Lesson plan: generate PDFs from activity markdowns, redesign cover page with HTML template, and open via blob in new tab

[Copilot]

Lesson plan PDFs were concatenating the raw uploaded SOURCE_PDF documents instead of the LLM-generated markdown content, the cover/TOC page was built with bare iText API calls producing a plain unstyled layout, and the download forced a file save instead of opening in the browser.

Cover page — HTML template

Replaces the programmatic iText summary page with a new lesson-plan-cover.html template rendered via HtmlConverter, matching the existing activity PDF design language (LEARN-Hub logo, #2f70b3 blue headings, styled tables with alternating rows).

Placeholders populated at render time: {{logoDataUri}}, {{downloadDate}}, {{totalDuration}}, {{activityCount}}, {{activitiesTableRows}}, {{searchCriteriaSection}}, {{breaksSection}}.

Activity PDFs — request-map markdowns first, with DB and SOURCE_PDF fallback

The frontend already sends full activity objects (including markdowns[].content and markdowns[].landscape) in the lesson plan request body. PDFService.getActivityPdfs() now uses these request-map markdowns as the primary source, rendering each section (Deckblatt → Artikulationsschema → Hintergrundwissen) via MarkdownToPdfService. Falls back to DB markdowns if the request map is empty, then to SOURCE_PDF for older activities.

// 1. Use markdown content from the request map (fastest – already in memory)
List<byte[]> markdownPdfs = generatePdfsFromRequestMarkdowns(activityMap, activityName);

// 2. If not in request, look them up in the DB
if (markdownPdfs.isEmpty()) { ... markdownPdfs = generatePdfsFromMarkdowns(dbActivity); }

// 3. Fallback: SOURCE_PDF (from request documents array or DB)

getLessonPlanInfo — checks request markdowns first

The availability check now inspects inline markdown content in the request map first, before performing DB lookups, and returns canGenerateLessonPlan = true for any activity that has markdowns (in request or DB) or a SOURCE_PDF.

Open in new tab via blob URL

LessonPlanModal.handleDownload() now uses window.URL.createObjectURL(blob) + window.open(url, "_blank"), matching the openBlobInNewTab pattern from ActivityDetails.tsx. The blob URL is revoked after 10 seconds to release memory. The lesson plan controller response disposition is changed from attachment to inline.

Other

• MarkdownToHtmlService.getLogoDataUri() getter added so the cover page renderer can embed the logo without duplicating the loading logic.
• All user-supplied strings are HTML-escaped before injection into the cover page template.

---

📍 Connect Copilot coding agent with Jira, Azure Boards or Linear to delegate work to Copilot in one click without leaving your project management tool.

[milljoniaer]

@copilot , so right now only the cover page of the lesson plan gets rendered, the concateneted merged pdfs are missing. also please dont download directly, but do it via this blob we also have in the activity download section

[Copilot]

@copilot , so right now only the cover page of the lesson plan gets rendered, the concateneted merged pdfs are missing. also please dont download directly, but do it via this blob we...

Fixed in commit 971b862.

Missing activity PDFs (root cause): getActivityPdfs() was only looking up markdowns from the DB. The frontend already sends the full activity objects (including markdowns[].content and markdowns[].landscape) in the request body. The new logic uses those request-map markdowns as the primary source, falls back to DB markdowns if the map is empty, then falls back to the SOURCE_PDF as a last resort. getLessonPlanInfo() now also checks request-map markdowns first.

Blob / open in new tab: LessonPlanModal.handleDownload() now uses window.URL.createObjectURL(blob) + window.open(url, "_blank"), matching the openBlobInNewTab pattern from ActivityDetails.tsx. The URL is revoked after 10 s to avoid memory leaks.

Disposition: Lesson plan controller changed from "attachment" to "inline" so the PDF opens directly in the browser.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• jfrog-prod-use1-shared-virginia-main.s3.amazonaws.com
  • Triggering command: /usr/lib/jvm/temurin-21-jdk-amd64/bin/java /usr/lib/jvm/temurin-21-jdk-amd64/bin/java --enable-native-access=ALL-UNNAMED -classpath /home/REDACTED/.m2/wrapper/dists/apache-maven-3.9.11/a2d47e15/boot/plexus-classworlds-2.9.0.jar -Dclassworlds.conf=/home/REDACTED/.m2/wrapper/dists/apache-maven-3.9.11/a2d47e15/bin/m2.conf -Dmaven.home=/home/REDACTED/.m2/wrapper/dists/apache-maven-3.9.11/a2d47e15 -Dlibrary.jansi.path=/home/REDACTED/.m2/wrapper/dists/apache-maven-3.9.11/a2d47e15/lib/jansi-native -Dmaven.multiModuleProjectDirectory=/home/REDACTED/work/LEARN-Hub/LEARN-Hub/server org.codehaus.plexus.classworlds.launcher.Launcher clean compile -q (dns block)
  • Triggering command: /usr/lib/jvm/temurin-17-jdk-amd64/bin/java /usr/lib/jvm/temurin-17-jdk-amd64/bin/java -Dfile.encoding=UTF-8 -Xms256M -Xss8M -Djdk.util.zip.disableZip64ExtraFieldValidation=true -Dsun.misc.unsafe.memory.access=allow --add-opens java.base/sun.reflect.annotation=ALL-UNNAMED -Xmx14580m -classpath /opt/hostedtoolcache/CodeQL/2.24.3/x64/codeql/java/tools/semmle-extractor-java.jar:/opt/hostedtoolcache/CodeQL/2.24.3/x64/codeql/java/tools/lombok-javac-extend.jar com.semmle.extractor.java.JavaExtractor --jdk-version 17 --javac-args @@@/tmp/codeql-scratch-21fe057196e5a270/dbs/java/log/ext/javac.args (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)