Document text

The client will extract all text from HL7 CCD files to construct a plain text document to send to the NLP API. It handles basic HTML formatting such as paragraphs and tables. This text is what will be stored in the database, and spans in results from the NLP API reference this text rather than the input XML file.

The client may make minor changes to text in order to help the NLP API pick up useful information, for example for section handling.