src/cli.ts

#!/usr/bin/env node
import fs from "fs/promises";
import path from "path";
import { Command } from "commander";
import { runCollect, compareItems, exportCsv } from "./core/pipeline.js";
import { startServer } from "./server-start.js";
import type { CollectOptions, DedupeMode, Platform, RawProductItem } from "./types.js";

const program = new Command();
program.name("pricebot").description("电商商品价格采集与对比工具").version("0.1.0");

function parsePlatforms(v: string): Platform[] {
  const parts = v
    .split(",")
    .map((x) => x.trim())
    .filter(Boolean);
  const ok = new Set<Platform>(["jd", "tb", "pdd"]);
  const out: Platform[] = [];
  for (const p of parts) {
    if (!ok.has(p as Platform)) throw new Error(`未知平台: ${p}`);
    out.push(p as Platform);
  }
  return out.length ? out : ["jd", "tb", "pdd"];
}

async function writeJson(p: string, data: unknown) {
  const out = path.resolve(process.cwd(), p);
  await fs.mkdir(path.dirname(out), { recursive: true });
  await fs.writeFile(out, `${JSON.stringify(data, null, 2)}\n`, "utf-8");
}

program
  .command("collect")
  .requiredOption("--q <keyword>", "关键词")
  .option("--platform <list>", "平台（逗号分隔：jd,tb,pdd）", "jd,tb,pdd")
  .option("--limit <n>", "每个平台最多抓取条数", "30")
  .option("--concurrency <n>", "平台并发数", "3")
  .option("--qps <n>", "单进程节流（请求/秒，0 为不限制）", "")
  .option("--retries <n>", "请求失败重试次数", "")
  .option("--log", "打印采集日志")
  .option("--dedupe <mode>", "去重模式：platform|cross", "platform")
  .option("--mode <mode>", "采集模式：mock|live", "mock")
  .option("--out <file>", "输出 JSON 路径", "out/collect.json")
  .action(async (opts) => {
    const options: CollectOptions = {
      q: String(opts.q),
      platforms: parsePlatforms(String(opts.platform)),
      limit: Number(opts.limit),
      concurrency: Number(opts.concurrency),
      qps: opts.qps === "" ? undefined : Number(opts.qps),
      retries: opts.retries === "" ? undefined : Number(opts.retries),
      dedupe: String(opts.dedupe) as DedupeMode,
      mode: String(opts.mode) as CollectOptions["mode"]
    };
    const res = await runCollect(options, {
      onEvent: opts.log
        ? (e) => {
            const p = e.platform ? `[${e.platform}] ` : "";
            process.stderr.write(`${e.t} ${e.level.toUpperCase()} ${p}${e.msg}\n`);
          }
        : undefined
    });
    await writeJson(String(opts.out), res);
    process.stdout.write(`${opts.out}\n`);
    if (res.warnings.length) process.stderr.write(`${res.warnings.join("\n")}\n`);
  });

program
  .command("compare")
  .requiredOption("--in <file>", "输入 JSON（collect 输出或 rawItems 数组）")
  .option("--dedupe <mode>", "去重模式：platform|cross", "platform")
  .option("--out <file>", "输出 JSON 路径", "out/compare.json")
  .option("--export <fmt>", "额外导出：csv", "")
  .action(async (opts) => {
    const inPath = path.resolve(process.cwd(), String(opts.in));
    const s = await fs.readFile(inPath, "utf-8");
    const data = JSON.parse(s) as { rawItems?: RawProductItem[] } | RawProductItem[];
    const rawItems = Array.isArray(data) ? data : (data.rawItems ?? []);
    const { items } = compareItems(rawItems, { dedupe: String(opts.dedupe) as DedupeMode });
    const out = { items };
    await writeJson(String(opts.out), out);
    if (String(opts.export).toLowerCase() === "csv") {
      const csvPath = String(opts.out).replace(/\.json$/i, ".csv");
      await exportCsv(items, csvPath);
      process.stdout.write(`${opts.out}\n${csvPath}\n`);
      return;
    }
    process.stdout.write(`${opts.out}\n`);
  });

program
  .command("serve")
  .option("--port <n>", "端口", "5175")
  .action(async (opts) => {
    await startServer({ port: Number(opts.port) });
  });

program.parseAsync(process.argv).catch((err: unknown) => {
  const msg = err instanceof Error ? err.message : String(err);
  process.stderr.write(`${msg}\n`);
  process.exit(1);
});