feat: prepare docs for langbot.app subpath

package-lock.json

@@ -1,10 +1,10 @@
 {
-  "name": "langbot-wiki",
+  "name": "langbot-docs",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
-      "name": "langbot-wiki",
+      "name": "langbot-docs",
       "dependencies": {
         "@base-ui/react": "^1.8.0",
         "@fumapress/mintlify": "^1.1.2",

package.json

@@ -1,5 +1,5 @@
 {
-  "name": "langbot-wiki",
+  "name": "langbot-docs",
   "private": true,
   "type": "module",
   "engines": {
@@ -11,14 +11,15 @@
     "prepare:fumapress": "node scripts/prepare-fumapress.mjs",
     "build": "fumapress build",
     "prebuild": "npm run prepare:fumapress",
-    "postbuild": "npm run generate:sitemap-alternates && npm run normalize:static-paths && npm run test:static",
+    "postbuild": "npm run finalize:base-path && npm run generate:sitemap-alternates && npm run normalize:static-paths && npm run test:static",
     "typecheck": "tsc --noEmit",
     "normalize:static-paths": "node scripts/normalize-static-paths.mjs",
     "test:static": "node --test tests/fumapress-static-build.test.mjs",
     "generate:sitemap-alternates": "node scripts/generate-alternate-sitemap.mjs",
     "sync:articles": "node scripts/sync-blog-articles.mjs",
     "check:articles": "node scripts/sync-blog-articles.mjs --check",
-    "test": "node --test tests/blog-articles.test.mjs tests/fumapress-migration.test.mjs tests/seo.test.mjs && python3 -m unittest discover -s tests -p 'test_*.py'"
+    "test": "node --test tests/blog-articles.test.mjs tests/fumapress-migration.test.mjs tests/seo.test.mjs && python3 -m unittest discover -s tests -p 'test_*.py'",
+    "finalize:base-path": "node scripts/finalize-base-path.mjs"
   },
   "dependencies": {
     "@base-ui/react": "^1.8.0",

press.config.tsx

@@ -13,7 +13,7 @@ import { Bot } from "lucide-react";
 import { readFileSync, readdirSync } from "node:fs";
 import path from "node:path";
 
-const SITE_URL = "https://docs.langbot.dev";
+const SITE_URL = "https://langbot.app/docs";
 const LOCALES = ["en", "zh", "ja"] as const;
 const MINTLIFY_LOCALE = { en: "en", zh: "cn", ja: "jp" } as const;
 const HREFLANG = { en: "en", zh: "zh-CN", ja: "ja" } as const;
@@ -405,6 +405,6 @@ export default defineConfig({
     robotsPlugin({
       rules: [{ userAgent: "*", allow: ["/", "/_next/"] }],
       sitemap: true,
-      additionalContent: "Sitemap: https://docs.langbot.dev/sitemap-alternates.xml",
+      additionalContent: "Sitemap: https://langbot.app/docs/sitemap-alternates.xml",
     }),
   );

robots.txt

@@ -1,6 +1,6 @@
 User-agent: *
-Allow: /
+Allow: /docs/
 Allow: /_next/
 
-Sitemap: https://docs.langbot.dev/sitemap.xml
-Sitemap: https://docs.langbot.dev/sitemap-alternates.xml
+Sitemap: https://langbot.app/docs/sitemap.xml
+Sitemap: https://langbot.app/docs/sitemap-alternates.xml

