web-search

import type { SearchHit } from "./types";

export interface SearchOptions {
  query: string;
  category?: string;
  language?: string;
  safesearch?: "0" | "1" | "2";
  timeRange?: "day" | "week" | "month" | "year";
  pageno?: number;
}

export interface SearchResponse {
  query: string;
  results: SearchHit[];
  answers: string[];
  infoboxes: Array<{ infobox: string; content?: string; urls?: Array<{ title: string; url: string }> }>;
  suggestions: string[];
  unresponsive_engines: string[];
}

export class SearxngError extends Error {
  constructor(message: string) {
    super(message);
    this.name = "SearxngError";
  }
}

export async function searxngSearch(
  instanceUrl: string,
  opts: SearchOptions,
  timeoutMs: number,
): Promise<SearchResponse> {
  if (!instanceUrl) {
    throw new SearxngError(
      "No SearXNG instance configured. Set the instance URL in the plugin's global configuration.",
    );
  }

  let base: URL;
  try {
    base = new URL(instanceUrl);
  } catch {
    throw new SearxngError(`Invalid SearXNG instance URL: "${instanceUrl}".`);
  }

  const search = new URL("search", base.toString().endsWith("/") ? base : new URL(base.toString() + "/"));
  search.searchParams.set("q", opts.query);
  search.searchParams.set("format", "json");
  if (opts.category) search.searchParams.set("categories", opts.category);
  if (opts.language && opts.language !== "auto") search.searchParams.set("language", opts.language);
  if (opts.safesearch !== undefined) search.searchParams.set("safesearch", opts.safesearch);
  if (opts.timeRange) search.searchParams.set("time_range", opts.timeRange);
  if (opts.pageno) search.searchParams.set("pageno", String(opts.pageno));

  const ctrl = new AbortController();
  const timer = setTimeout(() => ctrl.abort(), timeoutMs);
  let res: Response;
  try {
    res = await fetch(search, {
      headers: {
        Accept: "application/json",
        "User-Agent": "lms-plugin-web-search/1.0 (+https://lmstudio.ai/zexigh/web-search)",
      },
      signal: ctrl.signal,
    });
  } catch (e: unknown) {
    if (e instanceof Error && e.name === "AbortError") {
      throw new SearxngError(`SearXNG instance did not respond within ${timeoutMs} ms.`);
    }
    throw new SearxngError(`Could not reach SearXNG instance: ${e instanceof Error ? e.message : String(e)}`);
  } finally {
    clearTimeout(timer);
  }

  if (!res.ok) {
    throw new SearxngError(`SearXNG returned HTTP ${res.status} ${res.statusText}.`);
  }

  const ctype = res.headers.get("content-type") || "";
  if (!ctype.includes("application/json")) {
    throw new SearxngError(
      "SearXNG did not return JSON. The instance probably has the JSON format disabled in `settings.yml` (search.formats must include `json`).",
    );
  }

  const data = (await res.json()) as Record<string, unknown>;
  const rawResults = Array.isArray(data.results) ? (data.results as Record<string, unknown>[]) : [];
  const results: SearchHit[] = rawResults.map((r) => ({
    title: String(r.title ?? ""),
    url: String(r.url ?? ""),
    content: String(r.content ?? ""),
    engine: typeof r.engine === "string" ? r.engine : undefined,
    category: typeof r.category === "string" ? r.category : undefined,
    published_date: typeof r.publishedDate === "string" ? r.publishedDate : undefined,
  }));

  return {
    query: String(data.query ?? opts.query),
    results,
    answers: Array.isArray(data.answers) ? (data.answers as string[]) : [],
    infoboxes: Array.isArray(data.infoboxes) ? (data.infoboxes as SearchResponse["infoboxes"]) : [],
    suggestions: Array.isArray(data.suggestions) ? (data.suggestions as string[]) : [],
    unresponsive_engines: Array.isArray(data.unresponsive_engines)
      ? (data.unresponsive_engines as unknown[]).map((e) => (Array.isArray(e) ? String(e[0]) : String(e)))
      : [],
  };
}