pdf_encode_string adds encoding elements for entire text string rather than a font change may happen in a miidle of the string due to glyph variations. pdf_process_string to be called after pdf_encode_string, when we know an exact part of the string rendered with the current PDF font.
I'll better remove pdf_encode_string at all, because it's name is misleading. It was inherited from the old code, which re-encodes text.
Bump the priority since the bug 687660 depends on this.
Remove the dependency to 687660, since we fixed that one, and restore the old priority.
The file Bug687660a.ps converts to a PDF with multiple embedded Type 3 fonts with empty CharProcs.
A test case is Bug687472.ps
Patch to HEAD : http://ghostscript.com/pipermail/gs-cvs/2006-January/006258.html
One more patch to HEAD : http://ghostscript.com/pipermail/gs-cvs/2006-January/006259.html