Hmmm where to begin...

How about with reading the documentation?

The examples make it pretty clear that the parse method parses the table out into something the author calls "table states" and that those are available via $te->table_states() which returns a list of them. Then you have to use yet another method to get at the rows in that "table state." There is a shortcut whereby you can invoke the row method on the original TableExtract object (after calling the parse method) and it will return a list of the rows in the first table in the document.

I've never used the module. On first inspection, I think its interface could be improved.

-sauoq
"My two cents aren't worth a dime.";