Tables missing after page 1 - page Textract async results with NextToken

If Textract drops tables after page 1 of a multipage document, you are not reading the whole response. Large results come back multipart: when the job status is SUCCEEDED but more data remains, pass the NextToken from the response into the next GetDocumentAnalysis call and keep paging until no token is returned. Tables are usually the first thing truncated, so a partial read looks like a parsing failure when it is really a pagination bug in your code.

Context: Stack Overflow 76677782 (accepted answer): Textract returned table data only for page 1 of a multipage PDF, while text and form key-values came back fine for all pages. The accepted answer found the fix in the response pagination.

Source: Stack Overflow 76677782 (accepted answer): Textract returned table data only for page 1 of a multipage PDF, while text and form key-values came back fine for all pages. The accepted answer found the fix in the response pagination.

Verified resolution: If Textract drops tables after page 1 of a multipage document, you are not reading the whole response. Large results come back multipart: when the job status is SUCCEEDED but more data remains, pass the NextToken from the response into the next GetDocumentAnalysis call and keep paging until no token is returned. Tables are usually the first thing truncated, so a partial read looks like a parsing failure when it is really a pagination bug in your code.

Source URL: https://stackoverflow.com/questions/76677782