|
ArchivePoster
Guest
|
 |
« on: 2005-09-07, 15:51:00 » |
|
Posted by: mvelanka I am trying to use VTT for hinting devanagari font.
Majority of the letters in devanagari have a vertical line at the right hand side. This line makes the letter 'complete'. The letter without the vertical line is 'half'. Thus three glyphs would normally be required - one for half letter, one for the vertical line and one for the complete letter. I am trying to be smart in that I am making the complete letter(CL) as a composite glyph of half letter(HL) and the vertical line (VL). Thus the CL has a width = sum total of widths of HL and VL. Since VL will be common in many such CLs, this strategy would be better to control the features. It works well without a single mistake.
However hinting the CL is a problem. Literature says that in composite glyphs, the individual components are hinted first and then assembled to form the composite.
My experience in hinting the devanagari font using VTT is far from this.
None of the components (HL or VL) seem to get hinted when they participate in CL. The only difference in the composites in the literature and these CLs is that CLs do not have metrics of any of the components. I did not find similar example in literature. The metrics of CL are sum total of metrics of HL and VL in X direction.
Is this the reason why my HL and VL are not hinted when they take part in CL? Am am missing any step anywhere?
I tried to work around this by copying the assembly code generated in glyf prog of the HL and VL and pasting it in the assembly code of CL. (of course by changing the point numbers on the VL). This worked well.
However I am not sure whether this is the right strategy, or there is a smarter or more appropriate way to achieve this!
Pl guide me.
-Mahesh
|