snapshot of https://github.com/tldraw/tldraw.git (main, 2026-09-23; blobs > 131072B omitted)

internal/scripts/extract-draft-changelog.tsx

@@ -0,0 +1,258 @@
+import { join } from 'path'
+import { exec } from './lib/exec'
+import { REPO_ROOT, writeStringFile } from './lib/file'
+import { nicelog } from './lib/nicelog'
+
+function convertPRLinks(text: string): string {
+	// Convert (#1234) to ([#1234](https://github.com/tldraw/tldraw/issues/1234))
+	return text.replace(/\(#(\d+)\)/g, '([#$1](https://github.com/tldraw/tldraw/issues/$1))')
+}
+
+function extractRelevantSections(message: string): string {
+	// Extract Release Notes and API Changes sections (case-insensitive, flexible markdown heading levels)
+	const lines = message.split('\n')
+	const relevantSections: string[] = []
+	let inRelevantSection = false
+	let currentSection: string[] = []
+
+	for (const line of lines) {
+		// Check if this is a heading line (one or more #s followed by text)
+		const headingMatch = line.match(/^(#{1,6})\s+(.+)$/i)
+		if (headingMatch) {
+			const headingText = headingMatch[2].trim().toLowerCase()
+
+			// If we were in a relevant section, save it
+			if (inRelevantSection && currentSection.length > 0) {
+				relevantSections.push(currentSection.join('\n'))
+				currentSection = []
+			}
+
+			// Check if this is a Release Notes or API Changes heading
+			if (headingText.includes('release note') || headingText.includes('api change')) {
+				inRelevantSection = true
+				// Include the heading itself in the section
+				currentSection.push(line)
+			} else {
+				inRelevantSection = false
+			}
+		} else if (inRelevantSection) {
+			// We're in a relevant section, collect the line
+			currentSection.push(line)
+		}
+	}
+
+	// Don't forget the last section
+	if (inRelevantSection && currentSection.length > 0) {
+		relevantSections.push(currentSection.join('\n'))
+	}
+
+	// If we found relevant sections, return them joined together
+	if (relevantSections.length > 0) {
+		// Remove the section headings themselves (first line of each section)
+		const sectionsWithoutHeadings = relevantSections
+			.map((section) => {
+				const sectionLines = section.split('\n')
+				// Remove the first line (the heading) and keep the rest
+				return sectionLines.slice(1).join('\n').trim()
+			})
+			.filter((section) => section.length > 0)
+
+		return sectionsWithoutHeadings.join('\n\n').trim()
+	}
+
+	// Otherwise, return empty string (no content to include)
+	return ''
+}
+
+async function getLatestMinorRelease() {
+	// otherwise get the latest minor release
+	const tags = (await exec('git', ['tag', '--list', `v*.*.0`, '--sort=-version:refname']))
+		.toString()
+		.trim()
+	return tags.split('\n')[0]
+}
+
+export async function extractChangelog(startRef: string, endRef: string): Promise<string> {
+	nicelog(`Extracting changelog from ${startRef} to ${endRef}`)
+
+	const commitHashes = (await exec('git', ['cherry', startRef, endRef]))
+		.trim()
+		.split('\n')
+		.filter((line) => !line.startsWith('-'))
+		.map((line) => line.split(' ')[1])
+
+	nicelog(`Found ${commitHashes.length} commits`)
+
+	// Extract full commit messages for each commit
+	const entries: Array<{
+		firstLine: string
+		restOfMessage: string
+		changeType: string
+		changedFolders: string[]
+	}> = []
+
+	for (const hash of commitHashes) {
+		const fullMessage = await exec('git', ['log', '-1', '--pretty=format:%B', hash], {
+			processStdoutLine: () => {}, // Suppress output
+			processStderrLine: () => {},
+		})
+
+		const lines = fullMessage.trim().split('\n')
+		const firstLine = lines[0] || hash
+
+		// Skip commits with [skip ci] in the title
+		if (firstLine.includes('[skip ci]')) {
+			continue
+		}
+
+		// Get changed files for this commit
+		const changedFiles = await exec('git', ['show', '--name-only', '--pretty=format:', hash], {
+			processStdoutLine: () => {}, // Suppress output
+			processStderrLine: () => {},
+		})
+
+		// Extract unique folders (max 2 levels deep) from changed files
+		const folders = new Set<string>()
+		changedFiles
+			.split('\n')
