Python 解释器在迭代完成之前退出 for 循环：

Question

我正在构建一个 Python 脚本来从任意一组子目录中选择和打开文件。为此，我正在构建一个带有 for 循环的函数，以列出一个目录中的所有文件，以及该目录中所有子目录的另一个列表。然后另一个 for 循环通过递归调用相同的函数来更改每个子目录。我在调试模式下反复执行了这个函数，并非常仔细地观察了每个变量，因为它一步一步地变化。

我完全不知所措，因为每个循环都运行了一点点，但它没有完成就退出了。通常在编程中，事情完全可以工作或根本不工作，但它们很少只工作一半。两个循环都这样做，当对函数的第二次调用退出时，它不会返回第一次调用以完成第二次循环，而是完全退出对函数的所有调用。这是代码，非常感谢您的帮助：

def descend(directory, depth, dive=-1): # dive=-1 before we have touched the surface 
    global choices                      # list of files to be chosen from later  
    os.chdir(directory)                 # change to target directory 
    dirlist = []                        # empty the list of directories within the directory   
    baseList = os.listdir(directory)    # list of all files and directories

    if dive < depth:                    # depth corresponds to the levels of sub-directories to be searched
        dive += 1                       # effectively starts at 0

        fileIndex=0                     # index as we step through the baseList
        for x in baseList:              # why doesn't this loop complete?
            if os.path.isfile(x) == True:   # if it's a regular file
                path = directory + '/' + baseList.pop(fileIndex)    # create its full path
                choices.append(path)                                # and add that to our list of choices
            elif os.path.isdir(x) == True:                  # if it's a directory
                dirlist.append(baseList.pop(fileIndex))     # add to list of directories to recursively search
            fileIndex += 1      # add 1 to the file index

        for x in dirlist:       # for each item in the list of directories
            nextdir = directory + '/' + dirlist.pop()   # make a variable of its full path
            descend(nextdir, depth, dive)               # pass that back to this function and repeat

我什至尝试过一些愚蠢的解决方案，例如将 if len(baseList) != 0: continue 添加到循环的底部，这不应该被告知两次，但这对他们中的任何一个都没有帮助。我完全被难住了。

Answer 1

好的，我将在这里回答我自己的问题，因为我很好奇并做了更多的研究，评论太长了。也许这将有助于为其他人澄清它。在通过玩 IDLE 弄清楚发生了什么之后，我在 Pydoc 中找到了这个：

注意：当序列被 循环（这只能发生在可变序列中，例如列表）。一个 内部计数器用于跟踪接下来使用哪个项目， 并且每次迭代都会增加。当这个计数器有 达到循环终止的序列长度。这意味着 如果套件从 序列，下一个项目将被跳过（因为它得到的索引 当前已处理的项目）。同样，如果 套件在当前项目之前插入序列中的项目， 当前项目将在下一次循环中再次处理。 这可能会导致讨厌的错误，可以通过制作 使用整个序列的一部分进行临时复制，例如，

 for x in a[:]:
     if x < 0: a.remove(x)

我的 IDLE 实验的输入和输出：

>>> baselist=[0,1,2,3,4,5,6,7,8,9]
>>> listeven=[]
>>> listodd=[]
>>> index=0
>>> for x in baselist:
        print('baselist is ', baselist)
        print('listeven is ', listeven)
        print('listodd is ', listodd)
        print('x is ', x)
        print('index is ', index)
        if x % 2 == 0:
            print('appending ', x, ' to listeven by popping from baselist')
            listeven.append(baselist.pop(index))
        else:
            print('appending ', x, ' to listodd by popping from baselist')
            listodd.append(baselist.pop(index))
        index += 1

baselist is  [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
listeven is  []
listodd is  []
x is  0
index is  0
appending  0  to listeven by popping from baselist
baselist is  [1, 2, 3, 4, 5, 6, 7, 8, 9]
listeven is  [0]
listodd is  []
x is  2
index is  1
appending  2  to listeven by popping from baselist
baselist is  [1, 3, 4, 5, 6, 7, 8, 9]
listeven is  [0, 2]
listodd is  []
x is  4
index is  2
appending  4  to listeven by popping from baselist
baselist is  [1, 3, 5, 6, 7, 8, 9]
listeven is  [0, 2, 4]
listodd is  []
x is  6
index is  3
appending  6  to listeven by popping from baselist
baselist is  [1, 3, 5, 7, 8, 9]
listeven is  [0, 2, 4, 6]
listodd is  []
x is  8
index is  4
appending  8  to listeven by popping from baselist
>>> baselist
[1, 3, 5, 7, 9]

在循环中从列表中删除项目会导致解释器认为它已提前完成迭代，因为其隐藏的内部索引告诉它下一个索引超出范围，这意味着某些值被跳过！

现在这个函数看起来像：

def descend(directory, depth, dive=-1): 
global files
os.chdir(directory)
subBranches = []
baseList = os.listdir(directory)
if dive < depth:
    dive += 1
    for x in baseList:
        if os.path.isfile(x) == True:
            path = directory + '/' + x
            files.append(path)
        elif os.path.isdir(x) == True:
            subBranches.append(x)
    for x in subBranches:
        nextdir = directory + '/' + x
        descend(nextdir, depth, dive)