MKT-67: Add blog markdown content pipeline

apps/web-roo-code/content/blog/.gitkeep

@@ -0,0 +1,2 @@
+# This directory contains blog post markdown files.
+# See docs/blog.md for the specification.

apps/web-roo-code/package.json

@@ -24,6 +24,7 @@
 		"embla-carousel-autoplay": "^8.6.0",
 		"embla-carousel-react": "^8.6.0",
 		"framer-motion": "12.15.0",
+		"gray-matter": "^4.0.3",
 		"lucide-react": "^0.518.0",
 		"next": "~15.2.8",
 		"next-themes": "^0.4.6",

apps/web-roo-code/src/lib/blog/content.ts

@@ -0,0 +1,211 @@
+import fs from "fs"
+import path from "path"
+import matter from "gray-matter"
+import { ZodError } from "zod"
+import type { BlogPost } from "./types"
+import { blogFrontmatterSchema } from "./types"
+import { getNowPt } from "./pt-time"
+import { isPublished } from "./publishing"
+
+/**
+ * Path to the blog content directory (relative to project root).
+ */
+const CONTENT_DIR = "content/blog"
+
+/**
+ * Get the absolute path to the blog content directory.
+ */
+function getContentDir(): string {
+	return path.join(process.cwd(), CONTENT_DIR)
+}
+
+/**
+ * Error thrown when blog content validation fails.
+ */
+export class BlogContentError extends Error {
+	constructor(
+		message: string,
+		public filename?: string,
+	) {
+		super(filename ? `[${filename}] ${message}` : message)
+		this.name = "BlogContentError"
+	}
+}
+
+/**
+ * Parse a single markdown file into a BlogPost object.
+ *
+ * @param filename - Name of the markdown file (e.g., "my-post.md")
+ * @returns Parsed BlogPost object
+ * @throws BlogContentError if frontmatter is invalid
+ */
+function parseMarkdownFile(filename: string): BlogPost {
+	const filePath = path.join(getContentDir(), filename)
+	const fileContent = fs.readFileSync(filePath, "utf8")
+
+	// Parse frontmatter using gray-matter
+	const { data, content } = matter(fileContent)
+
+	// Validate frontmatter with zod
+	try {
+		const frontmatter = blogFrontmatterSchema.parse(data)
+
+		// Verify slug matches filename (without .md extension)
+		const expectedSlug = filename.replace(/\.md$/, "")
+		if (frontmatter.slug !== expectedSlug) {
+			throw new BlogContentError(
+				`Slug mismatch: frontmatter slug "${frontmatter.slug}" does not match filename "${expectedSlug}"`,
+				filename,
+			)
+		}
+
+		return {
+			...frontmatter,
+			content,
+			filename,
+		}
+	} catch (error) {
+		if (error instanceof ZodError) {
+			const issues = error.issues.map((i) => `  - ${i.path.join(".")}: ${i.message}`).join("\n")
+			throw new BlogContentError(`Invalid frontmatter:\n${issues}`, filename)
+		}
+		throw error
+	}
+}
+
+/**
+ * Load all markdown files from the content directory.
+ *
+ * @returns Array of all parsed blog posts (including drafts)
+ * @throws BlogContentError if any file has invalid frontmatter or duplicate slugs
+ */
+function loadAllPosts(): BlogPost[] {
+	const contentDir = getContentDir()
+
+	// Check if content directory exists
+	if (!fs.existsSync(contentDir)) {
+		return []
+	}
+
+	// Get all .md files
+	const files = fs.readdirSync(contentDir).filter((file) => file.endsWith(".md"))
+
+	// Parse all files
+	const posts: BlogPost[] = []
+	const slugToFilename = new Map<string, string>()
+
+	for (const filename of files) {
+		const post = parseMarkdownFile(filename)
+
+		// Check for duplicate slugs
+		const existingFilename = slugToFilename.get(post.slug)
+		if (existingFilename) {
+			throw new BlogContentError(
+				`Duplicate slug "${post.slug}" found in files: "${existingFilename}" and "${filename}"`,
+			)
+		}
+		slugToFilename.set(post.slug, filename)
+
+		posts.push(post)
+	}
+
+	return posts
+}
+
+/**
+ * Options for getAllBlogPosts.
+ */
+export interface GetAllBlogPostsOptions {
+	/**
+	 * Include draft posts in the results.
+	 * @default false
+	 */
+	includeDrafts?: boolean
+}
+
+/**
+ * Get all blog posts, optionally filtered by publish status.
+ *
+ * By default, only returns published posts that are past their scheduled
+ * publish time (evaluated at request time in Pacific Time).
+ *
+ * @param options - Options for filtering posts
+ * @returns Array of blog posts, sorted by publish_date (newest first)
+ *
+ * @example
+ * ```ts
+ * // Get only published posts (default)
+ * const posts = getAllBlogPosts();
+ *
+ * // Include drafts (e.g., for preview in CMS)
+ * const allPosts = getAllBlogPosts({ includeDrafts: true });
+ * ```
+ */
+export function getAllBlogPosts(options: GetAllBlogPostsOptions = {}): BlogPost[] {
+	const { includeDrafts = false } = options
+
+	const allPosts = loadAllPosts()
+	const nowPt = getNowPt()
+
+	// Filter posts based on publish status
+	const filteredPosts = includeDrafts ? allPosts : allPosts.filter((post) => isPublished(post, nowPt))
+
+	// Sort by publish_date (newest first), then by publish_time_pt
+	return filteredPosts.sort((a, b) => {
+		// Compare dates first (descending)
+		const dateCompare = b.publish_date.localeCompare(a.publish_date)
+		if (dateCompare !== 0) {
+			return dateCompare
+		}
+		// Same date - compare times (descending)
+		return parsePublishTimePt(b.publish_time_pt) - parsePublishTimePt(a.publish_time_pt)
+	})
+}
+
+/**
+ * Get a single blog post by its slug.
+ *
+ * Only returns the post if it's published and past its scheduled publish time.
+ * Draft posts and future-scheduled posts will return null.
+ *
+ * @param slug - The URL slug of the post
+ * @returns The blog post if found and published, null otherwise
+ *
+ * @example
+ * ```ts
+ * const post = getBlogPostBySlug('my-great-article');
+ * if (post) {
+ *   // Render the post
+ * } else {
+ *   // Show 404
+ * }
+ * ```
+ */
+export function getBlogPostBySlug(slug: string): BlogPost | null {
+	const allPosts = loadAllPosts()
+	const nowPt = getNowPt()
+
+	const post = allPosts.find((p) => p.slug === slug)
+
+	// Post not found
+	if (!post) {
+		return null
+	}
+
+	// Check if published
+	if (!isPublished(post, nowPt)) {
+		return null
+	}
+
+	return post
+}
+
+/**
+ * Get all valid slugs for published posts.
+ * Useful for generating static paths or sitemaps.
+ *
+ * @returns Array of slugs for published posts
+ */
+export function getPublishedSlugs(): string[] {
+	return getAllBlogPosts().map((post) => post.slug)
+}

