Extending Carota

[mpage38]

Hello Daniel,
really an awesome work.
I am especially interested in the rendering part of Carota. The reason is that there is currently no satisfactory solution to draw HTML text on a canvas.

1. There is html2canvas (http://html2canvas.hertzen.com/) but it fails on simple <ol> and <ul> lists.
2. In the future, SVG foreignObject will do the job, but it is not implemented in IE, including IE11.
3. There is also a solution based on Blob and createObjectURL but on IE<10 it fails as well.

From what I see, Carota is the best candidate to do the job.

In the short term, I would like to extend it to make it able to render other HTML tags, and other styling properties.
For the HTML part, I would like to implement <ol> and <ul> rendering. For CSS, I would like to implement margin properties, as well as line-height

To achieve this, I have a few questions.

1. I guess I first need to implement some kind of box model into Carota. Am I right ?
2. If yes, I imagine that I need to replace the line data structure with something that is more structured (box) ?
3. I don't fully understand why line is made of words. I would have thought that a line is made of runs, i.e. text chunks with homogenous styling. Is it because of the justification algorithm ?
   Best,
   Michel