python脚本语言作为转换数据的工具示例

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