web-access-cli 0.1.0

package/dist/cli.js

@@ -0,0 +1,197 @@
+#!/usr/bin/env node
+import { realpathSync } from "node:fs";
+import { pathToFileURL } from "node:url";
+import { Command, CommanderError, InvalidArgumentError, Option, } from "commander";
+import { loadConfig } from "./config/config.js";
+import { executeConfigEdit } from "./config/edit.js";
+import { executeDoctor, executeProviders } from "./core/diagnostics.js";
+import { asWebAccessError, WebAccessError } from "./core/errors.js";
+import { errorEnvelope, executeExtract, executeSearch } from "./core/router.js";
+import { COMMANDS } from "./core/types.js";
+import { normalizeDomains } from "./providers/common.js";
+import { VERSION } from "./version.js";
+function collect(value, previous) {
+    return [...previous, value];
+}
+function integer(value) {
+    const parsed = Number(value);
+    if (!Number.isSafeInteger(parsed) || parsed < 1)
+        throw new InvalidArgumentError("必须是大于 0 的整数");
+    return parsed;
+}
+function httpUrl(value) {
+    let url;
+    try {
+        url = new URL(value);
+    }
+    catch {
+        throw new WebAccessError("invalid_input", "url 必须是合法的绝对 URL");
+    }
+    if (url.protocol !== "http:" && url.protocol !== "https:")
+        throw new WebAccessError("invalid_input", "url 必须使用 http 或 https");
+    return url.toString();
+}
+function exitCode(envelope) {
+    if (envelope.ok)
+        return 0;
+    if (envelope.error.code === "aborted")
+        return 130;
+    if ([
+        "invalid_input",
+        "config_error",
+        "provider_unknown",
+        "provider_disabled",
+    ].includes(envelope.error.code))
+        return 2;
+    return 1;
+}
+function writeEnvelope(envelope, pretty) {
+    process.stdout.write(`${JSON.stringify(envelope, null, pretty ? 2 : undefined)}\n`);
+}
+export function createProgram(run, dependencies = {}) {
+    const program = new Command();
+    const runConfigEdit = dependencies.executeConfigEdit ?? executeConfigEdit;
+    program
+        .name("web-access")
+        .description("Agent-neutral 的网页搜索与内容提取 CLI")
+        .version(VERSION)
+        .option("--config <path>", "指定 JSON 配置文件")
+        .option("--pretty", "格式化 JSON 输出", false)
+        .showSuggestionAfterError();
+    program.configureOutput({ writeErr: () => { }, outputError: () => { } });
+    program.exitOverride();
+    program
+        .command("search")
+        .description("搜索网页")
+        .argument("<query>", "搜索关键词")
+        .option("-p, --provider <id>", "provider instance id，或 auto", "auto")
+        .option("-l, --limit <number>", "结果数量（1-20）", integer)
+        .addOption(new Option("--freshness <window>", "时间范围").choices([
+        "day",
+        "month",
+        "year",
+    ]))
+        .option("--include-domain <domain>", "仅包含域名，可重复", collect, [])
+        .option("--exclude-domain <domain>", "排除域名，可重复", collect, [])
+        .option("--timeout <milliseconds>", "总超时毫秒数", integer)
+        .action((query, options) => {
+        run(async () => {
+            const globals = program.opts();
+            const loaded = loadConfig(globals.config);
+            const request = {
+                query: query.trim(),
+                provider: options.provider.toLowerCase(),
+                limit: options.limit ?? loaded.app.search.limit,
+                freshness: options.freshness,
+                includeDomains: normalizeDomains(options.includeDomain),
+                excludeDomains: normalizeDomains(options.excludeDomain),
+                timeoutMs: options.timeout,
+            };
+            if (!request.query)
+                throw new WebAccessError("invalid_input", "query 不能为空");
+            if (request.limit > 20)
+                throw new WebAccessError("invalid_input", "limit 不能超过 20");
+            return executeSearch(request, {
+                loaded,
+                signal: processSignal.signal,
+            });
+        });
+    });
+    program
+        .command("extract")
+        .description("提取网页正文并转换为 Markdown")
+        .argument("<url>", "HTTP(S) URL")
