Batch Process Multiple YouTube Links into Formatted YouTube Notes

[mjwSolver]

📝 Quick Introduction

Taking YouTube notes is a staple of my self-study journey and I found that the Templater templates shared in a discussion to be very useful in retrieving the information (such as channel name and duration) but would work on one video at a time, when I have 10s of them.

Check out the discussion that inspired me: Get Title, Channel Name and Upload Date for YouTube link

I don't think there's any Queue / Batch Processing made for multiple YouTube links yet, so I'd like to share the code I made but incorporates a loop to go through all listed YouTube links from a separate markdown file to create a YouTube note one by one.

💡 Feature Highlights

1. Converts all YouTube links stored in a markdown file into it's own note in a designated folder.
2. Checks for existing notes and prevents duplicates using the tp.file.find_tfile command in the destination folder.
3. Self-contained, everything is in one file (no dependencies from other notes).

☝️ Future Works

It's possible to extend this functionality to create multiple YouTube notes for videos inside a YouTube playlist, given the playlist link. It'll be a future side-project I would be working on.

The Template 📄

Prerequisites:

1. A target folder for output (sources/youtube works for me and you can change it in the template later).
2. A markdown file for storing a list of YouTube url (Queue for YouTube Link.md)
3. The following template as yt queue.md:

<%*
// --- CONFIGURATION ---
const outputFolder = "sources/youtube";
// --------------------

const notice = (msg) => new Notice(msg, 10000);

// --- Read all links from the file ---
const fileContent = tp.file.content;
if (!fileContent || fileContent.trim() === "") {
  notice("❌ The file is empty.");
  return;
}
const links = fileContent.split('\n').filter(link => link.trim() !== "");
if (links.length === 0) {
  notice("❌ Could not find any links.");
  return;
}

notice(`▶️ Starting to process ${links.length} links...`);
let createdCount = 0;
let skippedCount = 0;

for (const link of links) {
  // --- 1. FETCH AND PARSE ---
  let doc;
  try {
    // YOUR FIX APPLIED: Use the link directly without splitting it.
    const page = await tp.obsidian.request({ url: link });
    const p = new DOMParser();
    doc = p.parseFromString(page, "text/html");
  } catch (e) {
    notice(`❌ Failed to fetch page for ${link}. Skipping.`);
    skippedCount++;
    continue;
  }

  const $ = (s) => doc.querySelector(s);
  if (!$("meta[name='title']")) {
      notice(`❌ Could not find video data for ${link}. Skipping.`);
      skippedCount++;
      continue;
  }

  // --- 2. EXTRACT AND FORMAT ---
  const title = $("meta[name='title']").content;
  const cleanTitle = title.replaceAll(/[^a-zA-Z0-9 ]/g, "").trim();
  const fileName = `${cleanTitle}.md`;
  const filePath = outputFolder ? `${outputFolder}/${fileName}` : fileName;

  if (app.vault.getAbstractFileByPath(filePath)) {
    skippedCount++;
    continue;
  }

  const shortlinkUrl = $("link[rel='shortlinkUrl']").href;
  const durationStr = $("meta[itemprop='duration']").content.slice(2, -1);
  const uploadDate = $("meta[itemprop='uploadDate']").content;
  const authorName = $("span[itemprop='author'] > link[itemprop='name']").getAttribute("content");

  // Format Duration
  const timeStr = (time) => time.toString().padStart(2, '0');
  let [minutes, seconds] = durationStr.split("M");
  let durationSummary = "Seconds";
  let hours = Math.floor(Number(minutes) / 60);
  minutes = (Number(minutes) % 60);
  if (parseInt(minutes, 10) > 0) { durationSummary = "Minutes"; }
  let formattedDuration = `${timeStr(minutes)}:${timeStr(seconds)}`;
  let yamlDuration = "00:" + formattedDuration;
  if (hours > 0) {
    formattedDuration = `${timeStr(hours)}:` + formattedDuration;
    durationSummary = "Hours";
    yamlDuration = formattedDuration;
  }

  // Format Dates
  const formatDate = (date) => {
    let dateString = new Date(date.split('T')[0]).toDateString();
    let [dayString, month, dayNumber, year] = dateString.split(' ');
    let cleanDayNumber = dayNumber.replace(/^0+/, '');
    return `${month} ${cleanDayNumber}, ${year}`;
  };
  const formatDateToISO = (date) => date.split('T')[0];
  const finalUploadDate = formatDate(uploadDate);
  const isoUploadDate = formatDateToISO(uploadDate);