scripts/finalize-base-path.mjs

@@ -0,0 +1,82 @@
+#!/usr/bin/env node
+
+import { readFile, readdir, writeFile } from "node:fs/promises";
+import path from "node:path";
+import { fileURLToPath } from "node:url";
+
+const ROOT = path.resolve(path.dirname(fileURLToPath(import.meta.url)), "..");
+const PUBLIC_ROOT = path.join(ROOT, "dist/public");
+const ORIGIN = "https://langbot.app";
+const BASE_PATH = "/docs";
+const TEXT_EXTENSIONS = new Set([".html", ".md", ".txt", ".xml", ".json", ".js", ".css"]);
+
+async function collectTextFiles(directory) {
+  const files = [];
+  for (const entry of await readdir(directory, { withFileTypes: true })) {
+    const target = path.join(directory, entry.name);
+    if (entry.isDirectory()) files.push(...await collectTextFiles(target));
+    else if (TEXT_EXTENSIONS.has(path.extname(entry.name)) || target.endsWith("/api/search")) files.push(target);
+  }
+  return files;
+}
+
+function prefixPublicAssets(content) {
+  return content
+    .replaceAll('"/images/', `"${BASE_PATH}/images/`)
+    .replaceAll("'/images/", `'${BASE_PATH}/images/`)
+    .replaceAll('"/favicon.ico"', `"${BASE_PATH}/favicon.ico"`)
+    .replaceAll("'/favicon.ico'", `'${BASE_PATH}/favicon.ico'`)
+    .replaceAll('"/rss.xml"', `"${BASE_PATH}/rss.xml"`)
+    .replaceAll("'/rss.xml'", `'${BASE_PATH}/rss.xml'`);
+}
+
+export async function finalizeBasePath(publicRoot = PUBLIC_ROOT) {
+  const sitemapPath = path.join(publicRoot, "sitemap.xml");
+  const originalSitemap = await readFile(sitemapPath, "utf8");
+  const sourceUrls = [...originalSitemap.matchAll(/<loc>(https:\/\/langbot\.app\/[^<]+)<\/loc>/g)]
+    .map((match) => match[1].replace(/\/$/, ""));
+  const replacements = sourceUrls
+    .map((source) => [source, `${ORIGIN}${BASE_PATH}${source.slice(ORIGIN.length)}`])
+    .sort((left, right) => right[0].length - left[0].length);
+
+  let changed = 0;
+  for (const file of await collectTextFiles(publicRoot)) {
+    let content = await readFile(file, "utf8");
+    const before = content;
+    for (const [source, destination] of replacements) content = content.replaceAll(source, destination);
+    content = content
+      .replaceAll(`${ORIGIN}/sitemap.xml`, `${ORIGIN}${BASE_PATH}/sitemap.xml`)
+      .replaceAll(`${ORIGIN}/sitemap-alternates.xml`, `${ORIGIN}${BASE_PATH}/sitemap-alternates.xml`);
+    content = prefixPublicAssets(content);
+    if (content !== before) {
+      await writeFile(file, content);
+      changed += 1;
+    }
+  }
+
+  const redirectsPath = path.join(publicRoot, "_redirects");
+  const redirects = await readFile(redirectsPath, "utf8");
+  const scopedRedirects = redirects
+    .split("\n")
+    .map((line) => {
+      if (!line.trim()) return line;
+      const [source, destination, status, ...rest] = line.trim().split(/\s+/);
+      const scope = (value) => value.startsWith("/") ? `${BASE_PATH}${value === "/" ? "" : value}` : value;
+      return [scope(source), scope(destination), status, ...rest].filter(Boolean).join(" ");
+    })
+    .join("\n");
+  await writeFile(redirectsPath, scopedRedirects);
+  return { changed, routes: replacements.length };
+}
+
+async function main() {
+  const result = await finalizeBasePath();
+  console.log(`Finalized ${result.routes} documentation routes under ${BASE_PATH} across ${result.changed} files`);
+}
+
+if (process.argv[1] && path.resolve(process.argv[1]) === fileURLToPath(import.meta.url)) {
+  main().catch((error) => {
+    console.error(error.stack || error.message);
+    process.exitCode = 1;
+  });
+}

scripts/generate-alternate-sitemap.mjs

@@ -5,7 +5,7 @@ import path from "node:path";
 import { fileURLToPath } from "node:url";
 
