Regex escape special character in AWK if statement

Shell Programming and Scripting
1

I am having issues escaping special characters in my AWK script as follows:

for id in `cat file`
do
grep $id in file2 | awk '\
    BEGIN {var=""} \
    { if ( /stringwith+'|'+'50'chars/ ) {
        echo "do this"
    } else if ( /anotherString/ ) {
        echo "do that"
    } else {
        echo "do other"
    } }'
done

I am not sure how I am supposed to escape the multitude of characters in the regex string "stringwith+'|'+'50'chars"

2

Either use:

awk -vW="stringwith[+]'[|]'[+]'50'chars" 'BEGIN {var=""}
    {
      if ($0 ~ W) {
        print "do this"
    } else if ( /anotherString/ ) {
        print "do that"
    } else {
        print "do other"
    } }'

or

awk 'BEGIN {var=""}
    {
      if ( /stringwith[+]'\''[|]'\''[+]'\''50'\''chars/ ) {
        print "do this"
    } else if ( /anotherString/ ) {
        print "do that"
    } else {
        print "do other"
    } }'
3

How about you start by telling us what is in your input file and what are you trying to get from it. It makes all the difference.

4

This works with modern awk - POSIX character classes

awk  '/stringwith+[[:punct:]]|[[:punct:]]+[[:punct:]]50[[:punct:]]chars/' somefile

mirni is right - what are you trying to do, not how you decided to do it....

5

I have a file that is a parsed together from several other files. Each line has a unique identifier that establishes which file it came from:

data data UID data ...

So, for each of these UID's I want to grab some value (value1). Once I find that value there are 3 potential outcomes I can obtain depending on particular Strings I find which succeed the original value:

data data UID data=data=value1...
...
data data UID data data (xyzstring)...

So, when I find xyzstring as "stringwith+'|'+'50'chars" I want to append certain text to a csv file, similarly with "anotherString", giving me the following shell script:

grep 'string' file2 | awk '{print $3}' > file
echo "this,that" > file.csv
for id in `cat file`
do
grep $id in file2 | awk '\
    BEGIN {var=""} \
    /knownString/ {split($0, tmp, "="); var=tmp[3]} \
    { if ( /stringwith+'|'+'50'chars/ ) {
        echo var",This" >> file.csv
    } else if ( /anotherString/ ) {
        echo var",That" >> file.csv
    } else {
        echo var",Other" >> file.csv
    } }'
done

I am fairly new to scripting and would be open to any suggestions.