We’re building a tool that reads aloud what users type in Google Docs using text-to-speech. It needs to work on most platforms and browsers. We’re using Google Apps Script with a sidebar that can play audio.
The Snag
Getting the text from Google Docs is tricky. We can’t access it directly from the sidebar. Our current method:
Sidebar checks our cloud script regularly
Cloud script looks at the synced doc
Changes are sent back to sidebar
Sidebar plays the audio
But there’s a big delay - 1 to 3 seconds from typing to hearing. That’s too slow for us.
What We’re After
Is there a quicker way to get text changes? Can we bypass the cloud somehow?
Update
Looks like it’s not doable right now. A Chrome Extension might work by grabbing text from the Google Docs page, but it’s not easy.
have u considered using the Google Docs realtime API? it might give u faster updates than ur current setup. another idea is to use a websocket connection between the sidebar and a custom server. this could bypass some of the delay ur seeing with apps script. just my 2 cents!
Having worked on similar projects, I can attest to the challenges you’re facing. One potential solution worth exploring is leveraging the Google Docs Realtime API. This API provides near-instantaneous updates and could significantly reduce your latency issues.
Another approach that’s yielded good results in my experience is implementing a client-side text diff algorithm. By comparing the current document state with the previous one locally, you can identify changes much faster than relying solely on server-side comparisons.
If you’re open to alternative platforms, consider investigating collaborative editing frameworks like Yjs or Automerge. These offer real-time synchronization capabilities that might better suit your needs.
Ultimately, achieving true real-time performance within the constraints of Google Docs and Apps Script is challenging. You might need to consider a hybrid approach, combining multiple techniques to get as close to real-time as possible.
I’ve faced similar challenges when working on real-time document collaboration tools. From my experience, the latency you’re encountering is indeed a common hurdle with Google Apps Script.
One approach that yielded better results for us was using WebSockets for more immediate communication between the client and server. While it’s not natively supported in Apps Script, you can set up a separate WebSocket server and have your sidebar connect to it. This way, you can push updates to the client much faster.
Another option we explored was using the Google Drive API directly instead of Apps Script. It allows for more frequent polling and can reduce latency significantly. However, this requires more setup and might not be as user-friendly to deploy.
Lastly, if you’re open to moving away from the Google ecosystem, consider platforms like Etherpad or ShareDB. They’re designed for real-time collaborative editing and offer better control over the update frequency.
Remember, each approach has its trade-offs in terms of complexity and user experience. It’s worth prototyping a few options to see what works best for your specific use case.