壓縮檔縮圖 python 實現

可以支援zip，rar,7zip,tar.gz等等各式

輸入壓縮檔名稱，返回一個xml檔案結構

#!usr/bin/python

#coding=utf-8
import uuid
import os
import zipfile
import re
import sys
import officetoimage
import rarfile
import tarfile
import subprocess

########SET CODE TO DEAL CHINESE############
default_encoding = 'utf-8'
if sys.getdefaultencoding()!=default_encoding:
    reload(sys)
    sys.setdefaultencoding(default_encoding)

#######TO GENERATE XML USING LIST##########
#######DIR IS LIST OF FILE STRUCT##########
def xml_generate(dir):
    body = ''
    for i in dir:
        if i.partition('/')[2] == '':            
            if(i.endswith('/')):
                body +="<directory name=\""+i.partition('/')[0] +'\">'
                tmp =[ j for j in dir if j.startswith(i) and j!=i]
                for j in range(len(tmp)):
                    tmp[j] = re.sub('^'+i,'',tmp[j])                                   
                if len(tmp)>0:
                    body += xml_generate(tmp)                   
                body +="</directory>"
            else:
                body +="<file>"+i.partition('/')[0]+"</file>"           
    return body

            
#######THIS IS A START TO GENERATE XML########## 
#######XML WILL BE WRITE TO FILE NAMED ZIP FILENAME##########
def xml_start(dir,outname):
    body = xml_generate(dir)            
    head = """<?xml version="1.0" encoding="UTF-8" ?>
<source>%s</source>""" %body
    tmp = outname.split('/')[0]
    if os.path.exists(tmp) is not True:
        os.makedirs(tmp,0755)
    out = open(outname,'w')
    out.write(head)
    out.close()
    
   
#######TO START DEAL ZIP##########        
def zip_deal(filename,readname,pwd):
    
    try:
        file_to_deal = zipfile.ZipFile(filename,'r')
    except IOError:
        print 'error zip'
        sys.exit(0)
   
    file_to_deal.setpassword(pwd)
    try:
        document=file_to_deal.read(readname)
    except:
        out = str(uuid.uuid1())
        command = '7z e -otmp/'+out+' -p'+pwd+' '+filename+ ' '+readname
        os.popen(command)    
        return 'tmp/'+out+'/'+readname.split('/')[-1]
        
    file_to_deal.close()
    
    return document

#######TO START DEAL RAR##########        
def rar_deal(filename,readname,pwd):
    try:
        file_to_deal = rarfile.RarFile(filename)
    except IOError:
        print 'error rar open'
        sys.exit(0)
        
    if file_to_deal.needs_password():
        file_to_deal.setpassword(pwd)
    try:
        document=file_to_deal.read(readname)
    except:
        print 'error rar read'
        sys.exit(0)
        
    file_to_deal.close()   
    return document

#######TO START DEAL TAR##########        
def tar_deal(filename,readname):
    try:
        file_to_deal = tarfile.open(filename)
    except IOError:
        print 'error'
        sys.exit(0)
   
    try:
        do = file_to_deal.extractfile(readname)
    except:
        print 'error'
        sys.exit(0)
    document = do.read()   
    file_to_deal.close()   
    return document

#######TO START DEAL ZIP##########        
def z7_deal(filename,readname,pwd):
    uuid_s = str(uuid.uuid1())
    command = '7z e -otmp/'+uuid_s+' -p'+pwd+' '+filename+ ' '+readname
    os.popen(command)    
    return 'tmp/'+uuid_s+'/'+readname.split('/')[-1]

#######FILE TO ZIP AND GET THE FILE STRUCT########## 
#######IF IS FILE RETURN PDF########## 
def file_to_zip(filename,pwd):
    out = 'tmp/' + str(uuid.uuid1())+'.xml'
    file_in = filename.partition('/')
    try:
        file_to_deal = zipfile.ZipFile(file_in[0],'r')
    except IOError:
        print 'error'
        sys.exit(0)
        
    search_list = file_to_deal.namelist()
    file_to_deal.close()
    if filename.endswith('.zip'):
        xml_start(search_list,out)  
        return out,'success'
    else :
        dir = 'tmp/'+str(uuid.uuid1())
        outname = dir+'/'+file_in[-1].split('/')[-1]   
        if os.path.exists(dir) is not True:
            os.makedirs(dir,0755)
        path= file_in[-1]
        file_content = zip_deal(file_in[0],path.decode('utf-8'),pwd)
        out = open(outname,'w')
       
        out.write(file_content)
        out.close()
        name_pdf = officetoimage.file_to_pdf(outname)
        return outname,name_pdf
    
