source-academy / frontend / 31349079291

45%

push

github

Fix Stepper tab staying stuck on "Loading the stepper..." on SICP pages (#4278)

* Fix Stepper tab staying stuck on "Loading the stepper..." on SICP pages

sideContentManager is a single app-wide singleton keyed by one "current"
workspace location (see its own getTabs()/setWorkspaceLocation()). The only
place that ever called setWorkspaceLocation() was Run's own
getPreparedConductorSaga - so on a SICP page (workspace location 'sicp'), a
conductor tab (e.g. the Stepper's) registered by a mere preload - selecting
the Stepper tab before ever pressing Run - forwarded into sideContentManager
correctly, but stayed invisible: getTabs('sicp') still saw whichever location
the singleton was last left on, typically the default 'playground'. The tab
only appeared once a Run finally called setWorkspaceLocation('sicp').

This is distinct from the activation-ordering race fixed in #4277: that fix
was necessary but not sufficient, since even a conductor tab that activates
and registers perfectly can still be invisible if the shared manager's
location doesn't match yet.

Fix: sync sideContentManager's workspace location eagerly, as soon as a
Playground instance knows its own location, rather than waiting for a Run.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RsEaJ8SQRH4QkXir4JSs4S

* Make the new SICP regression test's singleton fixture deterministic

Address CodeRabbit review on #4278: reset sideContentManager to the
'playground' workspace location before rendering (so the test can't pass
for the wrong reason if a prior test left it at 'sicp'), and move cleanup
into a finally block so it still runs if the assertion fails.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RsEaJ8SQRH4QkXir4JSs4S

---------

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>

3117 of 7854 branches covered (39.69%)

Branch coverage included in aggregate %.

2 of 2 new or added lines in 1 file covered. (100.0%)

7073 of 14598 relevant lines covered (48.45%)

30.8 hits per line