Closed lizadaly closed 1 year ago
Got some test work to do here, re-drafting.
Merging #1996 (7b36d67) into develop (e62aa6b) will increase coverage by
0.00%
. The diff coverage is81.57%
.
@@ Coverage Diff @@
## develop #1996 +/- ##
========================================
Coverage 76.92% 76.93%
========================================
Files 60 62 +2
Lines 6839 6871 +32
========================================
+ Hits 5261 5286 +25
- Misses 1578 1585 +7
Impacted Files | Coverage Δ | |
---|---|---|
web/main/urls.py | 100.00% <ø> (ø) |
|
web/reporting/admin/usage_dashboard.py | 100.00% <ø> (ø) |
|
web/reporting/views.py | 64.28% <64.28%> (ø) |
|
web/reporting/matomo.py | 83.33% <89.47%> (+0.61%) |
:arrow_up: |
web/main/test/test_permissions.py | 97.16% <100.00%> (+0.05%) |
:arrow_up: |
web/reporting/urls.py | 100.00% <100.00%> (ø) |
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.
Something I should've done a long time ago for @cath9 but fortunately for her now it was annoying me, so I'm improving this.
In the H2O reporting dashboard, high-level stats from Matomo were being fetched synchronously during the request cycle, which made the dashboard page load slowly. For awhile I was considering having the stats fetched and cached via a recurring celery job, but this is simpler, if more ridiculous.
Now the page loads instantly, which gives you immediate access to the Django-side reports, while the Matomo reporting is fetched async and pops in after a few seconds.
This involved writing some wee little Django scaffolding for the endpoint, and porting the Django template code to JS which modifies the DOM once the response comes back. There are also a few mechanical changes to make the Matomo results pickleable; before they could be Django model instances.
Two small functional changes:
Institution
model rather than the older user-suppliedaffiliation
field; this should help identify where authors might need an institution applied.