https://codereview.appspot.com/569700043/diff/554040049/lily/include/lazy-skyline-pair.hh
File lily/include/lazy-skyline-pair.hh (right):
https://codereview.appspot.com/569700043/diff/554040049/lily/include/lazy-skyline-pair.hh#newcode30
lily/include/lazy-skyline-pair.hh:30: CW = UP,
On 2020/05/08 15:49:18, lemzwerg wrote:
>
https://codereview.appspot.com/569700043/diff/573820043/lily/freetype.cc
> File lily/freetype.cc (right):
>
>
https://codereview.appspot.com/569700043/diff/573820043/lily/freetype.cc#newcode129
> lily/freetype.cc:129: else if (ctags[j] & 1)
> 1 →
[forgot to send it previously]
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc
File lily/freetype.cc (right):
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc#newcode126
lily/freetype.cc:126: else if (outline->tags[j] & 1)
On 2020/05/08 08:15:58,
https://codereview.appspot.com/569700043/diff/573820043/lily/freetype.cc
File lily/freetype.cc (right):
https://codereview.appspot.com/569700043/diff/573820043/lily/freetype.cc#newcode129
lily/freetype.cc:129: else if (ctags[j] & 1)
1 → FT_CURVE_TAG_ON
On Fri, May 8, 2020 at 4:28 PM wrote:
> > seems to make the assertion output comparatively useless. Wouldn't it
> be better
> > to do such things as
> >
> > else {
> > assert(condition);
> > ...
> > }
> >
> > instead?
>
> I'm used to the following idiom from LLVM:
> assert(0 && "Unknown
On 2020/05/08 14:24:18, dak wrote:
> On 2020/05/08 14:14:47, hahnjo wrote:
>
> > assert(false) that none of the previous cases was true? That should
detect a
> > breaking change in FT_Outline pretty quickly.
>
> else if (condition)
> { ... }
> else assert(false);
>
> seems to make the assertion
On 2020/05/08 14:14:47, hahnjo wrote:
> assert(false) that none of the previous cases was true? That should
detect a
> breaking change in FT_Outline pretty quickly.
else if (condition)
{ ... }
else assert(false);
seems to make the assertion output comparatively useless. Wouldn't it
be better
On Fri, May 8, 2020 at 4:14 PM wrote:
>
> On 2020/05/08 14:03:23, hanwenn wrote:
> >
> https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc#newcode162
> > lily/freetype.cc:162: }
> > On 2020/05/08 08:15:59, hahnjo wrote:
> > > This code looks very familiar. Don't we already
On 2020/05/08 14:03:23, hanwenn wrote:
>
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc#newcode162
> lily/freetype.cc:162: }
> On 2020/05/08 08:15:59, hahnjo wrote:
> > This code looks very familiar. Don't we already have it in
> > make_draw_bezier_boxes?
>
> yes.
So
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc
File lily/freetype.cc (right):
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc#newcode105
lily/freetype.cc:105: bool is_tt = (0 == strcmp ("TrueType",
FT_Get_Font_Format (face)));
On 2020/05/08
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc
File lily/freetype.cc (right):
https://codereview.appspot.com/569700043/diff/582060043/lily/freetype.cc#newcode105
lily/freetype.cc:105: bool is_tt = (0 == strcmp ("TrueType",
FT_Get_Font_Format (face)));
Can we have this
LGTM
https://codereview.appspot.com/569700043/
Reviewers: ,
Message:
goes on top of https://codereview.appspot.com/561730043/
Description:
Split glyph contours in up/down segments for skylines
This reduces the amount of skyline input coming from glyph outlines by
2x.
benchmark for arguments: input/regression/mozart-hrn-3
raw data:
13 matches
Mail list logo
