ziprar.py
__author__ = 'williezh'
#!/usr/bin/env python3
import os
import sys
import time
import shutil
import zipfile
from zipfile import ZIP_DEFLATED
#Zip文件处理类
class ZFile(object):
def __init__(self, fname, mode='r', basedir=''):
self.fname = fname
self.mode = mode
if self.mode in ('w', 'a'):
self.zfile = zipfile.ZipFile(fname, mode, compression=ZIP_DEFLATED)
else:
self.zfile = zipfile.ZipFile(fname, self.mode)
self.basedir = basedir
if not self.basedir:
self.basedir = os.path.dirname(fname)
def addfile(self, path, arcname=None):
path = path.replace('//', '/')
if not arcname:
if path.startswith(self.basedir):
arcname = path[len(self.basedir):]
else:
arcname = ''
self.zfile.write(path, arcname)
def addfiles(self, paths):
for path in paths:
if isinstance(path, tuple):
self.addfile(*path)
else:
self.addfile(path)
def close(self):
self.zfile.close()
def extract_to(self, path):
for p in self.zfile.namelist():
self.extract(p, path)
def extract(self, fname, path):
if not fname.endswith('/'):
fn = os.path.join(path, fname)
ds = os.path.dirname(fn)
if not os.path.exists(ds):
os.makedirs(ds)
with open(fn, 'wb') as f:
f.write(self.zfile.read(fname))
#创建Zip文件
def createZip(zfile, files):
z = ZFile(zfile, 'w')
z.addfiles(files)
z.close()
#解压缩Zip到指定文件夹
def extractZip(zfile, path):
z = ZFile(zfile)
z.extract_to(path)
z.close()
#解压缩rar到指定文件夹
def extractRar(zfile, path):
rar_command1 = "WinRAR.exe x -ibck %s %s" % (zfile, path)
rar_command2 = r'"C:\WinRAR.exe" x -ibck %s %s' % (zfile, path)
try:
res = os.system(rar_command1)
if res == 0:
print("Path OK.")
except:
try:
res = os.system(rar_command2)
if res == 0:
print("Success to unrar the file {}.".format(path))
except:
print('Error: can not unrar the file {}'.format(path))
# 解压多个压缩文件到一个临时文件夹
def extract_files(file_list):
newdir = str(int(time.time()))
for fn in file_list:
subdir = os.path.join(newdir, fn)
if not os.path.exists(subdir):
os.makedirs(subdir)
if fn.endswith('.zip'):
extractZip(fn, subdir)
elif fn.endswith('.rar'):
extractRar(fn, subdir)
return newdir
# 查找一个文件夹中的某些文件, 返回文件内容包含findstr_list中所有字符串的文件
def findstr_at(basedir, file_list, findstr_list):
files = []
for r, ds, fs in os.walk(basedir):
for fn in fs:
if fn in file_list:
with open(os.path.join(r, fn)) as f:
s = f.read()
if all(i in s for i in findstr_list):
files.append(os.path.join(r, fn))
return files
if __name__ == '__main__':
files = [i for i in sys.argv[1:] if not i.startswith('-')]
unzipfiles = [i for i in files if i.endswith('.zip') or i.endswith('.rar')]
xmlfiles = [i for i in files if i.endswith('.xml')]
save_unzipdir = True if '-s' in sys.argv else False
findstr = [i.split('=')[-1] for i in sys.argv if i.startswith('--find=')]
findstring = ', '.join(['`{}`'.format(i) for i in findstr])
newdir = extract_files(unzipfiles)
result = findstr_at(newdir, xmlfiles, findstr)
if not result:
msg = 'None of the file(s) contain the given string {}.'
print(msg.format(findstring))
else:
msg = '{} file(s) contain the given string {}:'
print(msg.format(len(result), findstring))
print('\n'.join([i.replace(newdir+os.sep, '') for i in sorted(result)]))
if not save_unzipdir:
shutil.rmtree(newdir)$ python3 ziprar.py aaa.zip aaa2.zip aaa3.zip aaa.xml aaa1.xml aaa2.xml --find="It was" --find="when"
None of the file(s) contain the given string `It was`, `when`.
$ python3 ziprar.py aaa.zip aaa2.zip aaa3.zip aaa.xml aaa1.xml aaa2.xml --find="It was" --find="I"
2 file(s) contain the given string `It was`, `I`:
aaa.zip/aaa2.xml
aaa2.zip/aaa2.xml
$ python3 ziprar.py aaa.zip aaa2.zip aaa3.zip aaa.xml aaa1.xml aaa2.xml --find="It was"
2 file(s) contain the given string `It was`:
aaa.zip/aaa2.xml
aaa2.zip/aaa2.xml