Restructure project files

lib/api/lodestone.js

@@ -0,0 +1,32 @@
+// Utilities for scraping data from the Lodestone
+
+/**
+ * @see https://stackoverflow.com/a/6969486
+ * @param {string}
+ * @returns {string}
+ */
+const regExpEscape = str => str.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+
+/**
+ * Creates a regular expression that matches a link to the named item and
+ * captures its EDB ID from the matched link's `href` attribute.
+ * @param {string} name
+ * @returns {RegExp}
+ */
+const itemLinkRegExp = name => new RegExp(`<a href="/lodestone/playguide/db/item/(?<id>[a-z0-9]+)[^"]+"[^>]*>(?<name>${regExpEscape(name)})</a>`, 'i');
+
+/**
+ * Gets the ID of the named item in Eorzea Database.
+ * @param {string} name
+ * @returns {Promise<string | undefined>}
+ */
+export async function findItemEDBID (name) {
+	// execute a search for the item's name
+	const searchURL = `https://na.finalfantasyxiv.com/lodestone/playguide/db/item/?q=${encodeURIComponent(name.replace(/\([^)]+\)|[&-]/g, ' '))}`;
+	const response = await fetch(searchURL);
+	const body = await response.text();
+	// find an `<a>` in the HTML response whose text exactly matches the name
+	const match = body.match(itemLinkRegExp(name));
+	// return the ID parsed from the URL in the `href` attribute
+	return match?.groups.id;
+}

