[Bug] It seems that fail to fetch LTE data from CelesTrak

[Hantingwithoutheart]

Problem Description

When running the TLE constellation tests, the program may crash during constellation generation because today’s JSON TLE data is not available, but the code still tries to write/read it as if it exists.
(1) To Reproduce

Run the code following the instruction.

1. install uv
2. create the venv
3. remove the # before main in StarPerf.py
4. commenting out the code related to XML_testing(optional)
5. uv run StarPerf.py

(3) Actual Behavior

Starting StarPerf...
Starting TLE Constellations Testing
Test(01/16) : constellation generation
Traceback (most recent call last):
File "c:\Users\davis\Desktop\code\python\starperf\StarPerf_Simulator\StarPerf.py", line 90, in

main()
File "c:\Users\davis\Desktop\code\python\starperf\StarPerf_Simulator\StarPerf.py", line 26, in
ation_generation_test
constellation = constellation_configuration.constellatiTLE_constellation_test_cases() on_configuration(dT, constellation_name) f_Simulator\samples\TLE_constellation
File "c:\Users\davis\Desktop\code\python\starperf\StarPertion_test_casesf_Simulator\src\constellation_generation\by_TLE\constellatitest()on_configuration.py", line 24, in constellation_configuratif_Simulator\samples\TLE_constellation\on line 16, in constellation_generation_
DOWNLOAD_TLE_DATA.download_TLE_data(constellation_name)
on_configuration(dT, constellation_nam
File "c:\Users\davis\Desktop\code\python\starperf\StarPerf_Simulator\src\constellation_generation\by_TLE\download_TLf_Simulator\src\constellation_generatiE_data.py", line 92, in download_TLE_data tellation_configuration
file[formatted_date].create_dataset(formatted_date + "-json", data=json_TLE) f_Simulator\src\constellation_generati
UnboundLocalError: local variable 'json_TLE' referenced befataore assignment

Alternatives

copilot tells:
问题链路在 download_TLE_data.py 很清楚：

1. 代码先请求 FORMAT=2le
2. 紧接着又请求同一组的 FORMAT=json
3. json_TLE 只在 response2.status_code == 200 时才赋值，见 download_TLE_data.py
4. 但后面无论 JSON 请求是否成功，都会写 formatted_date + "-json"，见 download_TLE_data.py

所以一旦出现这种情况：
2LE 请求成功了，但紧接着 JSON 请求因为“刚下载过、2 小时内不更新”被 403 拒绝，

就会发生你看到的异常：
json_TLE 没赋值，却被拿去写 HDF5，于是报 UnboundLocalError。

更关键的是，本地缓存已经把这个结论坐实了。你现在的 tle.h5 里，今天的组 20260416 只有 20260416-2LE，没有 20260416-json。这正说明当时流程是：

1. 2LE 已经下载成功并写入
2. JSON 没写进去
3. 程序在写 -json 数据集时崩了

而后续 satellite_to_shell_mapping.py 又强依赖当天的 -json 数据集，所以这个“半成品”状态还会影响你当天后面的重跑。

结论是：
根本原因不是 StarPerf.py，也不是测试入口本身，而是 download_TLE_data.py 的下载策略和异常处理有问题。它假设同一时刻连续下载 2le 和 json 一定都能成功，但 CelesTrak 实际上会因为 2 小时更新限制返回 403。

如果你要，我下一步可以直接帮你修这个根因，通常会做这几件事：

1. 给 JSON 请求也加明确的失败处理，不再让 json_TLE 未定义
2. 避免写出只有 -2LE 没有 -json 的半成品组
3. 如果当天本地缓存已经完整存在，就直接复用，不重复请求
4. 在远端返回 403 且提示“2 小时未更新”时，优先读取本地已有数据而不是崩溃

[snakefood3232]

The crash occurs because json_TLE is referenced before assignment when today's JSON TLE data is unavailable. I'd handle this by adding a conditional check for the API response before attempting to write the dataset. Can have a PR up in a few hours. I've built several scrapers and data pipelines handling similar missing data edge cases.