feat: index the text of attached PDFs

- jz_bm_material_file_texts caches the pdftotext output per file
- the River job jz.bm.extract_material_text extracts and resyncs
- the management multipart attach queues it inside its transaction
- file_texts in the document and a file snippet with its name
This commit is contained in:
Jakub Zych
2026-11-08 20:11:19 +01:00
parent dd13795e6f
commit 412821ec56
21 changed files with 747 additions and 24 deletions

View File

@@ -251,13 +251,21 @@ func (h *Handlers) ManagedMaterials(w http.ResponseWriter, r *http.Request) {
writeOpaque500(w)
return
}
var files map[uint][]models.IndexedFile
if cands != nil {
files, err = snippetFiles(db, materialIDs(rows), cands.Hits)
if err != nil {
writeOpaque500(w)
return
}
}
mcp := viaMCP(r)
out := make([]any, 0, len(rows))
for _, m := range rows {
row := newManagedMaterialJSON(m, people)
if cands != nil {
if hit, ok := cands.Hits[m.ID]; ok {
row.Snippet = snippetValue(pickSnippet(hit, m), mcp)
row.Snippet = snippetValue(pickSnippet(hit, m, files[m.ID]), mcp)
}
}
out = append(out, managedMaterialView(row, mcp, includeBody(r)))