谁能帮我处理如下的文件？(以文件的某一行排序后重组)谢谢了!!

skycity0210

第一个文件是以含“NUM=”的行来排序，如下
TOOLS {
    NUM=1
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=12
    DRILL_SIZE=17.71653543307087
    SLOT_LENGTH=0
}

TOOLS {
    NUM=2
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=28
    DRILL_SIZE=33.46456692913386
    SLOT_LENGTH=0
}

TOOLS {
    NUM=3
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=22
    DRILL_SIZE=27.55905511811024
    SLOT_LENGTH=0
}

TOOLS {
    NUM=4
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=40
    DRILL_SIZE=45.2755905511811
    SLOT_LENGTH=0
}

TOOLS {
    NUM=5
    SHAPE=HOLE
    TYPE=NON_PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=35
    DRILL_SIZE=41.33858267716536
    SLOT_LENGTH=0
}
第二个文件要以含“DRILL_SIZE=”来排序，并重组“NUM=”的那行，效果如下
TOOLS {
    NUM=1
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=12
    DRILL_SIZE=17.71653543307087
    SLOT_LENGTH=0
}

TOOLS {
    NUM=2
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=22
    DRILL_SIZE=27.55905511811024
    SLOT_LENGTH=0
}

TOOLS {
    NUM=3
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=28
    DRILL_SIZE=33.46456692913386
    SLOT_LENGTH=0
}

TOOLS {
    NUM=4
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=35
    DRILL_SIZE=41.33858267716536
    SLOT_LENGTH=0
}

TOOLS {
    NUM=5
    SHAPE=HOLE
    TYPE=NON_PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=40
    DRILL_SIZE=45.2755905511811
    SLOT_LENGTH=0
}
谢谢老大了!

skycity0210

好像只用awk有点不够用，awk的排序函数还不太会用，有哪位老大有好的思路？

dbcat

1. cat -n file |sed -n '/.\{2\}\..\{10,\}$/p' | sort -t"=" -k2 | cut -f1 | xargs -i awk 'NR>{}- 10&&NR<{}+3{print}' file >tmp

复制代码

1. seq 1 5 |xargs -i echo "NUM={}"|xargs -i  sed 's/NUM\=[1-5]\{1\}$/{}/' tmp | awk '{k=0;while(k<6){if(NR < 60*k
   
2. + 12*k +13  && NR > 60*k + 12*k ){print;}k++;}}'

复制代码

$ seq 1 5 |xargs -i echo "NUM={}"|xargs -i  sed 's/NUM\=[1-5]\{1\}$/{}/' tmp | awk '{k=0;while(k<6){if(NR < 60*k + 12*k +13  && NR > 60*k + 12*k ){print;}k++;}}'
TOOLS {
    NUM=1
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=12
    DRILL_SIZE=17.71653543307087
    SLOT_LENGTH=0
}
TOOLS {
    NUM=2
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=22
    DRILL_SIZE=27.55905511811024
    SLOT_LENGTH=0
}
TOOLS {
    NUM=3
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=28
    DRILL_SIZE=33.46456692913386
    SLOT_LENGTH=0
}
TOOLS {
    NUM=4
    SHAPE=HOLE
    TYPE=NON_PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=35
    DRILL_SIZE=41.33858267716536
    SLOT_LENGTH=0
}
TOOLS {
    NUM=5
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=40
    DRILL_SIZE=45.2755905511811
    SLOT_LENGTH=0
}

[ 本帖最后由 dbcat 于 2005-11-16 22:26 编辑 ]

skycity0210

谢谢老大了，能不能大概解释一下思路，看的有点晕，^_^

dbcat

First,以含“DRILL_SIZE=”来排序:

1. cat -n file |sed -n '/.{2}..{10,}$/p' | sort -t"=" -k2 | cut -f1 | xargs -i awk 'NR>{}- 10&&NR<{}+3{print}' file >tmp

复制代码

$cat tmp
TOOLS {
    NUM=1
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=12
    DRILL_SIZE=17.71653543307087
    SLOT_LENGTH=0
}
TOOLS {
    NUM=3
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=22
    DRILL_SIZE=27.55905511811024
    SLOT_LENGTH=0
}
TOOLS {
    NUM=2
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=28
    DRILL_SIZE=33.46456692913386
    SLOT_LENGTH=0
}
TOOLS {
    NUM=5
    SHAPE=HOLE
    TYPE=NON_PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=35
    DRILL_SIZE=41.33858267716536
    SLOT_LENGTH=0
}
TOOLS {
    NUM=4
    SHAPE=HOLE
    TYPE=PLATED
    TYPE2=STANDARD
    MIN_TOL=3.15
    MAX_TOL=3.15
    BIT=
    FINISH_SIZE=40
    DRILL_SIZE=45.2755905511811
    SLOT_LENGTH=0
}

Second,重组“NUM=”的那行a

1. seq 1 5 |xargs -i echo "NUM={}"|xargs -i  sed 's/NUM=[1-5]{1}$/{}/' tmp | awk '{k=0;while(k<6){if(NR < 60*k + 12*k +13  && NR > 60*k + 12*k ){print;}k++;}}'

复制代码

****ps:Shell Code Too long

skycity0210

seq 是什么命令？我用的c-shell,没见过seq，请老大指教!
代码实用就行^_^

dbcat

csh............
This script is for bash ^_^
$seq 1 5
1
2
3
4
5

In csh try:
??

[ 本帖最后由 dbcat 于 2005-11-16 22:36 编辑 ]

技安

awk ' BEGIN { FS="[\n=]" ; RS=""} { print $3 } ' a | sort -n > c

for i in `cat c ` ; do

awk ' BEGIN { FS="[\n=]" ; RS=""} { a[$3]=$0 } END { print a['$i'] }  ' a

done

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

如果 num 有相同的就不行了、、、，

[[i] 本帖最后由 技安 于 2005-11-16 22:55 编辑 [/i]]

技安

晕、我。。。。。。。。。。。。还以为、第一个文件，也是乱的呢。

技安

awk -F= ' $1 ~ /DRILL_SIZE/ { print $2 } ' a | sort -n > c

for i in ` cat c ` ; do

awk ' BEGIN { FS="[\n=]" ; RS=""} {  a[$19]=$0 } END { print a["'$i'"] } ' a

done

[[i] 本帖最后由 技安 于 2005-11-16 23:09 编辑 [/i]]