Need help formatting text input from OCR Model

[anthropaulogy]

Hi all!

I am creating a fitness app for myself, and I am trying to take a screenshot from my smart scale app, run it through an OCR and then be able to record the 13 or so metrics automatically.  The problem is the my smart scale is recording answers such as "165.6lb" or "31.3%", which is causing the OCR to either not capture the data properly if the fields are Numbers, Decimals, or Percents.  If the field is text, it captures the data properly, but now it cannot be used as a Decimal or Percent in my data. 

My current thoughts are to have the app import to the Raw Measurements tables that is all text fields, and when saved, triggers an action to create a row on the Formatted Measurements table, and formats the data into into the appropriate formats.  I am not sure how to create a formula that looks for the "lb" or "%" in a text field, and removes it.  Any ideas?

[Suvrutt_Gurjar]

---

@anthropaulogy wrote:

My current thoughts are to have the app import to the Raw Measurements tables that is all text fields, and when saved, triggers an action to create a row on the Formatted Measurements table, and formats the data into into the appropriate formats.  I am not sure how to create a formula that looks for the "lb" or "%" in a text field, and removes it.  Any ideas?

---

It could be possible depending on how the text string is. AppSheet has a fairly good collection of TEXT() and EXTRACT() functions that could be used. However point to note is that the text string will need to have fairly stable format because splitting of text will depend on certain text position based logic etc. However feasibility can be more clearly ascertained if you could share a sample text string.

To give you an idea, some sample solutions for similar requirements below

Solved: Extract text and numbers - Google Cloud Community

Solved: Extract text - country from address - Google Cloud Community

Solved: Extract some text from List - Google Cloud Community