  // --- 3. BUILD THE NOTE CONTENT STRING ---
  const newNoteContent = `---
aliases: ${title}
channel_name: ${authorName}
duration: ${yamlDuration}
uploaded: ${isoUploadDate}
---

# by ${authorName}
## ${title}
[${title}](${shortlinkUrl})
${formattedDuration} ${durationSummary} / ${finalUploadDate}
`;

  // --- 4. CREATE THE NOTE ---
  try {
    await app.vault.create(filePath, newNoteContent);
    createdCount++;
  } catch (e) {
    notice(`❌ ERROR creating note for ${title}. Aborting.`);
    return;
  }
}

// --- FINAL SUMMARY ---
notice(`✅ Finished! Created ${createdCount} new notes. Skipped ${skippedCount}.`);

// Optional: Clear the queue file after processing
// await app.vault.modify(tp.file.file, "");
%>

Feel free to make changes to suit your template, but note that you need to define the properties in JavaScript outside of the string template, you can't include the Templater starting and ending tags (I don't know what they're called, but you get the idea).

🧑‍💻 How to use

1. Move the yt queue.md into your Templater folder
2. Create a new MD file titled Queue for YouTube Link.md and insert a few YouTube links (I've got some links prepped in the code block below for you to test it with)
3. Command + P / Control + P > Templater: Open insert template modal
4. Type yt queue and press enter.
5. The links should be parsed one by one and the output is in the target folder. Notices on top right should indicate it's progress and status.

Here's a few YouTube links that should help you test it out:

1. FromSergio video about the templater plugin
2. Ben's Cyber Life video about his Obsidian setup
3. Jonathan .W. Pritchard video for the Tasks plugin

https://youtu.be/QpJbeP8f55A
https://youtu.be/ZQTj8ZSDFw4
https://youtu.be/t6OLlwekVI0

💭 After thought

I previously experimented by using the tp.file.include() so I can dynamically include the different templates I want to use, for example my template for YouTube shorts and long form YouTube videos aren't the same, but I wasn't successful in implementing it, as passing the URL to the template was difficult, perhaps a tp.user function as a work around - but I didn't want to stretch it that far.

✌ Best of luck!

I hope this template helps out other people as it did for me. Thanks and best of luck in your endeavors!

Edit:

• Previously mentioned to "Find the Queue for YouTube Link file" and changed to "Create a new MD file"

[filmgal]

Thank you so much for this! It is exactly what I was looking for and works perfectly.
I'm no programmer but I'll see if I can figure out how to use this process to create a film file using my film template for each link in a film.md file. Wish me luck.

[filmgal]

Hmm. I can't find how the template knows to look at the /sources/youtube/Queue for YouTube Link.md file.
There is no mention of that file anywhere in the template. It's not like it is the only md file in that folder because it is the output folder for all the youtube links turned to md pages.

[mjwSolver]

Hi @filmgal, thank you for your comment.

The Queue for YouTube Link.md file is a blank-new markdown file that's created in your vault. (Ctrl + N or ⌘ + N)
It can be located anywhere in there and I didn't exactly mention this as part of my original post. My bad.

You then use the yt queue.md template with the Templater command "Open insert template modal" on that file (Queue for YouTube Link.md) and assuming you have valid YouTube links in the file. It should work.

Apologies for any troubles caused, I should probably revise my entry to explicitly state that you'll need to create the file and not "find it". That's on me hahaha.

Hope I've cleared that up for you. 👍

[filmgal]

Hi @mjwSolver, I appreciate the response. Thank you.
So I don't have to have a file called "Queue for YouTube Link.md" as per Step 2 in the instructions?
Create a new MD file titled Queue for YouTube Link.md and insert a few YouTube links (I've got some links prepped in the code block below for you to test it with)?
Cause I had that file and just kept using that same file, replacing the existing urls with new ones.
You are saying that your program will just look to the most recently created file for urls in its content? Or am I misunderstanding your reply?

[mjwSolver]

Sort of, you can rename the file "Queue for YouTube Link" with any other name or even use an existing markdown file containing just YouTube links whichever best suits you.

It sounds like you're already using template correctly though, based on your statement:
Cause I had that file and just kept using that same file, replacing the existing urls with new ones.

Short answer, No.
The program doesn't look at the most recently created file as the reference. The program works on any open and active file the moment the template is inserted. If it contains valid YouTube links then the program will run smoothly.