Script to extract queries from JSON

Question

We have a pipeline to extract queries from PDF documents, async. Here's a sample from teh JSON
"  block_type='QUERY_RESULT', relationships=None, confidence=77.0, text='1 NON-SENSITIVE', "
This confirms that our AWS pipeline is working for us. 
However, now matter what combination of the sample AWS scripts we use, we get errors. 
Anybody out there have an idea on how to format the extraction script so that it will work as AWS intends?

Input In [29], in ()
     16 print('d  =', d)
     18 #get_query_answers
---> 19 query_answers = d.get_query_answers(page=page)
     21 #for x in query_answers:
     22 #    print(f"{image_filename},{x[1]},{x[2]}")
     24 print(tabulate(query_answers, tablefmt="github"))

File ~/anaconda3/envs/aws-local/lib/python3.9/site-packages/trp/trp2.py:569, in TDocument.get_query_answers(self, page)
    567 if answers:
    568     for answer in answers:
--> 569         result_list.append([query.query.text, query.query.alias, answer.text])
    570 else:
    571     result_list.append([query.query.text, query.query.alias, ""])

AttributeError: 'NoneType' object has no attribute 'text'

Answer

Could you please post the entire json response you are getting from Textract Queries API before it reaches the post-processing, extraction script? It would help in our debugging efforts. If you cannot post that publicly on a forum post, please open a customer support ticket with us and attach the image/pdf of concern, your entire code file, and the entire stack trace. Thanks.

Answer

Does this work for you? https://aws.amazon.com/blogs/machine-learning/specify-and-extract-information-from-documents-using-the-new-queries-feature-in-amazon-textract/

Script to extract queries from JSON

Relevant content