python开发之基于thread线程搜索本地文件的方法

yipeiwu_com6年前Python基础

本文实例讲述了python开发之基于thread线程搜索本地文件的方法。分享给大家供大家参考,具体如下:

先来看看运行效果图:

利用多个线程处理搜索的问题,我们可以发现他很快....

下面是代码部分:

# A parallelized "find(1)" using the thread module.
# This demonstrates the use of a work queue and worker threads.
# It really does do more stats/sec when using multiple threads,
# although the improvement is only about 20-30 percent.
# (That was 8 years ago. In 2002, on Linux, I can't measure
# a speedup. :-( )
# I'm too lazy to write a command line parser for the full find(1)
# command line syntax, so the predicate it searches for is wired-in,
# see function selector() below. (It currently searches for files with
# world write permission.)
# Usage: parfind.py [-w nworkers] [directory] ...
# Default nworkers is 4
import sys
import getopt
import time
import os
from stat import *
import _thread as thread
# Work queue class. Usage:
#  wq = WorkQ()
#  wq.addwork(func, (arg1, arg2, ...)) # one or more calls
#  wq.run(nworkers)
# The work is done when wq.run() completes.
# The function calls executed by the workers may add more work.
# Don't use keyboard interrupts!
class WorkQ:
  # Invariants:
  # - busy and work are only modified when mutex is locked
  # - len(work) is the number of jobs ready to be taken
  # - busy is the number of jobs being done
  # - todo is locked iff there is no work and somebody is busy
  def __init__(self):
    self.mutex = thread.allocate()
    self.todo = thread.allocate()
    self.todo.acquire()
    self.work = []
    self.busy = 0
  def addwork(self, func, args):
    job = (func, args)
    self.mutex.acquire()
    self.work.append(job)
    self.mutex.release()
    if len(self.work) == 1:
      self.todo.release()
  def _getwork(self):
    self.todo.acquire()
    self.mutex.acquire()
    if self.busy == 0 and len(self.work) == 0:
      self.mutex.release()
      self.todo.release()
      return None
    job = self.work[0]
    del self.work[0]
    self.busy = self.busy + 1
    self.mutex.release()
    if len(self.work) > 0:
      self.todo.release()
    return job
  def _donework(self):
    self.mutex.acquire()
    self.busy = self.busy - 1
    if self.busy == 0 and len(self.work) == 0:
      self.todo.release()
    self.mutex.release()
  def _worker(self):
    time.sleep(0.00001)   # Let other threads run
    while 1:
      job = self._getwork()
      if not job:
        break
      func, args = job
      func(*args)
      self._donework()
  def run(self, nworkers):
    if not self.work:
      return # Nothing to do
    for i in range(nworkers-1):
      thread.start_new(self._worker, ())
    self._worker()
    self.todo.acquire()
# Main program
def main():
  nworkers = 4
  #print(getopt.getopt(sys.argv[1:], '-w:'))
  opts, args = getopt.getopt(sys.argv[1:], '-w:')
  for opt, arg in opts:
    if opt == '-w':
      nworkers = int(arg)
  if not args:
    #print(os.curdir)
    args = [os.curdir]
  wq = WorkQ()
  for dir in args:
    wq.addwork(find, (dir, selector, wq))
  t1 = time.time()
  wq.run(nworkers)
  t2 = time.time()
  sys.stderr.write('Total time %r sec.\n' % (t2-t1))
# The predicate -- defines what files we look for.
# Feel free to change this to suit your purpose
def selector(dir, name, fullname, stat):
  # Look for world writable files that are not symlinks
  return (stat[ST_MODE] & 0o002) != 0 and not S_ISLNK(stat[ST_MODE])
# The find procedure -- calls wq.addwork() for subdirectories
def find(dir, pred, wq):
  try:
    names = os.listdir(dir)
  except os.error as msg:
    print(repr(dir), ':', msg)
    return
  for name in names:
    if name not in (os.curdir, os.pardir):
      fullname = os.path.join(dir, name)
      try:
        stat = os.lstat(fullname)
      except os.error as msg:
        print(repr(fullname), ':', msg)
        continue
      if pred(dir, name, fullname, stat):
        print(fullname)
      if S_ISDIR(stat[ST_MODE]):
        if not os.path.ismount(fullname):
          wq.addwork(find, (fullname, pred, wq))
# Call the main program
main()

希望本文所述对大家Python程序设计有所帮助。

相关文章

Python使用正则匹配实现抓图代码分享

内涵:正则匹配,正则替换,页面抓取,图片保存 。 实用的第一次 Python 代码 参考 #!/usr/bin/env python import urllib import re...

使用Pandas的Series方法绘制图像教程

使用Pandas的Series方法绘制图像教程

通常绘制二维曲线的时候可以使用matplotlib,不过如果电脑上安装了pandas的话可以直接使用Series的绘图方法进行图像的绘制。 pandas绘制图像其实也是给予matplot...

Python AES加密实例解析

本文主要是对aes加密技术做一个简要分析,然后使用Python实现,具体介绍如下。 AES,是美国联邦政府采用的一种加密技术,AES有几个模式,其中CBC模式是公认的安全性最好的模式,被...

python中copy()与deepcopy()的区别小结

python中copy()与deepcopy()的区别小结

前言 copy()与deepcopy()之间的区分必须要涉及到python对于数据的存储方式。 深复制被复制对象完全再复制一遍作为独立的新个体单独存在。所以改变原有被复制对象不会对已经复...

python的正则表达式re模块的常用方法

1.re的简介 使用python的re模块,尽管不能满足所有复杂的匹配情况,但足够在绝大多数情况下能够有效地实现对复杂字符串的分析并提取出相关信息。python 会将正则表达式转化为字节...