Python按自定义字段模式解析文件名

时间:2013-03-05 19:45:04

标签: python regex parsing csv

本周开始学习python,所以我想我会用excel来解析文件路径中的一些字段。

我有大约3000个文件都符合命名约定。 /Household/LastName.FirstName.Account.Doctype.Date.extension

例如,可以命名其中一个文件:Cosby.Bill..Profile.2006.doc 并且完整路径是/ Volumes / HD / Organized Files / Cosby,Bill / Cosby.Bill..Profile.2006.doc

在这种情况下:

考斯比,比尔将成为家庭

家庭(Cosby,Bill)是实际文件的封闭文件夹

比尔将是第一个名字

Cosby 将是姓氏

帐户字段已省略

个人资料是doctype

2006 是日期

doc 是扩展程序

所有这些文件都位于此目录/ Volumes / HD / Organized Files /我使用终端和ls将所有文件的列表放到桌面上的.txt文件中,我试图解析来自文件路径分为类似上面的示例。理想情况下,我想输出到csv,每个类别都有一列。这是我丑陋的代码:

def main():
    file = open('~/Desktop/client_docs.csv', "rb")
    output = open('~/Desktop/client_docs_parsed.txt', "wb")
    for line in file:
        i = line.find(find_nth(line, '/', 2))
        beghouse = line[i + len(find_nth(line, '/', 2)):]
        endhouse = beghouse.find('/')
        household = beghouse[:endhouse]
        lastn = (line[line.find(household):])[(line[line.find(household):]).find('/') + 1:(line[line.find(household):]).find('.')]
        firstn = line[line.find('.') + 1: line.find('.', line.find('.') + 1)]
        acct = line[line.find('{}.{}.'.format(lastn,firstn)) + len('{}.{}.'.format(lastn,firstn)):line.find('.',line.find('{}.{}.'.format(lastn,firstn)) + len('{}.{}.'.format(lastn,firstn)))]
        doctype_beg = line[line.find('{}.{}.{}.'.format(lastn, firstn, acct)) + len('{}.{}.{}.'.format(lastn, firstn, acct)):]
        doctype = doctype_beg[:doctype_beg.find('.')]
        date_beg = line[line.find('{}/{}.{}.{}.{}.'.format(household,lastn,firstn,acct,doctype)) + len('{}/{}.{}.{}.{}.'.format(household,lastn,firstn,acct,doctype)):]
        date = date_beg[:date_beg.find('.')]
        print '"',household, '"','"',lastn, '"','"',firstn, '"','"',acct, '"','"',doctype, '"','"',date,'"'

def find_nth(body, s_term, n):
    start = body[::-1].find(s_term)
    while start >= 0 and n > 1:
        start = body[::-1].find(s_term, start+len(s_term))
        n -= 1
    return ((body[::-1])[start:])[::-1]

if __name__ == "__main__": main()

它似乎工作正常,但是当有另一个封闭文件夹时我会遇到问题,然后它会转移所有我的字段...例如,而不是文件驻留在

/ Volumes / HD / Organized Files / Cosby,Bill /

at at / Volumes / HD / Organized Files / Resigned / Cosby,Bill /

我知道必须采用不那么笨重的方式来解决这个问题。

3 个答案:

答案 0 :(得分:1)

这是一个比您的函数find_nth()更实用的工具:
rstrip()

def find_nth(body, s_term, n):
    start = body[::-1].find(s_term)
    print '------------------------------------------------'
    print 'body[::-1]\n',body[::-1]
    print '\nstart == %s' % start
    while start >= 0 and n > 1:
        start = body[::-1].find(s_term, start+len(s_term))
        print 'n == %s    start == %s' % (n,start)
        n -= 1
    print '\n (body[::-1])[start:]\n',(body[::-1])[start:]
    print '\n((body[::-1])[start:])[::-1]\n',((body[::-1])[start:])[::-1]
    print '---------------\n'
    return ((body[::-1])[start:])[::-1]


def cool_find_nth(body, s_term, n):
    assert(len(s_term)==1)
    return body.rsplit(s_term,n)[0] + s_term


ss = 'One / Two / Three / Four / Five / Six / End'
print 'the string\n%s\n' % ss

print ('================================\n'
       "find_nth(ss, '/', 3)\n%s" % find_nth(ss, '/', 3) )

print '================================='
print "cool_find_nth(ss, '/', 3)\n%s" % cool_find_nth(ss, '/', 3)

