Shell Programming and Scripting

Hi,
I am working on a ksh script and I´m stuck on the following:
I have to get the pthread_id from a procstack file for a particular tid#.
---------- tid# 1274057 (pthread ID: 1800) ----------
---------- tid# 1736913 (pthread ID: 4019) ----------
---------- tid# 1478705 (pthread ID: 4269) ----------
---------- tid# 1876087 (pthread ID: 4465) ----------
---------- tid# 1912849 (pthread ID: 3086) ----------
---------- tid# 425991 (pthread ID: 2828) ----------
---------- tid# 1954017 (pthread ID: 1543) ----------
---------- tid# 2064419 (pthread ID: 1286) ----------
---------- tid# 1491095 (pthread ID: 1029) ----------
---------- tid# 2265101 (pthread ID: 772) ----------
---------- tid# 418027 (pthread ID: 515) ----------
---------- tid# 1867819 (pthread ID: 258) ----------
---------- tid# 2281679 (pthread ID: 1) ----------
---------- tid# 1851497 (pthread ID: 2571) ----------
---------- tid# 2199627 (pthread ID: 2314) ----------
---------- tid# 450699 (pthread ID: 2057) ----------

I can grep the output file by "tid#" or by the tid# number itself, but what I would like to obtain is the pthread ID. Lets say, I have tid# 1876087 to grep by.

Variable pthread identifies currently the whole line that contains my $TID:

echo pthread=`cat "${OM}_procstack_${PID}.txt" | grep $TID`
---------- tid# 1876087 (pthread ID: 4465) ----------

Can somebody help me to cut the 4465 from this line? I would like to put it into a variable so I can use it.

---------- tid# 1876087 (pthread ID: 4465) ----------

Both numbers on the line (tid# and pthread ID) can vary in length, and I can´t just awk and call it $6 because I do not want the ')' after it. Also, I think you awk files, not lines nor variables, right?

I do not want *everything from ':' to ')' * as there are a varying number of spaces, so I´m not sure how to go about this. I am sure one of you is going to make this sound so easy-peasy in a minute.. /

Thanks in advance!
T.

This is yet another Useless Use of Cat.

Creative use of the field separator can help get you what you want without resorting to external commands like cut, sed, and grep:

Code:

$ IFS=" ()" read A B C D E F G H < data
$ echo "A='$A'" "B='$B'" "C='$C'" "D='$D'" "E='$E'" "F='$F'" "G='$G'" "H='$H'"
A='----------' B='tid#' C='1274057' D='pthread' E='ID:' F='1800' G='----------' H=''
$

Code:

awk -F'[ )]' -v tid="$TID" '$3==tid{print $6}' "${OM}_procstack_${PID}.txt"

Code:

grep "tid# 1876087" yourfile | sed 's/.*ID: //g' | sed -e 's/\(^[0-9].*\)).*/\1/g'

Hi, jostber:

Using plain grep to filter sed input isn't necessary. Also, you can execute multiple commands with a single sed invocation. A translation of your pipeline:

Code:

sed '/tid# 1876087/!d; s/.*ID: //g; s/\(^[0-9].*\)).*/\1/g' yourfile

Regards,
Alister

sample is the your datafile

Code:

cat sample | cut -f2 -d':' |cut -f1 -d")"

Yet another Useless Use Of Cat for the collection A version without the blatant waste:

Code:

cut -f2 -d':' < sample | cut -f1 -d")"