os.walk目錄遍歷<a id="sec-1" name="sec-1"></a>
每個月都有那么幾天想劃水，又到劃水的日子了，今天分享的是剛在處理遍歷目錄相關用到的相關方法。
os.walk<a id="sec-1-1" name="sec-1-1"></a>
os.walk的參數如下:
os.walk(top, topdown=True, onerror=None, followlinks=False)
其中：
- top是要遍歷的目錄。
- topdown是代表要從上而下遍歷還是從下往上遍歷。
- onerror可以用來設置當便利出現錯誤的處理函數(該函數接受一個OSError的實例作為參數)，設置為空則不作處理。
- followlinks表示是否要跟隨目錄下的鏈接去繼續遍歷，要注意的是，os.walk不會記錄已經遍歷的目錄，所以跟隨鏈接遍歷的話有可能一直循環調用下去。
os.walk返回的是一個3個元素的元組 (root, dirs, files) ，分別表示遍歷的路徑名，該路徑下的目錄列表和該路徑下文件列表。注意目錄列表和文件列表不是具體路徑，需要具體路徑(從root開始的路徑)的話可以用 os.path.join(root,dir) 和 os.path.join(root,dir) 。
例子<a id="sec-1-2" name="sec-1-2"></a>
假設現在存在如下的文件和目錄結構:
? test_os_walk git:(master) ? tree
.
├── a.py
├── b.py
├── c.py
├── dir1
│ ├── dir4
│ │ ├── g.py
│ │ └── h.py
│ ├── dirx
│ │ ├── diry
│ │ │ └── k.py
│ │ └── z.py
│ ├── e.py
│ ├── f.py
│ └── g.py
├── dir2
│ ├── dira
│ │ └── dirb
│ │ └── dirc
│ │ └── aha.py
│ ├── k.py
│ ├── l.py
│ └── m.py
└── dir3
├── dir5
│ └── z.py
├── x.py
└── y.py
10 directories, 17 files
測試topdown<a id="sec-1-2-1" name="sec-1-2-1"></a>
當我用 os.walk 遍歷這個目錄時，程序和輸出如下:
import os
path = '/Users/nisen/Projects/python_advanced_class/test/test_os_walk'
for root, dirs, files in os.walk(path, True):
print 'root: %s' % root
print 'dirs: %s' % dirs
print 'files: %s' % files
print ''
結果如下，從root的路徑可以看出遍歷是自上而下的：
? test git:(master) ? python test11.py
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk
dirs: ['dir1', 'dir2', 'dir3']
files: ['a.py', 'b.py', 'c.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1
dirs: ['dir4', 'dirx']
files: ['e.py', 'f.py', 'g.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1/dir4
dirs: []
files: ['g.py', 'h.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1/dirx
dirs: ['diry']
files: ['z.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1/dirx/diry
dirs: []
files: ['k.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2
dirs: ['dira']
files: ['k.py', 'l.py', 'm.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2/dira
dirs: ['dirb']
files: []
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2/dira/dirb
dirs: ['dirc']
files: []
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2/dira/dirb/dirc
dirs: []
files: ['aha.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir3
dirs: ['dir5']
files: ['x.py', 'y.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir3/dir5
dirs: []
files: ['z.py']
而當設置os.walk的topdown為False時，結果如下, 可以看出他是自上而下遍歷的：
? test git:(master) ? python test11.py
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1/dir4
dirs: []
files: ['g.py', 'h.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1/dirx/diry
dirs: []
files: ['k.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1/dirx
dirs: ['diry']
files: ['z.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir1
dirs: ['dir4', 'dirx']
files: ['e.py', 'f.py', 'g.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2/dira/dirb/dirc
dirs: []
files: ['aha.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2/dira/dirb
dirs: ['dirc']
files: []
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2/dira
dirs: ['dirb']
files: []
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir2
dirs: ['dira']
files: ['k.py', 'l.py', 'm.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir3/dir5
dirs: []
files: ['z.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk/dir3
dirs: ['dir5']
files: ['x.py', 'y.py']
root: /Users/nisen/Projects/python_advanced_class/test/test_os_walk
dirs: ['dir1', 'dir2', 'dir3']
files: ['a.py', 'b.py', 'c.py']
運行時修改遍歷目錄<a id="sec-1-2-2" name="sec-1-2-2"></a>
當topdown設置為True時，可以在處理時修改返回的 dirs 列表，這樣可以遍歷下面的目錄時會根據修改后的 dirs 來遍歷。比如下面的例子，在遍歷的時候不把"CSV"目錄包括在內:
import os
from os.path import join, getsize
for root, dirs, files in os.walk('python/Lib/email'):
print root, "consumes",
print sum(getsize(join(root, name)) for name in files),
print "bytes in", len(files), "non-directory files"
if 'CVS' in dirs:
dirs.remove('CVS') # don't visit CVS directories