+        .option("-p, --provider <id>", "provider instance id，或 auto", "auto")
+        .option("--timeout <milliseconds>", "总超时毫秒数", integer)
+        .action((url, options) => {
+        run(async () => {
+            const globals = program.opts();
+            const loaded = loadConfig(globals.config);
+            const request = {
+                url: httpUrl(url),
+                provider: options.provider.toLowerCase(),
+                timeoutMs: options.timeout,
+            };
+            return executeExtract(request, {
+                loaded,
+                signal: processSignal.signal,
+            });
+        });
+    });
+    program
+        .command("providers")
+        .description("列出 provider instance、route 与配置状态")
+        .action(() => run(() => executeProviders(loadConfig(program.opts().config))));
+    program
+        .command("doctor")
+        .description("检查本地配置与已启用 provider 的可用性")
+        .action(() => run(() => executeDoctor(loadConfig(program.opts().config))));
+    const config = program.command("config").description("管理配置文件");
+    config
+        .command("edit")
+        .description("创建默认配置文件并用系统默认应用打开")
+        .action(() => run(() => runConfigEdit({
+        explicitPath: program.opts().config,
+    })));
+    return program;
+}
+function envelopeCommand(args) {
+    const nested = args.slice(0, 2).join(".");
+    if (COMMANDS.includes(nested))
+        return nested;
+    const command = args[0];
+    return COMMANDS.includes(command)
+        ? command
+        : null;
+}
+const processSignal = new AbortController();
+process.once("SIGINT", () => processSignal.abort());
+async function main() {
+    let pending;
+    const program = createProgram((task) => {
+        pending = Promise.resolve().then(task);
+    });
+    if (process.argv.length <= 2) {
+        const envelope = errorEnvelope(new WebAccessError("invalid_input", "必须指定命令"));
+        writeEnvelope(envelope, false);
+        process.exitCode = 2;
+        return;
+    }
+    try {
+        await program.parseAsync(process.argv);
+        if (!pending)
+            throw new WebAccessError("invalid_input", "必须指定命令");
+        const envelope = await pending;
+        writeEnvelope(envelope, !!program.opts().pretty);
+        process.exitCode = exitCode(envelope);
+    }
+    catch (caught) {
+        if (caught instanceof CommanderError &&
+            caught.code === "commander.helpDisplayed")
+            return;
+        if (caught instanceof CommanderError && caught.code === "commander.version")
+            return;
+        const normalized = caught instanceof CommanderError
+            ? new WebAccessError("invalid_input", caught.code === "commander.help" && program.args[0] === "config"
+                ? "必须指定 config 子命令"
+                : caught.message)
+            : asWebAccessError(caught);
+        const envelope = errorEnvelope(normalized, envelopeCommand(program.args));
+        writeEnvelope(envelope, !!program.opts().pretty);
+        process.exitCode = exitCode(envelope);
+    }
+}
+export function isMainModule(moduleUrl, entryPath) {
+    if (!entryPath)
+        return false;
+    try {
+        return moduleUrl === pathToFileURL(realpathSync(entryPath)).href;
+    }
+    catch {
+        return false;
+    }
+}
+if (isMainModule(import.meta.url, process.argv[1])) {
+    await main();
+}
+//# sourceMappingURL=cli.js.map

package/dist/cli.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"cli.js","sourceRoot":"","sources":["../src/cli.ts"],"names":[],"mappings":";AACA,OAAO,EAAE,YAAY,EAAE,MAAM,SAAS,CAAC;AACvC,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AACzC,OAAO,EACN,OAAO,EACP,cAAc,EACd,oBAAoB,EACpB,MAAM,GACN,MAAM,WAAW,CAAC;AACnB,OAAO,EAAE,UAAU,EAAE,MAAM,oBAAoB,CAAC;AAChD,OAAO,EAAE,iBAAiB,EAAE,MAAM,kBAAkB,CAAC;AACrD,OAAO,EAAE,aAAa,EAAE,gBAAgB,EAAE,MAAM,uBAAuB,CAAC;AACxE,OAAO,EAAE,gBAAgB,EAAE,cAAc,EAAE,MAAM,kBAAkB,CAAC;AACpE,OAAO,EAAE,aAAa,EAAE,cAAc,EAAE,aAAa,EAAE,MAAM,kBAAkB,CAAC;AAQhF,OAAO,EAAE,QAAQ,EAAE,MAAM,iBAAiB,CAAC;AAC3C,OAAO,EAAE