feat(Replay): Add URLs Visited to EAP Replay Query

[cliffordxing]

Currently when executing the EAP replay query, we see the following error:

This is because the frontend expects the urls visited to be passed in the data, but the EAP query previously did not query this data. Thus, this PR adds a new query to post process the URLs visited to the replay data. Specifically, the query groups by the URL column, selecting the URL itself along with the earliest time stamp. We then sort this afterwards to ensure correctness of ordering.

Unit tests are also added to assert correctness of URL retrieval, ordering, and deduping.

[codecov]

Codecov Report

❌ Patch coverage is 83.33333% with 3 lines in your changes missing coverage. Please review.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
...y/replays/endpoints/organization_replay_details.py	83.33%	3 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           master   #104538      +/-   ##
===========================================
+ Coverage   80.51%    80.53%   +0.01%     
===========================================
  Files        9352      9353       +1     
  Lines      400281    400874     +593     
  Branches    25700     25700              
===========================================
+ Hits       322274    322829     +555     
- Misses      77539     77577      +38     
  Partials      468       468              

[phacops]

Can't we sort in the query already?

[cliffordxing]

seems tricky to sort by the "first_seen" alias I created since its a column we are selecting with the min function. When we try to:

ORDER BY first_seen

This transpiles to:

ORDER BY attributes_float_17['first_seen']

Which gives this error:

Column attributes_float_17 is not under aggregate function and not in GROUP BY keys error.

Seemed easier to sort afterwards, what do you think?

[phacops]

I think the solution here is to sort on the alias of the aggregation instead. If you call your first_seen aggregation differently, you can then sort on first_seen_aggregation and not hit this issue I believe.

Otherwise, we'll have to change something on the EAP.