mjhjmtu

The command sed $'s/[^[:print:]t]//g' is printing each line of my .txt as-is.

What is it used for? I could not find it on google.

Eg:

 sed $'s/[^[:print:]t]//g' *.txt | wc -l
 15909

 cat *.txt | wc -l
 15909

There are no edits done in the files whatsoever. What does this command do

The sed command removes any character that is not printable, and not tabs, from the content of each line of input. The number of lines would not change (the newline characters are not removed as they are not part of the content of the line), but the number of characters or bytes might. Test again with wc -c for the number of bytes, wc -m for the number of characters.

The regular expression [^...] matches any single character (actually collating element) not within the [...]. In this case, it's the character class [:print:] and the tab character. The $ at the start of the string makes bash replace t with a literal tab character before calling sed.

The character class [:print:] matches characters that are printable in the current locale, i.e. alphanumeric characters, punctuation characters and space (but not tab which is a control character).

In other words, it deletes everything that is not a alphanumeric character, punctuation character, space or tab.

To write the result back to the file (an "in-place edit"), some sed implementations have a -i option for that; use sed -i or sed -i '' depending on the implementation. Be sure that the correct output is produced first though, or you may destroy your data.