[build-tools] Separate iOS ccache by destination

[sjchmiela]

Why

Looks like ccache cache built for iOS simulator is not helpful for iOS device builds and vice versa.

Context: https://exponent-internal.slack.com/archives/C08PBDW7ZC3/p1787242902697509?thread_ts=1787171142.251299&cid=C08PBDW7ZC3

How

Separated ios-device-ccache- and ios-simulator-ccache- ccaches.

Test Plan

• Before: the simulator build restored the device build’s cache, got 0 hits, and could not save its own cache because the key already existed.
• After: the device and simulator builds used separate keys. The simulator build got 48 direct hits.

[codecov]

Codecov Report

❌ Patch coverage is 30.43478% with 16 lines in your changes missing coverage. Please review.
✅ Project coverage is 63.72%. Comparing base (ce93a92) to head (aeed5f8).
⚠️ Report is 3 commits behind head on main.

Files with missing lines	Patch %	Lines
...ild-tools/src/steps/functions/restoreBuildCache.ts	0.00%	7 Missing ⚠️
.../build-tools/src/steps/functions/saveBuildCache.ts	12.50%	7 Missing ⚠️
packages/build-tools/src/utils/cacheKey.ts	75.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #4238      +/-   ##
==========================================
- Coverage   63.73%   63.72%   -0.01%     
==========================================
  Files        1030     1030              
  Lines       47448    47464      +16     
  Branches     9977     9989      +12     
==========================================
+ Hits        30237    30241       +4     
- Misses      17110    17122      +12     
  Partials      101      101              

☔ 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.

[github-actions]

⏩ The changelog entry check has been skipped since the "no changelog" label is present.