 const ROOT = path.resolve(path.dirname(fileURLToPath(import.meta.url)), "..");
-const BASE = "https://docs.langbot.dev";
+const BASE = "https://langbot.app/docs";
 const LOCALES = ["en", "zh", "ja"];
 const HREFLANG = { en: "en", zh: "zh-CN", ja: "ja" };
 

tests/fumapress-static-build.test.mjs

@@ -38,7 +38,7 @@ function staticHtmlForCanonicalDocument(document) {
 }
 
 function staticHtmlForUrlPath(urlPath) {
-  const parts = urlPath.replace(/^\//, "").replace(/\/$/, "").split("/")
+  const parts = urlPath.replace(/^\/docs(?:\/|$)/, "").replace(/^\//, "").replace(/\/$/, "").split("/")
     .map((segment) => decodeURIComponent(segment));
   return path.join(publicRoot, ...parts, "index.html");
 }
@@ -90,9 +90,9 @@ test("representative localized guides and canonical host are emitted", async ()
   }
   const sitemap = await readFile(path.join(publicRoot, "sitemap.xml"), "utf8");
   const robots = await readFile(path.join(publicRoot, "robots.txt"), "utf8");
-  assert.match(sitemap, /https:\/\/docs\.langbot\.dev\//);
+  assert.match(sitemap, /https:\/\/langbot\.app\/docs\//);
   assert.doesNotMatch(sitemap, /https:\/\/docs\.langbot\.app\//);
-  assert.match(robots, /https:\/\/docs\.langbot\.dev\/sitemap\.xml/);
+  assert.match(robots, /https:\/\/langbot\.app\/docs\/sitemap\.xml/);
   assert.doesNotMatch(robots, /https:\/\/docs\.langbot\.app\//);
 });
 
@@ -107,7 +107,7 @@ test("platform navigation renders logo assets instead of icon path text", async
     const sidebar = html.slice(asideStart, asideEnd);
     assert.match(
       sidebar,
-      /<img[^>]+src="\/images\/platforms\/discord\.svg"[^>]*>/,
+      /<img[^>]+src="\/docs\/images\/platforms\/discord\.svg"[^>]*>/,
       `${locale} sidebar is missing the Discord logo`,
     );
     assert.match(sidebar, />Discord</, `${locale} sidebar is missing the platform title`);
@@ -137,7 +137,7 @@ test("nested bot groups preserve the logos declared in docs.json", async () => {
     const escapedLogo = logo.replace(".", "\\.");
     assert.match(
       sidebar,
-      new RegExp(`<img[^>]+src="/images/platforms/${escapedLogo}"[^>]*>\\s*${escapedTitle}<svg`),
+      new RegExp(`<img[^>]+src="/docs/images/platforms/${escapedLogo}"[^>]*>\\s*${escapedTitle}<svg`),
       `zh sidebar is missing the ${logo} logo for ${title}`,
     );
   }
@@ -170,7 +170,7 @@ test("all locales emit the OpenAPI surface with endpoint semantics", async () =>
     const representative = path.join(apiRoot, representativePath, "index.html");
     const html = await readFile(representative, "utf8");
     for (const semantic of semantics) assert.ok(html.includes(semantic), `${locale} page is missing ${semantic}`);
-    assert.ok(sitemap.includes(`https://docs.langbot.dev/${locale}/api-reference/${encodedRepresentativePath}`));
+    assert.ok(sitemap.includes(`https://langbot.app/docs/${locale}/api-reference/${encodedRepresentativePath}`));
 
