import { readFileSync, writeFileSync } from "node:fs";
import { resolve } from "node:path";

/**
 * Removes specific Markdown formatting from a given string content,
 * leaving a plain text representation. This function is intended to strip
 * all Markdown syntax for purposes like cleaning article titles.
 * @param content The Markdown content as a string.
 * @returns The cleaned content string.
 */
export function cleanMarkdownString(content: string): string {
	let cleanedContent = content;

	// --- Only remove inline Markdown syntax and heading markers ---

	// 1. Remove heading markers (e.g., # Heading, ## Subheading), keeping heading text
	cleanedContent = cleanedContent.replace(/^(#+)\s*/gm, "");

	// 2. Remove inline code (e.g., `code`)
	cleanedContent = cleanedContent.replace(/`([^`]+?)`/g, "$1");

	// 3. Remove images (e.g., ![alt text](url)), keeping alt text
	cleanedContent = cleanedContent.replace(/!\[(.*?)\]\(.*?\)/g, "$1");

	// 4. Remove links (e.g., [link text](url)), keeping link text
	cleanedContent = cleanedContent.replace(/\[(.*?)\]\(.*?\)/g, "$1");

	// 5. Remove bold/italic formatting (e.g., **bold**, *italic*), keeping content
	// Order matters for overlapping syntax (e.g., ***bold-italic*** should be handled by bold/italic first)
	cleanedContent = cleanedContent.replace(/\*\*(.*?)\*\*/g, "$1"); // **bold**
	cleanedContent = cleanedContent.replace(/__(.*?)__/g, "$1"); // __bold__
	cleanedContent = cleanedContent.replace(/\*(.*?)\*/g, "$1"); // *italic*
	cleanedContent = cleanedContent.replace(/_(.*?)_/g, "$1"); // _italic_

	// 6. Remove remaining common Markdown punctuation that might appear in titles
	cleanedContent = cleanedContent.replace(/[*_`[\]()]/g, "");

	// --- Post-processing / Normalization ---

	// Replace multiple spaces/tabs with single space
	cleanedContent = cleanedContent.replace(/[ \t]+/g, " ");

	// Trim leading/trailing whitespace
	cleanedContent = cleanedContent.trim();

	return cleanedContent;
}

/**
 * Reads a Markdown file, cleans its content, and overwrites the original file.
 * This function is intended for full file cleaning, but its usage should be reviewed
 * if only parts of the file need cleaning. For this project, it might be deprecated
 * in favor of targeted string cleaning.
 * @param filePath The path to the Markdown file.
 */
export function cleanMarkdownFile(filePath: string) {
	try {
		const fullPath = resolve(process.cwd(), filePath);
		const content = readFileSync(fullPath, "utf8");

		// NOTE: This now uses a less aggressive cleanMarkdownString.
		// If full file markdown stripping is needed in the future,
		// cleanMarkdownString would need a mode or a separate function.
		const cleanedContent = cleanMarkdownString(content);

		writeFileSync(fullPath, cleanedContent, "utf8");
		console.log(`Successfully cleaned: ${filePath}`);
	} catch (error: unknown) {
		if (
			error instanceof Error &&
			"code" in error &&
			(error as NodeJS.ErrnoException).code === "ENOENT"
		) {
			console.error(`Error: File not found - ${filePath}`);
		} else {
			console.error(
				`Error cleaning file ${filePath}:`,
				error instanceof Error ? error.message : String(error),
			);
		}
		process.exit(1);
	}
}

// Get file path from command line arguments when executed directly
if (import.meta.main) {
	const filePath = process.argv[2];

	if (!filePath) {
		console.error("Usage: bun clean-md <file-path>");
		process.exit(1);
	}

	cleanMarkdownFile(filePath);
}