Rerun failed jobs by error message

[janosh]

would be great if jobs could be rerun by regex matching the recorded stacktrace. e.g.

jf job rerun --err-regex | -er "ModuleNotFoundError: No module named 'idna'"

to rerun all jobs with this kind stacktrace

│ created_on = '2024-09-13 19:23'                                                            │
│      db_id = '20229'                                                                       │
│   end_time = '2024-09-16 12:00'                                                            │
│      error = Traceback (most recent call last):   
...
│              "/lambdafs/janosh/.venv/py312/lib/python3.12/site-packages/emmet/core/settin… │
│              line 10, in <module>                                                          │
│                  import requests                                                           │
│                File                                                                        │
│              "/lambdafs/janosh/.venv/py312/lib/python3.12/site-packages/requests/__init__… │
│              line 151, in <module>                                                         │
│                  from . import packages, utils                                             │
│                File                                                                        │
│              "/lambdafs/janosh/.venv/py312/lib/python3.12/site-packages/requests/packages… │
│              line 9, in <module>                                                           │
│                  locals()[package] = __import__(package)                                   │
│                                      ^^^^^^^^^^^^^^^^^^^                                   │
│              ModuleNotFoundError: No module named 'idna'                                   │
│      index = 1                                                                             │
│   metadata = {}