A script question

Lemming

Hey guys,

doing a little shel scripting here and having a couple of little dilemmas.

I'm trying todo is search for every instance of a pattern in a file (in this case partial names from email addresses). Each instance is placed into a 1D array. Since the search returns the entire line for each instance
(eg. To: John Doe <john.doe@blah.com>)

I am then trying to search through each entry in the array for patterns with '@' in them, and then stripping out any '<' or '>' characters if they exist.
(I incidentally have NO clue how to go about that last bit)

I'm having problems as sooin as I reach attempting to search each array entry, since grep would appear to only work with files and directories.

here's the code:

function release
{
        uname=$1
        file=$2
        declare -a list[]
        count=0

        for a in `grep "$uname" "$file"`
        do
                list[$count]="$a"
                count=`expr $count + 1`
        done

        echo "Name: $uname"
        echo "File: $file"
       

        for((j=0; $j < $count; j++))
        do
                if [ `grep "@" ${list[$j]}` ]; then
                        echo "${list[$j]}"
                fi
        done

        #blah
        #blah
}

and here's the current output for this function:

Name: james
File: gAEFXj200027062
grep: To:: No such file or directory
grep: Systems: No such file or directory
grep: Solutions: No such file or directory
grep: <john.doe@blah.com>: No such file or directory
grep: cc:: No such file or directory
grep: johnd@blah.com: No such file or directory


Any ideas?

Lemming

Stupid stupid stupid ...

just used a filter fior my first problem.

eg. echo var[n] |grep "@"

anyway, all I need to do now is figure out how to get rid of the likes of the '<' and '>' tags in an array element

eg. <john.doe@blah.com>

Ronin

be lazy and use sed to stript off the offending characters that you don't want.

Ronin

Lemming

Ok,

I'm using awk to search for certain characters at certain points of a string (in this case the start/end)

But I can't for the life of me figure out how the f*ck to use awk (or sed for that matter) to get rid of these characters, as opposed to deleting the entire line.

here's an example I created to try and figure out how awk works:

#!/bin/sh

declare -a list[]
counter=5

list[0]="<john.doe@blah.com>"
list[1]="john.doe@blah2.com"
list[2]="\"john.doe@blah3.com\""
list[3]="<john.doe@blah4.com"
list[4]="john.doe@blah5.com>"

for((i=0; $i < $counter; i++))
do
        echo ${list[$i]} |awk '/^[<"]|[>"]$/'

done

What this does is lists all examples which have any of the listed characters in them. What I'd like to do is show ALL, but strip the listed characters from those that have them at the start and/or end .

The awk/sed mans just aren't sinking in at the moment

davej

Use sed:

sed 's/targetText/replacementText/'

eg

echo \<\"john@microsoft.com\"\> | sed 's/["<>]//g'

gives

john@microsoft.com

davej