apps/web-roo-code/src/lib/blog/index.ts

@@ -0,0 +1,37 @@
+/**
+ * Blog content pipeline for roocode.com/blog
+ *
+ * This module provides functions to load and manage blog posts from
+ * markdown files with frontmatter.
+ *
+ * @see docs/blog.md for the full specification
+ *
+ * @example
+ * ```ts
+ * import { getAllBlogPosts, getBlogPostBySlug, formatPostDatePt } from '@/lib/blog';
+ *
+ * // Get all published posts
+ * const posts = getAllBlogPosts();
+ *
+ * // Get a specific post
+ * const post = getBlogPostBySlug('my-article');
+ *
+ * // Format date for display
+ * const displayDate = formatPostDatePt(post.publish_date);
+ * // "2026-01-29"
+ * ```
+ */
+
+// Types
+export type { BlogPost, BlogFrontmatter, PtMoment } from "./types"
+export { blogFrontmatterSchema, SLUG_PATTERN, PUBLISH_TIME_PT_PATTERN, MAX_TAGS } from "./types"
+
+// Content loading
+export { getAllBlogPosts, getBlogPostBySlug, getPublishedSlugs, BlogContentError } from "./content"
+export type { GetAllBlogPostsOptions } from "./content"
+
+// PT timezone helpers
+export { getNowPt, parsePublishTimePt, formatPostDatePt } from "./pt-time"
+
+// Publishing helpers
+export { isPublished } from "./publishing"

