[gnodet]

Summary

• Fix off-by-one error in KeyMap.unbind() where bounds checks on lines 631 and 641 used > instead of >=, causing ArrayIndexOutOfBoundsException when a character's value equals mapping.length (128)
• The bind() method already correctly uses >= at line 672; this aligns unbind() with the same pattern
• Add test covering the boundary condition

Test plan

• New testUnbindBoundaryChar test verifies unbinding a key sequence with char value == KEYMAP_LENGTH returns gracefully without throwing
• All existing reader module tests pass (200 tests, 0 failures)

[augmentcode]

🤖 Augment PR Summary

Summary: Fixes an off-by-one bounds check in KeyMap.unbind() that could throw ArrayIndexOutOfBoundsException for key sequences containing a char equal to mapping.length.

Changes:

• Updated both bounds checks in KeyMap.unbind(..) from > to >=, matching the existing bind(..) behavior.
• Added a JUnit test covering unbinding with a boundary char value (128) to ensure it returns gracefully.

Notes: This aligns unbind/bind behavior around the fixed-size keymap table (KEYMAP_LENGTH).

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. 1 suggestion posted.

Comment augment review to trigger a new review at any time.

[augmentcode]

map.unbind("a" + boundaryChar) doesn’t currently exercise the “final char is at the boundary” path: after map.bind("action", "a"), map.mapping['a'] is not a KeyMap, so KeyMap.unbind(..) returns early before checking the last character’s bounds. This makes the test comment a bit misleading and may leave the multi-char-final-char boundary case untested.

Severity: low

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud