70 lines
2 KiB
TypeScript
70 lines
2 KiB
TypeScript
|
/**
|
||
|
|
* /api/chunk?doc=<doc-id>&chunk=<c0042>
|
||
|
|
*
|
||
|
|
* Returns one chunk's payload (page, type, bbox, content_en, content_pt,
|
||
|
|
* classification, anomaly flags). Used by client-side InlineCitation
|
||
|
|
* component to expand `[[doc/p007#c0042]]` links into rich cards.
|
||
|
|
*
|
||
|
|
* Tries DB first (with embeddings already indexed), falls back to filesystem
|
||
|
|
* (raw/<doc>--subagent/chunks/cXXXX.md) so the UX works even before the
|
||
|
|
* indexer has run on a particular doc.
|
||
|
|
*/
|
||
|
|
import { NextRequest } from "next/server";
|
||
|
|
import { getChunk } from "@/lib/retrieval/hybrid";
|
||
|
|
import { readChunk as readChunkFs } from "@/lib/chunks";
|
||
|
|
|
||
|
|
export const runtime = "nodejs";
|
||
|
|
export const dynamic = "force-dynamic";
|
||
|
|
|
||
|
|
function json(data: unknown, status = 200) {
|
||
|
|
return new Response(JSON.stringify(data), {
|
||
|
|
status,
|
||
|
|
headers: {
|
||
|
|
"content-type": "application/json",
|
||
|
|
"cache-control": "public, max-age=300",
|
||
|
|
},
|
||
|
|
});
|
||
|
|
}
|
||
|
|
|
||
|
|
export async function GET(req: NextRequest) {
|
||
|
|
const u = new URL(req.url);
|
||
|
|
const doc = u.searchParams.get("doc")?.trim();
|
||
|
|
const chunk = u.searchParams.get("chunk")?.trim();
|
||
|
|
if (!doc || !chunk) return json({ error: "doc and chunk required" }, 400);
|
||
|
|
|
||
|
|
// 1. Try DB
|
||
|
|
try {
|
||
|
|
const c = await getChunk(doc, chunk);
|
||
|
|
if (c) {
|
||
|
|
return json({
|
||
|
|
source: "db",
|
||
|
|
chunk_id: c.chunk_id,
|
||
|
|
doc_id: c.doc_id,
|
||
|
|
page: c.page,
|
||
|
|
type: c.type,
|
||
|
|
bbox: c.bbox,
|
||
|
|
classification: c.classification,
|
||
|
|
content_en: c.content_en,
|
||
|
|
content_pt: c.content_pt,
|
||
|
|
});
|
||
|
|
}
|
||
|
|
} catch {
|
||
|
|
// db unavailable → fall through to fs
|
||
|
|
}
|
||
|
|
|
||
|
|
// 2. Filesystem fallback
|
||
|
|
const fs = await readChunkFs(doc, chunk);
|
||
|
|
if (!fs) return json({ error: "not_found", doc_id: doc, chunk_id: chunk }, 404);
|
||
|
|
return json({
|
||
|
|
source: "fs",
|
||
|
|
chunk_id: fs.fm.chunk_id,
|
||
|
|
doc_id: doc,
|
||
|
|
page: fs.fm.page,
|
||
|
|
type: fs.fm.type,
|
||
|
|
bbox: fs.fm.bbox,
|
||
|
|
classification: fs.fm.classification,
|
||
|
|
content_en: fs.content_en,
|
||
|
|
content_pt: fs.content_pt,
|
||
|
|
});
|
||
|
|
}
|