在Python中使用os.walk为相应的文件夹和子文件夹创建txt文件

Question

I have been trying to create a function a that uses os.walk to go through a directory tree and create a text file that is named after the folder and only create it the text file if there is data files in a source folder and append the files to that txt file. It practically makes a backup of the data files on another mapped network drive

def copy_from_dest(src, dest):
    os.chdir(src)
    for root, dirs, files in os.walk(".",topdown=True):
        for name in dirs:
            newdir = (os.path.join(root, name))
            os.chdir(newdir)        
            filenames = glob.glob('*.txt')
            if len(filenames) > 1:
                for fname in filenames:
                    add_to_file(src + newdir, fname, dest + newdir, newdir + "_data.txt")
            os.chdir(src)

add_to_file is a function that appends the files to the text file and removes the afterwards. The problem I am facing is that the first branch of directories works perfectly the second branch (or subdirectory of subdirectory) makes the file in the second branch or second step in when it was suppose to make it in the third.

For example:

• Root/ folder works fine
• Root/folderstep1 works fine
• Root/folderstep1/folderstep2/folderstep3 incorrectly writes the txt file in folderstep2

Please can anyone assist? Thanks in advance!

Answer 1

The fourth parameter to your add_to_file function looks suspicious: It seems that you want to construct a file name like "folderstep3_data.txt" for the current directory, but instead you're passing something like ./folderstep1/folderstep2/folderstep3_data.txt . But since I don't know, how add_to_file works internally, I cannot tell.

Try logging the parameters to add_to_file :

print (src + newdir), fname, (dest + newdir), newdir + "_data.txt"

and see, if that reveals the problem.

Two additional points:

if len(filenames)>1: should probably be a >0 ? Or do you only want to write something, if at least two text files exist?

Also, since os.walk() already walks all the subdirectories for you and also gives you a list of filenames for every one, couldn't you simply thst, if files (the var you get from the os.walk() -loop) contains a text file (using str.endswith('.txt') and use only root as directory?

If you have a structure like

+ src
+--+ subfolder 1
|  +-- file1.txt
|  +-- file2.txt
|  +--+ subfolder1.2
|     +-- file3.txt
+--+ subfolder 2
   +-- file4.txt

then os.walk() will give you

root=src, dirs=['subfolder1','subfolder2'], files=[]
root=src\subfolder1, dirs=['subfolder1.2'], files=['file1.txt','file2.txt']
root=src\subfolder1\subfolder1.2, dirs=[], files=['file3.txt']
root=src\subfolder2, files=['file3.txt']

So it seems, that root and files should be enough information to make the call to add_file . So there would be no need for os.chdir() and glob.glob() .