Comments column not extracted

[CMCDragonkai]

This worked quite well for all the columns and rows, but for some reason the comments column wasn't extracted (it's all text of course). It looked like this:

[jeremybmerrill]

@CMCDragonkai, there's a known issue (related to #99) where the detected selection area is just a little bit too small. If you expand the selection on the right just a little bit -- so it includes the border, not just the text -- then the comments column should be extracted too.

[CMCDragonkai]

I see, does is the bounding box detection based on image processing or something else? Surely there must be a way to match what is contained.

[jeremybmerrill]

The bounding box detection -- for PDFs that support it -- uses the line elements in the PDF. So it totally should work, I just think we're at a bit of an impasse about the solution.