Fastlane fix for Full Description

[sidhant947]

Fastlane fix for full description , followed by Source 1 , Source 2

[farfromrefug]

@sidhant947 what bugs me there is that it is also used for playstores.
Not sure of the right format for it to work on all platofrms

[campbellkelly054-gif]

Question? Not to sound too stupid here but what are your thoughts on automatic detection of header vs body vs footer vs credentials? I like how you fixed this particular issue. I've got a few of my own projects and I'd like to pick your brain.

[farfromrefug]

@campbellkelly054-gif i took a in between approach #713. Let see if it works

[campbellkelly054-gif]

• First, thankyou for being so quick to respond. Means alot.

• I'll give it a shot. I knew it would be size related on an ocr image but it gets interesting when you have text blocks that dont honor size conventions like h1 h2 etc. Or if you are just scanning a body text and there is no header/subheading/body structure. I prefer to use code over llm especially in mobile work. It's an easy reach to these days but I think it's lazy. The work I'm doing relies on having no network connection. Maybe a lan node system, maybe not. Dynamic in nature with a mobile device being an edge node.