feat(milvus2): add auto search mode for search strategy inference

[kaijchen]

What type of PR is this?

feat

Check the PR title.

• This PR title match the format: <type>(optional scope): <description>
• The description of this PR title is user-oriented and clear enough for others to understand.
• Attach the PR updating the user documentation if the current PR requires user awareness at the usage level.

(Optional) Translate the PR title into Chinese.

feat(milvus2): 添加 auto 搜索模式，支持推断搜索策略

(Optional) More detailed description for this PR(en: English/zh: Chinese).

en:
This PR introduces an Auto Search Mode for the Milvus2 retriever that automatically determines the optimal search strategy based on configuration:

Key Changes:

1. New Auto Search Mode (Default)

   • Both VectorField and SparseVectorField configured -> Hybrid Search (RRF reranker)
   • Only VectorField configured -> Dense (Approximate) Search
   • Only SparseVectorField configured -> Sparse Search
   • SearchMode in RetrieverConfig is now optional; defaults to auto mode

2. Search Mode Refactoring

   • Moved all search mode implementations from search_mode/ subpackage to main milvus2/ package
   • Original search_mode/ package contains deprecated aliases for backward compatibility

3. SearchParams Builder

   • Added SearchParams field to RetrieverConfig for per-field search parameters
   • New NewSearchParams() builder for fluent parameter construction

zh(optional):
本PR为Milvus2检索器引入了自动搜索模式，根据配置自动选择最优搜索策略：

• 同时配置VectorField和SparseVectorField -> 混合搜索
• 仅配置VectorField -> 稠密向量搜索
• 仅配置SparseVectorField -> 稀疏向量搜索

(Optional) Which issue(s) this PR fixes:

(optional) The PR that updates user documentation: