我不清楚你的问题,你可能想重新阅读 os.walk 文档。 root
是被遍历的当前目录。 dirs
是 中的子目录root
,files
是 中的文件root
。由于您的代码现在是您计算相同的文件(从根目录)并将其记录为每个子目录中的文件数。
这就是我想出的。希望它接近你想要的。如果没有,请调整 :) 它会打印目录、目录中的文件数以及目录及其所有子目录中的文件数。
import os
import csv
# Open the csv and write headers.
with open("Subject_Task_Count.csv",'wb') as out:
outwriter = csv.writer(out)
outwriter.writerow(['Directory','FilesInDir','FilesIncludingSubdirs'])
# Track total number of files in each subdirectory by absolute path
totals = {}
# topdown=False iterates lowest level (leaf) subdirectories first.
# This way I can collect grand totals of files per subdirectory.
for path,dirs,files in os.walk('.',topdown=False):
files_in_current_directory = len(files)
# Start with the files in the current directory and compute a
# total for all subdirectories, which will be in the `totals`
# dictionary already due to topdown=False.
files_including_subdirs = files_in_current_directory
for d in dirs:
fullpath = os.path.abspath(os.path.join(path,d))
# On my Windows system, Junctions weren't included in os.walk,
# but would show up in the subdirectory list. this try skips
# them because they won't be in the totals dictionary.
try:
files_including_subdirs += totals[fullpath]
except KeyError as e:
print 'KeyError: {} may be symlink/junction'.format(e)
totals[os.path.abspath(path)] = files_including_subdirs
outwriter.writerow([path,files_in_current_directory,files_including_subdirs])