Support contains/matches for custom properties

[macobo]

Changes

This PR a filtering with contains/matches for custom properties, e.g. "Filter where (custom property) url contains foo"

First PR 🙌

Tests

• Automated tests have been added
• This PR does not require tests

Changelog

• Entry has been added to changelog
• This PR does not make a user-facing change

Documentation

• Docs have been updated
• This change does not need a documentation update

Dark mode

• The UI has been tested both in dark and light mode
• This PR does not change the UI

[aerosol]

There's something wrong with re-applying the filter when modifying the custom props breakdown query:

record-2024-01-15-16-45-30-acre.putt.blot.mp4

[aerosol]

On a freshly seeded db (you can repro via mix ecto.reset), Total Visits doubles when a non-existing property value is supplied to the query:

record-2024-01-15-16-49-48-hull.roth.john.mp4

I thought it was due to seed data being nonsense, but modifying the test you've added, also returns extra rows that shouldn't be included?

diff --git a/test/plausible_web/controllers/api/stats_controller/pages_test.exs b/test/plausible_web/controllers/api/stats_controller/pages_test.exs
index 339a4bab3..ba9aa0fed 100644
--- a/test/plausible_web/controllers/api/stats_controller/pages_test.exs
+++ b/test/plausible_web/controllers/api/stats_controller/pages_test.exs
@@ -103,7 +103,7 @@ defmodule PlausibleWeb.Api.StatsController.PagesTest do
         build(:pageview, pathname: "/5")
       ])
 
-      filters = Jason.encode!(%{props: %{"prop" => "~bar"}})
+      filters = Jason.encode!(%{props: %{"prop" => "~bar|pppp"}})
       conn = get(conn, "/api/stats/#{site.domain}/pages?period=day&filters=#{filters}")
 
       assert json_response(conn, 200) == [

[macobo]

@aerosol thank you for the great feedback. I've updated the branch.

(1) Regarding breakdowns, this was a legit bug in the code, see fix in 43772bc

(2) seems to be about how regex is internally escaped. I hoped that merging #3634 would fix this, but it did not. Given that this behavior already exists on master, I'll fix this in a separate PR, filed it in my worklog.