extract_codepoints.py

#!/usr/bin/env python2

import codecs
from struct import pack, unpack
import json, sys, os

import math
from PIL import Image

def extract_codepoints(font):
    version = unpack('B', font.read(1))[0]
    max_height = unpack('B', font.read(1))[0]
    numbers = unpack('H', font.read(2))[0]
    wildcard = unpack('H', font.read(2))[0]

    table_size = unpack('B', font.read(1))[0]
    codepoint_bytes = unpack('B', font.read(1))[0]

    print >>sys.stderr, 'Contains %d codepoints' % numbers
    print >>sys.stderr, 'Maximum height: %d' % max_height

    codepoints = []

    hash_table = {}
    for i in range(table_size):
        index = unpack('B', font.read(1))[0]
        size = unpack('B', font.read(1))[0]
        offset = unpack('H', font.read(2))[0]
        hash_table[index] = (size, offset)

    table_offset = 0x08 + table_size*0x04
    offset_tables = {}
    for index in hash_table:
        size, offset = hash_table[index]
        font.seek(table_offset+offset)

        offset_table = []
        for j in range(size):
            if codepoint_bytes == 4:
                codepoint, offset = unpack('<II', font.read(8))
            else:
                codepoint, offset = unpack('<HI', font.read(6))
            offset_table.append((codepoint, offset))
            codepoints.append(codepoint)
        offset_tables[index] = offset_table

    for index in hash_table:
        size, offset = hash_table[index]
        table_offset += (codepoint_bytes+4)*size

    gap = int(max_height*1.5)
    side = int(math.sqrt(numbers))
    if side*side < numbers:
        side += 1
    side += 1
    bitmap = Image.new('RGB', (side*gap, side*gap), 'white')

    metadata = []

    count = 0
    for index in hash_table:
        size, offset = hash_table[index]
        offset_table = offset_tables[index]
        for j in range(size):
            codepoint, offset = offset_table[j]
            font.seek(table_offset+offset)
            width, height, left, top = unpack('<BBbb', font.read(4))
            advance = unpack('b', font.read(1))[0]
            metadata.append({
                'codepoint': codepoint,
                'char': unichr(codepoint),
                'advance':advance,
                'left':left,
                'top':top,
            })
            charimg = Image.new('RGB', (width, height), 'white')

            size = width*height
            size = (size+7)/8
            data = font.read(size)

            for y in range(height):
                for x in range(width):
                    off = x+y*width
                    bit = ord(data[off/8]) & (0x01 << (off%8))
                    if bit > 0:
                        charimg.putpixel((x, y), (0,0,0))
                        if count/side*gap+x+left > 0 and count%side*gap+y+top > 0:
                            bitmap.putpixel((count/side*gap+x+left, count%side*gap+y+top), (0,0,0))

            charfile = '%s/%05X.png' % (dir_name, codepoints[count])
            if width and height:
                charimg.save(charfile)
            else:
                # just touch it
                open(charfile, 'wb').close()

            count += 1

    #bitmap.show()

    with codecs.open('%s/list.json' % dir_name, 'w', 'utf-8') as out:
        json.dump({
            'max_height': max_height,
            'codepoints': codepoints,
            'metadata': metadata,
        }, out, indent=4, ensure_ascii=False)

if __name__ == '__main__':
    if len(sys.argv) <= 2:
        print 'Usage: extract_codepoints.py fontfile output_dir'
        exit(1)

    file_name, dir_name = sys.argv[1:3]
    if os.path.exists(dir_name):
        print('Directory %s already exists!' % dir_name)
        exit(1)
    os.mkdir(dir_name)
    font = open(file_name, 'rb')
    extract_codepoints(font)