+			.filter((file) => file.trim())
+			.forEach((file) => {
+				const parts = file.split('/')
+				if (parts.length >= 2) {
+					folders.add(`${parts[0]}/${parts[1]}`)
+				} else if (parts.length === 1) {
+					folders.add(parts[0])
+				}
+			})
+
+		let restOfMessage = lines.slice(1).join('\n').trim()
+
+		// Extract change type from checkboxes
+		let changeType = 'other'
+		const changeTypes = ['bugfix', 'improvement', 'feature', 'api', 'other']
+
+		for (const type of changeTypes) {
+			if (restOfMessage.includes(`- [x] \`${type}\``)) {
+				changeType = type
+				break
+			}
+		}
+
+		// Remove the Change type section and all checkboxes
+		restOfMessage = restOfMessage
+			// remove cursor summary
+			.replace(/<!-- CURSOR_SUMMARY -->(.|\s)*/gm, '')
+			// remove Change type section
+			.replace(/### Change type[\s\n]*/g, '')
+			// remove checkboxes
+			.replace(/- \[[ x]\] `(bugfix|improvement|feature|api|other)`[\s\n]*/g, '')
+			// remove co-authored-by lines
+			.replace(/^co-authored-by: .*$/gim, '')
+			// Clean up excessive newlines
+			.replace(/\n{3,}/g, '\n\n')
+			// remove trailing horizontal rule
+			.trim()
+			.replace(/-{3,}$/, '')
+			.trim()
+
+		// Extract only Release Notes and API Changes sections
+		const extractedContent = extractRelevantSections(restOfMessage)
+
+		// Convert PR links like (#1234) to full URLs
+		const contentWithLinks = convertPRLinks(extractedContent)
+
+		// Only include entries that touch packages/ folder (but not internal packages)
+		const excludedPackages = ['packages/dotcom-shared', 'packages/worker-shared']
+		const touchesPackages = Array.from(folders).some(
+			(folder) => folder.startsWith('packages/') && !excludedPackages.includes(folder)
+		)
+		if (touchesPackages) {
+			entries.push({
+				firstLine: convertPRLinks(firstLine),
+				restOfMessage: contentWithLinks,
+				changeType,
+				changedFolders: Array.from(folders).sort(),
+			})
+		}
+	}
+
+	// Group entries by change type
+	const groupedEntries: Record<string, typeof entries> = {
+		bugfix: [],
+		improvement: [],
+		feature: [],
+		api: [],
+		other: [],
+	}
+
+	for (const entry of entries) {
+		groupedEntries[entry.changeType].push(entry)
+	}
+
+	// Generate output.md content
+	let output = `Generated from commits between \`${startRef}\` and \`${endRef}\`\n\n`
+
+	// Output each change type section
+	const changeTypeLabels = {
+		bugfix: 'Bug Fixes',
+		improvement: 'Improvements',
+		feature: 'Features',
+		api: 'API Changes',
+		other: 'Other Changes',
+	}
+
+	for (const [changeType, label] of Object.entries(changeTypeLabels)) {
+		const entriesForType = groupedEntries[changeType]
+		if (entriesForType.length === 0) continue
+
+		output += `## ${label}\n\n`
+
+		for (const entry of entriesForType) {
+			output += `### ${entry.firstLine}\n\n`
+
+			if (entry.restOfMessage) {
+				output += `${entry.restOfMessage}\n\n`
+			}
+		}
+	}
+
+	nicelog(`Extracted ${entries.length} changelog entries`)
+	return output
+}
+
+async function main() {
+	const args = process.argv.slice(2)
+
+	await exec('git', ['fetch', 'origin'])
+	await exec('git', ['fetch', 'origin', '--tags'])
+
+	if (args.includes('--help') || args.includes('-h')) {
+		console.error('Usage: yarn extract-draft-changelog [start-ref] [end-ref]')
+		console.error('Example: yarn extract-draft-changelog v2.0.0 HEAD')
+		console.error('Example: yarn extract-draft-changelog v2.0.0  # uses current HEAD')
+		process.exit(1)
+	}
+
+	const startRef = args[0] || (await getLatestMinorRelease())
+	const endRef = args[1] || 'origin/production'
+
+	try {
+		const output = await extractChangelog(startRef, endRef)
+
+		// Write to changelog.md in the repo root
+		const outputPath = join(REPO_ROOT, 'changelog.md')
+		await writeStringFile(outputPath, output)
+
+		nicelog(`Draft changelog written to changelog.md`)
+	} catch (error) {
+		console.error('Error extracting changelog:', error)
+		process.exit(1)
+	}
+}
+
+if (require.main === module) {
+	main().catch((error) => {
+		console.error('Unhandled error:', error)
+		process.exit(1)
+	})
+}