perf: reuse query embeddings in hierarchical retriever

[mildred522]

Description

Related Issue

N/A

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Performance improvement
• Test update

Changes Made

• Preserve existing retrieval logic and output ordering.

• I have added tests that prove my fix is effective or that my feature works

• New and existing unit tests pass locally with my changes

• I have tested this on the following platforms:

  • Linux
  • macOS
  • Windows

Checklist

• My code follows the project's coding style
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published

Additional Notes

• Micro-benchmark (200 runs, fake embedder with 5ms sleep):
  • embed calls per retrieve: 2.0 → 1.0
  • avg per-call time: ~10.93ms → ~5.45ms (~50% reduction)
    This indicates the expected latency/cost improvement from eliminating duplicate embeddings.

———

[CLAassistant]

All committers have signed the CLA.

[qin-ctx]

内部默认都是 colleciton 是 context，不过这里检验下也没问题

[qin-ctx]

good job