feat(search): P3.1 filters and multi-query merge

Native include/exclude domains, time_range, topic, exact_match, min_score,
and up to 4 queries fused with RRF. Honesty notes stay off the Warn/fail
path. Tavily keeps include_answer false and sets auto_parameters false.
Cache key is canonical JSON. CLI flags and skill recipe included.
This commit is contained in:
2026-09-29 15:30:19 -07:00
parent 5c6a4a40ec
commit 8b26776cdd
21 changed files with 1606 additions and 195 deletions
+1 -99
View File
@@ -2,7 +2,6 @@ package search
import (
"context"
"encoding/json"
"log/slog"
"sort"
"sync"
@@ -90,104 +89,7 @@ func (e *Engine) timeout() time.Duration {
}
func (e *Engine) Run(ctx context.Context, query string, k int, doRerank bool, caller, class string) RunResult {
start := e.now()
k = ClampK(k, e.DefaultK)
if e.DefaultK <= 0 {
k = ClampK(k, DefaultK)
}
if class == "" {
class = dcache.ClassFleet
}
attempted := e.attemptedNames(class)
key := dcache.SearchKey(query, attempted, k, e.now(), class)
if b, ok := e.Cache.GetSearch(key); ok {
var hits []Hit
if json.Unmarshal(b, &hits) == nil {
e.mu.Lock()
e.CacheHit++
e.mu.Unlock()
return RunResult{Hits: clampHits(hits, k), Cached: true, Attempted: attempted, SearchMS: e.now().Sub(start).Milliseconds()}
}
}
e.mu.Lock()
e.CacheMiss++
e.mu.Unlock()
type one struct {
name string
out SourceOut
skip bool
}
outs := make([]one, len(e.Sources))
var wg sync.WaitGroup
for i, src := range e.Sources {
name := src.Name()
if name == NameTavily && e.tavilyCapped(class) {
outs[i] = one{name: name, out: SourceOut{Warn: "tavily: daily cap"}, skip: true}
e.addCapped(NameTavily)
continue
}
wg.Add(1)
go func(i int, src Source, name string) {
defer wg.Done()
sctx, cancel := context.WithTimeout(ctx, e.timeout())
defer cancel()
outs[i] = one{name: name, out: src.Search(sctx, query, k)}
}(i, src, name)
}
wg.Wait()
var lists []namedHits
var hits []Hit
var warns []string
var okNames, failNames, att []string
degraded := false
for _, o := range outs {
if o.skip {
if o.out.Warn != "" {
warns = append(warns, o.out.Warn)
}
continue
}
att = append(att, o.name)
e.addQuery(caller, o.name)
if o.out.Warn != "" {
warns = append(warns, o.out.Warn)
failNames = append(failNames, o.name)
degraded = true
e.addErr(o.name, o.out.Warn)
continue
}
okNames = append(okNames, o.name)
lists = append(lists, namedHits{name: o.name, hits: o.out.Hits})
if o.out.Credits > 0 {
e.addCredits(caller, o.name, o.out.Credits)
if o.name == NameTavily && e.Cache != nil {
e.Cache.AddSearchCredits(tavilyCapSource(class), e.now().UTC().Format("2006-01-02"), o.out.Credits)
}
}
}
hits = mergeSearch(e, lists, doRerank)
var rerankMS int64
hits = clampHits(hits, k)
if !degraded && e.Cache != nil && len(att) > 0 {
payload, _ := json.Marshal(hits)
ck := dcache.SearchKey(query, att, k, e.now(), class)
e.Cache.PutSearch(ck, payload, e.ttl())
}
return RunResult{
Hits: hits,
Warnings: warns,
SourcesOK: okNames,
SourcesFailed: failNames,
SearchMS: e.now().Sub(start).Milliseconds(),
RerankMS: rerankMS,
Attempted: att,
}
return e.RunFilter(ctx, []string{query}, k, doRerank, caller, class, Filter{})
}
func (e *Engine) attemptedNames(class string) []string {