how can i output the number 2641569270623 from a text file called checkpoint.txt that is a one line file (generated by a maths sieving program) that looks like:
pmin=2641569270623,factor_count=8,cpu_secs=1705.793,frac_done=0.002592,elapsed_secs=1710.844
i tried sed -n "/pmin=/,/,factor/ p" checkpoint.txt | sed "$ d" which ouputs nothing and sed -n "/pmin=/,/,factor/ p" checkpoint.txt which output the whole line
Try this:
sed '/^pmin=/s/pmin=\([^,]*\),.*/\1/' file
Regards
Ranges in sed are ranges of entire lines -- /pmin=/,/,factor/ means print each line starting with a line containing "pmin=" through a different line containing ",factor" (and starting over again if a subsequent line matches "pmin=" again).
You could do
sed -e 's/.*pmin=//;s/,factor.*//' checkpoint.txt
which will replace from start of line through "pmin=" with empty line, and ditto from ",factor" through to end of line. The result is still printed regardless of whether the resulting line is empty or not.
both arguments work PERFECTLY. thanks alot guys!
is it possible to use this output to replace text from another file in one command?
ie: replace the "720" in line 1 of a different file with text comtaining "Sieved to 720" with the 2641569270623 we extracted from the original file?
sed -e 's/.*pmin=/s%Sieved to [0-9]*%Sieved to /;s/,factor.*/%/' checkpoint.txt |
sed -f - otherfile
If your sed won't do "-f -" you will have to save to a temporary file.
The repetition of "Sieved to " can be avoided, but is left as an exercise.
You can also include the whole sed command within the sed command and pipe it to sh:
sed '/^pmin=/s/pmin=\([^,]*\),.*/sed '"'"'s%Sieved to [0-9]*%Sieved to \1%'"'"' other_file/' file | sh
until you are more familiar with regular expressions, you can use field based operations
awk -F"," '{ split($1,a,"="); print a[2] }' file
awk -F'[=,]' '{print $2}' file
sed 's#^[^=][^=]*=\([^,][^,]*\).*#\1#' file