shell进行文件换行和替换处理的问题(有实际案例)

szqimao

小弟现在银行工作，有一个文件需要在AIX下进行格式处理，麻烦大虾帮忙看看有没有办法用shell命令解决。
该文件正常情况下是以:LST:开头，换行符\x0A结束，但是有些情况下，正常的一行中可能会存在换行符就变成两行或者三行，那怎么才能把该文件整理成正常的以:LST:开头的呢？

我想的办法一是判断每行开头如果不是:LST:，则将此行与上一行连在一起，也就是去掉上一行的换行符，不知道这命令该怎么写！
办法二是先去掉所有行的换行符，把各行连成一行，再从二个:LST:开始替换成换行符+:LST:，但是怎么从第二个:LST:开始替换? 而且在AIX下用sed "s/:LST:/\r:LST:/g"，sed "s/:LST:/\n:LST:/g"替换换行,回车都不行  ，还请大家出下高招。

:LST:00000006000000000010000005840301584000016  301584000016  陈荣场
:LST:00000084000000000025000005840301584000016  30158
4000016  姚军凯
:LST:00000149000000000070000005840301584000016  301584000016  郑猛发
:LST:00000260000000000006000005840301584000016  301584000016  刘韦君
:LST:000002810
0000000001000000584030158400
0016  301584000016  卜静文
:LST:00000290000000000015000005840301584000016  301584000016  苏锦  
:LST:00000296000000000024000005840301584000016  301584000016  蔡炳沂

yestreenstars

1. awk '{printf $0}' urfile | sed 's/:LST:/\n&/g'

复制代码

第一行会多出一个空行，你手动把它删掉就可以了，最后一行不会换行，你手动换一下行就可以了。

dn833

1. awk 'BEGIN{RS=":LST:"}NF{$1=$1;print RS$0}'

复制代码

blackold

回复 1# szqimao


sed:

1. sed '/^:LST:/{:a;N;/\n:LST:/!{s/\n//;ba};P;D}' urfile

复制代码

awk:

1. awk -v RS=':LST:' '$0{gsub("\n","",$0);print RS $0}' urfile

复制代码

seesea2517

表示比较关心在银行工作这件事。

L_kernel

楼主泄露了用户隐私。哈哈~~

1. awk 'BEGIN{RS="";FS=":LST:"}{for(i=2;i<=NF;i++)print FS,gensub(/\n/,"","G",$i)}' 1.txt
   
2. :LST: 00000006000000000010000005840301584000016  301584000016  陈荣场
   
3. :LST: 00000084000000000025000005840301584000016  301584000016  姚军凯
   
4. :LST: 00000149000000000070000005840301584000016  301584000016  郑猛发
   
5. :LST: 00000260000000000006000005840301584000016  301584000016  刘韦君
   
6. :LST: 00000281000000000010000005840301584000016  301584000016  卜静文
   
7. :LST: 00000290000000000015000005840301584000016  301584000016  苏锦  
   
8. :LST: 00000296000000000024000005840301584000016  301584000016  蔡炳沂
   

复制代码

lovegod1206

(1)awk '{printf "%s",$0}' file>file1  #一行
(2)awk -F':LST:' -v OFS=':LST:' '{for(i=2;i<NF;i++) printf (":LST:%s\n",$i)}' file1 #重新按:LST:分隔符换行

szqimao

这几天忙，今天刚一打开论坛贴子就收到这么多热心回复，真心感谢各位的热情回复。
经过我在系统中逐一测试，lovegod1206 的回答的语句执行后，输出的结果完全满足要求，
另外，blackold 的回答的第二条语句也已经基本接近最终处理结果，可能是blackold觉得没必要再留下最简单的最后一个步骤了吧。
不管怎样感谢各位了，现将所有回复命令的执行结果公布如下：

awk '{printf $0}' urfile | sed 's/:LST:/\n&/g'

n:LST:00000006000000000010000005840301584000016  301584000016  陈荣场n:LST:00000084000000000025000005840301584000016  301584000016  姚军凯n:LST:00000149000000000070000005840301584000016  301584000016  郑猛发n:LST:00000260000000000006000005840301584000016  301584000016  刘韦君n:LST:00000281000000000010000005840301584000016  301584000016  卜静文n:LST:00000290000000000015000005840301584000016  301584000016  苏锦  n:LST:00000296000000000024000005840301584000016  301584000016  蔡炳沂

