If I have a file with for example the line:
hai hello test hey hi
I then want to search for the lines with test in it, and return the two surrounding strings hello and hey.
I can get the lines in which the search word is found using:
grep "test" $file"
But I can't figure out how to only get the two surrounding strings and do further things with the strings
With grep -o you can print only that part of a line that matches the given regex. Now you need a regex that matches test and its two surrounding words. To come up with something like this you have to know what a word is supposed to be, that is, what makes it different from any other string. Here we go with: "A word is a sequence of letters A-Za-z surrounded by anything that is not a letter"
grep -Eo '[A-Za-z]+[^A-Za-z]+test[^A-Za-z]+[A-Za-z]+' file
Note that this ignores all test that are not surrounded be words, for instance the lines test, test b, and a test. If you give a more detailed explanation of what you need I can adjust the regex.
oat the end ofhello, the space betweenhelloandtest, or even the empty string are strings too,