python使用正则搜索字符串或文件中的浮点数代码实例

用python和numpy处理数据次数比较多,写了几个小函数,可以方便地读写数据:

# -*- coding: utf-8 -*-
#----------------------------------------------------------------------
# filename:gettxtdata.py
#功能:读取字符串和文件中的数值数据(浮点数)
#主要提供类似matlab中的dlmread和dlmwrite函数
#同时提供loadtxtdata和savetxtdata函数
#data: 2013-1-10
#author:吴徐平
#----------------------------------------------------------------------
import numpy
#----------------------------------------------------------------------
def stringtodoublearray(string):
  """
  #将字符串中的所有非double类型的字符全部替换成空格
  #以'#'开头注释直至行尾,都被清空
  #返回一维numpy.array数组

  """ 
  from stringio import stringio
  import re
  
  dataarray=numpy.empty([0],numpy.float64)

  if len(string.strip())>0:
    #清空注释行,都是以'#'开头子字符
    doublestring=re.sub('#.*$', " ", string, count=0, flags=re.ignorecase)
    #删除非数字字符      
    doublestring=re.sub('[^0-9.e+-]', " ", doublestring, count=0, flags=re.ignorecase)
    #去掉不正确的数字格式(代码重复是有必要的)
    doublestring=re.sub('[.e+-](?=\s)', " ", doublestring, count=0, flags=re.ignorecase)
    doublestring=re.sub('[.e+-](?=\s)', " ", doublestring, count=0, flags=re.ignorecase)
    doublestring=re.sub('[e+-]$', " ", doublestring, count=0, flags=re.ignorecase)
    doublestring=re.sub('[e+-]$', " ", doublestring, count=0, flags=re.ignorecase)
    #去掉首尾空格
    doublestring=doublestring.strip()
    if len(doublestring)>0:
      striods=stringio(doublestring)
      dataarray= numpy.genfromtxt(striods)
  
  return dataarray

#----------------------------------------------------------------------
def getdoublelistfromstring(string):
  """
  #使用换行符分割字符串
  #将字符串中的所有非double类型的字符全部替换成空格
  #以'#'开头注释直至行尾,都被清空
  #将每一行转换成numpy.array数组
  #返回numpy.array数组的列表

  """ 
  from stringio import stringio
  import re
 
  doublelist=[]
  stringlist=string.split('\n')#使用换行符分割字符串
  for line in stringlist:
    if len(line.strip())>0:
      #清空注释行,都是以'#'开头子字符
      doublestring=re.sub('#.*$', " ", line, count=0, flags=re.ignorecase)
      #删除非数字字符      
      doublestring=re.sub('[^0-9.e+-]', " ", doublestring, count=0, flags=re.ignorecase)
      #去掉不正确的数字格式(代码重复是有必要的)
      doublestring=re.sub('[.e+-](?=\s)', " ", doublestring, count=0, flags=re.ignorecase)
      doublestring=re.sub('[.e+-](?=\s)', " ", doublestring, count=0, flags=re.ignorecase)
      doublestring=re.sub('[e+-]$', " ", doublestring, count=0, flags=re.ignorecase)
      doublestring=re.sub('[e+-]$', " ", doublestring, count=0, flags=re.ignorecase)
      #去掉首尾空格
      doublestring=doublestring.strip()
      if len(doublestring)>0:
        striods=stringio(doublestring)
        doublelist.append(numpy.genfromtxt(striods))   
  return doublelist
  
#----------------------------------------------------------------------
def getdoublelistfromfile(filename):
  """
  #将文本文件中的所有double类型的字符全部替换成numpy.array数组
  #每一行都是numpy.array数组
  ##返回numpy.array数组的列表
  #注意:返回列表的每个元素又都是一个numpy.array数组
  #注意:返回列表的每个元素(或文件每行)可以包含不同多个数的数字

  """ 
  file=open(filename, 'r')
  read_file = file.read()
  file.close() 
  doublelist=getdoublelistfromstring(read_file)
  return doublelist

def dlmread(filename,dtype=numpy.float64):
  """
  #load data from txt-file.
  #分隔符默认是:";",",",空格类 (包括\t)等等
  #以#开头的被认为是注释,不会被读取
  #return value:二维数值数组(numpy.ndarray)
  #对文本中数据的排列格式要求最低,且容许出现注释字符,智能化程度最高,但速度较慢
  """
  doublelist=getdoublelistfromfile(filename)
  dlsize=[]#每一行数组的大小
  for dl in doublelist:
    dlsize.append(dl.size)
    
  mincolumnsize=min(dlsize)#数组的最大列数
  maxcolumnsize=max(dlsize)#数组的最小列数
  #数组创建和赋值
  doublearray=numpy.empty([len(doublelist),mincolumnsize],dtype=dtype)
  
  row=range(0,len(doublelist))
  colum=range(0,mincolumnsize)
  
  for i in row:
    for j in colum:
      doublearray[i][j]=doublelist[i][j] 
    
  return doublearray
#----------------------------------------------------------------------

def loadtxtdata(filename,delimiter=""):
  """
  #load data from txt-file with delimiter.
  #分隔符默认是:";",",",空格类 (包括\t)和自定义的delimiter等
  #return value:  二维数值数组(numpy.ndarray)
  #对文本中数据的排列格式要求较高,且不容许出现注释字符,智能化程度较低,但速度较快
  """
  from stringio import stringio
  import re
  
  file_handle=open(filename,'r')
  linesall=file_handle.read()#读入字符串
  file_handle.close()
  
  delimiterall=delimiter+",;"#分隔符
  spacestring=" "#空格
  for rchar in delimiterall:
    linesall=linesall.replace(rchar,spacestring)
    
  return numpy.genfromtxt(stringio(linesall))
  
#----------------------------------------------------------------------  
def savetxtdata(filename, x, fmt='%.8e', delimiter=' ', newline='\n'):
  """
  save data to txt-file.
  """
  numpy.savetxt(filename, x, fmt=fmt, delimiter=delimiter, newline=newline)   
  return true
  
#----------------------------------------------------------------------
def dlmwrite(filename, x, fmt='%.8e', delimiter=' ', newline='\n'):
  """
  save data to txt-file.
  """
  numpy.savetxt(filename, x, fmt=fmt, delimiter=delimiter, newline=newline)   
  return true
  
#----------------------------------------------------------------------
#测试程序 
#----------------------------------------------------------------------
if __name__ == '__main__':
  #生成随机数
  data=numpy.random.randn(3,4)
  filename='d:/x.txt'
  #写入文件
  dlmwrite(filename,data)
  x=getdoublelistfromfile(filename)
  print(x)
  print(dlmread(filename))
  y=stringtodoublearray('79l890joj')
  print(y)
  z=loadtxtdata(filename)
  print(z)

我只在python2.7中试过,如果要在python3.x中使用,可自行测试.