,gBAAgB,EAAE,MAAM,uBAAuB,CAAC;AACzD,OAAO,EAAE,OAAO,EAAE,MAAM,cAAc,CAAC;AAWvC,SAAS,OAAO,CAAC,KAAa,EAAE,QAAkB;IACjD,OAAO,CAAC,GAAG,QAAQ,EAAE,KAAK,CAAC,CAAC;AAC7B,CAAC;AAED,SAAS,OAAO,CAAC,KAAa;IAC7B,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC;IAC7B,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,MAAM,GAAG,CAAC;QAC9C,MAAM,IAAI,oBAAoB,CAAC,aAAa,CAAC,CAAC;IAC/C,OAAO,MAAM,CAAC;AACf,CAAC;AAED,SAAS,OAAO,CAAC,KAAa;IAC7B,IAAI,GAAQ,CAAC;IACb,IAAI,CAAC;QACJ,GAAG,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,CAAC;IACtB,CAAC;IAAC,MAAM,CAAC;QACR,MAAM,IAAI,cAAc,CAAC,eAAe,EAAE,kBAAkB,CAAC,CAAC;IAC/D,CAAC;IACD,IAAI,GAAG,CAAC,QAAQ,KAAK,OAAO,IAAI,GAAG,CAAC,QAAQ,KAAK,QAAQ;QACxD,MAAM,IAAI,cAAc,CAAC,eAAe,EAAE,uBAAuB,CAAC,CAAC;IACpE,OAAO,GAAG,CAAC,QAAQ,EAAE,CAAC;AACvB,CAAC;AAED,SAAS,QAAQ,CAAC,QAAwB;IACzC,IAAI,QAAQ,CAAC,EAAE;QAAE,OAAO,CAAC,CAAC;IAC1B,IAAI,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,SAAS;QAAE,OAAO,GAAG,CAAC;IAClD,IACC;QACC,eAAe;QACf,cAAc;QACd,kBAAkB;QAClB,mBAAmB;KACnB,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,CAAC;QAE/B,OAAO,CAAC,CAAC;IACV,OAAO,CAAC,CAAC;AACV,CAAC;AAED,SAAS,aAAa,CAAC,QAAwB,EAAE,MAAe;IAC/D,OAAO,CAAC,MAAM,CAAC,KAAK,CACnB,GAAG,IAAI,CAAC,SAAS,CAAC,QAAQ,EAAE,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI,CAC7D,CAAC;AACH,CAAC;AAED,MAAM,UAAU,aAAa,CAC5B,GAAmE,EACnE,eAAgC,EAAE;IAElC,MAAM,OAAO,GAAG,IAAI,OAAO,EAAE,CAAC;IAC9B,MAAM,aAAa,GAAG,YAAY,CAAC,iBAAiB,IAAI,iBAAiB,CAAC;IAC1E,OAAO;SACL,IAAI,CAAC,YAAY,CAAC;SAClB,WAAW,CAAC,8BAA8B,CAAC;SAC3C,OAAO,CAAC,OAAO,CAAC;SAChB,MAAM,CAAC,iBAAiB,EAAE,cAAc,CAAC;SACzC,MAAM,CAAC,UAAU,EAAE,aAAa,EAAE,KAAK,CAAC;SACxC,wBAAwB,EAAE,CAAC;IAC7B,OAAO,CAAC,eAAe,CAAC,EAAE,QAAQ,EAAE,GAAG,EAAE,GAAE,CAAC,EAAE,WAAW,EAAE,GAAG,EAAE,GAAE,CAAC,EAAE,CAAC,CAAC;IACvE,OAAO,CAAC,YAAY,EAAE,CAAC;IAEvB,OAAO;SACL,OAAO,CAAC,QAAQ,CAAC;SACjB,WAAW,CAAC,MAAM,CAAC;SACnB,QAAQ,CAAC,SAAS,EAAE,OAAO,CAAC;SAC5B,MAAM,CAAC,qBAAqB,EAAE,6BAA6B,EAAE,MAAM,CAAC;SACpE,MAAM,CAAC,sBAAsB,EAAE,YAAY,EAAE,OAAO,CAAC;SACrD,SAAS,CACT,IAAI,MAAM,CAAC,sBAAsB,EAAE,MAAM,CAAC,CAAC,OAAO,CAAC;QAClD,KAAK;QACL,OAAO;QACP,MAAM;KACN,CAAC,CACF;SACA,MAAM,CAAC,2BAA2B,EAAE,WAAW,EAAE,OAAO,EAAE,EAAE,CAAC;SAC7D,MAAM,CAAC,2BAA2B,EAAE,UAAU,EAAE,OAAO,EAAE,EAAE,CAAC;SAC5D,MAAM,CAAC,0BAA0B,EAAE,QAAQ,EAAE,OAAO,CAAC;SACrD,MAAM,CACN,CACC,KAAa,EACb,OAOC,EACA,EAAE;QACH,GAAG,CAAC,KAAK,IAAI,EAAE;YACd,MAAM,OAAO,GAAG,OAAO,CAAC,IAAI,EAAiB,CAAC;YAC9C,MAAM,MAAM,GAAG,UAAU,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;YAC1C,MAAM,OAAO,GAAkB;gBAC9B,KAAK,EAAE,KAAK,CAAC,IAAI,EAAE;gBACnB,QAAQ,EAAE,OAAO,CAAC,QAAQ,CAAC,WAAW,EAAE;gBACxC,KAAK,EAAE,OAAO,CAAC,KAAK,IAAI,MAAM,CAAC,GAAG,CAAC,MAAM,CAAC,KAAK;gBAC/C,SAAS,EAAE,OAAO,CAAC,SAAS;gBAC5B,cAAc,EAAE,gBAAgB,CAAC,OAAO,CAAC,aAAa,CAAC;gBACvD,cAAc,EAAE,gBAAgB,CAAC,OAAO,CAAC,aAAa,CAAC;gBACvD,SAAS,EAAE,OAAO,CAAC,OAAO;aAC1B,CAAC;YACF,IAAI,CAAC,OAAO,CAAC,KAAK;gBACjB,MAAM,IAAI,cAAc,CAAC,eAAe,EAAE,YAAY,CAAC,CAAC;YACzD,IAAI,OAAO,CAAC,KAAK,GAAG,EAAE;gBACrB,MAAM,IAAI,cAAc,CAAC,eAAe,EAAE,eAAe,CAAC,CAAC;YAC5D,OAAO,aAAa,CAAC,OAAO,EAAE;gBAC7B,MAAM;gBACN,MAAM,EAAE,aAAa,CAAC,MAAM;aAC5B,CAAC,CAAC;QACJ,CAAC,CAAC,CAAC;IACJ,CAAC,CACD,CAAC;IAEH,OAAO;SACL,OAAO,CAAC,SAAS,CAAC;SAClB,WAAW,CAAC,qBAAqB,CAAC;SAClC,QAAQ,CAAC,OAAO,EAAE,aAAa,CAAC;SAChC,MAAM,CAAC,qBAAqB,EAAE,6BAA6B,EAAE,MAAM,CAAC;SACpE,MAAM,CAAC,0BAA0B,EAAE,QAAQ,EAAE,OAAO,CAAC;SACrD,MAAM,CAAC,CAAC,GAAW,EAAE,OAA+C,EAAE,EAAE;QACxE,GAAG,CAAC,KAAK,IAAI,EAAE;YACd,MAAM,OAAO,GAAG,OAAO,CAAC,IAAI,EAAiB,CAAC;YAC9C,MAAM,MAAM,GAAG,UAAU,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;YAC1C,MAAM,OAAO,GAAmB;gBAC/B,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC;gBACjB,QAAQ,EAAE,OAAO,CAAC,QAAQ,CAAC,WAAW,EAAE;gBACxC,SAAS,EAAE,OAAO,CAAC,OAAO;aAC1B,CAAC;YACF,OAAO,cAAc,CAAC,OAAO,EAAE;gBAC9B,MAAM;gBACN,MAAM,EAAE,aAAa,CAAC,MAAM;aAC5B,CAAC,CAAC;QACJ,CAAC,CAAC,CAAC;IACJ,CAAC,CAAC,CAAC;IAEJ,OAAO;SACL,OAAO,CAAC,WAAW,CAAC;SACpB,WAAW,CAAC,kCAAkC,CAAC;SAC/C,MAAM,CAAC,GAAG,EAAE,CACZ,GAAG,CAAC,GAAG,EAAE,CACR,gBAAgB,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,EAAiB,CAAC,MAAM,CAAC,CAAC,CAClE,CACD,CAAC;IAEH,OAAO;SACL,OAAO,CAAC,QAAQ,