Advanced Features

JSON Export

Sticks the raw TranscriptData as JSON inside the HTML file so you can extract it later programmatically.

Option: includeJson: true

const html = generateTranscriptHtml(data, {
  includeJson: true,
});

The JSON is embedded as a <script> tag with type application/json:

<script type="application/json" id="discord-transcript-json">
  {"channel":{"id":"123","name":"general",...},"messages":[...]}
</script>

Extracting the data

In a browser:

const el = document.getElementById("discord-transcript-json");
const data = JSON.parse(el.textContent);
console.log(data.messages.length);

With Node.js (parsing the HTML file):

import { readFileSync } from "fs";

const html = readFileSync("transcript.html", "utf-8");
const match = html.match(
  /<script type="application\/json" id="discord-transcript-json">([\s\S]*?)<\/script>/
);
if (match) {
  const data = JSON.parse(match[1]);
  console.log(`${data.messages.length} messages`);
}

---

Pagination

Breaks the transcript into pages with Previous/Next buttons at the bottom.

Option: messagesPerPage: <number>

const html = generateTranscriptHtml(data, {
  messagesPerPage: 50,
});

• Splits messages into chunks of the given size
• Shows Previous / Page X of Y / Next at the bottom
• Only the current page is visible
• Works alongside search and other features

tip

For very large transcripts (1000+ messages), pagination significantly improves browser performance since only one page of messages is rendered at a time.

Choosing a page size

Messages	Recommended messagesPerPage
< 200	No pagination needed
200–500	100
500–2000	50
2000+	25–50

---

Custom Component Renderers

Let you swap out how specific component types get rendered. Handy if you want custom branding or need to handle a component differently.

Option: customRenderers: Record<number, (component: DiscordComponent) => string>

import { generateTranscriptHtml } from "@getoeteter/discord-transcripts";
import type { DiscordComponent } from "@getoeteter/discord-transcripts";

const html = generateTranscriptHtml(data, {
  customRenderers: {
    // Override button rendering (type 2)
    2: (component: DiscordComponent) => {
      const label = component.label || "Button";
      return `<div class="my-custom-button">${label}</div>`;
    },

    // Override container rendering (type 17)
    17: (component: DiscordComponent) => {
      return `<div class="my-container">${component.content || ""}</div>`;
    },
  },
});

How it works

1. Before rendering a component, we check if you've provided a custom renderer for that type
2. If you have, we call it with the full DiscordComponent object
3. Your function needs to return an HTML string
4. Child components are not automatically rendered - your custom renderer handles the full output

Component type reference

Type	Name	Use case for override
1	ActionRow	Custom layout
2	Button	Custom styling or click handlers
3	StringSelect	Custom dropdown rendering
9	Section	Custom section layout
10	TextDisplay	Custom text rendering
12	MediaGallery	Custom gallery layout
17	Container	Custom container styling

caution

Custom renderers receive raw component data. You are responsible for sanitizing any user content in your output. Use HTML escaping for any text you render.

Embed Media

Discord CDN links expire after a while, so images, avatars, and other media in your transcripts can break over time. Enable embedMedia to download all media and embed it directly in the HTML as base64 data URIs - the transcript becomes fully self-contained.

Option: embedMedia: true

With createTranscript (discord.js)

const result = await createTranscript(channel, {
  embedMedia: true,
  theme: "dark",
});

With generateTranscriptHtml (standalone)

import {
  generateTranscriptHtml,
  resolveMediaUrls,
} from "@getoeteter/discord-transcripts";

const mediaMap = await resolveMediaUrls(data);
const html = generateTranscriptHtml(data, {
  _resolvedMedia: mediaMap,
});

What gets embedded

• User avatars and guild icons
• Image, video, and audio attachments
• Embed images, thumbnails, author icons, and footer icons
• Custom emoji in messages, reactions, buttons, and polls
• Stickers
• V2 component media (thumbnails, galleries, files)

Caveats

• Output files will be much larger since all media is base64-encoded (roughly 33% larger than the original files).
• Each resource has a 50 MB size limit and 30 second download timeout. Failed downloads gracefully fall back to the original URL.
• Media is downloaded with up to 10 concurrent requests.