perf: optimize String.compare

[hargoniX]

This PR optimizes String.compare to turn it into 1 instead of 2 memcmp calls.

[hargoniX]

!bench

[leanprover-radar]

Benchmark results for 9bbe198 against 34df732 are in. There are significant results. @hargoniX

• ✅ build//instructions: -20.1G (-0.18%)

Large changes (1✅)

• ✅ compiled/ilean_roundtrip//instructions: -624.9M (-2.68%)

Medium changes (2✅)

• ✅ compiled/watchdogRss//instructions: -288.1M (-1.14%)
• ✅ interpreted/identifier_completion//instructions: -941.1M (-1.41%)

Small changes (26✅)

Too many entries to display here. View the full report on radar instead.

[mathlib-lean-pr-testing]

Mathlib CI status (docs):

• ❗ Batteries/Mathlib CI will not be attempted unless your PR branches off the nightly-with-mathlib branch. Try git rebase 34df7320668250fa586c2bbb706637896581a3f9 --onto 43ef70db63e2c1c1720b84fd153125c96a00c5d2. You can force Mathlib CI using the force-mathlib-ci label. (2026-05-20 13:33:22)

[leanprover-bot]

Reference manual CI status:

• ❗ Reference manual CI will not be attempted unless your PR branches off the nightly-with-manual branch. Try git rebase 34df7320668250fa586c2bbb706637896581a3f9 --onto 803553a556fd82fa1060efb0c43eda542130cb16. You can force reference manual CI using the force-manual-ci label. (2026-05-20 13:33:24)