[awk] 设定匹配规则文件？

iamkey9

有多个文件，含有如下语句，行有可能乱序

filenamexxx.txt
...
The clock shows "12:00:00"
The month is "Jan" and the day is "31"
The-location-is-Beijing-(China)                             
...

我想定义规则文件给 awk 运行时候读入
#keywords     IFS   fieldNo
#===================
clock               " "     $4
month            " "      $4
day                " "      $9
location         "-"      $4

想要达到的目的是，匹配 keywords, 用IFS分割此行，然后依次打引fieldNo所代表的字段

求读入匹配文件的函数

yinyuemi

回复 1# iamkey9


    这样行不？

1. $ echo 'The clock shows "12:00:00"
   
2. The month is "Jan" and the day is "31"
   
3. The-location-is-Beijing-(China) ' |\
   
4. awk -F'[" -]' '/clock|month/{print $2": "$5}/location/{print "day: "$4}/day/{print $2": "$(NF-1)}'
   
5. clock: 12:00:00
   
6. month: Jan
   
7. month: 31
   
8. day: Beijing

复制代码

cu_little_bird

大神在楼上

mpstat

回复 2# yinyuemi


    不知道他这个文件会不规则到什么程度,说不定下一条记录里local又用引号day用-了....
   我觉得能生成这么不规则的文件也不容易啊

yinyuemi

回复 4# mpstat


    从lz的描述来看，我理解的是，只是行乱序，对于特定的行，格式应该是固定的

iamkey9

回复 5# yinyuemi


谢谢，其实我是想把AWK和规则文件分开，这样以后看不懂AWK的人也可以改规则文件，另外，行格式是基本固定的（其实不固定也没关系，添加新规则就可以了）

yinyuemi

回复 6# iamkey9

1. head format.txt data
   
2. ==> format.txt <==
   
3. #keywords     IFS   fieldNo
   
4. #===================
   
5. clock               " "     4
   
6. month            " "      4
   
7. day                " "      9
   
8. location         "-"      4
   
9. 
   
10. ==> data <==
    
11. The clock shows "12:00:00"
    
12. The month is "Jan" and the day is "31"
    
13. The-location-is-Beijing-(China)
    
14. 
    
15. awk  'NR>2&&NR==FNR{a[gensub(/ /,"","g",$1)]=$2;b[gensub(/ /,"","g",$1)]=gensub(/ /,"","g",$3)}
    
16. NR>FNR{if(/location/){t=a["location"];split($0,c,t);print c[2],": "c[b[c[2]]]};
    
17. if(/month/){t=a["month"];l=split($0,c,t);print c[2],": "c[b[c[2]]] RS "day: "c[b["day"]]};
    
18. if(/clock/){t=a["clock"];split($0,c,t);print c[2],": "c[b[c[2]]]};delete c}
    
19. ' FS='"' format.txt  data
    
20. clock : "12:00:00"
    
21. month : "Jan"
    
22. day: "31"
    
23. location : Beijing
    

复制代码

jason680

有多个文件，含有如下语句，行有可能乱序

filenamexxx.txt
...
The clock shows "12:00:00"
The mont ...
iamkey9 发表于 2011-07-28 11:41

$ cat awk_rule
# keywords FieldNo IFS
#====================
clock 4 " "
month 4 " "
day   9 " "
location 4 "-"
test 2 "[=,]"
name 4 "[=,]"

$ cat f1.txt
...
The clock shows "12:00:00"
The month is "Jan" and the day is "31"
The-location-is-Beijing-(China)
test=123,name=John
...

$ awk 'FNR==NR{if(!/^[ \t]*#/){if(NF<3){print "error:"$0;exit}kw[n]=$1;no[n]=$2;s[n]=$3;if(NF>3){for(a=4;a<=NF;a++)s[n]=s[n]" "$a;}}s[n]=substr(s[n],2,length(s[n])-2);n++}NR>FNR{for(n in kw){if(match($0,kw[n])){split($0,m,s[n]);print kw[n]"="m[no[n]]}}}' awk_rule f1.txt
clock="12:00:00"
day="31"
month="Jan"
location=Beijing
test=123
name=John