学步园

      都说python是解释形语言，意思是读取一行执行一行，现在就有一段python程序摆在面前，其功能是“将一个csv文件名以及按文件中的内容转换成多个以内容为目录的linux下目录结构”； 这段话应该比较的模型化，用一个具体的例子来说会容易理解点，这里就假设有这样的一个小需求：现有一个压缩文件 JiangSu-1949-10-01-HumanScience.csv.gz，该压缩文件包含了江苏省1949年10月1日前建立的人文科学专业大学的详细名单，需要转换为一组文件夹，每个文件夹单独存放从签署大学总录中包含的一个大学详细信息。（注：
以下仅为实现该小需求的示例python， 参数文件为 JiangSu-1949-10-01-HumanScience.csv.gz）

#!/usr/bin/python<br /> import csv, gzip, os, sys, re<br /> def ResolveUnivercity(TotalUniversities, pattern):<br /> SpecifiedUnivercity = pattern.search(TotalUniversities)</p> <p> if SpecifiedUnivercity is None:<br /> return None<br /> return SpecifiedUnivercity.groups()<br /> def TransformData(TotalUniversities, RootUnivercity_Path, pattern):<br /> SpecifiedUniversity = ResolveUnivercity(TotalUniversities, pattern)<br /> if SpecifiedUniversity is None:<br /> print 'Cannot resolved TotalUniversities file'<br /> return 1<br /> else:<br /> print 'UniversityName: [%s] Born: [%s-%s-%s] Area: [%s]' % SpecifiedUniversity<br /> if not SpecifiedUniversity[4] in ["NatureScience", "HumanScience"]:<br /> print "system only analyze Nature or Human Science"<br /> return 1<br /> if RootUnivercity_Path is None:<br /> RootUnivercity_Path = '%s' % SpecifiedUniversity[0]<br /> gfile = gzip.open(TotalUniversities, "rb")<br /> rownum = 0<br /> HasNext = False</p> <p> SpecifiedUnivercity_Content = {}<br /> SpecifiedUnivercity_DetailContent = None<br /> for row in csv.reader(gfile):<br /> if rownum == 0:<br /> header = row<br /> else:</p> <p> SpecifiedUniversity_Name = row[0]<br /> if len(SpecifiedUniversity_Name) > 0:<br /> HasNext = True</p> <p> if SpecifiedUnivercity_Content.has_key(SpecifiedUniversity_Name):<br /> SpecifiedUnivercity_DetailContent = SpecifiedUnivercity_Content[SpecifiedUniversity_Name][1]</p> <p> else:<br /> rownum = 0<br /> # create new file<br /> SpecifiedUniversity_File = '%s_%s_%s.csv' % (SpecifiedUniversity[0], SpecifiedUniversity_Name, SpecifiedUniversity[2])</p> <p> AbsolutePath = os.path.join(RootUnivercity_Path, SpecifiedUniversity_Name)<br /> AbsolutePathName = os.path.join(AbsolutePath, SpecifiedUniversity_File)</p> <p> if not os.path.exists(AbsolutePath):<br /> os.makedirs(AbsolutePath, 0777)<br /> outfile = open(AbsolutePathName, 'wb')</p> <p> SpecifiedUnivercity_DetailContent = csv.writer(outfile, Token=',')<br /> # insert into SpecifiedUniversity<br /> SpecifiedUnivercity_Content[SpecifiedUniversity_Name] = [outfile, SpecifiedUnivercity_DetailContent, False]<br /> else:<br /> HasNext = False</p> <p> if SpecifiedUnivercity_DetailContent is not None:<br /> SpecifiedUnivercity_DetailContent.writerow(row)<br /> else:<br /> print "row %d has no owner file" % rownum<br /> rownum += 1<br /> gfile.close()<br /> for key in SpecifiedUnivercity_Content.keys():<br /> SpecifiedUnivercity_Content[key][1].close()<br /> SpecifiedUnivercity_Content[key][0].close()</p> <p>if __name__ == "__main__":<br /> #JiangSu-1949-10-01-HumanScience.csv.gz<br /> ResolvedSpecofiedUnivercity = re.compile(r'(/w{2,})-(/d{4})-(/d{2})-(/d{2})-(/w{1,})/.')</p> <p> for file in sys.argv[1:]:<br /> TransformData(file, None, ResolvedSpecofiedUnivercity)<br />