python实现类似awk的简单功能

命令行的awk很方便，但处理灵活一点的话对awk不熟，深入学习又没太大必要，用python做个简单的，复杂的话也用python脚本实现，程序的一致性更好。

#!/usr/bin/python
#coding:utf-8

import os
import sys
from getopt import getopt

#处理协程
def deallines(dealer,instream,paras):
    sep = paras.get("sep","\t")
    for line in instream:
        fields = line.strip("\r\n").split(sep)
        for result in dealer(line,fields) or ():
            if type(result) == str:
                yield result
            elif type(result) == int:
                yield str(result)+"\n"
            else:
                yield sep.join([str(f) for f in result])+"\n"

#定义处理函数
dealerid = 0
def definedealer(dealstr,paras):
    global dealerid
    dealerid += 1
    funcname = "dealer_"+str(dealerid)
    exec "def "+funcname+"(line,f): "+dealstr
    return vars()[funcname]

#打印结果
def printer(instream):
    for line in instream:
        print line,

#处理字符转义
def convertchar(s):
    if s == "\\t":
        return "\t"
    elif s == "\\r":
        return "\r"
    elif s == "\\n":
        return "\n"
    elif s == "\\\\":
        return "\\"
    elif s == "\\'":
        return "\'"
    elif s == '\\"':
        return '\"'
    elif s == "\\a":
        return "\a"
    elif s == "\\b":
        return "\b"
    elif s == "\\e":
        return "\e"
    elif s == "\\v":
        return "\v"
    elif s == "\\f":
        return "\f"
    elif len(s) == 4 and s.startswith("\\x"):
        return chr(eval('0x'+s[2:4]))
    else:
        return s

#转换参数
def convertparas(paras):
    result = {"sep":"\t"}
    for key,value in paras:
        if key == '-F':
            result["sep"] = convertchar(value)
    return result
<pre name="code" class="python"><pre name="code" class="python">

FNR = 0def record(instream): global FNR for r in instream: FNR += 1 yield r

#主函数def main(args): paras,deals = getopt(args[1:],'F:') paras = convertparas(paras) dealers = [definedealer(dealstr,paras) for dealstr in deals] stream = record(sys.stdin) for dealer in dealers: stream = deallines(dealer,stream,paras) printer(stream)if __name__
 =='__main__': main(sys.argv)


#输入数据，姓名 年龄 身高 体重


cat person.txt

zs      10      1.3     35.0
ls      12      1.4     36.0
#print直接输出

cat person.txt | dealline 'print line,'
zs      10      1.3     35.0
ls      12      1.4     36.0
#分两步，第一步提取姓名 身高 体重，第二步计算体重指数=体重/身高/身高
cat person.txt | dealline 'yield f[0],f[2],f[3]' 'yield f[0],float(f[2])/float(f[1])/float(f[1])'

zs      20.7100591716

ls      18.3673469388
注：-F分隔符缺省为\t，line为整行，f为按分隔符split后的数组，FNR为记录号