CAAC;SACjB,WAAW,CAAC,0BAA0B,CAAC;SACvC,MAAM,CAAC,GAAG,EAAE,CACZ,GAAG,CAAC,GAAG,EAAE,CACR,aAAa,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,EAAiB,CAAC,MAAM,CAAC,CAAC,CAC/D,CACD,CAAC;IAEH,MAAM,MAAM,GAAG,OAAO,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,CAAC,QAAQ,CAAC,CAAC;IAC/D,MAAM;SACJ,OAAO,CAAC,MAAM,CAAC;SACf,WAAW,CAAC,oBAAoB,CAAC;SACjC,MAAM,CAAC,GAAG,EAAE,CACZ,GAAG,CAAC,GAAG,EAAE,CACR,aAAa,CAAC;QACb,YAAY,EAAE,OAAO,CAAC,IAAI,EAAiB,CAAC,MAAM;KAClD,CAAC,CACF,CACD,CAAC;IAEH,OAAO,OAAO,CAAC;AAChB,CAAC;AAED,SAAS,eAAe,CAAC,IAAc;IACtC,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC1C,IAAI,QAAQ,CAAC,QAAQ,CAAC,MAAyB,CAAC;QAC/C,OAAO,MAAyB,CAAC;IAClC,MAAM,OAAO,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;IACxB,OAAO,QAAQ,CAAC,QAAQ,CAAC,OAA0B,CAAC;QACnD,CAAC,CAAE,OAA2B;QAC9B,CAAC,CAAC,IAAI,CAAC;AACT,CAAC;AAED,MAAM,aAAa,GAAG,IAAI,eAAe,EAAE,CAAC;AAC5C,OAAO,CAAC,IAAI,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,aAAa,CAAC,KAAK,EAAE,CAAC,CAAC;AAEpD,KAAK,UAAU,IAAI;IAClB,IAAI,OAA4C,CAAC;IACjD,MAAM,OAAO,GAAG,aAAa,CAAC,CAAC,IAAI,EAAE,EAAE;QACtC,OAAO,GAAG,OAAO,CAAC,OAAO,EAAE,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACxC,CAAC,CAAC,CAAC;IACH,IAAI,OAAO,CAAC,IAAI,CAAC,MAAM,IAAI,CAAC,EAAE,CAAC;QAC9B,MAAM,QAAQ,GAAG,aAAa,CAC7B,IAAI,cAAc,CAAC,eAAe,EAAE,QAAQ,CAAC,CAC7C,CAAC;QACF,aAAa,CAAC,QAAQ,EAAE,KAAK,CAAC,CAAC;QAC/B,OAAO,CAAC,QAAQ,GAAG,CAAC,CAAC;QACrB,OAAO;IACR,CAAC;IACD,IAAI,CAAC;QACJ,MAAM,OAAO,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;QACvC,IAAI,CAAC,OAAO;YAAE,MAAM,IAAI,cAAc,CAAC,eAAe,EAAE,QAAQ,CAAC,CAAC;QAClE,MAAM,QAAQ,GAAG,MAAM,OAAO,CAAC;QAC/B,aAAa,CAAC,QAAQ,EAAE,CAAC,CAAC,OAAO,CAAC,IAAI,EAAiB,CAAC,MAAM,CAAC,CAAC;QAChE,OAAO,CAAC,QAAQ,GAAG,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACvC,CAAC;IAAC,OAAO,MAAM,EAAE,CAAC;QACjB,IACC,MAAM,YAAY,cAAc;YAChC,MAAM,CAAC,IAAI,KAAK,yBAAyB;YAEzC,OAAO;QACR,IAAI,MAAM,YAAY,cAAc,IAAI,MAAM,CAAC,IAAI,KAAK,mBAAmB;YAC1E,OAAO;QACR,MAAM,UAAU,GACf,MAAM,YAAY,cAAc;YAC/B,CAAC,CAAC,IAAI,cAAc,CAClB,eAAe,EACf,MAAM,CAAC,IAAI,KAAK,gBAAgB,IAAI,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,KAAK,QAAQ;gBAC/D,CAAC,CAAC,iBAAiB;gBACnB,CAAC,CAAC,MAAM,CAAC,OAAO,CACjB;YACF,CAAC,CAAC,gBAAgB,CAAC,MAAM,CAAC,CAAC;QAC7B,MAAM,QAAQ,GAAG,aAAa,CAAC,UAAU,EAAE,eAAe,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC;QAC1E,aAAa,CAAC,QAAQ,EAAE,CAAC,CAAC,OAAO,CAAC,IAAI,EAAiB,CAAC,MAAM,CAAC,CAAC;QAChE,OAAO,CAAC,QAAQ,GAAG,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACvC,CAAC;AACF,CAAC;AAED,MAAM,UAAU,YAAY,CAC3B,SAAiB,EACjB,SAA6B;IAE7B,IAAI,CAAC,SAAS;QAAE,OAAO,KAAK,CAAC;IAC7B,IAAI,CAAC;QACJ,OAAO,SAAS,KAAK,aAAa,CAAC,YAAY,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI,CAAC;IAClE,CAAC;IAAC,MAAM,CAAC;QACR,OAAO,KAAK,CAAC;IACd,CAAC;AACF,CAAC;AAED,IAAI,YAAY,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,EAAE,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;IACpD,MAAM,IAAI,EAAE,CAAC;AACd,CAAC"}

package/dist/config/config.js

@@ -0,0 +1,463 @@
+import { existsSync, readFileSync } from "node:fs";
+import { homedir } from "node:os";
+import { isAbsolute, join, resolve } from "node:path";
+import { WebAccessError } from "../core/errors.js";
+import { PROVIDER_TYPES, SEARCH_FILTER_MODES, } from "../core/types.js";
+export const CONFIG_ENV = "WEB_ACCESS_CONFIG";
+export const DEFAULT_SEARCH_PROVIDERS = [
+    "tavily",
+    "exa",
+    "brave",
+    "searxng",
+    "anysearch",
+    "xcrawl",
+];
+export const DEFAULT_EXTRACT_PROVIDERS = [
+    "firecrawl",
+    "jina",
+    "exa",
+    "anysearch",
+    "xcrawl",
+    "http",
+];
+export const DEFAULT_MAX_RESPONSE_BYTES = 5 * 1024 * 1024;
+const DEFAULT_SEARCH_CONFIG = {
+    providers: [...DEFAULT_SEARCH_PROVIDERS],
+    limit: 5,
+    timeoutMs: 60_000,
+    attemptTimeoutMs: 20_000,
+    maxResponseBytes: DEFAULT_MAX_RESPONSE_BYTES,
+};
+const DEFAULT_EXTRACT_CONFIG = {
+    providers: [...DEFAULT_EXTRACT_PROVIDERS],
+    timeoutMs: 120_000,
+    attemptTimeoutMs: 45_000,
+    maxResponseBytes: DEFAULT_MAX_RESPONSE_BYTES,
+    minContentCharacters: 500,
+};
+const SEARCH_TYPES = new Set([
