF fails to download/parse, or the analysis service errors.

Source

Thrown at Co-creation-projects/Apricity-InnocoreAI/api/routes/workflow.py:288

            paper_url=first_paper["url"],
            analysis_type=request.analysis_type
        ))
        
        results["steps"].append({
            "step": 2,
            "name": "分析论文",
            "status": "completed",
            "analysis": analysis_result
        })
        
        results["status"] = "completed"
        return results
        
    except HTTPException:
        raise
    except Exception as e:
        logger.error(f"搜索和分析失败: {str(e)}")
        raise HTTPException(status_code=500, detail=f"执行失败: {str(e)}")

@router.get("/status/{workflow_id}")
async def get_workflow_status(workflow_id: str):
    """获取工作流状态"""
    try:
        # 这里可以实现工作流状态跟踪
        # 暂时返回模拟状态
        return {
            "workflow_id": workflow_id,
            "status": "completed",
            "progress": 100,
            "message": "工作流已完成"
        }
    except Exception as e:
        logger.error(f"获取工作流状态失败: {str(e)}")
        raise HTTPException(status_code=500, detail=str(e))

View on GitHub (pinned to 606a07d341)

Solutions

  1. Check the '搜索和分析失败' log for the analyze_paper traceback.
  2. Call POST /analysis directly with the same paper URL to reproduce the analysis failure in isolation.
  3. Guard the URL: skip papers whose URL fails a quick HEAD/availability check and try the next paper.
  4. Loosen coupling by calling the underlying service rather than importing another route handler.

Example fix

// before
first_paper = papers[0]
analysis_result = await analyze_paper(PaperAnalysisRequest(paper_url=first_paper["url"], analysis_type=request.analysis_type))
// after
for candidate in papers[:3]:
    try:
        analysis_result = await analyze_paper(PaperAnalysisRequest(paper_url=candidate["url"], analysis_type=request.analysis_type))
        break
    except Exception:
        continue
else:
    raise HTTPException(status_code=502, detail="候选论文均无法分析")
Defensive patterns

Strategy: try-catch

Validate before calling

from urllib.parse import urlparse
def is_probably_fetchable(url: str) -> bool:
    return bool(urlparse(url).scheme in ('http','https') and urlparse(url).netloc)

Type guard

def is_probably_fetchable(url: str) -> bool:
    u = urlparse(url or "")
    return u.scheme in ("http", "https") and bool(u.netloc)

Try / catch

try:
    resp = await client.post("/workflow/search-and-analyze", json=payload, timeout=300)
except httpx.HTTPError:
    raise
if resp.status_code >= 500:
    # analysis of first paper failed; retry after dropping it or pick another
    payload['skip_urls'] = [first_url]; retry()

Prevention

When it happens

Trigger: POST /workflow/search-and-analyze where first_paper['url'] is dead, paywalled, or not a PDF; analyze_paper raising because the analysis backend rejects the type; PaperAnalysisRequest validation issues on the URL field.

Common situations: Search returns a link-abstract record whose URL requires subscription; PDF is scanned/non-text so parsing fails; direct function-call coupling means changes to analyze_paper's signature break this caller at runtime rather than being caught by route registration.

Related errors

AI-assisted analysis of datawhalechina/hello-agents@606a07d341 (2026-08-14). Data as JSON: /api/errors/e32243fafaab4e95. Report an issue: GitHub.