     const spec = JSON.parse(await readFile(path.join(root, `openapi/service-api-${locale}.json`), "utf8"));
     const tagOrder = [];
@@ -200,7 +200,7 @@ test("canonical sitemap routes match the legacy Mintlify directory structure", a
   // its URL as a redirect without treating it as public documentation.
   legacy.delete("/scripts/README-blog-articles");
   const sitemap = await readFile(path.join(publicRoot, "sitemap.xml"), "utf8");
-  const actual = new Set([...sitemap.matchAll(/<loc>https:\/\/docs\.langbot\.dev(\/[^<]+)<\/loc>/g)]
+  const actual = new Set([...sitemap.matchAll(/<loc>https:\/\/langbot\.app\/docs(\/[^<]+)<\/loc>/g)]
     .map((match) => decodeURIComponent(match[1]).replace(/\/$/, "")));
   assert.deepEqual([...actual].sort(), [...legacy].sort());
 });
@@ -321,7 +321,7 @@ test("every legacy navigation entry resolves to its exact sidebar URL", async ()
       const asideStart = html.indexOf('<aside id="nd-sidebar"');
       const asideEnd = html.indexOf("</aside>", asideStart);
       const sidebar = html.slice(asideStart, asideEnd);
-      if (!sidebar.includes(`href="/${route}"`)) failures.push({ page, htmlPath });
+      if (!sidebar.includes(`href="/docs/${route}"`)) failures.push({ page, htmlPath });
     }
   }
   assert.deepEqual(failures, []);
@@ -332,8 +332,8 @@ test("the Chinese quick-start navigation keeps the incumbent troubleshooting pag
   const asideStart = html.indexOf('<aside id="nd-sidebar"');
   const asideEnd = html.indexOf("</aside>", asideStart);
   const sidebar = html.slice(asideStart, asideEnd);
-  assert.match(sidebar, /href="\/zh\/insight\/troubleshooting"/);
-  assert.doesNotMatch(sidebar, /href="\/zh\/develop\/adapter\/discord\/troubleshooting"/);
+  assert.match(sidebar, /href="\/docs\/zh\/insight\/troubleshooting"/);
+  assert.doesNotMatch(sidebar, /href="\/docs\/zh\/develop\/adapter\/discord\/troubleshooting"/);
 });
 
 test("the exact 14 zh-only sources publish no fallback en or ja routes", async () => {
@@ -366,14 +366,14 @@ test("the exact 14 zh-only sources publish no fallback en or ja routes", async (
 
 test("every page emits a canonical and reciprocal available hreflangs", async () => {
   const sitemap = await readFile(path.join(publicRoot, "sitemap.xml"), "utf8");
-  const routes = new Set([...sitemap.matchAll(/<loc>https:\/\/docs\.langbot\.dev(\/[^<]+)<\/loc>/g)].map((match) => match[1].replace(/\/$/, "")));
+  const routes = new Set([...sitemap.matchAll(/<loc>https:\/\/langbot\.app\/docs(\/[^<]+)<\/loc>/g)].map((match) => match[1].replace(/\/$/, "")));
   const allFiles = await collectHtmlRoutes(publicRoot);
   const localized = allFiles.filter((file) => routes.has(routeFromHtml(file)));
   for (const file of localized) {
     const route = routeFromHtml(file);
     const html = await readFile(file, "utf8");
     const canonical = extractHeadLinks(html, "canonical");
-    assert.deepEqual(canonical, [{ href: `https://docs.langbot.dev${route}`, hreflang: undefined }], route);
+    assert.deepEqual(canonical, [{ href: `https://langbot.app/docs${route}`, hreflang: undefined }], route);
 
