🚀 go-pugleaf

In article <317F7B9F.3130@wpo.borland.com> Joe Hecht <jhecht@wpo.borland.com> writes:

   LAO Shihong wrote:
   > 
   > I want to convert some ps file to type1
   > ps font. Now I encountered one problem:
   > As some of the pathes are crossed, after
   > I have converted it to type1 font, those
   > crossed parts are shown as white holes.
   > Does type1 font use eofill only?
   > 
   > Are there any algorithmes can be used
   > to comvert those crossed pathes to outlines?
   > 
   > Thanks in advance. Any advice will be appreciated.
   > Feel free to reply in Japanese, Chinese or English.
   > 
   > ----
   > LAO Shihong (Firstname is surname)

   Types 1's use the Non-Zero Winding Rule.

Type 1 font rasterizing involves a `fill/eofill' operator that is quite 
different from the `top level' PostScript `fill/eofill' operator.
(For a start, it does not add an average of 1/2 pixel thickness all the 
way around).

However, from the point of view of which areas are black and which are 
white, it is in ATM equivalent to `eofill' except at large sizes where 
it is equivalent to `fill'.  

On some clone PS interpreters it acts like `fill' while on Adobe RIPs
it acts like `eofill'.

No, there is no way to change this.  You have just fix your outlines
by taking the union of the point sets defined by the pieces that overlap.

However, when using the `seac' operator to make composites, each part is
drawn separately, so you will not get white in the overlapping areas.
So if you can decompose your glyph into base and accent where the names
of the base and accent are in Adobe Standard Encoding you can use this trick.
You may need to `lie' about what the pieces are in order to use this trick.
------------------------------------------------------------------------------
Berthold K.P. Horn
Cambridge, Massachusetts, USA
------------------------------------------------------------------------------