Swirl’s team is working hard to create a simple, safe, and user-friendly way for businesses to search for information. We’re creating a secure, quick solution that uses top-notch technology to help you get more done at work which is Swirl Search. With this, we are excited to share that we now have the new AutomaticPayloadMapperResultProcessor. The Automatic Payload Mapper Result Processor automatically organizes the data from search providers into the Swirl system, so you don’t have to map out the results manually. It automatically maps fields/columns returned from SearchProviders to the Swirl schema without needing to use manual result_mappings.
This has just been merged into the develop branch for evaluation purposes. To try it out… first, verify that you are on the correct branch:
The response should confirm you are on develop, e.g., Your branch is up to date with ‘origin/develop.’ From there, do:
Finally, modify a Swirl SearchProvider as follows:
Remove the result_mappings, including especially NO_PAYLOAD
Add the AutomaticPayloadMapperResultProcessor to the result_processors list
Run a search against the SearchProvider, e.g., mytag:. Swirl will match response columns with the Swirl fields (URL, title, body, and date_published) automatically, as follows:
For the title, it looks for the most populated text field with a median length of 50 or the most populated number (int or float) as a fallback.
For the body, it selects the longest, most populated text field
For the date_published, it will select the most populated date field.
For the URL, it will select the most populated text field that holds only URLs (that start with http:// or https://)
See below examples of successful responses from Google PSE, BigQuery, and ThoughtSpot, all done without result_mappings!
Thank you for reading our post. If you want to try Swirl, please visit our GitHub repository at https://github.com/swirlai/swirl-search. Should you need help or support, or if you’d like the SearchProviders. Please join our Swirl community on Slack by following this link: Swirl Community Slack. We appreciate your interest and are here to assist you!
Sign up for our Newsletter
Bringing AI to the Data
Stay in the loop with the SWIRL Community get the latest news, articles and updates about AI.