类CatalogWalker可以遍历目录列表、计算和报告一些统计信息。

Question

仅供参考，也许对某人有用

from os import walk
from time import time

from abc import ABCMeta, abstractmethod

class Reporter(object):
    __metaclass__=ABCMeta

    @abstractmethod
    def report(self, catalogs_cntr, files_ctnr, speed):
        """ output catalogs_cntr, files_ctnr, speed to somewhere """
        #pass

class ConsoleIndicator(Reporter):
    def __init__(self, out_splitter):
        self.__out_splitter=out_splitter

    def report(self, catalogs_cntr, files_ctnr, speed):
        if self.__out_splitter and self.__out_splitter.pipes:
            _result = "\rcataloges:" + repr(catalogs_cntr).rjust(8) 
            _result += "\tfiles:" + repr(files_ctnr).rjust(13) 
            _result += "\tspeed: %12.3f"%speed + " files/s"

            self.__out_splitter.write(_result)
            self.__out_splitter.flush()        

class CatalogsWalker(object):
    """ """

    def __init__(self, catalogs, reporter=None):
        """ """

        self.__files_cntr, self.__catalogs_cntr, self.__start_time, self.__speed = 0, 0, 0, 0
        self.__reporter = reporter 
        self.__catalogs = catalogs if hasattr(catalogs, "__iter__") else [catalogs]

    def __iter__(self):
        self.__files_cntr, self.__catalogs_cntr, self.__start_time, self.__speed = 0, 0, 0, 0
        self.__start_time=time()
        for catalog in self.__catalogs:
            for p,d,ns in walk(catalog): 
                _ = d;
                self.__catalogs_cntr+=1
                for n in ns:
                    self.__files_cntr+=1
                    self.__speed=self.__files_cntr/(time()-self.__start_time)
                    if self.__reporter:
                        self.__reporter.report(self.__catalogs_cntr, self.__files_cntr, self.__speed)
                    yield p,n

    @property
    def reporter(self):
        return self.__reporter

    @property
    def catalog_counter(self):
        return self.__catalogs_cntr    

    @property
    def file_counter(self):
        return self.__files_cntr

    @property
    def start_time(self):
        return self.__start_time

    @property
    def speed(self):
        return self.__speed

Answer 1

我想谈谈一个微妙的问题。首先，看看此页面解释了可迭代和迭代器之间的区别。。

现在，CatalogsWalker是可迭代的，但是它有跟踪迭代状态的实例属性。如果有人在同一个对象上并行地重复多次，这个跟踪就会出错。这个类更适合作为迭代器。

下面是如何重新排列代码以将CatalogsWalker转换为迭代器的方法。我还从jonrsharpe的出色答案中得到了一些建议。

class CatalogsWalker(object):
    def __init__(self, catalogs, reporter=None):
        self._files_cntr, self._catalogs_cntr, self._start_time, self._speed = 0, 0, 0, 0
        self._reporter = reporter 
        self.__catalogs = catalogs if hasattr(catalogs, "__iter__") else [catalogs]
        self._walker = self._walk()

    def __iter__(self):
        return self

    # def __next__(self):  # Python 3
    def next(self):    # Python 2    
        return next(self._walker)

    def _walk(self):
        self._start_time=time()
        for catalog in self._catalogs:
            for p,_,ns in walk(catalog): 
                self._catalogs_cntr+=1
                for n in ns:
                    self._files_cntr+=1
                    self._speed=self._files_cntr/(time()-self._start_time)
                    if self._reporter:
                        self._reporter.report(self._catalogs_cntr, self._files_cntr, self._speed)
                        yield p,n

    @property
    def reporter(self):
        return self._reporter

    @property
    def catalog_counter(self):
        return self._catalogs_cntr    

    @property
    def file_counter(self):
        return self._files_cntr

    @property
    def start_time(self):
        return self._start_time

    @property
    def speed(self):
        return self._speed