结果

the string
One / Two / Three / Four / Five / Six / End

------------------------------------------------
body[::-1]
dnE / xiS / eviF / ruoF / eerhT / owT / enO

start == 4
n == 3    start == 10
n == 2    start == 17

 (body[::-1])[start:]
/ ruoF / eerhT / owT / enO

((body[::-1])[start:])[::-1]
One / Two / Three / Four /
---------------

================================
find_nth(ss, '/', 3)
One / Two / Three / Four /
=================================
cool_find_nth(ss, '/', 3)
One / Two / Three / Four /

编辑1

这是另一个非常实用的工具:正则表达式

import re

reg = re.compile('/'
                 '([^/.]*?)/'
                 '([^/.]*?)\.'
                 '([^/.]*?)\.'
                 '([^/.]*?)\.'
                 '([^/.]*?)\.'
                 '([^/.]*?)\.'
                 '[^/.]+\Z')

def main():
    #file = open('~/Desktop/client_docs.csv', "rb")
    #output = open('~/Desktop/client_docs_parsed.txt', "wb")
    li = ['/Household/LastName.FirstName.Account.Doctype.Date.extension',
          '- /Volumes/HD/Organized Files/Cosby, Bill/Cosby.Bill..Profile.2006.doc']
    for line in li:
        print "line == %r" % line
        household,lastn,firstn,acct,doctype,date = reg.search(line).groups('')       
        print ('household == %r\n'
               'lastn     == %r\n'
               'firstn    == %r\n'
               'acct      == %r\n'
               'doctype   == %r\n'
               'date      == %r\n'
               % (household,lastn,firstn,acct,doctype,date))

if __name__ == "__main__": main() 

结果

line == '/Household/LastName.FirstName.Account.Doctype.Date.extension'
household == 'Household'
lastn     == 'LastName'
firstn    == 'FirstName'
acct      == 'Account'
doctype   == 'Doctype'
date      == 'Date'

line == '- /Volumes/HD/Organized Files/Cosby, Bill/Cosby.Bill..Profile.2006.doc'
household == 'Cosby, Bill'
lastn     == 'Cosby'
firstn    == 'Bill'
acct      == ''
doctype   == 'Profile'
date      == '2006'

编辑2

当我发布上一次编辑时,我想知道我的大脑在哪里。以下工作也是如此:

rig = re.compile('[/.]')
rig.split(line)[-7:-1] 

答案 1 :(得分:1)

从我可以收集的信息来看,我相信这将作为一种解决方案,不依赖于以前编译的文件列表

import csv
import os, os.path

# Replace this with the directory where the household directories are stored.
directory = "home"
output = open("Output.csv", "wb")
csvf = csv.writer(output)

headerRow = ["Household", "Lastname", "Firstname", "Account", "Doctype", 
              "Date", "Extension"]

csvf.writerow(headerRow)

for root, households, files in os.walk(directory):
    for household in households:
        for filename in os.listdir(os.path.join(directory, household)):
            # This will create a record for each filename within the "household"
            # Then will split the filename out, using the "." as a delimiter
            # to get the detail
            csvf.writerow([household] + filename.split("."))
output.flush()
output.close()

这使用os库来“遍历”住户列表。然后,对于每个“家庭”,它将收集文件列表。这需要这个列表,在csv文件中生成记录,使用句点作为分隔符来分隔文件的名称。

它利用csv库生成输出,看起来有点像;

"Household,LastName,Firstname,Account,Doctype,Date,Extension"

如果不需要扩展名,则可以通过更改行来省略它:

csvf.writerow([household] + filename.split("."))

csvf.writerow([household] + filename.split(".")[-1])

告诉它只使用文件名的最后一部分,然后从headerRow中删除“Extension”字符串。

希望这会有所帮助

答案 2 :(得分:0)

有点不清楚问题是什么,但同时,这是让你开始的事情:

#!/usr/bin/env python

import os
import csv

with open("f1", "rb") as fin:
    reader = csv.reader(fin, delimiter='.')
    for row in reader:
        # split path
        row = list(os.path.split(row[0])) + row[1:]
        print ','.join(row)

输出:

/Household,LastName,FirstName,Account,Doctype,Date,extension

另一种解释是您希望将每个字段存储在参数中 还有一条额外的道路搞砸了......

这是for循环中row的样子:

['/Household/LastName', 'FirstName', 'Account', 'Doctype', 'Date', 'extension']

然后解决方案可能是倒退。

row[-1]分配给extensionrow[-2]分配给date,依此类推。