##language:zh
#pragma section-numbers off
##含有章节索引导航的 ZPyUG 文章通用模板
<<TableOfContents>>

## 默许导航,请保留
<<Include(ZPyUGnav)>>

= 提取文档中的图象标识信息 =
{{{
Shuguang Yang <iyangsg@gmail.com>
reply-to        python-cn@googlegroups.com
to      python-cn`CPyUG`华蟒用户组 <python-cn@googlegroups.com>
date    Wed, Aug 20, 2008 at 15:50
subject [CPyUG:62863] 用python提取文档中的图象，图象的标识信息
}}}
##startInc
应用 Python 解决一些实际问题（

 * '''http://www.ibm.com/developerworks/cn/linux/tips/l-python/'''
的提取嵌入在文档中的图像部分，原文中的程序思路正确但代码不够简洁，不够pythonic。
改进后的代码如下：

{{{
#!python
import sys
import os
import string
headers = [('JFIF', 6, 'jpg'), ('GIF', 0, 'gif'), ('PNG', 1, 'png')] #不同图片格式的标识信息
marker = []
filename = '/path/to/your/file'
try:
   fid = open(filename, 'rb')
except:
   sys.exit(1)
s = 0
for line in fid: #按行迭代
   for flag, offset, ext in headers:
       index = string.find(line, flag)
       if index > 0:
           pos = s + index - offset
           marker.append((pos, ext))
   s += len(line)
fid.seek(0)
imgnum = 0
if len(marker) == 0:
   print 'No images included in this document'
   sys.exit(1)
for info in marker:
   thispos = info[0]
   thisext = info[1]
   index = marker.index(info)
   try:
       nextinfo = marker[index + 1]
       nextpos = nextinfo[0]
       gap = nextpos - thispos
   except IndexError:
       nextpos = s
       gap = nextpos - thispos
   fid.seek(thispos)
   data = fid.read(gap)
   imgname = 'imgname%02d.%s' % (index, thisext)
   fid1 = open(imgname, 'wb')
   fid1.write(data)
   fid1.close()
   imgnum += 1
fid.close()
print '%02d images have been extracted' % imgnum
}}}
##endInc
----
 '''反馈'''

创建 by -- ZoomQuiet [<<DateTime(2008-08-20T07:55:11Z)>>]