On Sun, Oct 30, 2022 at 2:26 AM Chris Sherlock <chris.sherloc...@gmail.com> wrote:
> > > Sent from my iPhone > > On 23 Oct 2022, at 5:20 am, Caolán McNamara <caol...@redhat.com> wrote: > > On Sat, 2022-10-22 at 01:42 +1100, Chris Sherlock wrote: > > Test name: VclLogicalFontInstanceTest::testglyphboundrect > > equality assertion failed > > - Expected: 51x82@(7,-80) > > - Actual : 51x83@(7,-80) > > > I’ve had another look at the code, with some help from Khaled. The > > issue is the way in which we do our final rounding. > > > On Freetype, it uses 26.6 fixed point, and thus needs it’s own > > special FreeType ceil and floor functions. CoreText appears to using > > floating point, but the std::ceil and std::floor functions give > > slight different results. > > > I presume that for mac its CoreTextStyle::ImplGetGlyphoundRect and the > CTFontGetBoundingRectsForGlyphs so could print the numbers we get there > vs the ones we get in FreeTypeFont::GetGlyphBoundRect to verify that. > > > I’ve added some debugging statements. the Freetype library does its own > floor and ceil calculations. it’s looking like it isn’t a risky a rounding > error, it is just that Freetype calculates one value and MacOS framework > for glyphs calculates it differently. > > Patch is https://gerrit.libreoffice.org/c/core/+/142006 if anyone would > like to review :-) > > Chris > A polite ping to the mailing list as people are now asking me for the rationale behind the patch, could someone please advise the way forward? Chris
