From c8b8ad20645333b8bf14f3b36facd8664142f7e7 Mon Sep 17 00:00:00 2001
From: Raiyn Aydin <aydinhrrs@gmail.com>
Date: Wed, 30 Sep 2026 17:06:31 +0800
Subject: [PATCH 1/2] fix(lightnovelwp): keep images and paragraph HTML in
 chapters

#2593 replaced the bottomnav-anchored regex with a walk that collected
only the text of <p> elements. Illustration chapters hold images and no
prose, so they came back empty, and ordinary chapters lost inline images
and formatting.

Select the first div.epcontent with cheerio and return its paragraphs as
HTML plus any images outside a paragraph, resolving lazy-load and
relative sources to absolute URLs. Scripts, styles and noscript are
dropped. A missing epcontent block still returns '' and bottomnav is not
needed. Bump the template version so every source picks up the fix.

Closes #2610
---
 plugins/multisrc/lightnovelwp/template.ts | 89 +++++++++--------------
 1 file changed, 35 insertions(+), 54 deletions(-)

diff --git a/plugins/multisrc/lightnovelwp/template.ts b/plugins/multisrc/lightnovelwp/template.ts
index 080b4a452..82d40c4ed 100644
--- a/plugins/multisrc/lightnovelwp/template.ts
+++ b/plugins/multisrc/lightnovelwp/template.ts
@@ -42,7 +42,7 @@ export class LightNovelWPPlugin implements Plugin.PluginBase {
     this.icon = `multisrc/lightnovelwp/${metadata.id.toLowerCase()}/icon.png`;
     this.site = metadata.sourceSite;
     const versionIncrements = metadata.options?.versionIncrements || 0;
-    this.version = `1.1.${10 + versionIncrements}`;
+    this.version = `1.1.${11 + versionIncrements}`;
     this.options = metadata.options ?? ({} as LightNovelWPOptions);
     this.filters = metadata.filters satisfies Filters;
 
@@ -438,64 +438,45 @@ export class LightNovelWPPlugin implements Plugin.PluginBase {
         throw error;
       }
     }
-    // The chapter text lives in the `epcontent` block. Older child themes
-    // closed it with a `bottomnav` div, but that wrapper is gone on current
-    // ones, so anchoring on it silently returned an empty chapter.
-    //
-    // Walk the markup with the parser rather than cutting the document with a
-    // regex: the block has to end at the element that closes it, and any
-    // landmark guessed from the surrounding text is wrong on some site. A
-    // <script> inside the block is prose-adjacent, not the end of it, so
-    // treating the first one as a boundary truncates the chapter.
-    let depth = 0;
-    let inside = false;
-    let found = false;
-    let isReadingParagraph = false;
-    let paragraph = '';
-    const paragraphs: string[] = [];
-
-    const parser = new Parser({
-      onopentag(name, attribs) {
-        if (name === 'div' && !inside) {
-          const className = attribs['class'] || '';
-          if (/(^|\s)epcontent(\s|$)/.test(className)) {
-            inside = true;
-            found = true;
-            depth = 1;
-          }
-          return;
-        }
-        if (!inside) return;
-        if (name === 'div') depth++;
-        else if (name === 'p') isReadingParagraph = true;
-      },
-      ontext(text) {
-        if (inside && isReadingParagraph && text.trim()) {
-          paragraph += text;
-        }
-      },
-      onclosetag(name) {
-        if (!inside) return;
-        if (name === 'p') {
-          if (paragraph.trim()) paragraphs.push(paragraph.trim());
-          paragraph = '';
-          isReadingParagraph = false;
-        } else if (name === 'div') {
-          depth--;
-          if (depth === 0) inside = false;
-        }
-      },
-    });
-
-    parser.write(data);
-    parser.end();
+    // The chapter lives in the `epcontent` block. Older child themes closed
+    // it with a `bottomnav` div, but that wrapper is gone on current ones, so
+    // select the element itself rather than cutting the document between
+    // landmarks: a <script> inside the block then cannot end it early either.
+    const $ = load(data);
+    const content = $('div.epcontent').first();
 
