zoomscrape

Snippets for extracting the in-meeting chat out of a Zoom cloud-recording playback page when:

• The recording owner did not publish a separate chat.txt download, and
• You only have viewer / "Shared with me" access (no Zoom API), and
• The right-hand chat panel is a virtualized scroller — dragging to select across messages breaks as soon as you scroll, so the usual click-and-drag-then-Ctrl+C trick only captures the currently visible window.

The trick: the chat panel is just HTML. Scroll it programmatically from the DevTools console, harvest each batch of messages as the virtual list renders them, dedupe, and dump the result into a normal <textarea> on the page so you can copy it without fighting clipboard permissions.

Quick start

1. Open the Zoom recording playback page in your browser. Make sure the right-hand panel is on Audio Transcript or Chat Messages — whichever one you want to capture. (The script works for both; it doesn't care about the tab name, only about the scroller you anchor to.)
2. Right-click on one of the chat / transcript message bubbles in that panel → Inspect. This selects the element in DevTools so it's available in the Console as $0.
3. Open the Console tab (don't click anywhere else first — you need $0 to still be pointing at the message you inspected).
4. Paste the contents of scrape.js and press Enter.
5. A red-bordered textarea will appear in the top-left of the page containing every message it collected. Click into it, Ctrl+A, Ctrl+C, and paste wherever you need.

If you'd rather get a file instead of using the clipboard, after running scrape.js paste download.js and it will trigger a normal browser download of zoom_chat.txt.

How it works

• Walks up from $0 to find the row container (first ancestor with several similar siblings) — that's the virtual list.
• Walks further up to find the nearest scrollable ancestor (overflow-y: auto or scroll, with scrollHeight > clientHeight). That's the panel's scroller.
• Sets scrollTop = 0, then loops: scroll down ~one viewport, sleep ~350 ms so the virtualizer can render the next batch, harvest visible rows. Stops when scrolling stalls (already at the bottom).
• Dedupes by the first ~120 chars of each row's innerText.
• Dumps the joined result into <textarea id="__chat_dump__"> overlaid on the page, and also tries document.execCommand('copy') as a best-effort clipboard write.

Troubleshooting

Symptom	Likely cause	Fix
Right-click a chat message → Inspect first, then rerun.	$0 isn't pointing at the panel.	DevTools must be open; right-click inside the panel (on actual message text), then return to Console without clicking elsewhere.
No scrollable ancestor found	Anchored on a node that's outside the panel (e.g. a header/search box).	Right-click on the body of an actual message bubble, not the timestamp / search field.
Collected 0 messages	Row-parent walk-up landed on a wrapper with too few siblings.	Look at the Row parent (list) console log — confirm it's the list element. If not, click a different message and retry.
Far fewer messages than expected	Virtual list throttling, or scroll didn't reach bottom.	Re-run; or increase the sleep between scroll steps (350 → 700 ms) inside scrape.js.
Nothing in clipboard but the textarea has the text	execCommand('copy') was blocked because DevTools had focus.	Just click into the red-bordered textarea on the page and Ctrl+A / Ctrl+C manually. That's the supported path.

Why not just use the Zoom API / download the chat file?

In most cases you should — Zoom cloud recordings normally save the in-meeting chat as a separate meeting_saved_chat.txt next to the video, downloadable from the recording's file listing. This scraper only exists for the cases where:

• That file isn't offered (chat saving wasn't enabled, or the host stripped it before sharing), and
• You don't have API access to the host's Zoom account.

If either of those isn't true, prefer the official path.

Publishing this folder to nutanix-scratch/tools-scripts-docs

This directory was authored locally at C:\Users\KostadisRoussos\Documents\zoomscrape. To land it in the repo on the shadow-main branch:

git clone --branch shadow-main \
  https://github.com/nutanix-scratch/tools-scripts-docs.git
cp -r /c/Users/KostadisRoussos/Documents/zoomscrape \
  tools-scripts-docs/zoomscrape
cd tools-scripts-docs
git checkout -b add-zoomscrape
git add zoomscrape
git commit -m "Add zoomscrape: extract chat from Zoom recordings via DevTools"
git push -u origin add-zoomscrape
gh pr create --base shadow-main --fill

Adjust the branch / PR style to whatever the repo conventions are.