python文件排序的方法总结

站长资源 2024/11/24 佚名

2 0 1

在python环境中提供两种排序方案：用库函数sorted()对字符串排序，它的对象是字符；用函数sort()对数字排序，它的对象是数字，如果读取文件的话，需要进行处理（把文件后缀名‘屏蔽'）。

（1）首先：我测试的文件夹是/img/，里面的文件都是图片，如下图所示：

（2）测试库函数sorted()，直接贴出代码：

import numpy as np
import os
 
img_path='./img/'
 
img_list=sorted(os.listdir(img_path))#文件名按字母排序
img_nums=len(img_list)
for i in range(img_nums):
    img_name=img_path+img_list[i]
    print(img_name)

运行效果如下：

从图片可以清晰的看出，文件名是按字符排序的。

（3）测试函数sort()，代码：

import numpy as np
import os
img_path='./img/'
 
img_list=os.listdir(img_path)
img_list.sort()
img_list.sort(key = lambda x: int(x[:-4])) ##文件名按数字排序
img_nums=len(img_list)
for i in range(img_nums):
    img_name=img_path+img_list[i]
    print(img_name)

运行效果如下：

可以看出，文件名是按数字排序的；顺便提下，sort函数中用到了匿名函数(key = lambda x:int(x[:-4]))，其作用是将后缀名'.jpg'“屏蔽”（因为‘.jpg'是4个字符，所以[:-4]的含义是从文件名开始到倒数第四个字符为止），具体看python的匿名函数和数组取值方式。

实例扩展：

import gzip
import os
from multiprocessing import Process, Queue, Pipe, current_process, freeze_support
from datetime import datetime
def sort_worker(input,output):
 while True:
 lines = input.get().splitlines()
 element_set = {}
 for line in lines:
  if line.strip() == 'STOP':
   return
  try:
   element = line.split(' ')[0]
   if not element_set.get(element): element_set[element] = ''
  except:
   pass
 sorted_element = sorted(element_set)
 #print sorted_element
 output.put('\n'.join(sorted_element))
def write_worker(input, pre):
 os.system('mkdir %s'%pre)
 i = 0
 while True:
  content = input.get()
  if content.strip() == 'STOP':
   return
  write_sorted_bulk(content, '%s/%s'%(pre, i))
  i += 1
def write_sorted_bulk(content, filename):
 f = file(filename, 'w')
 f.write(content)
 f.close()
def split_sort_file(filename, num_sort = 3, buf_size = 65536*64*4):
 t = datetime.now()
 pre, ext = os.path.splitext(filename)
 if ext == '.gz':
  file_file = gzip.open(filename, 'rb')
 else:
  file_file = open(filename)
 bulk_queue = Queue(10)
 sorted_queue = Queue(10)
 NUM_SORT = num_sort
 sort_worker_pool = []
 for i in range(NUM_SORT):
  sort_worker_pool.append( Process(target=sort_worker, args=(bulk_queue, sorted_queue)) )
  sort_worker_pool[i].start()
 NUM_WRITE = 1
 write_worker_pool = []
 for i in range(NUM_WRITE):
  write_worker_pool.append( Process(target=write_worker, args=(sorted_queue, pre)) )
  write_worker_pool[i].start()
 buf = file_file.read(buf_size)
 sorted_count = 0
 while len(buf):
  end_line = buf.rfind('\n')
  #print buf[:end_line+1]
  bulk_queue.put(buf[:end_line+1])
  sorted_count += 1
  if end_line != -1:
   buf = buf[end_line+1:] + file_file.read(buf_size)
  else:
   buf = file_file.read(buf_size)
 for i in range(NUM_SORT):
  bulk_queue.put('STOP')
 for i in range(NUM_SORT):
  sort_worker_pool[i].join()
  
