XTF parser

xtf.py

#!/bin/env python3

import sys
import struct
import os

print('<?xml version="1.0" standalone="yes"?>')
print('<svg width="100%" height="100%" version="1.1" xmlns = "http://www.w3.org/2000/svg">')
print('  <defs>')
print('    <font id="font" horiz-adv-x="1000">')
print('      <font-face font-family="Super Sans" font-weight="normal" font-style="italic"')
print('          units-per-em="1000" cap-height="600" x-height="400"')
print('          ascent="700" descent="300"')
print('          alphabetic="0" mathematical="350" ideographic="400" hanging="500">')
print('        <font-face-src>')
print('          <font-face-name name="Super Sans Italic"/>')
print('        </font-face-src>')
print('      </font-face>')
print('      <missing-glyph><path d="M0,0h200v200h-200z"/></missing-glyph>')
#print('      <glyph unicode="@"><!-- Outline of @ glyph --></glyph>')

with open(sys.argv[1], 'rb') as f:
  magic = f.read(4)
  string_len = struct.unpack("I", f.read(4))[0]
  string = f.read(string_len).decode('ascii')

  # https://msdn.microsoft.com/en-us/library/dd144956%28v=vs.85%29.aspx
  cursor = f.tell()
  cbThis = struct.unpack("I", f.read(4))[0] #FIXME: Is this the one being exploited?! If so, log something!
  flAccel = struct.unpack("I", f.read(4))[0]
  assert(flAccel == 0)
  cGlyphsSupported = struct.unpack("I", f.read(4))[0]
  print("<!-- " + str(cGlyphsSupported) + " glyphs supported -->")
  cRanges = struct.unpack("I", f.read(4))[0]
  ranges = []
  for i in range(cRanges):
    wcLow = struct.unpack("H", f.read(2))[0]
    cGlyphs = struct.unpack("H", f.read(2))[0]
    ranges.append((wcLow, cGlyphs))
  assert(f.tell() == (cursor + cbThis))

  for i in range(cGlyphsSupported):

    # https://msdn.microsoft.com/en-us/library/windows/desktop/dd374209(v=vs.85).aspx
    #FIXME: Parse these!
    f.read(4)
    f.read(4)
    f.read(8)
    f.read(4)
    f.read(4)

    offset = struct.unpack("I", f.read(4))[0]

  glyphCount = 0
  for r in ranges:
    print("<!-- Decoding " + str(r[1]) + " symbols -->")
    glyphCount += r[1]
    for c in range(r[0], r[0] + r[1]):

      symbol = struct.pack('H', c).decode("utf-16")

      print("<!-- Decoding symbol " + str(c) + ": \"" + symbol + "\" -->")

      symbol = "&#x%04X;" % c

      index_count = struct.unpack("H", f.read(2))[0]
      vertex_count = struct.unpack("H", f.read(2))[0]
      indices = []
      for i in range(index_count):
        indices.append(struct.unpack("H", f.read(2))[0])
      vertices = []
      for i in range(vertex_count):
        vertices.append((struct.unpack("f", f.read(4))[0], struct.unpack("f", f.read(4))[0]))

      def strf(f):
        return str(int(f * 1000))

      path = ""

      assert(len(indices) % 3 == 0)
      for i in range(len(indices) // 3):
        path += "M " + strf(vertices[indices[i*3+0]][0]) + " " + strf(vertices[indices[i*3+0]][1]) + " "
        path += "L " + strf(vertices[indices[i*3+1]][0]) + " " + strf(vertices[indices[i*3+1]][1]) + " "
        path += "L " + strf(vertices[indices[i*3+2]][0]) + " " + strf(vertices[indices[i*3+2]][1]) + " "
        path += "z "

      print('      <glyph unicode="' + symbol + '" horiz-adv-x="300" d="' + path + '"></glyph>')

    # Hack to only export first range
    if False:
      f.seek(0, os.SEEK_END)
      break
  
  cursor = f.tell()
  f.seek(0, os.SEEK_END)
  assert(f.tell() == cursor)

print('      <!-- more glyphs -->')
print('    </font>')
print('  </defs>')
print('</svg>')