diff --git a/README.md b/README.md
index 7991910c35a833525aca484f728647676cadd362..b3035f5a19a4394372effd710bc95acc683586d9 100644
--- a/README.md
+++ b/README.md
@@ -42,14 +42,23 @@ $ client/FontMaker_Cli.exe
 > 目前这个命令行软件还不支持输入参数进行定制生成字库文件，运行后默认使用的参数如下：
 >
 > 字体：幼圆
+
 > 字号：16 像素
+
 > 字重：普通（不加粗、不斜体、无下划线）
+
 > 字符宽度：固定
+
 > 字符宽度：16 像素
+
 > 字符高度：16 像素
+
 > 水平偏移：0 像素
+
 > 垂直偏移：0 像素
+
 > 扫描方式：行扫描
+
 > 字节顺序：高位在前
 
 #### 使用电脑测试
diff --git a/abconfig b/abconfig
new file mode 100644
index 0000000000000000000000000000000000000000..70d320a14c48291594d533733d62e681e16520f6
--- /dev/null
+++ b/abconfig
@@ -0,0 +1,2 @@
+# client/combined.bin
+drivers/ssd1306.py
diff --git a/client b/client
index f41014c6c12440894152b05dac4c65843238d2ed..cc516ad64bbd2810db4422b0e431c2014fbda1e9 160000
--- a/client
+++ b/client
@@ -1 +1 @@
-Subproject commit f41014c6c12440894152b05dac4c65843238d2ed
+Subproject commit cc516ad64bbd2810db4422b0e431c2014fbda1e9
diff --git a/fontlib.py b/fontlib.py
index 1620d561ccb1d42c83a031b16bcf3399f728a34c..5ecae72b7feeeb5d55daf0c599bda1cbb8e3f5c7 100644
--- a/fontlib.py
+++ b/fontlib.py
@@ -72,7 +72,7 @@ class FontLib(object):
 		self.__header = None
 
 		with open(self.__font_filename, 'rb') as font_file:
-			self.__header = FontLibHeader(font_file.read(FontLibHeader.LENGTH))
+			self.__header = FontLibHeader(memoryview(font_file.read(FontLibHeader.LENGTH)))
 			self.__placeholder_buffer = self.__get_character_unicode_buffer(font_file, {ord('?')})[0][1]
 
 		gc.collect()
@@ -86,17 +86,24 @@ class FontLib(object):
 	def __get_character_unicode_buffer(self, font_file, unicode_set):
 		buffer_list = []
 
+		def __seek(data, target):
+			for index in range(0, 1000, 2):
+				if data[index: index + 2] == target:
+					return offset + index
+			else:
+				return 0
+
 		for unicode in unicode_set:
 			if self.__is_ascii(unicode):
 				char_offset = self.__header.ascii_start + (unicode - 0x20) * self.__header.data_size
 			elif self.__is_gb2312(unicode):
-				gb2312_index = struct.pack('<H', unicode)
+				gb2312_index = memoryview(struct.pack('<H', unicode))
 
-				for offset in range(self.__header.index_table_address, self.__header.ascii_start, 2):
+				for offset in range(self.__header.index_table_address, self.__header.ascii_start, 1000):
 					font_file.seek(offset)
 
-					if font_file.read(2) == gb2312_index:
-						char_index_offset = font_file.tell() - 2
+					char_index_offset = __seek(memoryview(font_file.read(1000)), gb2312_index)
+					if char_index_offset:
 						break
 				else:
 					buffer_list.append([unicode, self.__placeholder_buffer])