def file_to_rar(filename,pwd):
    out = 'tmp/'+str(uuid.uuid1())+'.xml'
    file_in = filename.partition('/')
    try:
        file_to_deal = rarfile.RarFile(file_in[0],'r')
    except IOError:
        print 'error rar'
        sys.exit(0)     
    if file_to_deal.needs_password():
        file_to_deal.setpassword(pwd)
    search_list = file_to_deal.namelist()    
   
    for f in file_to_deal.infolist():
        if f.isdir():
            i = search_list.index(f.filename)
            search_list[i] += '/'
            
    file_to_deal.close()    
    for i in range(len(search_list)):
        search_list[i] = search_list[i].replace('\\','/')
        
    if filename.endswith('.rar'):
        xml_start(search_list,out)  
        return out,'success'
    else :
        dir = 'tmp/'+str(uuid.uuid1())
        outname = dir+'/'+file_in[-1].split('/')[-1]   
        if os.path.exists(dir) is not True:
            os.makedirs(dir,0755)
        #outname = 'tmp/'+str(uuid.uuid1())+'/'+file_in[-1].split('/')[-1]         
        path= file_in[-1]
        file_content = rar_deal(file_in[0],path.decode('utf-8'),pwd)
        out = open(outname,'w')
        out.write(file_content)
        out.close()
        name_pdf = officetoimage.file_to_pdf(outname)
        return outname,name_pdf
    
def file_to_tar(filename,pwd):
    out = 'tmp/'+str(uuid.uuid1())+'.xml'
    file_in = filename.partition('/')
    try:
        file_to_deal = tarfile.open(file_in[0],'r')
    except IOError:
        print 'error'
        sys.exit(0)     
    
    search_list = file_to_deal.getnames() 
    for f in search_list:
        if file_to_deal.getmember(f).isdir():
            i = search_list.index(f)
            search_list[i] += '/'  
            
    file_to_deal.close()    
        
    if filename.endswith('.tar') or filename.endswith('.tar.gz'):
        xml_start(search_list,out)  
        return out,'success'
    else :
        dir = 'tmp/'+str(uuid.uuid1())
        outname = dir+'/'+file_in[-1].split('/')[-1]   
        if os.path.exists(dir) is not True:
            os.makedirs(dir,0755)
        #outname = 'tmp/'+str(uuid.uuid1())+'/'+file_in[-1].split('/')[-1]        
        path= file_in[-1]
        file_content = tar_deal(file_in[0],path.decode('utf-8'))
        out = open(outname,'w')
        out.write(file_content)
        out.close()
        name_pdf = officetoimage.file_to_pdf(outname)
        return outname,name_pdf
    
def file_to_7z(filename,pwd):
    #out = 'tmp/' + filename.split('.')[0]+'.xml'
    out = 'tmp/'+str(uuid.uuid1())+'.xml'
    #print out
    file_in = filename.partition('/')
    if pwd == None:
        pwd =''
    if filename.endswith('.7z') :
        command = '7z l -p'+pwd+' '+filename 
        f = os.popen(command)    
        content = f.readlines()
        if content[6].startswith('Error'):
            return 'error 7z'
        for i in range(len(content)):
            if content[i].startswith('-----'):
                content = content[i+1:-2]
                break
       
        search_list = []
        for s in content:
            if s.split(' ')[2][0] == 'D':
                search_list.append(s.split(' ')[-1][:-1]+'/')
            else:
                search_list.append(s.split(' ')[-1][:-1])
         
        xml_start(search_list,out)  
        return out,'success'
    else :       
        path= file_in[-1]
        file_path = z7_deal(file_in[0],path.decode('utf-8'),pwd)
        name_pdf = officetoimage.file_to_pdf(file_path)
        return file_path,name_pdf
    

##### FILENAME IS FILE.ZIP OR FILE.ZIP/FILE#######    
if __name__ == '__main__':     
  
    filename = sys.argv[1]
    try:
        pwd = sys.argv[2]
    except:
        pwd = ''
    
    if filename.split('/')[0].endswith('.zip'):
        file_path,dir_content = file_to_zip(filename,pwd)
    elif filename.split('/')[0].endswith('.rar'):
        file_path,dir_content = file_to_rar(filename,pwd)
    elif filename.split('/')[0].endswith('.tar.gz') or filename.split('/')[0].endswith('.tar') :
        file_path,dir_content = file_to_tar(filename,pwd)
    elif filename.split('/')[0].endswith('.7z'):
        file_path,dir_content = file_to_7z(filename,pwd)
    else:
        dir_content = 'error code'
    print file_path,dir_content