Updates on the issue log

[Will-Cooper]

Tackling a few issues on the tracker.
Updating bokeh (and corresponding static js)
Making requirements.txt a specific pin to exactly match what is on reclaim host.
Fixes #181 (commit typo, instead of #186 which doesn't exist yet); uses session instead of curdoc, so will work on multithreaded servers and doesn't have a weird inconsistent overlap with bokeh.
Solved the join.. on... sql queries which were making lists with duplicate column names; breaking the url generation.
#170 is indeed reproduced in the simplest case of doing a full text search providing a spectra table with a bulk spectra download option added (as with the solo result page)
Fixed #177 with the CAMD resetting behaviour to be more expected.

[kelle]

I launched a codespace to test this out and unfortunately got 404 errors on solo object results pages.

[Will-Cooper]

I launched a codespace to test this out and unfortunately got 404 errors on solo object results pages.

Just needed merging, this is because the SIMPLE-db upstream change had broken the website which was fixed by #188, all of which happened after this PR.

[kelle]

Testing out the bulk spectra download function (#170) I see that the filenames are changed. E.g.,
spex-prism_2MASSJ10073369-4555147_20061209_LOO07A.txt
gets downloaded as
spectra_292_spex-prism_2MASSJ16150413%2B1340079_20060902_LOO07A.txt
Can you say why this spectra_##_ prefix is necessary?

[kelle]

• not sure how to test/confirm fix for Write Spectra Fault #181
• I see the change in the CMD reset! Great!

[Will-Cooper]

Testing out the bulk spectra download function (#170) I see that the filenames are changed. E.g., spex-prism_2MASSJ10073369-4555147_20061209_LOO07A.txt gets downloaded as spectra_292_spex-prism_2MASSJ16150413%2B1340079_20060902_LOO07A.txt Can you say why this spectra_##_ prefix is necessary?

This is mostly to guarantee zero duplicates (e.g. same file name), which would break the code as it enters a python dictionary before being zipped, which as unique keys.

[Will-Cooper]

• not sure how to test/confirm fix for Write Spectra Fault #181

it's an awkward one: prove that an intermittent fault no longer occurs; I tested triggering the write spectra function a tonne of times and it had stopped randomly breaking.

[kelle]

I also see some files being saved as spectra_284_viewcontent. The one I checked is actually a fits file.

[kelle]

Opened a new issue about filenames (#193). I think this PR can be merged.