Linux: support mirrored charactors

third_party/harfbuzz/contrib/tables/mirroring-parse.py

Index: third_party/harfbuzz/contrib/tables/mirroring-parse.py
diff --git a/third_party/harfbuzz/contrib/tables/mirroring-parse.py b/third_party/harfbuzz/contrib/tables/mirroring-parse.py
new file mode 100644
index 0000000000000000000000000000000000000000..04a9431a7187e00163a8466bb2c1ccdd6e0d460b
--- /dev/null
+++ b/third_party/harfbuzz/contrib/tables/mirroring-parse.py
@@ -0,0 +1,44 @@
+import sys
+
+# http://www.unicode.org/Public/UNIDATA/auxiliary/BidiMirroring.txt

[tony, 2009/08/07 19:57:06]

Nit: Can you add a comment explaining what this does and what output file it
generates (and the need to check in said generated file). 

+
+def main(infile, outfile):
+  pairs = []
+  for line in infile:
+    line = line[:-1]
+    if len(line) == 0 or line[0] == '#':
+      continue
+    if '#' in line:
+      (data, _) = line.split('#', 1)
+    else:
+      data = line
+    if ';' not in data:
+      continue
+    (a, b) = data.split(';', 1)
+    a = int(a, 16)
+    b = int(b, 16)
+
+    pairs.append((a, b))
+
+  pairs.sort()
+
+  print >>outfile, '// Generated from Unicode Bidi Mirroring tables\n'
+  print >>outfile, '#ifndef MIRRORING_PROPERTY_H_'
+  print >>outfile, '#define MIRRORING_PROPERTY_H_\n'
+  print >>outfile, '#include <stdint.h>'
+  print >>outfile, 'struct mirroring_property {'
+  print >>outfile, '  uint32_t a;'
+  print >>outfile, '  uint32_t b;'
+  print >>outfile, '};\n'
+  print >>outfile, 'static const struct mirroring_property mirroring_properties[] = {'
+  for pair in pairs:
+    print >>outfile, '  {0x%x, 0x%x},' % pair
+  print >>outfile, '};\n'
+  print >>outfile, 'static const unsigned mirroring_properties_count = %d;\n' % len(pairs)
+  print >>outfile, '#endif  // MIRRORING_PROPERTY_H_'
+
+if __name__ == '__main__':
+  if len(sys.argv) != 3:
+    print 'Usage: %s <input .txt> <output .h>' % sys.argv[0]
+  else:
+    main(file(sys.argv[1], 'r'), file(sys.argv[2], 'w+'))