feat: search materials through beachcomber
- Material is a beachcomber Searchable (index jz_bm_materials) in the hand-written models/material_search.go; jz.bm requires summercms v0.6.4 - classes.MaterialIndexer asks Typesense for up to 250 candidates with body highlights; the app-wide Gate is on only for jz.bm-marked contexts - GET /_bm/api/v1/materials takes q: candidates are re-gated, counted and paged in SQL in rank order, each row with at most one plain snippet - search failure answers 503 Search is unavailable.; without q the list is unchanged
This commit is contained in:
123
controllers/api/material_search.go
Normal file
123
controllers/api/material_search.go
Normal file
@@ -0,0 +1,123 @@
|
||||
package api
|
||||
|
||||
import (
|
||||
"net/http"
|
||||
"strconv"
|
||||
"strings"
|
||||
"unicode/utf8"
|
||||
|
||||
"git.golem15.com/golem15/summercms/modules/beachcomber"
|
||||
"git.golem15.com/golem15/summercms/modules/wire"
|
||||
"git.golem15.com/jakub/sm-bm-plugin/models"
|
||||
)
|
||||
|
||||
// Material search: the engine (beachcomber, Typesense) answers candidate ids
|
||||
// only. Visibility, the filters, the total and the paging stay in SQL, and a
|
||||
// row's snippet is read only once SQL returned that row, so a material the
|
||||
// caller may not see never reaches the JSON, its count or a snippet.
|
||||
|
||||
// minSearchRunes is the shortest query that searches; a shorter one is
|
||||
// ignored, as the student app does.
|
||||
const minSearchRunes = 2
|
||||
|
||||
// searchQuery returns the trimmed q of in, or "" when it is shorter than
|
||||
// minSearchRunes.
|
||||
func searchQuery(in map[string]any) string {
|
||||
q := strings.TrimSpace(str(in, "q"))
|
||||
if utf8.RuneCountInString(q) < minSearchRunes {
|
||||
return ""
|
||||
}
|
||||
return q
|
||||
}
|
||||
|
||||
// writeSearchUnavailable answers a search the engine cannot serve: it is not
|
||||
// configured or the call failed. The list without q keeps working.
|
||||
func writeSearchUnavailable(w http.ResponseWriter) {
|
||||
wire.WriteJSON(w, http.StatusServiceUnavailable, map[string]any{"error": true, "message": "Search is unavailable."})
|
||||
}
|
||||
|
||||
// rankOrder orders rows by their position in ids, the engine's rank. The ids
|
||||
// are parsed integers, so the array literal carries no caller text.
|
||||
func rankOrder(ids []uint) string {
|
||||
parts := make([]string, len(ids))
|
||||
for i, id := range ids {
|
||||
parts[i] = strconv.FormatUint(uint64(id), 10)
|
||||
}
|
||||
return "array_position(ARRAY[" + strings.Join(parts, ",") + "]::bigint[], jz_bm_materials.id), jz_bm_materials.id DESC"
|
||||
}
|
||||
|
||||
// segmentJSON is one run of a snippet: the engine's plain text, verbatim,
|
||||
// and whether it matched the query. It is never markup; the client renders
|
||||
// it as text.
|
||||
type segmentJSON struct {
|
||||
Text string `json:"text"`
|
||||
Hit bool `json:"hit"`
|
||||
}
|
||||
|
||||
// snippetJSON is the one excerpt a search result shows. Source is "body" or
|
||||
// "file" (FileName set); Lead and Trail say whether text was cut before or
|
||||
// after the excerpt.
|
||||
type snippetJSON struct {
|
||||
Source string `json:"source"`
|
||||
FileName *string `json:"file_name"`
|
||||
Lead bool `json:"lead"`
|
||||
Trail bool `json:"trail"`
|
||||
Segments []segmentJSON `json:"segments"`
|
||||
}
|
||||
|
||||
// pickSnippet returns the snippet of m from hit: the first body snippet,
|
||||
// mapped one to one. It is nil when the hit has none. Call it only for a
|
||||
// row the SQL re-gate returned.
|
||||
func pickSnippet(hit beachcomber.Hit, m models.Material) *snippetJSON {
|
||||
for _, sn := range hit.Highlights {
|
||||
if sn.Field != "body" || len(sn.Segments) == 0 {
|
||||
continue
|
||||
}
|
||||
out := &snippetJSON{Source: "body", Segments: make([]segmentJSON, 0, len(sn.Segments))}
|
||||
var joined strings.Builder
|
||||
for _, seg := range sn.Segments {
|
||||
out.Segments = append(out.Segments, segmentJSON{Text: seg.Text, Hit: seg.Match})
|
||||
joined.WriteString(seg.Text)
|
||||
}
|
||||
body := ""
|
||||
if m.Body != nil {
|
||||
body = models.MaterialPlainBody(*m.Body)
|
||||
}
|
||||
text := strings.TrimSpace(joined.String())
|
||||
out.Lead = !strings.HasPrefix(body, text)
|
||||
out.Trail = !strings.HasSuffix(body, text)
|
||||
return out
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
// flattenSnippet is a snippet as plain text, for MCP: the segment texts
|
||||
// joined, with an ellipsis where text was cut.
|
||||
func flattenSnippet(s *snippetJSON) string {
|
||||
if s == nil {
|
||||
return ""
|
||||
}
|
||||
var b strings.Builder
|
||||
if s.Lead {
|
||||
b.WriteString("…")
|
||||
}
|
||||
for _, seg := range s.Segments {
|
||||
b.WriteString(seg.Text)
|
||||
}
|
||||
if s.Trail {
|
||||
b.WriteString("…")
|
||||
}
|
||||
return b.String()
|
||||
}
|
||||
|
||||
// snippetValue is the snippet key of a row: the struct, a plain string
|
||||
// through MCP, or nil when there is none (the key is then left out).
|
||||
func snippetValue(s *snippetJSON, mcp bool) any {
|
||||
if s == nil {
|
||||
return nil
|
||||
}
|
||||
if mcp {
|
||||
return flattenSnippet(s)
|
||||
}
|
||||
return s
|
||||
}
|
||||
Reference in New Issue
Block a user