+    "tavily",
+    "exa",
+    "brave",
+    "searxng",
+    "anysearch",
+    "xcrawl",
+]);
+const EXTRACT_TYPES = new Set([
+    "firecrawl",
+    "jina",
+    "exa",
+    "http",
+    "anysearch",
+    "xcrawl",
+]);
+const DEFAULT_INSTANCE_CONFIGS = [
+    { id: "tavily", type: "tavily" },
+    { id: "exa", type: "exa" },
+    { id: "brave", type: "brave" },
+    { id: "searxng", type: "searxng" },
+    { id: "firecrawl", type: "firecrawl" },
+    { id: "jina", type: "jina" },
+    { id: "http", type: "http" },
+    { id: "anysearch", type: "anysearch", searchFilterMode: "strict" },
+    { id: "xcrawl", type: "xcrawl", searchFilterMode: "strict" },
+];
+const STANDARD_KEY_ENV = {
+    tavily: "TAVILY_API_KEY",
+    exa: "EXA_API_KEY",
+    brave: "BRAVE_API_KEY",
+    firecrawl: "FIRECRAWL_API_KEY",
+    jina: "JINA_API_KEY",
+    anysearch: "ANYSEARCH_API_KEY",
+    xcrawl: "XCRAWL_API_KEY",
+};
+const STANDARD_BASE_ENV = {
+    tavily: "TAVILY_BASE_URL",
+    exa: "EXA_BASE_URL",
+    brave: "BRAVE_BASE_URL",
+    searxng: "SEARXNG_BASE_URL",
+    firecrawl: "FIRECRAWL_BASE_URL",
+    jina: "JINA_BASE_URL",
+    anysearch: "ANYSEARCH_BASE_URL",
+    xcrawl: "XCRAWL_BASE_URL",
+};
+const SEARCH_FILTER_TYPES = new Set(["anysearch", "xcrawl"]);
+const ALLOWED_INSTANCE_KEYS = new Set([
+    "id",
+    "type",
+    "apiKey",
+    "apiKeyEnv",
+    "baseUrl",
+    "baseUrlEnv",
+    "headers",
+    "searchFilterMode",
+]);
+const ALLOWED_SEARCH_KEYS = new Set([
+    "providers",
+    "limit",
+    "timeoutMs",
+    "attemptTimeoutMs",
+    "maxResponseBytes",
+]);
+const ALLOWED_EXTRACT_KEYS = new Set([
+    "providers",
+    "timeoutMs",
+    "attemptTimeoutMs",
+    "maxResponseBytes",
+    "minContentCharacters",
+]);
+export class ConfigError extends WebAccessError {
+    constructor(message, details) {
+        super("config_error", message, { details });
+        this.name = "ConfigError";
+    }
+}
+function isRecord(value) {
+    return !!value && typeof value === "object" && !Array.isArray(value);
+}
+function assertRecord(value, path) {
+    if (!isRecord(value))
+        throw new ConfigError(`${path} 必须是 JSON 对象`);
+}
+function assertKnownKeys(value, allowed, path) {
+    for (const key of Object.keys(value)) {
+        if (!allowed.has(key))
+            throw new ConfigError(`${path}.${key} 不是受支持的配置字段`);
+    }
+}
+function parsePositiveInt(value, path, defaultValue) {
+    if (value === undefined)
+        return defaultValue;
+    if (!Number.isInteger(value) || value < 1)
+        throw new ConfigError(`${path} 必须是大于 0 的整数`);
+    return value;
+}
+function parseLimit(value) {
+    const limit = parsePositiveInt(value, "search.limit", DEFAULT_SEARCH_CONFIG.limit);
+    if (limit > 20)
+        throw new ConfigError("search.limit 不能超过 20");
+    return limit;
+}
+function parseString(value, path) {
+    if (typeof value !== "string" || value.trim() === "")
+        throw new ConfigError(`${path} 必须是非空字符串`);
+    return value.trim();
+}
+function parseOptionalString(value, path) {
+    if (value === undefined)
+        return undefined;
+    return parseString(value, path);
+}
+function validateEnvName(value, path) {
+    if (!/^[A-Za-z_][A-Za-z0-9_]*$/.test(value))
+        throw new ConfigError(`${path} 不是合法环境变量名`);
+    return value;
+}
+function normalizeHeaders(value, path) {
+    if (value === undefined)
+        return undefined;
+    assertRecord(value, path);
+    const headers = {};
+    for (const [name, headerValue] of Object.entries(value)) {
+        if (!name.trim() || /[\r\n]/.test(name))
+            throw new ConfigError(`${path} 包含非法 header 名称`);
+        if (typeof headerValue !== "string" || /[\r\n]/.test(headerValue))
+            throw new ConfigError(`${path}.${name} 必须是无换行字符串`);
+        headers[name] = headerValue;
+    }
+    return headers;
+}
+function normalizeBaseUrl(value, path) {
+    let url;
+    try {
+        url = new URL(value);
+    }
+    catch {
+        throw new ConfigError(`${path} 必须是 HTTP(S) URL`);
+    }
