Search Results

The sed utility isn't very good for arithmetic. Try something like:
awk '
BEGIN { FS = OFS = "|" }
NR==1 { print }
$1=="HD" { next }
$1=="TR" { $NF = count; t = $0; next }
{ print; count++...

Adapting your approach, how about echo "HD|$(date +%m%d%Y)|$(date +%m%d%Y)" > outfile
sed '1d;$d' file1 >> outfile
sed '1d;$d' file2 >> outfile
echo "TR|$(date +%m%d%Y)|$(date +%m%d%Y)|$(($(wc...

Fixed:

awk > _new_ '/decalre|begin|exception/, /;/ {
sub(/^\//, "\\/")
}1' infile &&
mv -- _new_ infile

Indeed you did. I only went into detail about it in case scottn or anyone else was interested in the most likely reason why it did not work with his sed.

That said, unless GNU sed behavior...

awk '/decalre|begin/, /;/ {
sub(/\//, "\\/")
}1' infile

Or maybe:

awk '/decalre|begin|exception/, /;/ {
sub(/\//, "\\/")
}1' infile

Yes, it's the command I showed you, instead with single quotes, and two backslashes. More accurately put, it's radoulov's original solution with an extra backslash (and with $!s instead of 1s, since...

That's because the replacement text in radoulov's solution is using an implementation-defined sequence. Since the slash is no longer the regular expression delimiter, preceding it with a backslash...

To "exclude the last line alone", you mean change all lines except the last one?

Actually, radoulov's sed didn't work for me, strangely.


sed '$!s|^/|\\\/|' file

Assuming GNU sed (given your example usage):

sed -i '1s|^/|\/|' infile

pipe the command into xargs as follows:

ARCHIVE=somethingsomething.zip
sed ... | xargs zip a $ARCHIVE

Then use a MIME-compatible mail user agent (MUA) to send the zip file.

pine...

YOURPATH=/tmp/whatever
cat $(sed "s/:[^.]*/ /g;s:[.]/:$YOURPATH/:g" summary.txt) | mailx -s "your subject" user@domain.com

Something like
cat $(sed 's/:[^.]*/ /g' summary.txt) | mailx -s "your subject" user@domain.com