On Thu, Jun 23, 2011 at 11:28 AM, Ryosuke Niwa <rniwa@webkit.org> wrote: > On Thu, Jun 23, 2011 at 11:06 AM, Anne van Kesteren <annevk@opera.com>wrote: > >> On Wed, 11 May 2011 06:58:50 +0200, Robert O'Callahan < >> robert@ocallahan.org> wrote: >> >>> On Wed, May 11, 2011 at 3:43 AM, Ojan Vafai <ojan@chromium.org> wrote: >>> >>>> Similarly, does it need to disambiguate line-wrapped positions? >>>> >>>> "Line 1<wrap> >>>> Line 2" >>>> >>>> After Line 1 and before Line 2 are equivalent DOM positions. >>>> >>>> I think we might be able to have a single flag for both of these. The >>>> case of "Latin <wrap here>HEBREW", in which case the cursor could be at the >>>> end of the first line or the end of the second line. But being at the >>>> beginning of the first line would be a different DOM position. >>>> >>>> I expect there's a case I'm not thinking of and we might need two flags. >>>> >>> >>> We use a single flag for both, because the semantics match perfectly. >>> >> >> So suggestions for names are .attachment and .affinity neither of which >> sounds very boolean. We could do something like .isBeforeChar that returns a >> boolean or maybe .placement which either returns the string "before" or >> "after". Preferences? >> > > I still don't like the fact we're mixing before/after line break and bidi > caret. > Why? > Can we have two separate API/attribute for those? e.g. what does > beforeChar mean in the context of line break? > I agree that beforeChar is weird. "placement" seems fine to me.
Received on Thursday, 23 June 2011 18:59:34 UTC