ChatGPT Generates Regex for AIDC App

Regular Expressions (aka “regex”) are already commonly used to increase the accuracy and efficiency of data capture and identification tasks, reducing errors and improving productivity. The CodeREADr app for iOS and Android OS devices has enabled the use of regex for many years now.

ChatCPT is a new tool to write regex code. Simply describe to Chat GPT the pattern and results you need and in seconds it generates the code. The website Regex101 can be used to test the code before deploying it.

Using these tools will allow non-developers to add regex codes to improve their data capture and collection accuracy. They will save time and money, as well.

There are at least four (4) areas to use regex to improve accuracy with the CodeREADr barcode scanning software.

• Alter Scanned Values. Here regex is used to change the scanned value to match the required result in a database.
• Alter Responses to Scanned Values. Here regex is used to display a specific response to an app user when a certain pattern is detected.
• Export Templates to Normalize Data. Here regex is used to change the format of the scanned data to match the format required in the target database.
• Custom On-Device Validation (COV). Here regex is used to stop app users from scanning barcodes that don’t match specific patterns. COV is a sophisticated, edge computing technology based on no-code/low-code implementations on the scanning device itself.

We found with ChatGPT your query needs to be very explicit based on the result you need. For example, to use the result with CodeREADr the query would be asked like this:

“Please provide a regex pattern and substitution input using an online regex tester like regex101.com to display the values from a UPC barcode with the leading zero and check digit removed.”

That is how we received the correct response in our example below under “Remove the Leading Zero and Check-Digit from UPC Barcode.” Note that if an input string is multiline, ask ChatGPT to include the multiline flag in the regex pattern. We did that for the GS1 examples below.

Example: Remove the Leading Zero and Check-Digit from UPC Barcode

Scanned UPC Value: 012345678901

Regex from ChatGPT

Pattern: ^0(\d{10})\d$

Replacement: $1

Regex101: https://regex101.com/r/rbcdLi/1

Result: The value “1234567890” will now accurately match the value in the inventory database.

Example: Serial Number Extraction from GS1 Code-128

Scanned Value with serial number 12345678: (01)05410013111009(10)ABC123(21)12345678

Scanning Result

GTIN = (01)05410013111009

BATCH/LOT = (10)ABC123

SERIAL = (21)12345678

Regex from ChatGPT

Pattern: (?s).*\(21\)(\d+).*

Replacement: $1

Regex101: https://regex101.com/r/HjpEOL/1

Result: The serial number “12345678” can now be used to track where items have been shipped and deployed.

Example: Lot  Number Extraction from GS1 Code-128

Scanned Value with lot number 10003456789: (01)00314141999995(10)10003456789(21)20006500234

Scanning Result

GTIN = (01)00314141999995

BATCH/LOT = (10)10003456789

SERIAL = (21)20006500234

Regex from ChatGPT

Pattern: (?s).*\(10\)(\d+).*

Replacement: $1

Regex101: https://regex101.com/r/m93R0i/1

Result: The lot number “10003456789” can now be used to verify if it’s a recalled item.

The future of using ChatGPT with CodeREADr looks very promising, as both technologies can complement each other for various use cases, especially when it comes to data capture, data collection, data connections, and process automation. Please email support@codereadr.com with “ChatGPT” in the subject line to learn more.