mirror of
https://github.com/supabase/supabase.git
synced 2026-10-12 21:05:06 +03:00
* refactor: reading markdown docs files
Refactor how Markdown docs files are read:
- Reuses the same logic across search index generation & page generation
- Improves the indexed content for search:
- Stops removing MDX components, which often contain useful
information like Admonitions
- Denormalizes Partials and CodeSamples for more complete content
This is a prerequisite step for implementing the "Copy docs as Markdown"
functionality.
Only touches regular guides for now, not federated ones.
* fix: tailwind build error (#37728)
We changed to default to ESM imports a while ago, which means local
builds are now breaking because the Tailwind uses a require. Changed to
CJS for Tailwind config file. (I have no idea how this has been working
on Vercel all this time.)
* style: prettier
117 lines
3.0 KiB
TypeScript
117 lines
3.0 KiB
TypeScript
import { Octokit } from '@octokit/core'
|
|
import { BaseLoader, BaseSource } from './base.js'
|
|
import { createHash } from 'node:crypto'
|
|
|
|
const getBasename = (path: string) => path.split('/').at(-1)!.replace(/\.md$/, '')
|
|
|
|
export class LintWarningsGuideLoader extends BaseLoader {
|
|
type = 'markdown' as const
|
|
|
|
constructor(
|
|
source: string,
|
|
path: string,
|
|
public org: string,
|
|
public repo: string,
|
|
public branch: string,
|
|
public docsDir: string
|
|
) {
|
|
super(source, path)
|
|
}
|
|
|
|
async load() {
|
|
const octokit = new Octokit()
|
|
|
|
const response = await octokit.request('GET /repos/{owner}/{repo}/contents/{path}', {
|
|
owner: this.org,
|
|
repo: this.repo,
|
|
path: this.docsDir,
|
|
ref: this.branch,
|
|
headers: {
|
|
'X-GitHub-Api-Version': '2022-11-28',
|
|
},
|
|
})
|
|
|
|
if (response.status >= 400) {
|
|
throw Error(`Could not get contents of repo ${this.org}/${this.repo}`)
|
|
}
|
|
|
|
if (!Array.isArray(response.data)) {
|
|
throw Error(
|
|
'Reading a directory, not a file. Should not reach this, solely to appease Typescript.'
|
|
)
|
|
}
|
|
|
|
const lintsList = response.data.filter(({ path }) => /docs\/\d+.+\.md$/.test(path))
|
|
|
|
// Fetch all lint files and combine them into a single guide
|
|
const lints = await Promise.all(
|
|
lintsList.map(async ({ path }) => {
|
|
const fileResponse = await fetch(
|
|
`https://raw.githubusercontent.com/${this.org}/${this.repo}/${this.branch}/${path}`
|
|
)
|
|
|
|
if (fileResponse.status >= 400) {
|
|
throw Error(`Could not get contents of file ${this.org}/${this.repo}/${path}`)
|
|
}
|
|
|
|
const content = await fileResponse.text()
|
|
const basename = getBasename(path)
|
|
|
|
return {
|
|
path: basename,
|
|
content,
|
|
originalPath: path,
|
|
}
|
|
})
|
|
)
|
|
|
|
// Create a separate source for each lint file
|
|
return lints.map(
|
|
(lint) =>
|
|
new LintWarningsGuideSource(
|
|
this.source,
|
|
`${this.path}?queryGroups=lint&lint=${lint.path}`,
|
|
lint
|
|
)
|
|
)
|
|
}
|
|
}
|
|
|
|
export class LintWarningsGuideSource extends BaseSource {
|
|
type = 'markdown' as const
|
|
|
|
constructor(
|
|
source: string,
|
|
path: string,
|
|
public lint: {
|
|
path: string
|
|
content: string
|
|
originalPath: string
|
|
}
|
|
) {
|
|
super(source, path)
|
|
}
|
|
|
|
async process() {
|
|
this.checksum = createHash('sha256').update(this.lint.content).digest('base64')
|
|
|
|
this.meta = {
|
|
title: `Database Advisor: Lint ${this.lint.path}`,
|
|
}
|
|
|
|
this.sections = [
|
|
{
|
|
content: this.lint.content,
|
|
},
|
|
]
|
|
|
|
return { checksum: this.checksum, meta: this.meta, sections: this.sections }
|
|
}
|
|
|
|
extractIndexedContent(): string {
|
|
const sections = this.sections ?? []
|
|
const sectionText = sections.map(({ content }) => content).join('\n\n')
|
|
return `# Database Advisor: Lint ${this.lint.path}\n\nThis is a database lint rule for Supabase, targeting the lint ID ${this.lint.path}. Lint rules help enforce performance and security best practices for your Supabase database.\n\n${sectionText}`
|
|
}
|
|
}
|