##language:zh
#pragma section-numbers on
'''
用Python处理下载新浪下载下来的电子书
'''
::-- ZoomQuiet [<<DateTime(2007-06-01T02:45:05Z)>>]
<<TableOfContents>>
## 默许导航,请保留
<<Include(CPUGnav)>>


= Ben Luo.处理下载的电子书 =
{{{Ben Luo <benluo@gmail.com> 		 hide details	 9:57 am (16 minutes ago) 
	reply-to		python-chinese@lists.python.cn	 
	to		python-chinese@lists.python.cn	 
	date		Jun 1, 2007 9:57 AM	 
	subject		[python-chinese] 用Python处理下载新浪下载下来的电子书	 
}}}用Firefox的Download All 插件从新浪读书频道下载了文章。想在Treo650里看纯文本。然后就用了以下两个小程序。算是Unix思想的一种体现吧。小工具只做一件事情。大家把自己电脑里的小宝贝拿出来分享啊。或者在网站上多一个Wiki页面？


== html2txt.py ==
{{{#!python
#####################
#html2txt.py
#####################

from formatter import AbstractFormatter, NullWriter
from htmllib import HTMLParser

def _(str, in_encoder="gbk", out_encoder="utf8"):
    return unicode(str, in_encoder).encode(out_encoder)


class myWriter(NullWriter):
    def __init__(self):
        NullWriter.__init__(self)
        self._bodyText = []

    def send_flowing_data(self, str):
        self._bodyText.append(str)

    def _get_bodyText(self):
        return '\n'.join(self._bodyText)

    bodyText = property(_get_bodyText, None, None, 'plain text from body')

class myHTMLParser(HTMLParser):
    def do_meta(self, attrs):
        self.metas = attrs

def convertFile(filename):
    mywriter = myWriter()
    absformatter = AbstractFormatter(mywriter)
    parser = myHTMLParser(absformatter)
    parser.feed(open(filename).read())
    return ( _(parser.title), parser.formatter.writer.bodyText )

import os
import os.path

OUTPUTDIR = "./txt"
INPUTDIR = "."
if __name__ == "__main__":
    if not os.path.exists(OUTPUTDIR):
        os.mkdir(OUTPUTDIR)

    for file in os.listdir(INPUTDIR):
        if file[-4:] == '.htm' or file[-5:] == '.html':
            print "Coverting", file,
            outfilename = os.path.splitext(file)[0]
            a, text = convertFile(file)
            outfilename = outfilename + '.txt'
            outfullname = os.path.join(OUTPUTDIR, outfilename)
            open(outfullname, "wt").write(text)
            print "Done!"
}}}

== pickupcontent.py ==
{{{#!python
################################
#pickupcontent.py
################################

# -*- coding: utf-8 -*-

import sys
import glob
import os
import re

sys.argv[1:] = [item for arg in sys.argv[1:] for item in glob.glob(arg)]
startstr = u"^八十".encode("gb2312") # article title
endstr = u"^\[返回".encode("gb2312") #
tmp_start = re.compile(startstr)
tmp_end = re.compile(endstr)
for infile in sys.argv[1:]:
   # print infile
    f = open(infile,'r')
    #print f
    lines = f.readlines()
    fout = ''
    for index, line in enumerate(lines):
        if tmp_start.match(line):
            kstart = index
        if tmp_end.match(line):
            kend = index
            break
           
    f.close()
    fout = fout.join(lines[kstart:kend])
    tmp = open('tmp','w')
    tmp.write(fout)
    tmp.close()
    os.remove(infile)
    os.rename('tmp',infile)

}}}

== 反馈 ==