Tweets cleanup for sentiment analysis

Hi,
I have been testing nri-flex twitter monitor written by Sean Winters and everything works fine. But I noticed all the tweets that start with @username are not processed by sentiment analysis api.
Can you please advise on how I can cleanup the tweets? Python use regex, is it possible to use regex in nri-flex yml files? If yes, please provide syntax.

Regards,
Emile

Hey there @emile.amewoto,

I believe that you’re referring to the example found here: https://github.com/newrelic/nri-flex/blob/master/examples/twitter-monitor.yml? If so, the best place to receive support for that would be the repository’s Issues tab. Global Technical Support isn’t able to debug flex scripts for customers, but by filing an issue, the maintainer of those examples may be able to help you out. To file an issue, you’ll want to go here: https://github.com/newrelic/nri-flex/issues

As for the Regex, the Flex Integration itself doesn’t do any of the actual parsing. Instead, it runs the commands that you give it, and ingests the data that’s returned. If you were looking to do some Regex parsing, the best thing to do would be to write a separate script (Python if you wanted) that handles the parsing for you, and then have the Flex Integration run that script at a timed interval. For more information about building your own Flex Integration, you can check out our documentation here: New Relic Flex: Build your own integration | New Relic Documentation

Thank you Alex for the update