Re: [Libreoffice] [PUSHED] Re: SmNodeToTextVisitor Fixes

Hi Luke,


> Really sorry for not giving this any attention for so long.

No worry... I've been busy with exams for the most of January...
Though, I hope I had a break through with the newline-<?> bug, yesterday...

Removing the brackets from the SmOperNodes was a pretty bad, so they can
> just go back in.

Nice :)

Removing them from UnHor and BinHor also broke some stuff, however, I
> didn't want to add them back in here, so decided that I would add them
> with the other nodes types that they were breaking with which seemed to
> be BinVer nodes (also the BinDiagonal which is similar).

Okay, sounds like a great solution... That way we don't get too many
brackets...

Some of the tests are the same kind as the one from last time. It can be
> a bit difficult to get the right node tree for the test from the parser,
> which also seems to put in more expression nodes than doing it with the
> cursor, which confused me for a bit.

Really neat... Now we'll also get failures if we mess up in cursor...

--
Regards Jonas Finnemann Jensen.


On Sat, Jan 29, 2011 at 14:01, Luke Dixon <6b8b4567@gmail.com> wrote:

> Hi Jonas,
>
> Really sorry for not giving this any attention for so long.
>
> > > I've noticed other stuff I've messed up though, so I'm going to
> continue
> > > on this some more :)
> > Again I'm not sure you messed it up... It might as well have been
> > messed up the first time I wrote it...
> > But don't let that stop you from fixing it...
>
> I'm pretty sure I messed some of it up quite badly, I've attached a
> patch that adds some of the brackets back in.
>
> Removing the brackets from the SmOperNodes was a pretty bad, so they can
> just go back in.
>
> Removing them from UnHor and BinHor also broke some stuff, however, I
> didn't want to add them back in here, so decided that I would add them
> with the other nodes types that they were breaking with which seemed to
> be BinVer nodes (also the BinDiagonal which is similar).
>
> Some of the tests are the same kind as the one from last time. It can be
> a bit difficult to get the right node tree for the test from the parser,
> which also seems to put in more expression nodes than doing it with the
> cursor, which confused me for a bit.
>
> Regards,
> Luke