From 818f6fdcfbfe296dab032af4023d98a1ac566fcf Mon Sep 17 00:00:00 2001
From: Rikki <i@rikki.moe>
Date: Mon, 26 May 2025 14:25:17 +0800
Subject: init

---
 src/blog_search.ts | 77 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 77 insertions(+)
 create mode 100644 src/blog_search.ts

(limited to 'src/blog_search.ts')

diff --git a/src/blog_search.ts b/src/blog_search.ts
new file mode 100644
index 0000000..9855cd5
--- /dev/null
+++ b/src/blog_search.ts
@@ -0,0 +1,77 @@
+import { z } from 'zod'
+
+export const ROUTE_PATH = '/api/blog_search'
+
+const EMBEDDING_DIMENSION = 1024 as const
+const MAX_TEXT_LENGTH = 1_000
+const MAX_RETRIES = 5
+const WORKERS_AI_MAX_RETRIES = 3
+
+const RequestBodySchema = z.object({
+    text: z.string(),
+})
+
+const AIResponseSchema = z.object({
+    data: z.array(z.array(z.number()).length(EMBEDDING_DIMENSION)).length(1),
+})
+
+const textResponse = (msg: string, status = 400) => new Response(msg, { status })
+
+
+export async function handle(c: HonoContext) {
+    let json: unknown
+    try {
+        json = await c.req.json()
+    } catch {
+        return textResponse('Bad Request: Invalid JSON', 400)
+    }
+
+    const body = RequestBodySchema.safeParse(json)
+    if (!body.success) {
+        return textResponse(`Bad Request: ${body.error.message}`, 400)
+    }
+
+    const text = body.data.text.trim()
+    if (text.length > MAX_TEXT_LENGTH) {
+        return textResponse('Bad Request: Text is too long', 400)
+    }
+
+    let embedding: number[] | null = null
+    for (let attempt = 0; attempt < MAX_RETRIES; attempt++) {
+        try {
+            const raw = await c.env.AI.run(
+                '@cf/baai/bge-m3',
+                { text },
+                {
+                    gateway: {
+                        id: 'workers-embedding',
+                        retries: { maxAttempts: WORKERS_AI_MAX_RETRIES },
+                    },
+                },
+            )
+
+            const parsed = AIResponseSchema.safeParse(raw)
+            if (!parsed.success) {
+                return textResponse(`Internal Server Error: ${parsed.error.message}`, 500)
+            }
+
+            embedding = parsed.data.data[0]
+            break
+        } catch (err) {
+            console.error(`Embedding attempt ${attempt + 1} failed`, err)
+        }
+    }
+
+    if (!embedding) {
+        return textResponse('Internal Server Error: Embedding failed', 500)
+    }
+
+    // ─── Respond with binary vector ────────────────────────────────────────────
+    const buffer = new Float32Array(embedding).buffer
+
+    const headers: Record<string, string> = {
+        'Content-Type': 'application/octet-stream',
+    }
+
+    return new Response(buffer, { status: 200, headers })
+}
\ No newline at end of file
-- 
cgit v1.2.3