problem with arrays

[jupiler]

Hello,

While working on the previous script, I encountered a new problem. Given a file ("termlist.txt&quot and an inputfile ("textfile.txt" --> one sentence = one line), I need to write a script that states:

If a pattern in sentence X of "textfile.txt" matches a term in file "termlist.txt", print that sentence X to new file ("newfile.txt&quot. The problem however is the fact that a term can consist of more than one word. Hopefully, the examples will show you what I want to do.

Given "textfile.txt":

John went to the local pub
He saw a man sitting in the corner
The man gave him a chair
John drank a pint
He paid his drink, the value added tax included, and got out

Given "termlist.txt":

local pub
drink
value added tax

Gawk should only select:

John went to the LOCAL PUB
He paid his DRINK, the VALUE ADDED TAX included, and got out

The patterns need to be marked in the text as well. Can someone help me with this? Thanks,

Jupiler

 

[vgersh99]

something like that:

nawk -f term.awk textfile.txt

#------------- term.awk
BEGIN {
termlist="termlist.txt"

while (getline < termlist > 0) {
# skip comments in the config file
if ( $0 ~ /^[#].*/ ) continue;
termsARR[$0];
}
close(termlist);


}

{
found=0;
for(ind in termsARR)
if (match($0, ind)) {
found++;
gsub(ind, toupper(ind), $0);
}
if (found) print;
}

vlad
+---------------------------+
|#include<disclaimer.h> |
+---------------------------+

 

[marsd]

function strsearch(str,mat,cnt,alist, i) {
if (cnt >= length(str)) {
if (alist) {
return alist
} else {
return
}
}

if (substr(str,cnt,length(mat)) != mat) {
#print &quot;Not matched: &quot;, substr(str,cnt,length(mat)), &quot;at&quot;, cnt
i = cnt + 1
return strsearch(str,mat,i,alist)
} else if (substr(str,cnt,length(mat)) == mat) {
#print &quot;Matched: &quot;, substr(str,cnt,length(mat)) ,&quot;at&quot;, cnt
i = cnt + 1
alist = length(alist) < 2 ? substr(str,cnt,length(mat)) : alist&quot; &quot;substr(str,cnt,length(mat))
return strsearch(str,mat,i,alist)
}
}



BEGIN {
while ((getline < termlist) > 0) {
arr[a++] = $0
}
close(termlist)
}
{
for (x in arr) {
#print &quot;searching for : &quot;, arr[x]
all = strsearch($0,arr[x],1,&quot;0&quot
if (all) {
allx = toupper(all1)
sub(all,allx,$0)
}
print $0 > filename3
}
}

This may work in some pathological cases where match()
fails.

Good Luck.