[Fontforge-devel] Patch for pair-based kerning Python

[<Bar...@ch...>]

I only needed and tested the gpos_pair case in the following patch.

(I needed this for a script to restore mysteriously emptied
kerning-pair subtables. Fontforge seems to trash the contents of these
subtables sometimes--frequently, actually. Am I the only person
experiencing this? Am I doing something wrong?)


--- fontforge/fontforge/python.c        2007-07-26 19:18:06.000000000 +0000
+++ fontforge.EDITED/fontforge/python.c 2007-07-28 00:14:13.000000000 +0000
@@ -4658,7 +4658,7 @@
     temp.subtable = sub;
 
     if ( sub->lookup->lookup_type==gpos_single ) {
-       if ( !PyArg_ParseTuple(args,"(shhhh)", &subname,
+       if ( !PyArg_ParseTuple(args,"shhhh", &subname,
                &temp.u.pos.xoff, &temp.u.pos.yoff,
                &temp.u.pos.h_adv_off, &temp.u.pos.v_adv_off))
 return( NULL );
@@ -4666,7 +4666,7 @@
     } else if ( sub->lookup->lookup_type==gpos_pair ) {
        temp.type = pst_pair;
        temp.u.pair.vr = chunkalloc(sizeof(struct vr [2]));
-       if ( !PyArg_ParseTuple(args,"(sshhhhhhhh)", &subname, &other,
+       if ( !PyArg_ParseTuple(args,"sshhhhhhhh", &subname, &other,
                &temp.u.pair.vr[0].xoff, &temp.u.pair.vr[0].yoff,
                &temp.u.pair.vr[0].h_adv_off, &temp.u.pair.vr[0].v_adv_off,
                &temp.u.pair.vr[1].xoff, &temp.u.pair.vr[1].yoff,
@@ -4702,12 +4702,12 @@
        }
        temp.u.pair.paired = copy(other);
     } else if ( sub->lookup->lookup_type==gsub_single ) {
-       if ( !PyArg_ParseTuple(args,"(ss)", &subname, &other))
+       if ( !PyArg_ParseTuple(args,"ss", &subname, &other))
 return( NULL );
        temp.type = pst_substitution;
        temp.u.subs.variant = copy(other);
     } else {
-       if ( !PyArg_ParseTuple(args,"(sO)", &subname, &others))
+       if ( !PyArg_ParseTuple(args,"sO", &subname, &others))
 return( NULL );
        other = GlyphNamesFromTuple(others);
        if ( other==NULL )