User:Ioaxxere/auto-glossary.js

// This gadget works with the ] template to automatically create a glossary for a particular category via scraping.

mw.util.addCSS(`
	.mw-parser-output .auto-glossary a:empty::after {
		content: '';
		counter-increment: mw-numbered-ext-link;
	}
	.mw-parser-output .auto-glossary td:first-child {
		counter-reset: mw-numbered-ext-link; 
		text-align: center;
	}
	.mw-parser-output .auto-glossary ol {
		margin: 0 0 0 2em;
		padding: 0;
	}
`);

(async () => {
	for (let autoGlossary of document.querySelectorAll(".auto-glossary")) {
		let category = autoGlossary.getAttribute("data-category");
		let label = autoGlossary.getAttribute("data-label"); // The label parameter is currently unused.
		let lang = autoGlossary.getAttribute("data-language");
		let inlineOnly = autoGlossary.getAttribute("data-inline-only");

		// Get a list of words in the given category.
		let actionAPI = new mw.Api({ ajax: { headers: { "Api-User-Agent": "Gadget developed by ]" } } });
		let restAPI = new mw.Rest({ ajax: { headers: { "Api-User-Agent": "Gadget developed by ]" } } }); // Currently unused.
		let wordlist = ;
		let continueParam = "";
		while (true) {
			let response = await actionAPI.get({
				action: "query",
				list: "categorymembers",
				cmprop: "title",
				cmlimit: "max",
				cmtitle: "Category:" + category,
				cmcontinue: continueParam
			});

			// Only include entries from mainspace, appendix space, and reconstruction space.
			for (entry of response.query.categorymembers) {
				if (entry.ns === 0 || entry.ns === 100 || entry.ns === 118) {
					wordlist.push(entry.title);
				}
			}

			if (!response.continue) break;
			continueParam = response.continue.cmcontinue;
		}
		wordlist.sort((a, b) => (/^/.test(a) ? /^/.test(b) ? a.localeCompare(b, undefined, { sensitivity: "base" }) : -1 : /^/.test(b) ? 1 : 0)); // ChatGPTsort

		let loadingIndicator = document.createElement("div");
		let entriesLoaded = 0;
		autoGlossary.appendChild(loadingIndicator);

		let glossaryTable = document.createElement("table");
		glossaryTable.className = "wikitable";
		glossaryTable.style = "width: 100%; table-layout: fixed; display: table;";
		glossaryTable.innerHTML = "<thead><tr><th style='width:8em'>Terms</th><th>Definitions</th></tr></thead><tbody></tbody>";

		let batchSize = 500;
		for (let batchStart = 0; batchStart < wordlist.length; batchStart += batchSize) {
			if (batchStart == batchSize) {
				batchSize = 10e99; // Finish the rest of the list.
			}
			let ongoingRequests = 0;
			let glossaryQueries = await Promise.all(wordlist.slice(batchStart, batchStart + batchSize).map(async (word) => {
				// Limit concurrent requests, otherwise the browser might run of out memory.
				while (ongoingRequests >= 200) await new Promise(resolve => setTimeout(resolve, 10));
				ongoingRequests++;
				let response = await fetch("https://en.wiktionary.org/api/rest_v1/page/html/" + encodeURIComponent(word), {
					headers: { "User-Agent": "Gadget developed by ]." }
				});
				let responseText = await response.text();
				ongoingRequests--;

				let responseDocument = new DOMParser().parseFromString(responseText, "text/html");
				let row = null;

				try {
					let langID = lang.replaceAll(" ", "_");
					let L2_section = responseDocument.getElementById(langID).parentNode;
					glossaryDefs = Array.from(L2_section.getElementsByTagName("li")).filter(li => {
						// Check if the li element contains an inline category.
						if (Array.from(li.querySelectorAll(":scope > link, :scope > span > span > link"))
							.some(linkElement => linkElement.getAttribute("href").startsWith("./Category:" + category.replaceAll(" ", "_")))) {
							return true;
						}
						// Check if the headword element contains the category.
						let headwordLine = li.parentElement.previousElementSibling;
						if (!inlineOnly && headwordLine && headwordLine.tagName === "P" &&
							Array.from(headwordLine.querySelectorAll("link"))
								.some(linkElement => linkElement.getAttribute("href").startsWith("./Category:" + category.replaceAll(" ", "_")))) {
							return true;
						}
					});
					// If no definitions found: grab all top-level definitions.
					if (!inlineOnly && !glossaryDefs.length) {
						glossaryDefs = L2_section.querySelectorAll(".mw-parser-output section > ol > li");
					}
					// Insert a row if definitions were found.
					if (glossaryDefs.length) {
						row = document.createElement("tr");
						let entryCell = row.insertCell(0);
						let entryLink = document.createElement("a");
						entryLink.href = "https://en.wiktionary.orghttps://dictious.com/en/" + encodeURIComponent(word) + "#" + langID;
						entryLink.textContent = word;
						entryCell.appendChild(entryLink);

						let defCell = row.insertCell(1).appendChild(document.createElement("ol"));
						glossaryDefs.forEach(def => defCell.appendChild(def));

						// Convert URLs into absolute URLs.
						defCell.querySelectorAll("a").forEach(link => {
							link.href = new URL(link.getAttribute("href"), "https://en.wiktionary.orghttps://dictious.com/en/" + encodeURIComponent(word)).href;
						});

						// Clean HTML.
						defCell.querySelectorAll("link, .previewonly, .maintenance-line").forEach(elem => { elem.remove(); });
						defCell.querySelectorAll("*").forEach(elem => {
							Array.from(elem.attributes).forEach(attr => {
								if (attr.name != "class" && attr.name != "href" && attr.name != "style" && attr.name != "title" && attr.name != "rel") {
									elem.removeAttribute(attr.name);
								}
							});
						});
					}
				} catch {
					console.log("auto-glossary.js could not parse: " + word);
				}

				entriesLoaded++;
				loadingIndicator.textContent = `Loaded entry: ${entriesLoaded}/${wordlist.length}`;
				return row;
			}));

			if (batchStart == 0) {
				autoGlossary.appendChild(glossaryTable);
			}

			glossaryQueries.forEach(row => {
				if (row) {
					for (quote of document.querySelectorAll("ol > li")) {
						window.setupHiddenQuotes(quote);
					}
					glossaryTable.lastChild.appendChild(row); // Append each row into <tbody>.
				}
			});

		}
		// Clear out loading display.
		Array.from(autoGlossary.childNodes).forEach(elem => {
			if (elem != glossaryTable) {
				elem.remove();
			}
		});
	}
})();