Files
Abdulmuizz HamzatandRotimi Best b599f1a017 chore(docs): migrate the docs site from Fumadocs to Blume (#753)
* Update documentation for ClassroomIO: change Fumadocs to Blume, revise MDX format conventions, and clarify sidebar navigation requirements.

* docs: update README and scripts for improved preview and error handling

* docs: update Discord link in configuration and documentation

* docs: add token-auth route and update documentation formatting

* Fix formatting in auto-comment-issues.yml

* Fix JSON syntax by adding a comma

* fix(docs): declare Blume's Astro build deps so Cloudflare build resolves them

blume build generates an Astro project under apps/docs/.blume/ that imports
astro, @astrojs/mdx, @tailwindcss/vite, tailwindcss, @tailwindcss/typography,
@shikijs/twoslash and @scalar/astro. Under pnpm's isolated node_modules these
are nested inside blume and not resolvable from the docs app, so a clean
--frozen-lockfile build (Cloudflare) fails at 'Cannot find module astro/config'.
Declare them directly in @cio/docs, pinned to blume's resolved versions.

* refactor(docs): update API reference paths and documentation for clarity

* fix: change platofrm url to docs home page

* Refactor documentation for self-hosting ClassroomIO

- Removed the Supabase Cloud setup guide as it is no longer needed.
- Updated the Enterprise SSO documentation to include a reference link for SSO & licensing.
- Added new documentation for Content Security Policy (CSP) configuration.
- Created a new section for configuring SMTP for email notifications.
- Introduced a comprehensive configuration index for self-hosted ClassroomIO, detailing environment variables and their usage.
- Added documentation for optional integrations including Google OAuth, OpenAI, and Unsplash.
- Documented the generation and usage of required secrets for ClassroomIO.
- Created a section for SSO & licensing, detailing the requirements for Enterprise features.
- Added object storage configuration documentation, covering both bundled MinIO and external S3-compatible services.
- Documented versioning and upgrade processes for ClassroomIO.
- Updated the Coolify and Docker self-hosting guides to streamline configuration references and upgrade instructions.
- Enhanced the Railway self-hosting guide with links to relevant configuration sections.
- Updated token authentication documentation to reference SSO & licensing.

* Add comprehensive guides for ClassroomIO features and functionalities

- Introduced a new guide for the Student Dashboard, detailing navigation and features available to students.
- Added a guide on using Tags for organizing and filtering courses within ClassroomIO.
- Created a guide on how to take attendance in courses, including marking attendance and related functionalities.
- Developed a Terminology guide to clarify key terms used throughout ClassroomIO.
- Implemented a Token Auth guide for signing users into ClassroomIO from external applications using JWT.
- Added instructions for using math in the editor with LaTeX support.
- Created a guide for customizing the welcome email sent to students upon enrollment in courses.
- Added a guide explaining what ClassroomIO is and its key features.
- Updated the home documentation to provide a clearer overview of ClassroomIO and its resources.
- Enhanced self-hosted configuration documentation with a new guide for deploying ClassroomIO on Dokploy.
- Updated links in the self-hosted configuration documentation for better navigation.
- Added new images to support the Dokploy deployment guide.
- Updated asset packaging script to handle redirects for the new guide structure.
- Fixed footer link to point to the correct Getting Started guide.

* chore: uprage version

* feat(docs): add DigitalOcean deployment guide and flatten guides structure

- Add a guide for deploying ClassroomIO on DigitalOcean using Docker Compose
  and Caddy for automatic HTTPS.
- List DigitalOcean as a deployment option in the self-hosting index.
- Move guides out of content/docs/guides into content/docs (student-dashboard,
  tags, terminology, token-auth, what-is-classroomio, audience, cohorts,
  community, course, enterprise-sso, mcp, organization, quickstart/*).
- Rename home.mdx to index.mdx.
- Update package-assets script to emit redirects for the old guides paths.

---------

Co-authored-by: Rotimi Best <best@classroomio.com>
2026-07-22 22:30:10 +01:00

67 lines
2.2 KiB
JavaScript

/**
* Fetches the public API OpenAPI spec from R2 and enriches it with the bearer
* auth scheme, so Blume can render the API reference statically at build time.
*
* This replaces the old SSR route that fetched the spec on every request. The
* page now refreshes when the docs are rebuilt — .github/workflows/upload-openapi-spec.yml
* triggers a docs deploy after it publishes a new spec.
*
* If the fetch fails but a previously fetched spec is on disk, that one is
* reused: a network blip should not stop you building or previewing the docs.
* With no spec at all there is nothing to render, so the build fails.
*/
import { access, mkdir, writeFile } from 'node:fs/promises';
import { dirname, resolve } from 'node:path';
import { fileURLToPath } from 'node:url';
import { enrichPublicApiOpenApiSpec, PUBLIC_API_OPENAPI_URL } from '@cio/utils/openapi/public-api';
const OUTPUT = resolve(dirname(fileURLToPath(import.meta.url)), '../openapi/public-api.json');
const exists = async (path) => {
try {
await access(path);
return true;
} catch {
return false;
}
};
let response;
try {
response = await fetch(PUBLIC_API_OPENAPI_URL);
} catch (error) {
if (await exists(OUTPUT)) {
console.warn(
`[docs] Could not reach ${PUBLIC_API_OPENAPI_URL} (${error.cause?.code ?? error.message}). Reusing the spec already in openapi/.`
);
process.exit(0);
}
throw new Error(
`Failed to reach the OpenAPI spec at ${PUBLIC_API_OPENAPI_URL}, and no previously fetched copy exists in openapi/. The API reference cannot be built.`,
{ cause: error }
);
}
if (!response.ok) {
if (await exists(OUTPUT)) {
console.warn(
`[docs] ${PUBLIC_API_OPENAPI_URL} returned ${response.status} ${response.statusText}. Reusing the spec already in openapi/.`
);
process.exit(0);
}
throw new Error(
`Failed to fetch the OpenAPI spec from ${PUBLIC_API_OPENAPI_URL}: ${response.status} ${response.statusText}`
);
}
const spec = enrichPublicApiOpenApiSpec(await response.json());
await mkdir(dirname(OUTPUT), { recursive: true });
await writeFile(OUTPUT, `${JSON.stringify(spec, null, 2)}\n`);
console.log(`[docs] Wrote enriched OpenAPI spec to ${OUTPUT}`);