+    if (url.protocol !== "http:" && url.protocol !== "https:")
+        throw new ConfigError(`${path} 必须使用 http 或 https`);
+    if (url.username || url.password)
+        throw new ConfigError(`${path} 不允许包含 URL 用户名或密码`);
+    url.search = "";
+    url.hash = "";
+    url.pathname = url.pathname.replace(/\/+$/, "");
+    return url.toString().replace(/\/$/, "");
+}
+function parseInstance(value, index) {
+    const path = `providers[${index}]`;
+    assertRecord(value, path);
+    assertKnownKeys(value, ALLOWED_INSTANCE_KEYS, path);
+    const id = parseString(value.id, `${path}.id`).toLowerCase();
+    if (!/^[a-z][a-z0-9_-]{0,63}$/.test(id) || id === "auto")
+        throw new ConfigError(`${path}.id 不是合法 instance id`);
+    const type = parseString(value.type, `${path}.type`).toLowerCase();
+    if (!PROVIDER_TYPES.includes(type))
+        throw new ConfigError(`${path}.type 不受支持: ${type}`);
+    const apiKey = parseOptionalString(value.apiKey, `${path}.apiKey`);
+    const apiKeyEnvRaw = parseOptionalString(value.apiKeyEnv, `${path}.apiKeyEnv`);
+    const baseUrlRaw = parseOptionalString(value.baseUrl, `${path}.baseUrl`);
+    const baseUrl = baseUrlRaw
+        ? normalizeBaseUrl(baseUrlRaw, `${path}.baseUrl`)
+        : undefined;
+    const baseUrlEnvRaw = parseOptionalString(value.baseUrlEnv, `${path}.baseUrlEnv`);
+    const headers = normalizeHeaders(value.headers, `${path}.headers`);
+    const searchFilterMode = value.searchFilterMode === undefined
+        ? undefined
+        : parseString(value.searchFilterMode, `${path}.searchFilterMode`).toLowerCase();
+    if (searchFilterMode !== undefined &&
+        !SEARCH_FILTER_MODES.includes(searchFilterMode))
+        throw new ConfigError(`${path}.searchFilterMode 必须是 strict 或 best_effort`);
+    if (searchFilterMode !== undefined &&
+        !SEARCH_FILTER_TYPES.has(type))
+        throw new ConfigError(`${path}.searchFilterMode 仅适用于 anysearch 或 xcrawl`);
+    return {
+        id,
+        type: type,
+        ...(apiKey ? { apiKey } : {}),
+        ...(apiKeyEnvRaw
+            ? { apiKeyEnv: validateEnvName(apiKeyEnvRaw, `${path}.apiKeyEnv`) }
+            : {}),
+        ...(baseUrl ? { baseUrl } : {}),
+        ...(baseUrlEnvRaw
+            ? { baseUrlEnv: validateEnvName(baseUrlEnvRaw, `${path}.baseUrlEnv`) }
+            : {}),
+        ...(headers ? { headers } : {}),
+        ...(searchFilterMode
+            ? { searchFilterMode: searchFilterMode }
+            : {}),
+    };
+}
+function parseRoute(value, path) {
+    if (value === undefined)
+        return undefined;
+    if (!Array.isArray(value))
+        throw new ConfigError(`${path} 必须是 instance id 数组`);
+    const result = [];
+    for (const [index, item] of value.entries()) {
+        const id = parseString(item, `${path}[${index}]`).toLowerCase();
+        if (id === "auto" || result.includes(id))
+            throw new ConfigError(`${path} 包含重复或保留 id: ${id}`);
+        result.push(id);
+    }
+    return result;
+}
+function parseSearch(value) {
+    if (value === undefined)
+        return {};
+    assertRecord(value, "search");
+    assertKnownKeys(value, ALLOWED_SEARCH_KEYS, "search");
+    return {
+        providers: parseRoute(value.providers, "search.providers"),
+        limit: parseLimit(value.limit),
+        timeoutMs: parsePositiveInt(value.timeoutMs, "search.timeoutMs", DEFAULT_SEARCH_CONFIG.timeoutMs),
+        attemptTimeoutMs: parsePositiveInt(value.attemptTimeoutMs, "search.attemptTimeoutMs", DEFAULT_SEARCH_CONFIG.attemptTimeoutMs),
+        maxResponseBytes: parsePositiveInt(value.maxResponseBytes, "search.maxResponseBytes", DEFAULT_SEARCH_CONFIG.maxResponseBytes),
+    };
+}
+function parseExtract(value) {
+    if (value === undefined)
+        return {};
+    assertRecord(value, "extract");
+    assertKnownKeys(value, ALLOWED_EXTRACT_KEYS, "extract");
+    return {
+        providers: parseRoute(value.providers, "extract.providers"),
+        timeoutMs: parsePositiveInt(value.timeoutMs, "extract.timeoutMs", DEFAULT_EXTRACT_CONFIG.timeoutMs),
