将数值结果从单个文本文件隔离到Linux中的多个文件

Question

I have a data like this

#start
#gatherData
*ELEMENT_SHELL
48709       1   50614   50616   50618   50613
48710       1   50613   50618   50608   50609
48711       1   50616   50617   50619   50618
48712       1   50618   50619   50607   50608
48715       1   50589   50590   50620   50615
48716       1   50615   50620   50616   50614
48717       1   50590   50591   50621   50620
48721       1   50623   50625   50626   50622
48722       1   50622   50626   50610   50611
48723       1   50625   50614   50613   50626
*END
$PresentData
$RESULT OF strength
48709  1.0267261e-002
48710  1.0721873e-002
48711  1.1930415e-002
48712  1.2186395e-002
48715  9.7443219e-003
48716  1.0036242e-002
48717  1.1186538e-002
48721  7.9333931e-003
48722  8.6850608e-003
48723  8.9872172e-003

What I want to do is to check first of all the results under $RESULT OF strength

which numbers in the second column lie between 0 and 1e-002, then based on that search the number between *ELEMENT_SHELL AND *END and send the complete line to new text file test1.txt. If the number is between 1e-002 to 1e-003 to the next text file test2.txt and segregate this single file into two different files. Text1.text would have

48709       1   50614   50616   50618   50613
48710       1   50613   50618   50608   50609
48711       1   50616   50617   50619   50618
48712       1   50618   50619   50607   50608
48716       1   50615   50620   50616   50614
48717       1   50590   50591   50621   50620

Text2.txt would have

48721       1   50623   50625   50626   50622
48722       1   50622   50626   50610   50611
48723       1   50625   50614   50613   50626
48715       1   50589   50590   50620   50615

Can any expert suggest the way with SED, or AWk? I think final results could be piped easily but the segregation from the same file and find it again is problematic. Thanks in advance

Answer 1

As a basic solution, consider the following code:

[hamadhassan $] cat tri.awk
#!/usr/bin/gawk -f 

BEGIN{
    load_state=1; 
}


$0=="$RESULT OF strength"{
#    print "end of load state"
    load_state=0;
}

load_state==1 && NF==6{
#    print "storing "$0
    lut[$1]=$0; # store line in look up table:
}

load_state==0 && NF==2{
    if($2>0.0 && $2<1e-2){
    if($1 in lut){
        print lut[$1] > "Text2.txt";
    }
    }else{
    if($1 in lut){
        print lut[$1] > "Text1.txt";
    }
    }

}
[hamadhassan $]

which given your sample input:

[hamadhassan $] cat test.in
#start
#gatherData
*ELEMENT_SHELL
48709       1   50614   50616   50618   50613
48710       1   50613   50618   50608   50609
48711       1   50616   50617   50619   50618
48712       1   50618   50619   50607   50608
48715       1   50589   50590   50620   50615
48716       1   50615   50620   50616   50614
48717       1   50590   50591   50621   50620
48721       1   50623   50625   50626   50622
48722       1   50622   50626   50610   50611
48723       1   50625   50614   50613   50626
*END
$PresentData
$RESULT OF strength
48709  1.0267261e-002
48710  1.0721873e-002
48711  1.1930415e-002
48712  1.2186395e-002
48715  9.7443219e-003
48716  1.0036242e-002
48717  1.1186538e-002
48721  7.9333931e-003
48722  8.6850608e-003
48723  8.9872172e-003[hamadhassan $]

gives:

[hamadhassan $] ./tri.awk test.in
[hamadhassan $] cat Text2.txt
48715       1   50589   50590   50620   50615
48721       1   50623   50625   50626   50622
48722       1   50622   50626   50610   50611
48723       1   50625   50614   50613   50626
[hamadhassan $] cat Text1.txt
48709       1   50614   50616   50618   50613
48710       1   50613   50618   50608   50609
48711       1   50616   50617   50619   50618
48712       1   50618   50619   50607   50608
48716       1   50615   50620   50616   50614
48717       1   50590   50591   50621   50620
[hamadhassan $]

This was on CenTOS 6 with awk 3.1.7.

Answer 2

You can try with the following commands (assuming that the source file is txt.txt :

grep "$RESULT OF strength" -A1000 txt.txt | awk '$2>0.01' | cut -f 1 | xargs -I{} grep {} txt.txt | egrep "[0-9]+[[:blank:]]+1[[:blank:]]+" > test1.txt


grep "$RESULT OF strength" -A1000 txt.txt | awk '$2<0.01' | cut -f 1 | xargs -I{} grep {} txt.txt | egrep "[0-9]+[[:blank:]]+1[[:blank:]]+" > test2.txt

If the columns are separated by spaces, then it would be:

grep "$RESULT OF strength" -A1000 txt.txt | sed 's/[\s]{2,}/\t/g' | awk '$2>0.01' | cut -f 1 -d' ' | xargs -I{} grep {} txt.txt | egrep "[0-9]+[[:blank:]]+1[[:blank:]]+" > test1.txt

grep "$RESULT OF strength" -A1000 txt.txt | sed 's/[\s]{2,}/\t/g' | awk '$2<0.01' | cut -f 1 -d' ' | xargs -I{} grep {} txt.txt | egrep "[0-9]+[[:blank:]]+1[[:blank:]]+" > test2.txt