On Sunday, May 3, 2015 at 4:01:17 AM UTC-4, Axel Hecht wrote:
> On 5/2/15 3:07 AM, Zibi Braniecki wrote:
> > On Friday, May 1, 2015 at 3:47:57 PM UTC-4, Axel Hecht wrote:
> >> Maybe look at what treehugger does via setAnnotation?
> >> https://github.com/ajaxorg/treehugger/blob/master/lib/treehugger/tree.js
> > Interesting. So you think it would be better to give you col/line instead 
> > of pos? I like to think of the source as a long string, and avoid building 
> > any piece as white-space driven, so my idea would be rather to give you 
> > pos.start and pos.end.
> >
> > Would that work?
> In my experience, machines want the global offset, humans don't. Text 
> editors are "both", at least ace splits the doc into lines.
> 
> Which is a lengthy way of saying "depends on the use case", and the 
> optimizations for it.

True, my thinking is that the editor will format a source string into rows and 
cols, so it should be able to calculate it. Parser doesn't store row/cols, 
parser stores position in a source string, so it's easier for the parser to 
annotate it.

zb.
_______________________________________________
tools-l10n mailing list
[email protected]
https://lists.mozilla.org/listinfo/tools-l10n

Reply via email to