+        attemptTimeoutMs: parsePositiveInt(value.attemptTimeoutMs, "extract.attemptTimeoutMs", DEFAULT_EXTRACT_CONFIG.attemptTimeoutMs),
+        maxResponseBytes: parsePositiveInt(value.maxResponseBytes, "extract.maxResponseBytes", DEFAULT_EXTRACT_CONFIG.maxResponseBytes),
+        minContentCharacters: parsePositiveInt(value.minContentCharacters, "extract.minContentCharacters", DEFAULT_EXTRACT_CONFIG.minContentCharacters),
+    };
+}
+function parseRawConfig(value) {
+    assertRecord(value, "config");
+    assertKnownKeys(value, new Set(["$schema", "providers", "search", "extract"]), "config");
+    if (value.$schema !== undefined && typeof value.$schema !== "string")
+        throw new ConfigError("$schema 必须是字符串");
+    const rawInstances = value.providers === undefined ? [] : value.providers;
+    if (!Array.isArray(rawInstances))
+        throw new ConfigError("providers 必须是数组");
+    const instances = rawInstances.map(parseInstance);
+    const ids = new Set();
+    for (const instance of instances) {
+        if (ids.has(instance.id))
+            throw new ConfigError(`providers 中存在重复 id: ${instance.id}`);
+        ids.add(instance.id);
+    }
+    return {
+        instances,
+        search: parseSearch(value.search),
+        extract: parseExtract(value.extract),
+    };
+}
+function mergeInstances(overrides) {
+    const merged = DEFAULT_INSTANCE_CONFIGS.map((instance) => ({ ...instance }));
+    const byId = new Map(merged.map((instance) => [instance.id, instance]));
+    for (const override of overrides) {
+        const existing = byId.get(override.id);
+        if (existing) {
+            if (existing.type !== override.type)
+                throw new ConfigError(`默认 instance ${override.id} 的 type 不能改为 ${override.type}`);
+            Object.assign(existing, override);
+        }
+        else {
+            const copy = { ...override };
+            merged.push(copy);
+            byId.set(copy.id, copy);
+        }
+    }
+    return merged;
+}
+function validateRoutes(instances, route, capability) {
+    const allowed = capability === "search" ? SEARCH_TYPES : EXTRACT_TYPES;
+    const byId = new Map(instances.map((instance) => [instance.id, instance]));
+    const chosen = route ??
+        (capability === "search"
+            ? [...DEFAULT_SEARCH_PROVIDERS]
+            : [...DEFAULT_EXTRACT_PROVIDERS]);
+    for (const id of chosen) {
+        const instance = byId.get(id);
+        if (!instance)
+            throw new ConfigError(`${capability}.providers 引用了不存在的 instance: ${id}`);
+        if (!allowed.has(instance.type))
+            throw new ConfigError(`${capability}.providers 的 instance ${id} 不支持 ${capability}`);
+    }
+    return chosen;
+}
+export function getDefaultConfigPath() {
+    return join(homedir(), ".config", "web-access-cli", "config.json");
+}
+export function resolveConfigPath(explicitPath, env = process.env) {
+    const raw = explicitPath || env[CONFIG_ENV] || getDefaultConfigPath();
+    return isAbsolute(raw) ? raw : resolve(raw);
+}
+function resolveEnvValue(name, fallback, env) {
+    const custom = name ? env[name]?.trim() : undefined;
+    if (custom)
+        return { value: custom, source: "custom_env" };
+    const standard = fallback ? env[fallback]?.trim() : undefined;
+    if (standard)
+        return { value: standard, source: "standard_env" };
+    return { value: null, source: "missing" };
+}
+function resolveProvider(instance, env) {
+    const standardKeyEnv = instance.id === instance.type ? STANDARD_KEY_ENV[instance.type] : undefined;
+    const keyEnv = resolveEnvValue(instance.apiKeyEnv, standardKeyEnv, env);
+    const apiKey = keyEnv.value ?? instance.apiKey ?? null;
+    const credentialSource = apiKey
+        ? keyEnv.value
+            ? keyEnv.source
+            : "config"
+        : "missing";
+    const standardBaseEnv = STANDARD_BASE_ENV[instance.type];
+    const baseEnv = resolveEnvValue(instance.baseUrlEnv, instance.id === instance.type ? standardBaseEnv : undefined, env);
+    const rawBaseUrl = baseEnv.value ??
