binary-husky/gpt_academic · error · Exception

Failed to analyze query: {str(e)}

Error message

Failed to analyze query: {str(e)}

What it means

The outermost wrapper of QueryAnalyzer.analyze: every exception raised inside the method (errors 90-93, parameter parsing, any unexpected error) is caught and re-raised as Exception('Failed to analyze query: {str(e)}'). It preserves the original message but destroys the original exception type and traceback, so diagnosis requires reading the embedded text.

Source

Thrown at crazy_functions/review_fns/query_analyzer.py:398

            return SearchCriteria(
                query_type=query_type,
                main_topic=main_topic,
                sub_topics=sub_topics,
                start_year=start_year,
                end_year=end_year,
                arxiv_params=arxiv_params,
                semantic_params=semantic_params,
                pubmed_params=pubmed_params,
                crossref_params=crossref_params,
                paper_id=paper_id,
                paper_title=paper_title,
                paper_source=paper_source,
                original_query=query,
                adsabs_params=adsabs_params
            )

        except Exception as e:
            raise Exception(f"Failed to analyze query: {str(e)}")

    def _normalize_query_type(self, query_type: str, query: str) -> str:
        """规范化查询类型"""
        if query_type in ["review", "recommend", "qa", "paper"]:
            return query_type

        query_lower = query.lower()
        for type_name, keywords in self.valid_types.items():
            for keyword in keywords:
                if keyword in query_lower:
                    return type_name

        query_type_lower = query_type.lower()
        for type_name, keywords in self.valid_types.items():
            for keyword in keywords:
                if keyword in query_type_lower:
                    return type_name

View on GitHub (pinned to d6bde0fa54)

Solutions

1. Read the suffix after 'Failed to analyze query:' — it names the real cause; map it to errors 90-93 if it matches.
2. Fix the underlying issue (API key/model config, prompt format, retry on transient failure).
3. In your own code, catch this exception and degrade gracefully (skip analysis, use default search parameters) rather than crashing the request pipeline.
4. If you control the code, re-raise with `raise Exception(...) from e` or use logging.exception to keep the traceback.

Example fix

# before
except Exception as e:
    raise Exception(f'Failed to analyze query: {str(e)}')

# after
except Exception as e:
    logger.exception('Query analysis failed')
    raise Exception(f'Failed to analyze query: {str(e)}') from e

try:
    analyzed = analyzer.analyze(query)
except Exception as e:  # 'Failed to analyze query: ...'
    logger.warning('analysis failed (%s); using default search params', e)
    analyzed = AnalyzedQuery(query_type='paper', original_query=query)  # safe defaults

When it happens

Trigger: Any failure during multi-prompt LLM fan-out or downstream parsing: None responses, missing response slots, unparseable <query_type> tag, invalid arXiv/Semantic Scholar parameter blocks, or attribute errors while assembling the AnalyzedQuery result object.

Common situations: End users see this single message for every analyzer failure; typical underlying causes are LLM API misconfiguration (key/model), non-conforming model output, or transient API errors — all surfaced through this one wrapper.

Related errors

• Cannot convert response {i} to string
• 未收到第 {i + 1} 个响应
• 无法提取query_type标签内容
• 分析查询失败: {str(e)}
• Response {i} is None

AI-assisted analysis of binary-husky/gpt_academic@d6bde0fa54 (2026-08-14). Data as JSON: /api/errors/b7c94a2912176aa5. Report an issue: GitHub.