Report the whole fe80::/10 range from Address6.isLinkLocal()

[spokodev]

Address6.isLinkLocal() under-reports the IANA link-local range fe80::/10. It returns true only when the interface-prefix bits (10–63) are all zero, i.e. fe80::/64, so every other address in the range reads as non-link-local:

new Address6('fe80::1').isLinkLocal();     // true
new Address6('fe80:1::1').isLinkLocal();   // false  <-- fe80::/10, but reported false
new Address6('fe90::1').isLinkLocal();     // false
new Address6('fea0::1').isLinkLocal();     // false
new Address6('febf::1').isLinkLocal();     // false

This contradicts the library's own getType(), which already classifies all of those as Link-local unicast:

new Address6('fe80:1::1').getType();       // 'Link-local unicast'
new Address6('fe80:1::1').isLinkLocal();   // false — disagrees

getType() reaches that verdict through isHostInSubnet('fe80::/10'), so the inline comment on isLinkLocal() — "we can't check isHostInSubnet with 'fe80::/10'" — is not accurate.

The fix

Replace the hard-coded 64-bit prefix compare with isHostInSubnet(LINK_LOCAL_UNICAST_SUBNET), mirroring getType() and the sibling predicates (isULA() uses isHostInSubnet(ULA_SUBNET)). The fe80::/10 subnet is added to the existing block of module-level subnet constants. The IPv4-mapped delegation at the top of the method is preserved.

The change is a strict superset of the previous behaviour: every input that returned true still returns true, fec0::1 (site-local) and global addresses stay false.

History

getScope()/getType() were broadened to the full fe80::/10 in #200 (closing #122), and the boolean predicates were added in #197, but isLinkLocal() kept its narrower fe80::/64 check, so the two have disagreed since. This aligns them.

Tests

Added a case in the existing isLinkLocal block asserting the whole fe80::/10 range (including a zone-id form) reports true and matches getType(), plus a negative fec0::1. Fails before the change, passes after; full suite 3595 passing.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (5e3ceb7) to head (b0aedb1).
⚠️ Report is 12 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main      #223   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            6         6           
  Lines          370       373    +3     
  Branches       216       218    +2     
=========================================
+ Hits           370       373    +3     

☔ 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.

[beaugunderson]

Thanks, and sorry for the overlap: the same fix shipped in 10.5.1 through GHSA-rpw4-54j3-4h4q, which was filed a few days after you opened this. Closing as superseded.