Files
sift/internal/search/sources.go
T
kyle b6e1cdf825 p3: map Brave LLM Context; RRF merge; per-source no_rerank
Live llm/context hits live in grounding.generic plus a URL-keyed
sources map, not a sources array. Fan-in uses reciprocal-rank fusion
instead of cosine. no_rerank keeps that source's own order as a prefix.
2026-09-23 16:02:21 -07:00

261 lines
6.5 KiB
Go

package search
import (
"context"
"encoding/json"
"fmt"
"io"
"net/http"
"net/url"
"sort"
"strings"
"git.wntrmute.dev/kyle/sift/internal/search/dial"
)
func warnStatus(name string, status int) string {
if status == 0 {
return name + ": error"
}
return fmt.Sprintf("%s: %d", name, status)
}
func warnErr(name string, err error) string {
if err == nil {
return ""
}
if strings.Contains(strings.ToLower(err.Error()), "timeout") || strings.Contains(err.Error(), "deadline") {
return name + ": timeout"
}
return name + ": error"
}
func pinErr(ctx context.Context, c *dial.Client, name string) string {
if c == nil {
return name + ": error"
}
if err := c.Ensure(ctx); err != nil {
return warnErr(name, err)
}
return ""
}
func readJSON(resp *http.Response, dest any) error {
defer resp.Body.Close()
if resp.StatusCode != 200 {
_, _ = io.Copy(io.Discard, io.LimitReader(resp.Body, 512))
return fmt.Errorf("status %d", resp.StatusCode)
}
dec := json.NewDecoder(io.LimitReader(resp.Body, 1<<20))
return dec.Decode(dest)
}
// SearxNG is GET {url}/search?q=&format=json&pageno=1
type SearxNG struct {
Client *dial.Client
}
func (s *SearxNG) Name() string { return NameSearxNG }
func (s *SearxNG) Ready() bool { return s != nil && s.Client.Ready() }
func (s *BraveLLM) Ready() bool { return s != nil && s.Client.Ready() }
func (s *SearxNG) Search(ctx context.Context, query string, k int) SourceOut {
if s == nil || s.Client == nil {
return SourceOut{Warn: NameSearxNG + ": error"}
}
if w := pinErr(ctx, s.Client, NameSearxNG); w != "" {
return SourceOut{Warn: w}
}
base := strings.TrimRight(s.Client.URL.String(), "/")
u, err := url.Parse(base + "/search")
if err != nil {
return SourceOut{Warn: warnErr(NameSearxNG, err)}
}
q := u.Query()
q.Set("q", query)
q.Set("format", "json")
q.Set("pageno", "1")
u.RawQuery = q.Encode()
req, err := http.NewRequestWithContext(ctx, http.MethodGet, u.String(), nil)
if err != nil {
return SourceOut{Warn: warnErr(NameSearxNG, err)}
}
req.Header.Set("Accept", "application/json")
resp, err := s.Client.HTTP.Do(req)
if err != nil {
return SourceOut{Warn: warnErr(NameSearxNG, err)}
}
if resp.StatusCode != 200 {
_, _ = io.Copy(io.Discard, io.LimitReader(resp.Body, 256))
resp.Body.Close()
return SourceOut{Warn: warnStatus(NameSearxNG, resp.StatusCode)}
}
var env struct {
Results []struct {
URL string `json:"url"`
Title string `json:"title"`
Content string `json:"content"`
Score float64 `json:"score"`
} `json:"results"`
}
if err := readJSON(resp, &env); err != nil {
return SourceOut{Warn: warnErr(NameSearxNG, err)}
}
sort.SliceStable(env.Results, func(i, j int) bool { return env.Results[i].Score > env.Results[j].Score })
hits := make([]Hit, 0, len(env.Results))
for _, r := range env.Results {
if r.URL == "" {
continue
}
hits = append(hits, Hit{URL: r.URL, Title: r.Title, Snippet: r.Content, Score: r.Score, Source: NameSearxNG})
}
return SourceOut{Hits: hits}
}
// BraveLLM is GET {url}?q= with X-Subscription-Token.
