muddle/views/markdown.ts

import type { Loader, LoaderFile } from "../contracts/index.ts";

/**
 * Markdown support.
 *
 * A small, dependency free renderer that covers the subset used by content
 * collections: headings, paragraphs, lists, quotes, code and the usual inline
 * markup. The loader turns a markdown file into frontmatter and html.
 */

export interface Frontmatter {
	[key: string]: unknown;
}

/** Parses the frontmatter of a markdown file. */
export function parseFrontmatter(source: string): {
	frontmatter: Frontmatter;
	body: string;
} {
	const match = /^---\r?\n([\s\S]*?)\r?\n---\r?\n?/.exec(source);
	if (!match) return { frontmatter: {}, body: source };
	const frontmatter = parseYaml(match[1] as string);
	return { frontmatter, body: source.slice(match[0].length) };
}

/** Parses a small subset of yaml: scalars, lists and nested maps. */
export function parseYaml(source: string): Frontmatter {
	const result: Frontmatter = {};
	const lines = source.split("\n");
	let index = 0;

	while (index < lines.length) {
		const line = (lines[index] as string).trimEnd();
		index += 1;
		if (line.trim().length === 0 || line.trimStart().startsWith("#")) continue;

		const separator = line.indexOf(":");
		if (separator === -1) continue;
		const key = line.slice(0, separator).trim();
		const raw = line.slice(separator + 1).trim();

		if (raw.length === 0) {
			const items: unknown[] = [];
			while (index < lines.length) {
				const next = (lines[index] as string).trim();
				if (next.length === 0) {
					index += 1;
					continue;
				}
				if (!next.startsWith("- ")) break;
				items.push(parseScalar(next.slice(2).trim()));
				index += 1;
			}
			result[key] = items.length > 0 ? items : "";
			continue;
		}

		result[key] = parseScalar(raw);
	}

	return result;
}

function parseScalar(raw: string): unknown {
	if (raw.startsWith("[") && raw.endsWith("]")) {
		return raw
			.slice(1, -1)
			.split(",")
			.map((entry) => parseScalar(entry.trim()))
			.filter((entry) => entry !== "");
	}
	if (raw === "true") return true;
	if (raw === "false") return false;
	if (raw === "null") return null;
	if (raw !== "" && !Number.isNaN(Number(raw))) return Number(raw);
	return raw.replace(/^['"]|['"]$/g, "");
}

/** Escapes the html entities of a text. */
export function escapeHtml(value: string): string {
	return value
		.replace(/&/g, "&amp;")
		.replace(/</g, "&lt;")
		.replace(/>/g, "&gt;")
		.replace(/"/g, "&quot;");
}

/**
 * Renders markdown to html.
 *
 * The renderer is intentionally small, it covers the constructs used by the
 * content collections of a project.
 */
export function renderMarkdown(source: string): string {
	const lines = source.replace(/\r\n/g, "\n").split("\n");
	const output: string[] = [];
	let index = 0;

	while (index < lines.length) {
		const line = lines[index] as string;

		if (line.trim().length === 0) {
			index += 1;
			continue;
		}

		if (/^\s*\$\$\s*$/.test(line)) {
			const formula: string[] = [];
			index += 1;
			while (
				index < lines.length &&
				!/^\s*\$\$\s*$/.test(lines[index] as string)
			) {
				formula.push(lines[index] as string);
				index += 1;
			}
			if (index < lines.length) index += 1;
			output.push(
				`<div class="muddle-math">${renderLatex(formula.join("\n"), true)}</div>`,
			);
			continue;
		}

		const fence = /^\s*```(\w*)\s*$/.exec(line);
		if (fence) {
			const language = fence[1] ?? "";
			const code: string[] = [];
			index += 1;
			while (
				index < lines.length &&
				!/^\s*```\s*$/.test(lines[index] as string)
			) {
				code.push(lines[index] as string);
				index += 1;
			}
			index += 1;
			const className = language ? ` class="language-${language}"` : "";
			output.push(
				`<pre><code${className}>${escapeHtml(code.join("\n"))}</code></pre>`,
			);
			continue;
		}

		if (/^\s*@@@\s*$/.test(line)) {
			const graph: string[] = [];
			index += 1;
			while (
				index < lines.length &&
				!/^\s*@@@\s*$/.test(lines[index] as string)
			) {
				graph.push(lines[index] as string);
				index += 1;
			}
			index += 1;
			output.push(
				`<pre class="mermaid" data-muddle-animate="true">${escapeHtml(graph.join("\n"))}</pre>`,
			);
			continue;
		}

