fix: UseAnchoredLiteral regression in FindIndices

[kolkov]

Summary

• Add missing UseAnchoredLiteral case to FindIndices, FindIndicesAt, and findIndicesAtWithState
• Add findIndicesAnchoredLiteral and findIndicesAnchoredLiteralAt methods

Problem

Pattern ^/.*[\w-]+\.php$ fell through to slow NFA path because UseAnchoredLiteral was missing in switch statements.

Regression: 0.01ms → 408ms (40,000x slower)

Fix

After fix: 408ms → 0.5ms (O(1) anchored literal matching restored)

Test plan

• All tests pass
• Linter passes
• Local benchmark confirms fix: anchored_php: 566µs

Fixes #107

[github-actions]

Benchmark Comparison

Comparing main → PR #110

Summary: geomean 139.0n 138.8n -0.11%

⚠️ Potential regressions detected:

geomean                               ³                +0.00%               ³
geomean                               ³                +0.00%               ³
geomean                         ³                +0.00%               ³
geomean                         ³                +0.00%               ³
AnchoredLiteralVsStdlib/stdlib_no_match-4               600.0n ± ∞ ¹    610.1n ± ∞ ¹     +1.68% (p=0.032 n=5)
BranchDispatch_Coregex/Hex32-4                          7.569n ± ∞ ¹    7.620n ± ∞ ¹     +0.67% (p=0.008 n=5)
IPRegex_Find/coregex_1KB_sparse-4                       2.426µ ± ∞ ¹    4.405µ ± ∞ ¹    +81.57% (p=0.008 n=5)
IPRegex_Find/stdlib_1MB_sparse-4                        153.4µ ± ∞ ¹   1763.8µ ± ∞ ¹  +1049.94% (p=0.008 n=5)
IPRegex_Find/coregex_6MB_sparse-4                       3.332µ ± ∞ ¹    4.924µ ± ∞ ¹    +47.78% (p=0.008 n=5)
IPRegex_Find/stdlib_1MB_dense-4                         1.748µ ± ∞ ¹   68.276µ ± ∞ ¹  +3805.95% (p=0.008 n=5)

Full results available in workflow artifacts. CI runners have ~10-20% variance.
For accurate benchmarks, run locally: ./scripts/bench.sh --compare