From eb4b713e9d5c66bd05e99d3dc6ff5e1b7df93e7c Mon Sep 17 00:00:00 2001
From: nmemmert <67961292+nmemmert@users.noreply.github.com>
Date: Mon, 3 Aug 2026 10:13:56 -0400
Subject: [PATCH] =?UTF-8?q?Fix=20docx=20import=20missing=20lesson=20conten?=
 =?UTF-8?q?t=20=E2=80=94=20include=20h1-h6=20block=20elements?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

mammoth outputs week/quiz headings as <h2>/<h1>, not <p><strong>. The
block-element regex only matched <p>, so all headings were skipped and no
lessons were ever created. Expand regex to h[1-6]|p|ul|ol and treat any
h-tag as a heading unconditionally.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 server.js | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/server.js b/server.js
index 342ab7b..598f182 100644
--- a/server.js
+++ b/server.js
@@ -44,15 +44,15 @@ async function parseUnitDocxBuffer(buffer, filename) {
 
   // Walk the HTML and collect typed elements
   const elements = [];
-  for (const m of stripped.matchAll(/<(p|ul|ol)(?:\s[^>]*)?>[\s\S]*?<\/\1>/gi)) {
+  for (const m of stripped.matchAll(/<(h[1-6]|p|ul|ol)(?:\s[^>]*)?>[\s\S]*?<\/\1>/gi)) {
     const tag = m[0].match(/^<(\w+)/)[1].toLowerCase();
     const inner = m[0].replace(/^<[^>]+>/, '').replace(/<\/\w+>\s*$/, '');
 
-    if (tag === 'p') {
+    if (tag === 'p' || /^h[1-6]$/.test(tag)) {
       const text = htmlToText(inner);
       if (!text) continue;
-      // A paragraph whose only non-whitespace content is wrapped in <strong> is a heading
-      const isHeading = /^<strong>[^<]+<\/strong>\s*$/.test(inner.trim());
+      // h1-h6 are always headings; <p> is a heading only if its sole content is <strong>
+      const isHeading = /^h[1-6]$/.test(tag) || /^<strong>[^<]+<\/strong>\s*$/.test(inner.trim());
       elements.push({ kind: isHeading ? 'heading' : 'para', text });
     } else {
       const items = [...inner.matchAll(/<li[^>]*>([\s\S]*?)<\/li>/gi)]