fix(today): base "time till next class" on Singapore time

[TheMythologist]

The Today view computed the next-lesson countdown by placing the lesson start time on the viewer's local calendar day while comparing against the true current instant. Since lesson times are in Singapore time, overseas students saw a countdown that was off by their timezone offset.

Derive "now" from the SGT wall-clock (via toSingaporeTime) when reasoning about today, and anchor the lesson start time to that same reference so the displayed duration reflects the real time until the next class. No-op for viewers already on Singapore time.

Part of #3338

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

2 Skipped Deployments

Project	Deployment	Actions	Updated (UTC)
nusmods-export	Ignored	Preview	Jul 5, 2026 10:38pm
nusmods-website	Ignored	Preview	Jul 5, 2026 10:38pm

[greptile-apps]

Greptile Summary

This PR fixes an incorrect "time till next class" countdown for overseas NUS students by converting the viewer's local currentTime to a Singapore-time wall-clock representation via toSingaporeTime before any date arithmetic. The same conversion is applied to semester/week selection in mapStateToProps and to the getStartTimeAsDate anchor in BeforeLessonCard, making the entire Today view timezone-consistent.

• groupLessons and renderDay now derive currentTime from toSingaporeTime(this.props.currentTime), so day boundaries, lesson filtering, and the countdown all use SGT regardless of the viewer's local clock.
• mapStateToProps also converts to SGT before computing lastDay and getAcadWeekInfo, preventing semester mismatches for overseas viewers.
• One gap remains: isSameDay(date, holiday) at line 246 still compares the new fake-SGT date against a new Date("YYYY-MM-DD") holiday that is parsed as UTC midnight, which shows as the previous calendar day for western-timezone viewers — breaking holiday detection that worked coincidentally before this change.

Confidence Score: 4/5

The core countdown fix is sound, but a one-line holiday comparison still uses a raw UTC-midnight Date against the new SGT-based date, silently breaking holiday detection for viewers in western timezones.

The toSingaporeTime conversion is applied consistently across groupLessons, renderDay, and mapStateToProps, and BeforeLessonCard correctly anchors the lesson time to the same reference. The remaining gap is isSameDay(date, holiday) at line 246: holidays are stored as new Date("YYYY-MM-DD") (UTC midnight), which in western timezones resolves to the day before the intended SGT date — now that date is a fake-SGT Date the coincidental match that existed before is gone, so public holidays won't be detected for overseas users west of UTC.

website/src/views/today/TodayContainer/TodayContainer.tsx — specifically the isSameDay holiday check at line 246.

_{Reviews (5): Last reviewed commit: "fix(today): select semester in Singapore..." | Re-trigger Greptile}

[codecov]

Codecov Report

❌ Patch coverage is 75.00000% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 58.53%. Comparing base (988c6fd) to head (361be69).
⚠️ Report is 259 commits behind head on master.

Files with missing lines	Patch %	Lines
website/src/views/today/BeforeLessonCard.tsx	0.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #4442      +/-   ##
==========================================
+ Coverage   54.52%   58.53%   +4.01%     
==========================================
  Files         274      316      +42     
  Lines        6076     7296    +1220     
  Branches     1455     1794     +339     
==========================================
+ Hits         3313     4271     +958     
- Misses       2763     3025     +262     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[jloh02]

We should probably fix greptile's comments in this samePR while we're at it?

[TheMythologist]

We should probably fix greptile's comments in this samePR while we're at it?

You're right - fixed in 361be69