archy/packages/app/vite-web-search.ts

import type { Plugin } from 'vite'
import type { Connect } from 'vite'
import { loadEnv } from 'vite'
import { search as searchDuckDuckGo } from 'duck-duck-scrape'

export interface WebSearchResult {
  title: string
  url: string
  content?: string
  imgSrc?: string
  engine?: string
}

const FALLBACK_INSTANCES = [
  'https://searx.tiekoetter.com',
  'https://search.bus-hit.me',
  'https://paulgo.io',
]

async function fetchFromSearXNG(
  searchUrl: string,
  q: string,
): Promise<{ results?: { title?: string; url?: string; content?: string; img_src?: string; thumbnail?: string; engine?: string }[] } | null> {
  try {
    const searchRes = await fetch(searchUrl, {
      headers: { Accept: 'application/json', 'User-Agent': 'AIUI/1.0' },
      signal: AbortSignal.timeout(6000),
    })
    if (!searchRes.ok) return null
    return (await searchRes.json()) as { results?: { title?: string; url?: string; content?: string; img_src?: string; thumbnail?: string; engine?: string }[] }
  } catch {
    return null
  }
}

function createWebSearchMiddleware(searxUrl: string | undefined) {
  return async (req: Connect.IncomingMessage, res: any, next: () => void) => {
    if (req.method !== 'GET') return next()
    const url = new URL(req.url ?? '', `http://${req.headers?.host ?? 'localhost'}`)
    const q = url.searchParams.get('q')?.trim()
    if (!q) {
      res.writeHead(400, { 'Content-Type': 'application/json' })
      res.end(JSON.stringify({ error: 'Missing q (query)' }))
      return
    }

    const instances = searxUrl ? [searxUrl] : FALLBACK_INSTANCES

    let data: { results?: { title?: string; url?: string; content?: string; img_src?: string; thumbnail?: string; engine?: string }[] } | null = null
    let lastError = ''

    for (const baseUrl of instances) {
      const normalized = baseUrl.replace(/\/$/, '')
      const searchUrl = `${normalized}/search?${new URLSearchParams({ q, format: 'json', pageno: '1' })}`
      console.log('[web-search]', q.slice(0, 40), '→', normalized)
      data = await fetchFromSearXNG(searchUrl, q)
      if (data?.results && data.results.length > 0) {
        console.log('[web-search]', data.results.length, 'results from', normalized)
        break
      }
      lastError = data ? 'no results' : 'request failed'
      console.warn('[web-search]', normalized, lastError, '— trying next')
    }

    if (!data) {
      console.warn('[web-search] SearXNG failed, trying DuckDuckGo fallback')
      try {
        const ddg = await searchDuckDuckGo(q)
        if (ddg?.results?.length) {
          const results: WebSearchResult[] = ddg.results
            .filter((r: { title?: string; url?: string }) => r.title && r.url)
            .slice(0, 6)
            .map((r: { title: string; url: string; description?: string; icon?: string }) => ({
              title: r.title,
              url: r.url,
              content: r.description ?? undefined,
              imgSrc: r.icon ?? undefined,
            }))
          console.log('[web-search] DuckDuckGo fallback:', results.length, 'results')
          res.setHeader('Content-Type', 'application/json')
          res.setHeader('Access-Control-Allow-Origin', '*')
          res.setHeader('Cache-Control', 'public, max-age=300')
          res.end(JSON.stringify({ results }))
          return
        }
      } catch (ddgErr) {
        console.warn('[web-search] DuckDuckGo fallback failed:', ddgErr)
      }
      console.error('[web-search] All search backends failed:', lastError)
      res.writeHead(502, { 'Content-Type': 'application/json' })
      res.end(JSON.stringify({ error: 'Web search unavailable. Try SEARXNG_URL in .env.local.' }))
      return
    }

    try {
      const results: WebSearchResult[] = (data.results ?? [])
        .filter((r) => r.title && r.url)
        .slice(0, 6)
        .map((r) => ({
          title: r.title ?? '',
          url: r.url ?? '',
          content: r.content ?? undefined,
          imgSrc: r.img_src || r.thumbnail || undefined,
          engine: r.engine ?? undefined,
        }))

      res.setHeader('Content-Type', 'application/json')
      res.setHeader('Access-Control-Allow-Origin', '*')
      res.setHeader('Cache-Control', 'public, max-age=300')
      res.end(JSON.stringify({ results }))
    } catch (err) {
      console.error('[web-search]', err)
      res.writeHead(502, { 'Content-Type': 'application/json' })
      res.end(JSON.stringify({ error: String(err) }))
    }
  }
}

export function webSearchPlugin(): Plugin {
  let searxUrl: string | undefined

  return {
    name: 'aiui-web-search',
    configResolved(config) {
      const env = loadEnv(config.mode, process.cwd(), '')
      searxUrl = env.SEARXNG_URL ?? env.VITE_SEARXNG_URL
    },
    configureServer(server) {
      server.middlewares.use('/api/web-search', createWebSearchMiddleware(searxUrl))
    },
    configurePreviewServer(server) {
      server.middlewares.use('/api/web-search', createWebSearchMiddleware(searxUrl))
    },
  }
}