 for i in range(NUM_WRITE):
  sorted_queue.put('STOP')
 for i in range(NUM_WRITE):
  write_worker_pool[i].join()
 print 'elasped ', datetime.now() - t
 return sorted_count
from heapq import heappush, heappop
from datetime import datetime
from multiprocessing import Process, Queue, Pipe, current_process, freeze_support
import os
class file_heap:
 def __init__(self, dir, idx = 0, count = 1):
  files = os.listdir(dir)
  self.heap = []
  self.files = {}
  self.bulks = {}
  self.pre_element = None
  for i in range(len(files)):
   file = files[i]
   if hash(file) % count != idx: continue
   input = open(os.path.join(dir, file))
   self.files[i] = input
   self.bulks[i] = ''
   heappush(self.heap, (self.get_next_element_buffered(i), i))
 def get_next_element_buffered(self, i):
  if len(self.bulks[i]) < 256:
   if self.files[i] is not None:
    buf = self.files[i].read(65536)
    if buf:
     self.bulks[i] += buf
    else:
     self.files[i].close()
     self.files[i] = None
  end_line = self.bulks[i].find('\n')
  if end_line == -1:
   end_line = len(self.bulks[i])
  element = self.bulks[i][:end_line]
  self.bulks[i] = self.bulks[i][end_line+1:]
  return element
 def poppush_uniq(self):
  while True:
   element = self.poppush()
   if element is None:
    return None
   if element != self.pre_element:
    self.pre_element = element
    return element
 def poppush(self):
  try:
   element, index = heappop(self.heap)
  except IndexError:
   return None
  new_element = self.get_next_element_buffered(index)
  if new_element:
   heappush(self.heap, (new_element, index))
  return element
def heappoppush(dir, queue, idx = 0, count = 1):
 heap = file_heap(dir, idx, count)
 while True:
  d = heap.poppush_uniq()
  queue.put(d)
  if d is None: return
def heappoppush2(dir, queue, count = 1):
 heap = []
 procs = []
 queues = []
 pre_element = None
 for i in range(count):
  q = Queue(1024)
  q_buf = queue_buffer(q)
  queues.append(q_buf)
  p = Process(target=heappoppush, args=(dir, q_buf, i, count))
  procs.append(p)
  p.start()
 queues = tuple(queues)
 for i in range(count):
  heappush(heap, (queues[i].get(), i))
 while True:
  try:
   d, i= heappop(heap)
  except IndexError:
   queue.put(None)
   for p in procs:
    p.join()
   return
  else:
   if d is not None:
    heappush(heap,(queues[i].get(), i))
    if d != pre_element:
     pre_element = d
     queue.put(d)
def merge_file(dir):
 heap = file_heap( dir )
 os.system('rm -f '+dir+'.merge')
 fmerge = open(dir+'.merge', 'a')
 element = heap.poppush_uniq()
 fmerge.write(element+'\n')
 while element is not None:
  element = heap.poppush_uniq()
  fmerge.write(element+'\n')
class queue_buffer:
 def __init__(self, queue):
  self.q = queue
  self.rbuf = []
  self.wbuf = []
 def get(self):
  if len(self.rbuf) == 0:
   self.rbuf = self.q.get()
  r = self.rbuf[0]
  del self.rbuf[0]
  return r
 def put(self, d):
  self.wbuf.append(d)
  if d is None or len(self.wbuf) > 1024:
   self.q.put(self.wbuf)
   self.wbuf = []
def diff_file(file_old, file_new, file_diff, buf = 268435456):
 print 'buffer size', buf
 from file_split import split_sort_file
 os.system('rm -rf '+ os.path.splitext(file_old)[0] )
 os.system('rm -rf '+ os.path.splitext(file_new)[0] )
 t = datetime.now()
 split_sort_file(file_old,5,buf)
 split_sort_file(file_new,5,buf)
 print 'split elasped ', datetime.now() - t
 os.system('cat %s/* | wc -l'%os.path.splitext(file_old)[0])
 os.system('cat %s/* | wc -l'%os.path.splitext(file_new)[0])
 os.system('rm -f '+file_diff)
 t = datetime.now()
 zdiff = open(file_diff, 'a')
 old_q = Queue(1024)
 new_q = Queue(1024)
 old_queue = queue_buffer(old_q)
 new_queue = queue_buffer(new_q)
 h1 = Process(target=heappoppush2, args=(os.path.splitext(file_old)[0], old_queue, 3))
 h2 = Process(target=heappoppush2, args=(os.path.splitext(file_new)[0], new_queue, 3))
 h1.start(), h2.start()
 old = old_queue.get()
 new = new_queue.get()
 old_count, new_count = 0, 0
 while old is not None or new is not None:
  if old > new or old is None:
   zdiff.write('< '+new+'\n')
   new = new_queue.get()
   new_count +=1
  elif old < new or new is None:
   zdiff.write('> '+old+'\n')
   old = old_queue.get()
   old_count +=1
  else:
   old = old_queue.get()
   new = new_queue.get()
 print 'new_count:', new_count
 print 'old_count:', old_count
 print 'diff elasped ', datetime.now() - t
 h1.join(), h2.join()

python,文件排序

广告合作：本站广告合作请联系QQ：858582 申请时备注：广告合作（否则不回）
免责声明：本站资源来自互联网收集,仅供用于学习和交流,请遵循相关法律法规,本站一切资源不代表本站立场,如有侵权、后门、不妥请联系本站删除！

评论“python文件排序的方法总结”

暂无评论...

python文件排序的方法总结

通过实例了解pythonslots使用方法

python如何遍历指定路径下所有文件（按按照时间区间检索）

评论“python文件排序的方法总结”

《魔兽世界》大逃杀！60人新游玩模式《强袭风暴》3月21日上线

更新日志

友情链接

python文件排序的方法总结

通过实例了解python__slots__使用方法

python如何遍历指定路径下所有文件（按按照时间区间检索）

评论“python文件排序的方法总结”

《魔兽世界》大逃杀！60人新游玩模式《强袭风暴》3月21日上线

更新日志

友情链接

通过实例了解pythonslots使用方法