Restricting PDF text extraction to only specific coordinates

Yes, PDFTextStream can be used to extract content from specified coordinates on one or many pages in a document.

com.snowtide.pdf.RegionOutputTarget is an com.snowtide.pdf.OutputHandler implementation that allows you to define areas on a page where the desired content is positioned. Once a page is piped through the RegionOutputTarget instance, the content found in the specified regions is then available from the RegionOutputTarget , optionally mapped to user-defined names.

This is particularly handy for extracting form data where each field is known to be positioned in a specific location (perhaps on a specific page in each document of a particular type): each form fields' coordinates can be defined in the RegionOutputTarget and given a name ("address" or "ID No." or "field-553" – whatever is appropriate for your application). Then that RegionOutputTarget can be queried for the text values of those fields after a PDF Page has been piped through it.

Some example code is available on RegionOutputTarget's API reference page.