apps/web-roo-code/src/lib/blog/pt-time.ts

@@ -0,0 +1,111 @@
+import type { PtMoment } from "./types"
+import { PUBLISH_TIME_PT_PATTERN } from "./types"
+
+/**
+ * Pacific Time timezone identifier.
+ */
+const PT_TIMEZONE = "America/Los_Angeles"
+
+/**
+ * Get the current moment in Pacific Time.
+ *
+ * @returns PtMoment with date (YYYY-MM-DD) and minutes since midnight
+ *
+ * @example
+ * ```ts
+ * const now = getNowPt();
+ * // { date: '2026-01-29', minutes: 540 } // 9:00am PT
+ * ```
+ */
+export function getNowPt(): PtMoment {
+	const now = new Date()
+
+	// Format date as YYYY-MM-DD in PT
+	const dateFormatter = new Intl.DateTimeFormat("en-CA", {
+		timeZone: PT_TIMEZONE,
+		year: "numeric",
+		month: "2-digit",
+		day: "2-digit",
+	})
+	const date = dateFormatter.format(now)
+
+	// Get hours and minutes in PT
+	const timeFormatter = new Intl.DateTimeFormat("en-US", {
+		timeZone: PT_TIMEZONE,
+		hour: "numeric",
+		minute: "numeric",
+		hour12: false,
+	})
+	const timeParts = timeFormatter.formatToParts(now)
+	const hour = parseInt(timeParts.find((p) => p.type === "hour")?.value ?? "0", 10)
+	const minute = parseInt(timeParts.find((p) => p.type === "minute")?.value ?? "0", 10)
+	const minutes = hour * 60 + minute
+
+	return { date, minutes }
+}
+
+/**
+ * Parse a publish_time_pt string (h:mmam/pm) to minutes since midnight.
+ *
+ * @param time - Time string in h:mmam/pm format (e.g., "9:00am", "12:30pm")
+ * @returns Minutes since midnight (0-1439)
+ * @throws Error if the time format is invalid
+ *
+ * @example
+ * ```ts
+ * parsePublishTimePt('9:00am');   // 540  (9 * 60)
+ * parsePublishTimePt('12:30pm');  // 750  (12 * 60 + 30)
+ * parsePublishTimePt('12:00am');  // 0    (midnight)
+ * parsePublishTimePt('11:59pm');  // 1439 (23 * 60 + 59)
+ * ```
+ */
+export function parsePublishTimePt(time: string): number {
+	if (!PUBLISH_TIME_PT_PATTERN.test(time)) {
+		throw new Error(`Invalid publish_time_pt format: "${time}". Must be h:mmam/pm (e.g., "9:00am", "12:30pm")`)
+	}
+
+	// Extract components: "9:00am" -> ["9", "00", "am"]
+	const match = time.match(/^(\d{1,2}):(\d{2})(am|pm)$/)
+	if (!match || !match[1] || !match[2] || !match[3]) {
+		throw new Error(`Failed to parse publish_time_pt: "${time}"`)
+	}
+
+	let hour = parseInt(match[1], 10)
+	const minute = parseInt(match[2], 10)
+	const period = match[3]
+
+	// Convert 12-hour to 24-hour format
+	if (period === "am") {
+		// 12:xxam = 0:xx (midnight hour)
+		if (hour === 12) {
+			hour = 0
+		}
+	} else {
+		// pm
+		// 12:xxpm = 12:xx (noon hour)
+		// 1:xxpm = 13:xx, etc.
+		if (hour !== 12) {
+			hour += 12
+		}
+	}
+
+	return hour * 60 + minute
+}
+
+/**
+ * Format a publish_date for display.
+ * Returns the date as-is since it's already in YYYY-MM-DD format.
+ *
+ * @param publishDate - Date string in YYYY-MM-DD format
+ * @returns Formatted date string (YYYY-MM-DD)
+ *
+ * @example
+ * ```ts
+ * formatPostDatePt('2026-01-29'); // '2026-01-29'
+ * ```
+ */
+export function formatPostDatePt(publishDate: string): string {
+	// The publish_date is already in YYYY-MM-DD format (Pacific Time)
+	// Per spec, we display date only, no time shown to users
+	return publishDate
+}