Script help: divide one file to several and named the new file with pattern in original file

[patrick xi]

Dear Friends, I need your help.
I have a long one line text file. To simplize it I made a sample fileA, following is its content,
ABC+12345678P1P90+DEF+12345678P2P90+def+12345678P3P90+JKL+12345678P4P90+MNO+12345678P5P90+.......
I want to seperate it to several files using such rule,
1. when match texts in the set {ABC, DEF,def,JKL,MNO}, (case sensitive), create a new file with the match charactors and its following.
2. name the file with 13th-15th charactor(C13,C14,C15) counted from the matching point.
.

the output should like following files,
fileA_P1P:
ABC+12345678P1P90+
fileA_P2P:
DEF+12345678P2P90+
fileA_P3P:
def+12345678P3P90+
fileA_P4P:
JKL+12345678P4P90+
fileA_P5P:
MNO+12345678P5P90+

Hope I've written my mind. Many thanks for any help.

Patrick

[Steve Steel]

Hi

OLDIFS=$IFS
export IFS=+
xx=$(cat $1)
set $(echo $xx)
export IFS=$OLDIFS
set $(echo $1)
let params=${#@}
while [ "$params" -ge "2" ]
do
name=$(echo $1"+"$2"+"|cut -f2 -d"P")
echo $1"+"$2"+" > "fileA_P"$name"P"
shift 2
let params=$params-2
done


Steve Steel

[john korterman · Accepted solution]

Hi,
try the attached script, using your input file as $1

regards,
John K.

[Rodney Hills]

Pat,

Try this one line perl script...

perl -ne '$x=~s/((ABC|DEF|def|JKL|MNO))/\t$1/g;@a=split("\\t",$x);do { $b=substr($_,12,3); system("echo $_ >fileA_$b") if $b; } for (@a);' theonelinefile

This assumes their are no "tab" characters in the string.

HTH

-- Rod Hills

[Rodney Hills]

Minor change-

perl -ne 's/((ABC|DEF|def|JKL|MNO))/\t$1/g;@a=split("\\t",$_);do { $b=substr($_,12,3); system("echo $_ >fileA_$b") if $b; } for (@a);' theonelinefile

-- Rod Hills

[patrick xi]

Thank you all for the help.

Hi, John, your script is the solution. The only small bug is a file named fileA_ was created besides fileA_P1P to fileA_P5P.

The content of fileA_ is :
MNO++

Best regards,
Patrick

[patrick xi]

please close