[Libreoffice-commits] core.git: vcl/source
Khaled Hosny
khaledhosny at eglug.org
Thu Nov 9 08:16:19 UTC 2017
vcl/source/gdi/pdfwriter_impl.cxx | 14 ++++++++++----
1 file changed, 10 insertions(+), 4 deletions(-)
New commits:
commit fb72ec9b6222c8a18445c078c0367e7d507720af
Author: Khaled Hosny <khaledhosny at eglug.org>
Date: Thu Nov 9 00:15:53 2017 +0200
tdf#113428: Fix glyph positions in PDF export
Not sure what is going on, but PDF export is the only place we call
GetNextGlyphs() with nLen > 1 and it seems I broke something here in
b894104a0b02a9b074c76feb925389d7bee6a493. Instead of trying to figure out
what the old code was trying to do, lets just calculate glyph positions
for the extra glyphs as GetNextGlyphs() does for the first one, making
PDF export closer to how the rendering on screen works.
Change-Id: I3f516e954ccfbe38abcf6985efe3bb1f998e967f
Reviewed-on: https://gerrit.libreoffice.org/44514
Tested-by: Jenkins <ci at libreoffice.org>
Reviewed-by: Khaled Hosny <khaledhosny at eglug.org>
diff --git a/vcl/source/gdi/pdfwriter_impl.cxx b/vcl/source/gdi/pdfwriter_impl.cxx
index db6d337c0c2e..e9fa6ad99673 100644
--- a/vcl/source/gdi/pdfwriter_impl.cxx
+++ b/vcl/source/gdi/pdfwriter_impl.cxx
@@ -6773,16 +6773,22 @@ void PDFWriterImpl::drawLayout( SalLayout& rLayout, const OUString& rText, bool
for( int i = 0; i < nGlyphs; i++ )
{
+ // tdf#113428: calculate the position of the next glyphs the same
+ // way GetNextGlyphs() would do if we asked for a single glyph at
+ // time.
+ if (i > 0)
+ {
+ Point aPos = pGlyphs[i]->maLinearPos;
+ aPos.X() /= rLayout.GetUnitsPerPixel();
+ aPos.Y() /= rLayout.GetUnitsPerPixel();
+ aGNGlyphPos = rLayout.GetDrawPosition(aPos);
+ }
aGlyphs.emplace_back( aGNGlyphPos,
pGlyphWidths[i],
pGlyphs[i]->maGlyphId,
pMappedFontObjects[i],
pMappedGlyphs[i],
pGlyphs[i]->IsVertical() );
- if( bVertical )
- aGNGlyphPos.Y() += pGlyphs[i]->mnNewWidth/rLayout.GetUnitsPerPixel();
- else
- aGNGlyphPos.X() += pGlyphs[i]->mnNewWidth/rLayout.GetUnitsPerPixel();
}
}
More information about the Libreoffice-commits
mailing list