Avoid fail when file does not exist

[baturin]

Let's start sorting out issues from #141. That simple PR avoids montage from failure when importing renamed/removed/incorrect image files.

[codecov-commenter]

Codecov Report

Merging #175 into master will decrease coverage by 0.15%.
The diff coverage is 35.00%.

@@            Coverage Diff             @@
##           master     #175      +/-   ##
==========================================
- Coverage   76.52%   76.37%   -0.16%     
==========================================
  Files          19       19              
  Lines        3766     3775       +9     
  Branches      471      474       +3     
==========================================
+ Hits         2882     2883       +1     
- Misses        643      650       +7     
- Partials      241      242       +1     

Impacted Files	Coverage Δ
montage/labs.py	30.30% <0.00%> (ø)
montage/admin_endpoints.py	81.99% <20.00%> (-0.79%)	⬇️
montage/loaders.py	73.17% <36.36%> (-1.68%)	⬇️
montage/rdb.py	77.90% <100.00%> (ø)

---

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 a921c5d...9d1bb64. Read the comment docs.

[mahmoud]

Looks pretty good! One comment on the error message, and also one question: do you want to update the tests at all to hit these new branches? Now that we have Sentry, and this is already an exception path, I don't think it's a blocker, but figured I'd ask.

[mahmoud]

This is a good, actionable error message. However, we should keep an eye out for if there are many, many missing files. Assuming there are 0-10, I think the error display method defined in the JS above should be fine.

[baturin]

@mahmoud Yes, I think we should add some tests for situations like that (there are also several situations around that should be handled in a similar way). BTW, have you checked my previous test improvements #159?
As for displaying error message, that's also pretty good question, and I think yes, we should rework error display method to make it more structured and handle many errors (as there could be many files with issues like that). I'm going to think of a way once we handle all issues on backend.

[mahmoud]

Cool. So, what I'm understanding is that we can roll with this for now and will improve error display in a future PR?

As for #159, somehow I missed that! Left a comment. :)

[mahmoud]

@baturin just checking in on this, are we OK to merge + address errors in the next?

[baturin]

@mahmoud Yeah, let's continue improving error reporting in future PRs, and merge that one.