+        instance.baseUrl ??
+        DEFAULT_BASE_URLS[instance.type] ??
+        null;
+    const baseUrl = rawBaseUrl
+        ? normalizeBaseUrl(rawBaseUrl, baseEnv.value
+            ? (instance.baseUrlEnv ??
+                STANDARD_BASE_ENV[instance.type] ??
+                "baseUrl 环境变量")
+            : `providers.${instance.id}.baseUrl`)
+        : null;
+    const baseUrlSource = baseEnv.value
+        ? baseEnv.source
+        : instance.baseUrl
+            ? "config"
+            : DEFAULT_BASE_URLS[instance.type]
+                ? "default"
+                : "missing";
+    return {
+        id: instance.id,
+        type: instance.type,
+        apiKey,
+        credentialSource,
+        baseUrl,
+        baseUrlSource,
+        headers: { ...(instance.headers ?? {}) },
+        searchFilterMode: instance.searchFilterMode ??
+            (SEARCH_FILTER_TYPES.has(instance.type) ? "strict" : null),
+    };
+}
+const DEFAULT_BASE_URLS = {
+    tavily: "https://api.tavily.com",
+    exa: "https://api.exa.ai",
+    brave: "https://api.search.brave.com",
+    firecrawl: "https://api.firecrawl.dev",
+    jina: "https://r.jina.ai",
+    anysearch: "https://api.anysearch.com",
+    xcrawl: "https://run.xcrawl.com",
+};
+export function createDefaultAppConfig() {
+    return {
+        providers: DEFAULT_INSTANCE_CONFIGS.map((instance) => ({ ...instance })),
+        search: {
+            ...DEFAULT_SEARCH_CONFIG,
+            providers: [...DEFAULT_SEARCH_CONFIG.providers],
+        },
+        extract: {
+            ...DEFAULT_EXTRACT_CONFIG,
+            providers: [...DEFAULT_EXTRACT_CONFIG.providers],
+        },
+    };
+}
+export function loadConfig(explicitPath, env = process.env) {
+    const path = resolveConfigPath(explicitPath, env);
+    const defaults = createDefaultAppConfig();
+    let parsed = {};
+    const exists = existsSync(path);
+    if (exists) {
+        try {
+            parsed = JSON.parse(readFileSync(path, "utf8"));
+        }
+        catch (error) {
+            throw new ConfigError(`无法解析 ${path}: ${error instanceof Error ? error.message : String(error)}`);
+        }
+    }
+    else if (explicitPath || env[CONFIG_ENV]) {
+        throw new ConfigError(`配置文件不存在: ${path}`);
+    }
+    const raw = parseRawConfig(parsed);
+    const instanceConfigs = mergeInstances(raw.instances);
+    const searchProviders = validateRoutes(instanceConfigs, raw.search.providers, "search");
+    const extractProviders = validateRoutes(instanceConfigs, raw.extract.providers, "extract");
+    const search = {
+        providers: searchProviders,
+        limit: raw.search.limit ?? defaults.search.limit,
+        timeoutMs: raw.search.timeoutMs ?? defaults.search.timeoutMs,
+        attemptTimeoutMs: raw.search.attemptTimeoutMs ?? defaults.search.attemptTimeoutMs,
+        maxResponseBytes: raw.search.maxResponseBytes ?? defaults.search.maxResponseBytes,
+    };
+    const extract = {
+        providers: extractProviders,
+        timeoutMs: raw.extract.timeoutMs ?? defaults.extract.timeoutMs,
+        attemptTimeoutMs: raw.extract.attemptTimeoutMs ?? defaults.extract.attemptTimeoutMs,
+        maxResponseBytes: raw.extract.maxResponseBytes ?? defaults.extract.maxResponseBytes,
+        minContentCharacters: raw.extract.minContentCharacters ?? defaults.extract.minContentCharacters,
+    };
+    const app = { providers: instanceConfigs, search, extract };
+    return {
+        path,
+        exists,
+        app,
+        instances: instanceConfigs.map((instance) => resolveProvider(instance, env)),
+    };
+}
+export function capabilitySupports(type, capability) {
+    return capability === "search"
+        ? SEARCH_TYPES.has(type)
+        : EXTRACT_TYPES.has(type);
+}
+export function getRoute(config, capability) {
+    return capability === "search"
+        ? config.search.providers
+        : config.extract.providers;
+}
+export function getCapabilityConfig(config, capability) {
+    return capability === "search" ? config.search : config.extract;
+}
+export function providerRef(instance) {
+    return { id: instance.id, type: instance.type };
+}
+//# sourceMappingURL=config.js.map