Amazon Textract: Is there a way to get just key/value pair instead of AnalyzeDocumentResponse or AnalyzeDocumentResult using java

Question

Trying to extract key value pair from pdf document using Java examples listed https://docs.aws.amazon.com/textract/latest/dg/example_textract_AnalyzeDocument_section.html.

From console, when I analyze document, I was able to download the result zip file with the following files but I am only interested in downloading keyValues.csv. Is there an api to just get the keyValues.csv instead of AnalyzeDocumentResponse/AnalyzeDocumentResult? I felt the responses are being too difficult to parse the key/value.

analyzeDocResponse.json
keyValues.csv
layout.csv
queryAnswers.csv
rawText.txt
signatures.csv
table-1.csv
table-2.csv
table-3.csv
table-4.csv

Answer

Thanks for using Textract service! Unfortunately we don't have an API to get keyValues.csv, you still need to parse the response and extract KV pairs.
Hope below links could help you extract KV pairs.

* Form data introduction: https://docs.aws.amazon.com/textract/latest/dg/how-it-works-kvp.html
* Python example code to extract KV: https://docs.aws.amazon.com/textract/latest/dg/examples-extract-kvp.html, you may want to implement similar function in java.
* Python helper lib: https://github.com/aws-samples/amazon-textract-textractor. This lib is an easy to use lib for parse Textract response and extract interested fields. If you are comfortable to switch to python, you may want to try this lib.
* Other example or helper lib: https://docs.aws.amazon.com/textract/latest/dg/other-examples.html

Amazon Textract: Is there a way to get just key/value pair instead of AnalyzeDocumentResponse or AnalyzeDocumentResult using java

Relevant content