I have a platform that is built in-house that I am considering replacing with Retool. Retool meets all of the requirements except for one.
- A team of up to 25 resources in Hyderabad India that copy and past data from multi-page PDFs into an input for. In short, manual data extraction that cannot be automated due to the variety in the 10s of thousands of PDFs that we do at a given time.
- The platform that we are using in-house allows one of the resources mentioned about to highlight text in a PDF and then by clicking on the input field on our app - automatically pastes the highlighted text into the form field in our app.
- If my internal ouffshore development team has been able to build this, I am sure there is a way to have this done in Retool via a new component.
- Also open to suggestions on another approach to accomplish the same goal.
Hey @davemac! Interesting use case. I'm thinking through a couple of things that we have in the Retool ecosystem that might help:
- We have a PDF Viewer component, and expose a few properties:
pluginType - doesn't seem like any of these will help though
- The button component can be configured to copy a value to the clipboard (you can define the value via JS, reference it from another component, etc.)
- We can also create a JS query that updates a form field
It's possible you might be able to hack something together with these. If not, we do support custom components! Here's a link to our docs about 'em.
Curious to see your progress here, and if you have any more questions / need more guidance, we can schedule a call with an engineer.
@justin - maybe I didn’t explain it properly, but to clarify - there is a PDF that is open in one window - on the left side of the screen - and a form that would be in retool in a browser (Chrome of Chromium Edge) that would be on the right. There are up to 800 data points taken off of one PDF and currently, the researcher can highlight a segment of text in the PDF “on the left” and click into the browser form field (input text field) “on the right” and the highlighted text is automatically pasted in to the form field which makes the typical “CTRL C - change window - CTRL V” process more efficient. Speed is paramount as with the current project we are doing, we have about 25 offshore researchers parsing out up to 800 fields from over 19K documents or ~ 15MM datapoints. You would be surprised at how much time the “point and shoot” pasting saves us.
@davemac If I understand correctly here, you’d need to be able to read the user’s selection in another window from within Retool. Since all JS is run inside a sandbox in Retool for security, we aren’t able to log user input or monitor their actions outside of the environment. Like Justin is getting at above, you’d probably need to bring the PDFs inside of retool and parse that text value. The text annotation component might be an interesting route to explore in addition to the pdf viewer?
@alex that would work for me provided the I can highlight a selection in the PDF and "point and shoot" into the text input. If it's all in the same sandbox / environment - that's fine. We can also build some management around the PDF uploads into the same (re)tool