lib/api/mediawiki.js

@@ -0,0 +1,188 @@
+// Extremely basic API client for MediaWiki
+
+import makeFetchCookie from 'fetch-cookie';
+
+function formDataBody (entries) {
+	let data = new FormData();
+	for (const [key, value] of Object.entries(entries)) {
+		if (value != null && value != false) {
+			data.set(key, value);
+		}
+	}
+	return data;
+}
+
+export class MediaWikiClient {
+	/**
+	 * Creates a new client. Remember to also call `.login()`.
+	 * @param {string} wikiURL Target wiki's MediaWiki path (i.e. the path that
+	 * contains `index.php` and `api.php`) without a trailing slash. For example
+	 * for English Wikipedia this would be `'https://en.wikipedia.org/w'`.
+	 */
+	constructor (wikiURL) {
+		this.wikiURL = wikiURL;
+		this.fetch = makeFetchCookie(fetch);
+	}
+
+	/**
+	 * Makes a GET request against `index.php`.
+	 * @param {Record<string, string>} params Query string parameters
+	 * @param {RequestInit} [options] Additional fetch options
+	 * @returns {Promise<Response>}
+	 */
+	fetchIndexGet (params, options = {}) {
+		return this.fetch(`${this.wikiURL}/index.php?${new URLSearchParams(params)}`, {
+			...options,
+			method: 'GET',
+		});
+	}
+
+	/**
+	 * Makes a JSON GET request against `api.php`.
+	 * @param {Record<string, string>} params Query string parameters
+	 * @param {RequestInit} [options] Additional fetch options
+	 * @returns {Promise<any>}
+	 */
+	async fetchApiGet (params, options = {}) {
+		const response = await this.fetch(`${this.wikiURL}/api.php?${new URLSearchParams({
+			...params,
+			format: 'json',
+		})}`, {
+			...options,
+			method: 'GET',
+		});
+		const body = await response.json();
+		if (body.error) {
+			throw new Error(`[${body.error.code}] ${body.error.info}`);
+		}
+		return body;
+	}
+
+	/**
+	 * Makes a JSON POST request against `api.php`.
+	 * @param {Record<string, string>} params Form data body parameters
+	 * @param {RequestInit} [options] Additional fetch options
+	 * @returns {Promise<any>}
+	 */
+	async fetchApiPost (params, options = {}) {
+		const response = await this.fetch(`${this.wikiURL}/api.php`, {
+			...options,
+			method: 'POST',
+			body: formDataBody({
+				...params,
+				format: 'json',
+			}),
+		});
+		const body = await response.json();
+		if (body.error) {
+			throw new Error(`[${body.error.code}] ${body.error.info}`);
+		}
+		return body;
+	}
+
+	/**
+	 * Obtains a login token for authenticating.
+	 * @returns {Promise<string>}
+	 */
+	async getLoginToken () {
+		const body = await this.fetchApiGet({
+			action: 'query',
+			meta: 'tokens',
+			type: 'login',
+		});
+		return body.query.tokens.logintoken;
+	}
+
+	/**
+	 * Obtains a CSRF token for making edits.
+	 * @returns {Promise<string>}
+	 */
+	async getCSRFToken () {
+		const body = await this.fetchApiGet({
+			action: 'query',
+			meta: 'tokens',
+		});
+		return body.query.tokens.csrftoken;
+	}
+
+	/**
+	 * Logs in with the given bot credentials.
+	 * @param {string} username
+	 * @param {string} password
+	 * @returns {Promise<void>}
+	 */
+	async login (username, password) {
+		const loginToken = await this.getLoginToken();
+		const body = await this.fetchApiPost({
+			action: 'login',
+			lgname: username,
+			lgpassword: password,
+			lgtoken: loginToken,
+		});
+		if (body.login.result === 'Failed') {
+			throw new Error(body.login.reason);
+		}
+	}
+
+	/**
+	 * Gets the current contents of the named item's wiki page and returns the
+	 * contents with the infobox updated to use the given EDB item ID.
+	 * @param {string} name
+	 * @returns {Promise<string>}
+	 */
+	async readPage (title) {
+		const response = await this.fetchIndexGet({
+			action: 'raw',
+			title,
+		});
+		return response.text();
+	}
+
+	/**
+	 * Updates the named page to the given text.
+	 * @param {string} title
+	 * @param {string} text
+	 * @param {string} summary Edit summary
+	 * @param {boolean} [minor] If true, this is a minor edit
+	 * @returns {Promise<any>}
+	 */
+	async editPage (title, text, summary, minor = false) {
+		const csrfToken = await this.getCSRFToken();
+		const body = await this.fetchApiPost({
+			action: 'edit',
+			title,
+			text,
+			summary,
+			minor,
+			bot: true,
+			watchlist: 'nochange',
+			token: csrfToken,
+			format: 'json',
+		});
+		return body;
+	}
+
+	/**
+	 * Gets the list of wiki pages that belong to the given category.
+	 * @param {string} name Category name including the `Category:` namespace.
+	 * @param {number[] | '*'} namespaces Integer namespace ID(s) or the string
+	 * `'*'`. If namespace IDs are provided, only pages in those namespaces will
+	 * be returned.
+	 * @param {string} limit Maximum number of items to return. Must be 500 or
+	 * less. I'm lazy and not supporting API paging so deal with it.
+	 * @returns {Promise<{pageid: number; title: string}[]>}
+	 */
+	async listCategoryPages (name, namespaces = '*', limit = 50) {
+		if (Array.isArray(namespaces)) {
+			namespaces = namespaces.join('|');
+		}
+		const body = await this.fetchApiGet({
+			action: 'query',
+			list: 'categorymembers',
+			cmtitle: name,
+			cmlimit: limit,
+			cmnamespace: namespaces,
+		});
+		return body.query.categorymembers;
+	}
+}

lib/util/diff.js

@@ -0,0 +1,30 @@
+// Helper for displaying edit changes so they can be manually verified
+
+import {execSync} from 'node:child_process';
+
+/**
+ * Converts a string from UTF-8 to base64.
+ * @param {string} str
+ * @returns {string}
+ */
+const toBase64 = str => Buffer.from(str, 'utf-8').toString('base64');
+
+/**
+ * terrible terrible terrible string diff helper for debugging
+ * @param {string} a
+ * @param {string} b
+ */
+export function diff (a, b) {
+	// base64 input strings before passing to shell to avoid escaping issues
+	// https://stackoverflow.com/a/60221847
+	// use tail to cut off useless file info lines
+	execSync(String.raw`bash <<- EOF
+		diff --color=always -u \
+			<(echo "${toBase64(a)}" | base64 -d) \
+			<(echo "${toBase64(b)}" | base64 -d) \
+		| tail -n +3
+	EOF`, {
+		// display result directly in terminal
+		stdio: 'inherit',
+	});
+}