line_count = 0
 
def read_split(f):
    global line_count
    line_count += 1
    line = f.readline()
    if not line:
        return False
    return line[:-1].split()
 
with open('simsun-12.bdf', 'r') as f:
    chars = {}
    line = ['']
    try:
        # 获得字符总数
        while line[0] != 'CHARS':
            line = read_split(f)
        # 读取所有字符
        for i in range(int(line[1])):
            assert read_split(f)[0] == 'STARTCHAR'
            line = read_split(f)
            # 编码
            assert line[0] == 'ENCODING'
            char = unichr(int(line[1]))
            line = read_split(f)
            # 绘制参数
            assert line[0] == 'BBX'
            width, height, x1, y1 = [int(x) for x in line[1:]]
            x0, y0 = (1 + x1, 12 - y1 - height)
            # 准备绘制文字
            base_image = [[0 for x in range(14)] for y in range(14)]
            # 读取并绘制文字
            line = read_split(f)
            assert line[0] == 'BITMAP'
            for y in range(y0, y0 + height):
                line = read_split(f)[0]
                bits = int(line, 16)
                bits >>= len(line) * 4 - width
                for x in range(x0 + width - 1, x0 - 1, -1):
                    base_image[y][x] = bits & 1
                    bits >>= 1
            chars[char] = base_image
            # 结束这个文字
            line = read_split(f)
            assert line[0] == 'ENDCHAR'
        line = read_split(f)
        assert line[0] == 'ENDFONT'
    except AssertionError:
        print 'line', line_count
        raise

import marshal
marshal.dump(chars, open('base_image', 'wb'))

#!/usr/bin/python
# - * - coding: utf8 - * -
 
import marshal
import copy
 
base_chars = marshal.load(open('base_image', 'rb'))
chars = {}
for char, image in base_chars.iteritems():
    new_image = []
    for y in range(14):
        new_row = []
        for x in range(14):
            if image[y][x]:
                value = 81
            else:
                value = 0
            if y > 0 and image[y-1][x]:
                value += 4
            if y < 13 and image[y+1][x]:
                value += 4
            if x > 0 and image[y][x-1]:
                value += 4
            if x < 13 and image[y][x+1]:
                value += 4
            new_image.append(value)
    chars[char] = new_image
 
marshal.dump(chars, open('advanced_data', 'wb'))

#!/usr/bin/python
# - * - coding: utf8 - * -
 
import marshal
 
chars = marshal.load(open('advanced_data', 'rb'))
 
def image_diff(image1, image2):
    ret = 0
    for v1, v2 in zip(image1, image2):
        ret += (v1 - v2) ** 2
    return ret
 
remember_number = 5
try:
    searched = marshal.load(open('searched_chars', 'rb'))
except IOError:
    searched = {}
while True:
    input_string = unicode(raw_input('ID: '), 'utf8')
    if not input_string:
        break
    for char in input_string:
        if char not in chars:
            continue
        if char not in searched:
            diff_data = []
            image = chars[char]
            for c, v in chars.iteritems():
                if c == char:
                    continue
                diff_data.append((image_diff(v, image), c))
            diff_data.sort()
            searched[char] = diff_data[:remember_number]
        print char,
        for item in searched[char]:
            print u'({0}, {1})'.format(*item),
        print
    print
 
marshal.dump(searched, open('searched_chars', 'wb'))

import Image, ImageDraw, ImageFont
image = Image.new('L', [14, 14], 255)
draw = ImageDraw.Draw(image)
font = ImageFont.truetype('simsun.ttc', 12)
draw.text((1, 1), u'好', font=font)
image.save('hw.png')

import Image, ImageDraw, ImageFont
image = Image.new('L', [28, 28], 255)
draw = ImageDraw.Draw(image)
font = ImageFont.truetype('simsun.ttc', 24)
draw.text((2, 2), u'好', font=font)
# 抗锯齿方式缩放
image = image.resize((14, 14), Image.ANTIALIAS)
image.save('hw.png')

for i in range(0xB0, 0xF8):
    chr_i = chr(i)
    for j in range(0xA1, 0xFF):
        if i == 0xD7 and j >= 0xFA and j <= 0xFE:
            continue
        char = unicode(chr_i + chr(j), 'gb18030')
        print char,

#!/usr/bin/python
# - * - coding: UTF-8 - * -
 
import Image, ImageDraw, ImageFont, ImageFilter
 
font = ImageFont.truetype('simsun.ttc', 24)
 
def get_char_data(char):
    image = Image.new('L', [28, 28], 255)
    draw = ImageDraw.Draw(image)
    draw.text((2, 2), char, font=font)
    image = image.resize((14, 14), Image.ANTIALIAS)
    return list(image.getdata())
 
def diff_of_data(data1, data2):
    ret = 0
    for i in zip(data1, data2):
        ret += abs(i[0] - i[1])
    return ret
 
chars = {}
for i in range(0xB0, 0xF8):
    chr_i = chr(i)
    for j in range(0xA1, 0xFF):
        if i == 0xD7 and j >= 0xFA and j <= 0xFE:
            continue
        char = unicode(chr_i + chr(j), 'gb18030')
        chars[char] = get_char_data(char)
 
remember_number = 5
input_string = unicode(raw_input('ID: '), 'utf8')
for input_char in input_string:
    if input_char not in chars:
        print input_char
        continue
    input_data = chars[input_char]
    diff_list = []
    for char, data in chars.iteritems():
        if char == input_char:
            continue
        diff_list.append((diff_of_data(data, input_data), char))
    diff_list.sort()
    print input_char,
    for diff, char in diff_list[:5]:
        print u'({0}, {1})'.format(diff, char).encode('utf8').ljust(11),
    print

#!/usr/bin/python
# - * - coding: UTF-8 - * -
 
import Image, ImageDraw, ImageFont, ImageFilter
 
font = ImageFont.truetype('simsun.ttc', 24)
 
def get_char_data(char):
    image = Image.new('L', [28, 28], 255)
    draw = ImageDraw.Draw(image)
    draw.text((2, 2), char, font=font)
    image = image.resize((14, 14), Image.ANTIALIAS)
    # XXX 期待更好的处理方式
    image = image.filter(ImageFilter.SHARPEN).filter(ImageFilter.SMOOTH)
    return list(image.getdata())
 
def diff_of_data(data1, data2):
    ret = 0
    for i in zip(data1, data2):
        ret += abs(i[0] - i[1])
    return ret
 
chars = {}
for i in range(0xB0, 0xF8):
    chr_i = chr(i)
    for j in range(0xA1, 0xFF):
        if i == 0xD7 and j >= 0xFA and j <= 0xFE:
            continue
        char = unicode(chr_i + chr(j), 'gb18030')
        chars[char] = get_char_data(char)
 
remember_number = 5
while True:
    input_string = unicode(raw_input('ID: '), 'utf8')
    if not input_string:
        break
    for input_char in input_string:
        if input_char not in chars:
            print input_char
            continue
        input_data = chars[input_char]
        diff_list = []
        for char, data in chars.iteritems():
            if char == input_char:
                continue
            diff_list.append((diff_of_data(data, input_data), char))
        diff_list.sort()
        print input_char,
        for diff, char in diff_list[:5]:
            print u'({0}, {1})'.format(diff, char).encode('utf8').ljust(11),
        print
    print