type BraveLLM struct {
Client *dial.Client
Key string
}
func (s *BraveLLM) Name() string { return NameBraveLLM }
func (s *BraveLLM) Search(ctx context.Context, query string, k int) SourceOut {
if s == nil || s.Client == nil {
return SourceOut{Warn: NameBraveLLM + ": error"}
}
if w := pinErr(ctx, s.Client, NameBraveLLM); w != "" {
return SourceOut{Warn: w}
}
u, err := url.Parse(s.Client.URL.String())
if err != nil {
return SourceOut{Warn: warnErr(NameBraveLLM, err)}
}
q := u.Query()
q.Set("q", query)
u.RawQuery = q.Encode()
req, err := http.NewRequestWithContext(ctx, http.MethodGet, u.String(), nil)
if err != nil {
return SourceOut{Warn: warnErr(NameBraveLLM, err)}
}
req.Header.Set("Accept", "application/json")
req.Header.Set("X-Subscription-Token", s.Key)
resp, err := s.Client.HTTP.Do(req)
if err != nil {
return SourceOut{Warn: warnErr(NameBraveLLM, err)}
}
if resp.StatusCode != 200 {
code := resp.StatusCode
resp.Body.Close()
return SourceOut{Warn: warnStatus(NameBraveLLM, code)}
}
raw, err := io.ReadAll(io.LimitReader(resp.Body, 1<<20))
resp.Body.Close()
if err != nil {
return SourceOut{Warn: warnErr(NameBraveLLM, err)}
}
hits := mapBrave(raw)
credits := 0
if len(hits) >= 0 {
credits = 1
}
return SourceOut{Hits: hits, Credits: credits}
}
func mapBrave(raw []byte) []Hit {
var top map[string]any
if err := json.Unmarshal(raw, &top); err != nil {
return nil
}
var hits []Hit
seen := map[string]bool{}
add := func(u, title, snippet string) {
if u == "" || seen[u] {
return
}
seen[u] = true
hits = append(hits, Hit{URL: u, Title: title, Snippet: snippet, Source: NameBraveLLM})
}
// Live LLM Context: grounding.generic[] (and poi) with snippets[].
if g, ok := top["grounding"].(map[string]any); ok {
for _, key := range []string{"generic", "poi"} {
arr, _ := g[key].([]any)
for _, item := range arr {
m, _ := item.(map[string]any)
if m == nil {
continue
}
snip := joinSnippets(m["snippets"])
if snip == "" {
snip = strField(m, "snippet", "description", "content", "text")
}
add(strField(m, "url", "link", "href"), strField(m, "title", "name"), snip)
}
}
}
var walk func(any)
walk = func(v any) {
switch t := v.(type) {
case []any:
for _, x := range t {
walk(x)
}
case map[string]any:
u := strField(t, "url", "link", "href")
if u == "" {
// URL-keyed object: sources["https://…"] = {title, hostname}
for k, x := range t {
if strings.HasPrefix(k, "http://") || strings.HasPrefix(k, "https://") {
title, snip := "", ""
if m, ok := x.(map[string]any); ok {
title = strField(m, "title", "name", "site_name")
snip = strField(m, "description", "snippet")
}
add(k, title, snip)
continue
}
walk(x)
}
return
}
add(u, strField(t, "title", "name"), strField(t, "snippet", "description", "content", "text"))
}
}
for _, key := range []string{"sources", "results", "web"} {
if v, ok := top[key]; ok {
walk(v)
}
}
return hits
}
func joinSnippets(v any) string {
switch t := v.(type) {
case []any:
parts := make([]string, 0, len(t))
for _, x := range t {
if s, ok := x.(string); ok && s != "" {
parts = append(parts, s)
}
}
return strings.Join(parts, "\n\n")
case string:
return t
default:
return ""
}
}
func strField(m map[string]any, keys ...string) string {
for _, k := range keys {
if v, ok := m[k]; ok {
if s, ok := v.(string); ok && s != "" {
return s
}
}
}
return ""
}