amazon-textract-response-parser
amazon-textract-response-parser copied to clipboard
aws-samples
add confidence to list
Issue #: #73
Description of changes: add confidence to query answer
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.
That is a API breaking change. Should have added more initially, was a quick hack to get it out.... Maybe use the trp2.convert_queries_to_list_trp2 (https://github.com/aws-samples/amazon-textract-textractor/blob/d324b360dec724fc40bf46fe9f2441e8e403903f/prettyprinter/textractprettyprinter/t_pretty_print.py#L147)
Or we can add another method. Often customers want bounding box information as well, makes sense to output in different formats maybe.
oh didn't see that, we can maybe add a option with default to no in that function similar what was done in that function. https://github.com/aws-samples/amazon-textract-textractor/blob/d324b360dec724fc40bf46fe9f2441e8e403903f/prettyprinter/textractprettyprinter/t_pretty_print.py#L179
but yes also another object would work well. @noxqs what's your take on it?
I think that could work also.. I depend on this feature as confidence levels are important. Its one of the two methods I have to offer feedback to my client re failing OCR. For now I monkey patch (sorry) get_query_answers in my projects, I understand this request is API breaking but to me the OCR result is only as good as the confidence level.
@schadem are you fine with the suggested solution if so I will update the PR.
It should be a new method as it is an API breaking change and impact customers already using the existing method. @tb102122