     const [, locale, ...suffixParts] = route.split("/");
     const suffix = suffixParts.join("/");
@@ -386,14 +386,14 @@ test("every page emits a canonical and reciprocal available hreflangs", async ()
       assert.deepEqual(Object.keys(actual).sort(), ["en", "ja", "x-default", "zh-CN"]);
       assert.equal(actual["x-default"], actual.en);
       for (const href of Object.values(actual)) {
-        const target = new URL(href).pathname.replace(/\/$/, "");
+        const target = new URL(href).pathname.replace(/^\/docs/, "").replace(/\/$/, "");
         assert.ok(routes.has(target), `${route} links to missing alternate ${target}`);
       }
     } else {
       const expected = {};
       for (const candidate of locales) {
         const candidateRoute = `/${candidate}/${suffix}`.replace(/\/$/, "");
-        if (routes.has(candidateRoute)) expected[hreflangByLocale[candidate]] = `https://docs.langbot.dev${candidateRoute}`;
+        if (routes.has(candidateRoute)) expected[hreflangByLocale[candidate]] = `https://langbot.app/docs${candidateRoute}`;
       }
       expected["x-default"] = expected.en ?? expected["zh-CN"] ?? expected.ja;
       assert.deepEqual(actual, expected, route);
@@ -404,8 +404,8 @@ test("every page emits a canonical and reciprocal available hreflangs", async ()
 test("alternate sitemap is complete, reciprocal, and advertised", async () => {
   const sitemap = await readFile(path.join(publicRoot, "sitemap-alternates.xml"), "utf8");
   const robots = await readFile(path.join(publicRoot, "robots.txt"), "utf8");
-  assert.match(robots, /^Sitemap: https:\/\/docs\.langbot\.dev\/sitemap\.xml$/m);
-  assert.match(robots, /^Sitemap: https:\/\/docs\.langbot\.dev\/sitemap-alternates\.xml$/m);
+  assert.match(robots, /^Sitemap: https:\/\/langbot\.app\/docs\/sitemap\.xml$/m);
+  assert.match(robots, /^Sitemap: https:\/\/langbot\.app\/docs\/sitemap-alternates\.xml$/m);
 
   const blocks = [...sitemap.matchAll(/<url>([\s\S]*?)<\/url>/g)].map((match) => match[1]);
   const groups = new Map();
@@ -432,7 +432,7 @@ test("alternate sitemap is complete, reciprocal, and advertised", async () => {
     "articles/n8n-multi-platform-ai-chatbot",
     "articles/dify-langbot-rag-knowledge-base",
     "articles/welcome",
-  ]) assert.ok(sitemap.includes(`https://docs.langbot.dev/en/${slug}`), `alternate sitemap missing ${slug}`);
+  ]) assert.ok(sitemap.includes(`https://langbot.app/docs/en/${slug}`), `alternate sitemap missing ${slug}`);
 });
 
 test("generated pages contain no internal .html route links", async () => {

tests/seo.test.mjs

@@ -74,11 +74,11 @@ test("custom robots policy is served from the Mintlify project root", async () =
   assert.match(robots, /^Allow: \/_next\/$/m);
   assert.match(
     robots,
-    /^Sitemap: https:\/\/docs\.langbot\.dev\/sitemap\.xml$/m,
+    /^Sitemap: https:\/\/langbot\.app\/docs\/sitemap\.xml$/m,
   );
   assert.match(
     robots,
-    /^Sitemap: https:\/\/docs\.langbot\.dev\/sitemap-alternates\.xml$/m,
+    /^Sitemap: https:\/\/langbot\.app\/docs\/sitemap-alternates\.xml$/m,
   );
   assert.doesNotMatch(robots, /https:\/\/docs\.langbot\.app/);
 });
@@ -92,7 +92,7 @@ test("alternate sitemap declares reciprocal en, zh-CN, ja and x-default links",
     (match) => match[1],
   );
   assert.ok(blocks.length > 0);
-  assert.match(sitemap, /https:\/\/docs\.langbot\.dev\//);
+  assert.match(sitemap, /https:\/\/langbot\.app\/docs\//);
   assert.doesNotMatch(sitemap, /https:\/\/docs\.langbot\.app\//);
 
   const groups = new Map();

vite.config.ts

@@ -4,5 +4,5 @@ import { fumadocsMdx } from "fumadocs-mdx/vite";
 import { defineConfig } from "vite";
 
 export default defineConfig({
-  plugins: [press(), fumadocsMdx(), tailwindcss()],
+  plugins: [press({ basePath: "/docs/" }), fumadocsMdx(), tailwindcss()],
 });