From fa1801c979290007181cd2efbe01d88e3ad049c9 Mon Sep 17 00:00:00 2001
From: "brobert (aider)" <borja@brobert.net>
Date: Tue, 1 Apr 2025 15:03:48 +0200
Subject: [PATCH] fix: extract text content properly in epub generation

---
 generate-epub.ts | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/generate-epub.ts b/generate-epub.ts
index ffa92ba..229c33a 100644
--- a/generate-epub.ts
+++ b/generate-epub.ts
@@ -31,10 +31,14 @@ async function generateEpub() {
 
   // Convert each post to ePub chapter format
   for (const post of posts) {
-    const { document } = parseHTML(post.content);
+    // Parse the content HTML and extract text nodes
+    const { document } = parseHTML(`<div>${post.content}</div>`);
     
-    // Clean up content - remove unwanted elements if needed
-    const content = document.body.innerHTML;
+    // Get all text content from paragraphs and other text containers
+    const paragraphs = [...document.querySelectorAll('p, div, span')]
+      .map(el => el.textContent?.trim())
+      .filter(Boolean)
+      .join('\n\n');
 
     options.content.push({
       title: post.title,
@@ -42,7 +46,7 @@ async function generateEpub() {
       data: `
         <h1>${post.title}</h1>
         <p><em>Por ${post.author}</em></p>
-        ${content}
+        ${paragraphs || 'No content available'}
         <p><a href="${post.url}">Publicación original</a></p>
       `
     });