     // No epcontent block means this was not a chapter page (a login wall or
     // an error page served with 200). Returning the whole document here would
     // show that page's text as the chapter, so report nothing instead.
-    if (!found) return '';
+    if (!content.length) return '';
+
+    content.find('script, style, noscript').remove();
+    content.find('img').each((_, el) => {
+      const img = $(el);
+      let src = img.attr('src');
+      if (!src || src.startsWith('data:')) {
+        src = img.attr('data-lazy-src') || img.attr('data-src') || src;
+      }
+      if (!src) return;
+      if (src.startsWith('//')) src = 'https:' + src;
+      else if (src.startsWith('/')) src = this.site.replace(/\/$/, '') + src;
+      img.attr('src', src);
+    });
+
+    // Keep paragraphs as HTML so inline formatting and images inside them
+    // survive, plus images outside any paragraph: illustration chapters hold
+    // only images, sometimes in bare <div>s with no <p> at all.
+    const blocks: string[] = [];
+    content.find('p, img').each((_, el) => {
+      const node = $(el);
+      if (node.is('img')) {
+        if (!node.parents('p').length) blocks.push($.html(node));
+      } else if (node.text().trim() || node.find('img').length) {
+        blocks.push($.html(node));
+      }
+    });
 
-    return paragraphs.join('\n');
+    return blocks.join('\n');
   }
   async searchNovels(
     searchTerm: string,

From 2baad3b5543e83b1585e76ad5fe1858f4254a0cd Mon Sep 17 00:00:00 2001
From: Raiyn Aydin <aydinhrrs@gmail.com>
Date: Wed, 30 Sep 2026 17:12:52 +0800
Subject: [PATCH 2/2] fix(lightnovelwp): strip script attributes and resolve
 relative images

Drop on* attributes and javascript:, vbscript: and non-image data: URLs
from href/src in the returned chapter HTML, which the app renders as is.
Resolve every relative image path against the chapter page URL instead
of only root-relative ones.
---
 plugins/multisrc/lightnovelwp/template.ts | 34 +++++++++++++++++++++--
 1 file changed, 32 insertions(+), 2 deletions(-)

diff --git a/plugins/multisrc/lightnovelwp/template.ts b/plugins/multisrc/lightnovelwp/template.ts
index 82d40c4ed..ce450b846 100644
--- a/plugins/multisrc/lightnovelwp/template.ts
+++ b/plugins/multisrc/lightnovelwp/template.ts
@@ -451,6 +451,7 @@ export class LightNovelWPPlugin implements Plugin.PluginBase {
     if (!content.length) return '';
 
     content.find('script, style, noscript').remove();
+    const pageUrl = this.site + chapterPath;
     content.find('img').each((_, el) => {
       const img = $(el);
       let src = img.attr('src');
@@ -458,11 +459,40 @@ export class LightNovelWPPlugin implements Plugin.PluginBase {
         src = img.attr('data-lazy-src') || img.attr('data-src') || src;
       }
       if (!src) return;
-      if (src.startsWith('//')) src = 'https:' + src;
-      else if (src.startsWith('/')) src = this.site.replace(/\/$/, '') + src;
+      // Resolve relative paths against the chapter page, as a browser would.
+      if (!/^[a-z][a-z\d+.-]*:/i.test(src)) {
+        try {
+          src = new URL(src, pageUrl).href;
+        } catch {
+          return;
+        }
+      }
       img.attr('src', src);
     });
 
+    // The app renders this HTML as is, so drop event handlers and URLs that
+    // would run script.
+    content.find('*').each((_, el) => {
+      const node = $(el);
+      for (const name of Object.keys(el.attribs)) {
+        if (/^on/i.test(name)) {
+          node.removeAttr(name);
+        } else if (name === 'href' || name === 'src') {
+          const value = el.attribs[name]
+            .split('')
+            .filter(c => c.charCodeAt(0) > 32)
+            .join('')
+            .toLowerCase();
+          if (
+            /^(javascript|vbscript):/.test(value) ||
+            (value.startsWith('data:') && !value.startsWith('data:image/'))
+          ) {
+            node.removeAttr(name);
+          }
+        }
+      }
+    });
+
     // Keep paragraphs as HTML so inline formatting and images inside them
     // survive, plus images outside any paragraph: illustration chapters hold
     // only images, sometimes in bare <div>s with no <p> at all.