		const heading = /^(#{1,6})\s+(.*)$/.exec(line);
		if (heading) {
			const level = (heading[1] as string).length;
			output.push(
				`<h${level}>${renderInline(heading[2] as string)}</h${level}>`,
			);
			index += 1;
			continue;
		}

		if (/^\s*(?:---+|\*\*\*+|___+)\s*$/.test(line)) {
			output.push("<hr>");
			index += 1;
			continue;
		}

		const quote = /^\s*>\s?(.*)$/.exec(line);
		if (quote) {
			const quoted: string[] = [];
			while (index < lines.length) {
				const next = /^\s*>\s?(.*)$/.exec(lines[index] as string);
				if (!next) break;
				quoted.push(next[1] as string);
				index += 1;
			}
			output.push(
				`<blockquote>${renderMarkdown(quoted.join("\n"))}</blockquote>`,
			);
			continue;
		}

		const bullet = /^\s*[-*+]\s+(.*)$/.exec(line);
		const ordered = /^\s*\d+[.)]\s+(.*)$/.exec(line);
		if (bullet || ordered) {
			const tag = bullet ? "ul" : "ol";
			const items: unknown[] = [];
			while (index < lines.length) {
				const current = lines[index] as string;
				const match = bullet
					? /^\s*[-*+]\s+(.*)$/.exec(current)
					: /^\s*\d+[.)]\s+(.*)$/.exec(current);
				if (!match) break;
				items.push(match[1] as string);
				index += 1;
			}
			const rendered = items
				.map((item) => `<li>${renderInline(String(item))}</li>`)
				.join("");
			output.push(`<${tag}>${rendered}</${tag}>`);
			continue;
		}

		const paragraph: string[] = [];
		while (index < lines.length) {
			const current = lines[index] as string;
			if (current.trim().length === 0) break;
			if (/^(#{1,6})\s+/.test(current)) break;
			if (/^\s*([-*+]\s|\d+[.)]\s|>)/.test(current)) break;
			paragraph.push(current.trim());
			index += 1;
		}
		output.push(`<p>${renderInline(paragraph.join(" "))}</p>`);
	}

	return output.join("\n");
}

/** Renders the inline markup of a text. */
export function renderInline(source: string): string {
	const codes: string[] = [];
	const maths: string[] = [];
	let text = source.replace(/`([^`]+)`/g, (_match, code: string) => {
		codes.push(code);
		return `\uE000${codes.length - 1}\uE000`;
	});
	text = text.replace(/(?<!\\)\$([^$\n]+?)\$/g, (_match, formula: string) => {
		maths.push(renderLatex(formula));
		return `\uE001${maths.length - 1}\uE001`;
	});

	text = escapeHtml(text);
	text = text.replace(
		/!\[([^\]]*)\]\(([^)\s]+)\)/g,
		'<img src="$2" alt="$1" loading="lazy">',
	);
	text = text.replace(/\[([^\]]+)\]\(([^)\s]+)\)/g, '<a href="$2">$1</a>');
	text = text.replace(/\*\*([^*]+)\*\*/g, "<strong>$1</strong>");
	text = text.replace(/__([^_]+)__/g, "<strong>$1</strong>");
	text = text.replace(/(^|[^*])\*([^*\n]+)\*/g, "$1<em>$2</em>");
	text = text.replace(/~~([^~]+)~~/g, "<del>$1</del>");

	text = text.replace(/\uE000(\d+)\uE000/g, (_match, index: string) => {
		return `<code>${escapeHtml(codes[Number(index)] ?? "")}</code>`;
	});
	return text.replace(
		/\uE001(\d+)\uE001/g,
		(_match, index: string) => maths[Number(index)] ?? "",
	);
}

/** Renders a trusted LaTex formula to KaTeX HTML and MathML. */
export function renderLatex(source: string, displayMode = false): string {
	return escapeHtml(
		`${displayMode ? "$$" : "$"}${source}${displayMode ? "$$" : "$"}`,
	);
}

export interface MarkdownLoaderOptions {
	/** Directory the loader scans, relative to the project root. */
	dir?: string;
}

/**
 * Loader for markdown files.
 *
 * The loader turns a markdown file into its frontmatter and its rendered
 * html, exactly like a content collection needs it.
 */
export function markdownLoader(_options: MarkdownLoaderOptions = {}): Loader {
	return {
		name: "markdown",
		extensions: [".md", ".markdown"],
		load(file: LoaderFile) {
			const { frontmatter, body } = parseFrontmatter(file.source);
			return {
				data: frontmatter,
				content: renderMarkdown(body),
			};
		},
	};
}