Search for element after a designated string

Question

I have a script that reads from a file and sorts out certain lines by a keyword. I need to be able to pick out a certain element from each of these discovered lines. The lines in the file are set up lie;

$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...

I am trying to figure out how I would read the lines with the keyword and then write only the element_needed to a file.

Example:

the line would be

$(eval $(call CreateUvmTest, keyword, run_test_file, $(run_test_file.exten) tc_run_test_file.sv

and I am hoping to write this to another file called listfile containing only run_test_file

It is very important that it takes lines only containing keyword in it cause there could be identical lines like this;

$(eval $(call CreateUvmTest, NOT_keyword, run_test_file, $(run_test_file.exten) tc_run_test_file.sv

Answer 1

Considering your other question, I guess you have to match the lines only if it contains $(eval $(call CreateTest, KEYWORD, I guess you may expect us to literally match the above and then extract the further element needed.

input.txt

$(eval $(call CreateUvmTest, NOT_keyword, run_test_file1, $(run_test_file.exten) tc_run_test_file.sv
$(eval $(call CreateUvmTest, keyword, run_test_file2, $(run_test_file.exten) tc_run_test_file.sv
$(eval $(call CreateUvmTest, NOT_keyword, run_test_file4, $(run_test_file.exten) tc_run_test_file.sv
$(eval $(call CreateUvmTest, NOT_keyword, run_test_file9, $(run_test_file.exten) tc_run_test_file.sv
$(eval $(call CreateUvmTest, keyword, run_test_file10, $(run_test_file.exten) tc_run_test_file.sv

extractElement.tcl

set fp [open input.txt r]
set data [split [read $fp] \n]
close $fp
set outfile [open listfile.txt w] 
foreach line $data {
    if {[regexp {\$\(eval \$\(call CreateUvmTest, keyword, ([^,]+)} $line match element]} {
        puts $element; # Printing in console
        puts $outfile $element; # This goes to the file
    }
}
close $outfile

Output :

run_test_file2
run_test_file10

Regex used :

{\$\(eval \$\(call CreateUvmTest, keyword, ([^,]+)}

Till keyword, it is matched literally, then [^,]+ is matching the element need to be extracted.

Note : This is written by the assumption that the data will be separated with comma in each line.

Answer 2

Given your input the best idea I have:

Given this input (with a line not matching the keyword):

> cat inputtcl.txt
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, iNOT_keyword, element_needed, morestuff...
$(eval $(call CreateUvmTest, keyword, element_needed, morestuff...

Using this script:

> cat ./extract.tcl
#!/usr/bin/tclsh
set infile [open "$path1" r]
set outfile [open "$path0/listfile.txt" w]
set input [read $infile]
regsub -all {(?w)^.*?,\s*?sanity,\s*(.*?),.*$} $input {\1} output
puts $outfile $output

I get this output:

> cat output.txt
element_needed
element_needed
element_needed
element_needed

The regex match anything before the first , then check any number of spaces and sanity followed by a , skip any number of spaces presents then capture anything until the next , then match anything until the end of line. the (?w) makes the ^ and $ match respectively start and end of line.