serializer can't read isd oops report file
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Python OOPS Date-dir repository |
Fix Released
|
High
|
Diogo Matsubara | ||
Bug Description
To reproduce:
In [1]: import bz2
In [2]: import oops_datedir_
In [3]: oops_datedir_
-------
AssertionError Traceback (most recent call last)
/home/matsubara
/home/matsubara
109 r'^(\d+
110 assert match is not None, (
--> 111 "Unable to interpret oops line: %s" % line)
112 start, end, db_id, statement = match.groups()
113 if db_id is not None:
AssertionError: Unable to interpret oops line: REQUEST_METHOD=POST
Oops file can be downloaded here:
https:/
That oops file was logged with a blank line between the request variables, causing the parser to break.
Related branches
- Robert Collins (community): Approve
-
Diff: 108 lines (+60/-20)2 files modifiedoops_datedir_repo/serializer_rfc822.py (+29/-20)
oops_datedir_repo/tests/test_serializer_rfc822.py (+31/-0)
| Robert Collins (lifeless) wrote | #1 |
| Changed in python-oops-datedir-repo: | |
| status: | New → In Progress |
| importance: | Undecided → High |
| assignee: | nobody → Diogo Matsubara (matsubara) |
| Changed in python-oops-datedir-repo: | |
| status: | In Progress → Fix Committed |
| Changed in python-oops-datedir-repo: | |
| status: | Fix Committed → Fix Released |
Ugh; do they still generate them like that, or was it crap data? Would they be impacted if we said 'we won't read those: we can do data migration instead' ?