1 files changed, 118 insertions, 0 deletions
diff --git a/examples/FLWriteXml.py b/examples/FLWriteXml.py
new file mode 100755
index 0000000..e174f6d
--- /dev/null
+++ b/examples/FLWriteXml.py
@@ -0,0 +1,118 @@
+#!/usr/bin/env python
+'''Outputs attachment point information and notes as XML file for TTFBuilder'''
+__url__ = 'http://github.com/silnrsi/pysilfont'
+__copyright__ = 'Copyright (c) 2015 SIL International (http://www.sil.org)'
+__license__ = 'Released under the MIT License (http://opensource.org/licenses/MIT)'
+__author__ = 'M Hosken'
+
+# user controls
+
+# output entries for all glyphs even those with nothing interesting to say about them
+all_glyphs = 1
+
+# output the glyph id as part of the information
+output_gid = 1
+
+# output the glyph notes
+output_note = 0
+
+# output UID with "U+" prefix
+output_uid_prefix = 0
+
+# print progress indicator
+print_progress = 0
+
+# no user serviceable parts under here!
+from xml.sax.saxutils import XMLGenerator
+import os
+
+def print_glyph(font, glyph, index):
+    if print_progress and index % 100 == 0:
+        print "%d: %s" % (index, glyph.name)
+      
+    if (not all_glyphs and len(glyph.anchors) == 0 and len(glyph.components) == 0 and
+        not (glyph.note and output_note)):
+        return
+    attribs = {}
+    if output_gid:
+        attribs["GID"] = unicode(index)
+    if glyph.unicode:
+        if output_uid_prefix:
+            attribs["UID"] = unicode("U+%04X" % glyph.unicode)
+        else:
+            attribs["UID"] = unicode("%04X" % glyph.unicode)
+    if glyph.name:
+        attribs["PSName"] = unicode(glyph.name)
+    xg.startElement("glyph", attribs)
+    
+    for anchor in (glyph.anchors):
+        xg.startElement("point", {"type":unicode(anchor.name), "mark":unicode(anchor.mark)})
+        xg.startElement("location", {"x":unicode(anchor.x), "y":unicode(anchor.y)})
+        xg.endElement("location")
+        xg.endElement("point")
+
+    for comp in (glyph.components):
+        g = font.glyphs[comp.index]
+        r = g.GetBoundingRect()
+        x0 = 0.5 * (r.ll.x * (1 + comp.scale.x) + r.ur.x * (1 - comp.scale.x)) + comp.delta.x
+        y0 = 0.5 * (r.ll.y * (1 + comp.scale.y) + r.ur.y * (1 - comp.scale.y)) + comp.delta.y
+        x1 = 0.5 * (r.ll.x * (1 - comp.scale.x) + r.ur.x * (1 + comp.scale.x)) + comp.delta.x
+        y1 = 0.5 * (r.ll.y * (1 - comp.scale.x) + r.ur.y * (1 + comp.scale.y)) + comp.delta.y
+
+        attribs = {"bbox":unicode("%d, %d, %d, %d" % (x0, y0, x1, y1))}
+        attribs["GID"] = unicode(comp.index)
+        if (g.unicode):
+            if output_uid_prefix:
+                attribs["UID"] = unicode("U+%04X" % g.unicode)
+            else:
+                attribs["UID"] = unicode("%04X" % g.unicode)
+        if (g.name):
+            attribs["PSName"] = unicode(g.name)
+        xg.startElement("compound", attribs)
+        xg.endElement("compound")
+        
+    if glyph.mark:
+        xg.startElement("property", {"name":unicode("mark"), "value":unicode(glyph.mark)})
+        xg.endElement("property")
+        
+    if glyph.customdata:
+        xg.startElement("customdata", {})
+        xg.characters(unicode(glyph.customdata.strip()))
+        xg.endElement("customdata")
+        
+    if glyph.note and output_note:
+        xg.startElement("note", {})
+        xg.characters(glyph.note)
+        xg.endElement("note")
+    xg.endElement("glyph")
+
+outname = fl.font.file_name.replace(".vfb", "_tmp.xml")
+fh = open(outname, "w")
+xg = XMLGenerator(fh, "utf-8")
+xg.startDocument()
+
+#fl.font.full_name is needed to get the name as it appears to Windows
+#fl.font.font_name seems to be the PS name. This messes up GenTest.pl when it generates WPFeatures.wpx
+xg.startElement("font", {'name':unicode(fl.font.full_name), "upem":unicode(fl.font.upm)})
+for i in range(0, len(fl.font.glyphs)):
+    print_glyph(fl.font, fl.font.glyphs[i], i)
+xg.endElement("font")
+
+xg.endDocument()
+fh.close()
+
+#somehow this enables UNC naming (\\Gutenberg vs i:) to work when Saxon is called with popen
+#without this, if outname is UNC-based, then drive letters and UNC volumes are invisible
+# if outname is drive-letter-based, then drive letters and UNC volumes are already visible
+if (outname[0:2] == r'\\'): 
+    os.chdir("c:")
+tidy = "tidy -i -xml -n -wrap 0 --char-encoding utf8 --indent-spaces 4 --quote-nbsp no --tab-size 4 -m %s"
+saxon = "saxon %s %s" % ('"' + outname + '"', r'"C:\Roman Font\rfs_font\10 Misc Utils\glyph_norm.xsl"') #handle spaces in file name
+f = os.popen(saxon, "rb")
+g = open(outname.replace("_tmp.xml", ".xml"), "wb")
+output = f.read()
+g.write(output)
+f.close()
+g.close()
+
+print "Done"