Full-text index Go API

search.FulltextIndexV2 provides the full-text index's single-document entry point. Construct it with an analyzer, or pass nil for the default analyzer. Go calls must supply the analysis mode explicitly:

index := search.NewFulltextIndexV2WithAnalyzer(nil)
index.Index(id, text, nil, nil, false)

With prepared set to false, the index analyzes the text with its configured analyzer and ignores the token and frequency arguments. Code that already used that same analyzer for admission or accounting can reuse its result:

terms := analyzer.Analyze(text)
frequency := make(map[string]int, len(terms))
for _, term := range terms {
    frequency[term]++
}
index.Index(id, text, terms, frequency, true)

When prepared is true, the caller must supply the exact tokens and their frequencies for the text using the analyzer configured on the index. This avoids analyzing the same document twice. IndexBatch remains the bulk-build API and analyzes its entries itself. Single and batch writes ignore empty IDs.