Add http://noveldex.io/ parser (#2471)

plugin/js/parsers/NoveldexParser.js

@@ -0,0 +1,200 @@
+"use strict";
+
+parserFactory.register("noveldex.io", () => new NoveldexParser());
+
+/**
+ * Parser for the http://noveldex.io/ site.
+ */
+class NoveldexParser extends Parser { // eslint-disable-line no-unused-vars
+    constructor() {
+        super();
+    }
+
+    /*
+     * NOTE: As of 2026-03, it seems that classes and ids on this site can't be
+     * trusted to remain stable. In a short timespan, over three separate
+     * occurences, multiple classes and at least one id has been changed.
+     */
+
+    /**
+     * Function for extracting a specific value from any of the schema/context
+     * "scripts" in the main page dom.
+     * 
+     * @param { Document } dom Source to extract from.
+     * @param { string } key The property to find (first found).
+     * @returns { unknown | null } The found property, or null.
+     * 
+     * @private
+     */
+    extractFromContext(dom, key) {
+        /** @type { NodeListOf<HTMLScriptElement> } */
+        const contextContainers = dom.querySelectorAll("script[type='application/ld+json']");
+
+        const property = Array.from(contextContainers, contextContainer => {
+            let context = null;
+
+            // Theoretically overkill but I have trust issues.
+            try {
+                context = JSON.parse(contextContainer.innerHTML);
+            } catch (e) {} // eslint-disable-line no-empty
+
+            return context?.[key] ?? null;
+        }).find(property => property != null);
+
+        return property;
+    }
+
+    /**
+     * @param { Document } dom
+     */
+    async getChapterUrls(dom) {
+        /**
+         * SVG check excludes all locked chapters.
+         * 
+         * @type { NodeListOf<HTMLAnchorElement> }
+         */
+        let chapterLinksElements = dom.querySelectorAll("div[id] div[data-state] a[href*='chapter']:not(:has(svg.lucide-lock))");
+
+        let chapterLinks = Array.from(chapterLinksElements, a => {
+            /**
+             * The separate the queries are needed because the first-of-type
+             * doesn't work in an all selector when the spans are are nested.
+             * 
+             * @type { NodeListOf<HTMLSpanElement> }
+             */
+            let titleSpans = a.querySelector("div:first-of-type").querySelectorAll("span");
+
+            return {
+                sourceUrl: a.href,
+                title: Array.from(titleSpans, span => span.innerText.trim()).join(" ")
+            };
+        });
+
+        return chapterLinks;
+    }
+
+    /**
+     * @param { Document } dom 
+     */
+    findContent(dom) {
+        // Use query to narrow search down to raw script tag contents.
+        const scripts = Array.from(dom.querySelectorAll("script:not([src])"), script => script.innerHTML);
+
+        // Look for blocks of text enclosed by invisible characters.
+        let mainRegex = /(?:[\uFEFF\u200B\u200C\u200D]+)(.+?)(?:[\uFEFF\u200B\u200C\u200D]+)/;
+
+        let foundContent = null;
+        for (const script of scripts) {
+            const match = mainRegex.exec(script);
+            if (match) {
+                foundContent = match[1];
+                break;
+            }
+        }
+
+        // FIXME: Error out?
+        if (!foundContent) return null;
+
+        // Undo JSON escape.
+        const unescaped = JSON.parse(`["${foundContent}"]`)[0];
+
+        // Parse partially unescaped HTML.
+        let elementified = document.createElement("div");
+        elementified.innerHTML = unescaped;
+
+        // Unescape HTML-escaped tags, and remove nested <p> tags.
+        Array.from(elementified.querySelectorAll("p"))
+            .filter(p => p.innerText.startsWith("<p>"))
+            .forEach(p => p.outerHTML = p.textContent);
+
+        return elementified;
+    }
+
+    /**
+     * @param { Document } dom
+     */
+    extractTitleImpl(dom) {
+        /** @type { HTMLHeadingElement } */
+        let storyTitle = dom.querySelector("main h1:first-of-type");
+
+        return storyTitle;
+    }
+
+    /**
+     * @param { Document } dom 
+     */
+    extractAuthor(dom) {
+        let author = this.extractFromContext(dom, "author");
+
+        /*
+         * Null-check is needed since typeof null is object for some
+         * unfathomable reason.
+         */
+        author = author != null && typeof author === "object" && author?.["name"] || null;
+
+        /*
+         * Translator team's name is stored in a massive escaped JSON blob under
+         * team -> name.
+         */
+        let translator = /(?:\\"team\\":.+?\\"name\\":\\")(.*?)(?:\\")/.exec(dom.body.innerHTML);
+
+        const combined = [author, translator[1]].filter(s => s != null && s !== "").join(", ");
+
+        return combined !== "" ? combined : super.extractAuthor(dom);
+    }
+
+    /**
+     * @param { Document } dom 
+     */
+    extractLanguage(dom) {
+        return dom.querySelector("html").getAttribute("lang");
+    }
+
+    /**
+     * @param { Document } dom
+     */
+    extractSubject(dom) {
+        /**
+         * @param { Document } dom
+         */
+        const extractGenres = (dom) => {
+            const genres = this.extractFromContext(dom, "genre");
+
+            if (!Array.isArray(genres)) return [];
+
+            return genres.filter(property => typeof property === "string");
+        };
+
+        /**
+         * @param { Document } dom 
+         */
+        const extractTags = (dom) => {
+            /** @type { NodeListOf<HTMLAnchorElement> } */
+            const tagAnchors = dom.querySelectorAll("a[href*='tag']");
+
+            return Array.from(tagAnchors, anchor => anchor.innerText);
+        };
+
+        return extractGenres(dom).concat(extractTags(dom)).join(", ");
+    }
+
+    /**
+     * @param { Document } dom 
+     */
+    extractDescription(dom) {
+        const description = this.extractFromContext(dom, "description");
+
+        return typeof description === "string" ? description : "";
+    }
+
+    /**
+     * @param { Document } dom 
+     */
+    findCoverImageUrl(dom) {
+        const url = this.extractFromContext(dom, "image");
+
+        if (typeof url !== "string") return null;
+
+        return url.startsWith("/") ? util.resolveRelativeUrl(dom.baseURI, url) : url;
+    }
+}

plugin/popup.html

@@ -794,6 +794,7 @@ <h3>Instructions</h3>
         <script src="js/parsers/NovelAllParser.js"></script>
         <script src="js/parsers/NovelcoolParser.js"></script>
         <script src="js/parsers/NovelCrushParser.js"></script>
+        <script src="js/parsers/NoveldexParser.js"></script>
         <script src="js/parsers/NovelFeverParser.js"></script>
         <script src="js/parsers/NovelfullParser.js"></script>
         <script src="js/parsers/NovelgoParser.js"></script>