Edit

Parser

The first stage of document processing is a Parser — an AutoML SDK component which removes HTML tags from training set documents, and then creates Elements in an AutoML SDK Document based on tags, defining the begin position and end position for each Element.

Elements created at this stage: Cell, Line, Page, Row, Table, Tag.

note

This step is configured automatically, there's no need to manually update anything.

Table Element is created with all the data within the <table> tag.

The following figure shows how the data from HTML code is parsed and which elements are created:

1. Two Row elements are created with text inside <tr> tags, because there are two rows in initial document.
2. Four Cell elements are created with text inside <td> tags, because there are four cells in initial document.
3. One Line element is created with text inside the <line> tag.