-------------------------------------------------------------------------------
awk 'BEGIN{RS=":LST:"}NF{$1=$1;print RS$0}'

:LST:LST
:LST:00000006000000000010000005840301584000016  301584000016  陈荣场

:LST:LST
:LST:00000084000000000025000005840301584000016  30158
4000016  姚军凯

:LST:LST
:LST:00000149000000000070000005840301584000016  301584000016  郑猛发

:LST:LST
:LST:00000260000000000006000005840301584000016  301584000016  刘韦君

:LST:LST
:LST:000002810
0000000001000000584030158400
0016  301584000016  卜静文

:LST:LST
:LST:00000290000000000015000005840301584000016  301584000016  苏锦  

:LST:LST
:LST:00000296000000000024000005840301584000016  301584000016  蔡炳沂

-------------------------------------------------------------------------------
sed '/^:LST:/{:a;N;/\n:LST:/!{s/\n//;ba};D}' urfile

sed: 0602-417 The label /^:LST:/{:a;N;/\n:LST:/!{s/\n//;ba};D} is greater than eight characters.[SZ_A_TEST] icsadm:/app/ics
/ly>


awk -v RS=':LST:' '$0{gsub("\n","",$0);print RS $0}' urfile
:LST:LST
:LST:00000006000000000010000005840301584000016  301584000016  陈荣场
:LST:LST
:LST:00000084000000000025000005840301584000016  301584000016  姚军凯
:LST:LST
:LST:00000149000000000070000005840301584000016  301584000016  郑猛发
:LST:LST
:LST:00000260000000000006000005840301584000016  301584000016  刘韦君
:LST:LST
:LST:00000281000000000010000005840301584000016  301584000016  卜静文
:LST:LST
:LST:00000290000000000015000005840301584000016  301584000016  苏锦  
:LST:LST
:LST:00000296000000000024000005840301584000016  301584000016  蔡炳沂

-------------------------------------------------------------------------------
awk 'BEGIN{RS="";FS=":LST:"}{for(i=2;i<=NF;i++)print FS,gensub(/\n/,"","G",$i)}'

:LST: awk: 0602-553 Function gensub is not defined.
The input line number is 1. The file is W1.
The source line number is 1.

-------------------------------------------------------------------------------

(1)awk '{printf "%s",$0}' file>file1  #一行

:LST:00000006000000000010000005840301584000016  301584000016  陈荣场:LST:00000084000000000025000005840301584000016  301584000
016  姚军凯:LST:00000149000000000070000005840301584000016  301584000016  郑猛发:LST:00000260000000000006000005840301584000016
  301584000016  刘韦君:LST:00000281000000000010000005840301584000016  301584000016  卜静文:LST:000002900000000000150000058403
01584000016  301584000016  苏锦  :LST:00000296000000000024000005840301584000016  301584000016  蔡炳沂


(2)awk -F':LST:' -v OFS=':LST:' '{for(i=2;i<NF;i++) printf (":LST:%s\n",$i)}' file1 #重新按:LST:分隔符换行

:LST:00000006000000000010000005840301584000016  301584000016  陈荣场
:LST:00000084000000000025000005840301584000016  301584000016  姚军凯
:LST:00000149000000000070000005840301584000016  301584000016  郑猛发
:LST:00000260000000000006000005840301584000016  301584000016  刘韦君
:LST:00000281000000000010000005840301584000016  301584000016  卜静文
:LST:00000290000000000015000005840301584000016  301584000016  苏锦

szqimao

lovegod1206 和blackold 能大概说一下语句的意思吗？对地我这样的初学者，真的不是太明白。

szqimao

seesea2517 发表于 2012-10-18 09:52
表示比较关心在银行工作这件事。

    不知道你关心啥？