Closed tschuettler closed 3 years ago
@tschuettler Not sure too... Have you checked what happens if you send something that is not a gzip file using this mime type? Maybe a better approach would be to not rely on the mime type and try to guess the file type?
@tonioo Thanks for your feedback. I tried to guess the file type with the help of https://pypi.org/project/python-magic/. I also added a test for a successfull report and an ignored report due to non-matching file type. Arguably the to be igored report may deserve its own handling. But it should not count as failed report.
If this PR is going to be merged the documentation needs to make sure that the required c library is installed: sudo apt-get install libmagic1
@tschuettler Could you update the README to include the library installation please?
Merging #53 (1b6afd3) into master (e1b272b) will increase coverage by
0.12%
. The diff coverage is100.00%
.
@@ Coverage Diff @@
## master #53 +/- ##
==========================================
+ Coverage 87.45% 87.57% +0.12%
==========================================
Files 19 19
Lines 502 507 +5
==========================================
+ Hits 439 444 +5
Misses 63 63
Impacted Files | Coverage Δ | |
---|---|---|
modoboa_dmarc/lib.py | 82.08% <100.00%> (+0.69%) |
:arrow_up: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update e1b272b...1b6afd3. Read the comment docs.
@tonioo Thanks for your feedback!
I had to modify the date of the report to be successfully processed, so that it wont conflict with merged test.
I do not have access to CentOS, so I could not verify that this package actually provides the required library, but searching for it brought this up. I have no idea about GHA, but assumed it would be reasonable to add the package to those stages.
Feel free to squash those commits or request for that.
@tonioo Hi there, can we do something to get this merged? Is there a timeframe for a new release containing this and #50? We would love to see this in production :)
@tschuettler Sorry, I've been pretty busy. I'll merge it now and try to release a new version this week.
Since the RFC only mentions GZIP as compression type, i have assumed that MIME parts with the content type
application/octet-stream
will always be an gzip archive. I'm not sure if that is a correct assumption.Fixes #52