Extracting a value from a string using sed

Question

I have the following code:

set MARKER_LIST = "t14_noTap_leftSlow t24_noTap_leftMedium t34_noTap_leftFast \
 t44_noTap_rightSlow t54_noTap_rightMedium t64_noTap_rightFast \
 t74_noTap_inphaseSlow t84_noTap_inphaseMedium t94_noTap_inphaseFast \
 t104_noTap_antiphaseSlow t114_noTap_antiphaseMedium t124_noTap_antiphaseFast"

foreach code (${MARKER_LIST})

set evt_code = `echo ${code} | sed 's/^.*t[0-9]*_noTap_//'`
echo "Extracted value: $evt_code"
end

I'm trying to extract the string after "tXX_noTap_" and "tXXX_noTap_" but it doesn't seem to be working. Would appreciate any help with this. Thanks!

Answer 1

Your problem is that you do not declare tcsh as the shell to interpret the script. So sh-aware code is expected. Start your script with the correct shebang in the first line:

#!/usr/bin/tcsh

---

Just beyond your question:

1. Your sed code itself is fine, but perhaps unnecessary complicated:

   sed 's/^.*_//' 
   

   should be enough, except you expect output with _ in the desired part of the string.

2. Please eliminate all the \ with their line breaks in ${MARKER_LIST}. Else the result will contain empty values at each appearance of \:

   Extracted value: leftSlow
   Extracted value: leftMedium
   Extracted value: leftFast
   Extracted value: 
   Extracted value: rightSlow
   Extracted value: rightMedium
   Extracted value: rightFast
   Extracted value: 
   Extracted value: inphaseSlow
   Extracted value: inphaseMedium
   Extracted value: inphaseFast
   Extracted value: 
   Extracted value: antiphaseSlow
   Extracted value: antiphaseMedium
   Extracted value: antiphaseFast
   

So this would be your script:

#!/usr/bin/tcsh

set MARKER_LIST = "t14_noTap_leftSlow t24_noTap_leftMedium t34_noTap_leftFast t44_noTap_rightSlow   t54_noTap_rightMedium t64_noTap_rightFast t74_noTap_inphaseSlow t84_noTap_inphaseMedium t94_noTap_inphaseFast t104_noTap_antiphaseSlow t114_noTap_antiphaseMedium t124_noTap_antiphaseFast"

foreach code (${MARKER_LIST})
    set evt_code = `echo ${code} | sed 's/^.*t[0-9]*_noTap_//'`
    echo "Extracted value: $evt_code"
end

Answer 2

I guess you were almost there

use the "g" option at the end of the sed statement to replace all occurences. And you don't want to prefix the regex with ^.* if you want to replace every occurence starting with tXX_noTap_

echo "t14_noTap_leftSlow t24_noTap_leftMedium t34_noTap_leftFast \
       t44_noTap_rightSlow t54_noTap_rightMedium t64_noTap_rightFast \
       t74_noTap_inphaseSlow t84_noTap_inphaseMedium t94_noTap_inphaseFast \
       t104_noTap_antiphaseSlow t114_noTap_antiphaseMedium t124_noTap_antiphaseFast" \
    | sed 's/t[0-9]*_noTap_//g'