PubTator API查询处理失败，求助排查500错误原因

开发者阿泽

2026-7-6

问题

我正尝试按照PubTator官方API文档提供的示例代码提交查询，但运行后最终得到500状态码，提示信息为"We have trouble processing your query"。以下是我运行的代码：

text = ('Humboldtia vahliana Wight. is an endangered species belonging to the '
'family Leguminosae (subfamily Caesalpinioideae). Though the plant propagates '
'through seeds, low seed setting, seed infestation, poor viability of seed, '
'reduction in natural regeneration of seedlings as well as anthropogenic '
'activities like overexploitation, habitat destruction and fragmented '
'distribution in localities are the major factors hindering the survival of the '
'species. Hence, strategies must be devised urgently for the conservation of this '
'species as these recalcitrant seeds do not contribute significantly to the seed '
'bank. The present attempt was to understand the seed physiology and biochemistry '
'during embryogenesis and embryo desiccation. H. vahliana seeds took 120 days '
'after anthesis to acquire full maturity. Immature seeds had higher moisture '
'content (87.40%) which gradually reduced during maturity and reached 55.42% '
'at the time of seed shed a true recalcitrant behavior of the seeds. Freshly '
'fallen mature seeds showed an optimal germination percentage of 82.32% which '
'was severely affected by the decrease in seed moisture content and the critical '
'moisture content was found to be 33.63% in which the percentage of germination '
'was only 30%. Cell membrane damage of seed was found to cause quick loss of '
'seed viability. The LC-MS/MS analysis showed insignificant amounts of ribose, '
'arabinose and trehalose but a significant accumulation of fructose in the mature '
'embryos rather than glucose and sucrose. Embryo drying significantly reduced the '
'level of these sugars including the stress related trehalose indicating the lack '
'of biosynthetic machinery to counter desiccation stress in these recalcitrant '
'seeds.')

import requests

r = requests.post(
    'https://www.ncbi.nlm.nih.gov/research/pubtator-api/annotations/annotate/submit/All',
    data=text.encode('utf-8'))
session_num = r.text

status_code = 404
while status_code == 404:
    result = requests.get(
    'https://www.ncbi.nlm.nih.gov/research/pubtator-api/annotations/annotate/retrieve/' + session_num)
    status_code = result.status_code

print(status_code, result.text)

注：API文档说明服务器在处理完成前会返回404状态码，因此编写了while循环持续检查状态码，且确认严格遵循了示例代码的请求格式，未找到该错误的相关说明。

分析与解决建议

请求格式不规范：直接提交文本内容时，需显式指定Content-Type为text/plain，否则服务器可能无法正确解析请求体。修改POST请求如下：

headers = {'Content-Type': 'text/plain'}
r = requests.post(
    'https://www.ncbi.nlm.nih.gov/research/pubtator-api/annotations/annotate/submit/All',
    data=text,
    headers=headers)

会话ID有效性验证：未检查POST请求是否成功就直接提取会话ID，若提交请求本身失败（如返回非200状态码），会导致后续用错误内容作为会话ID查询，引发500错误。添加验证步骤：
```
if r.status_code != 200:
    print(f"提交请求失败: {r.status_code} {r.text}")
    exit()
session_num = r.text.strip()
```
文本内容或长度问题：提交的文本过长，或包含特殊字符（如百分比符号、括号）可能触发服务器解析异常。先提交一段简短的测试文本（如"Humboldtia vahliana is a plant."），确认接口正常响应后再逐步扩展内容。
服务器临时故障：500错误多为服务器内部问题，可能是负载过高或临时维护。间隔1-2分钟后重试请求，或等待一段时间再测试。

内容的提问来源于stack exchange，提问作者SLotreck