In order to make a school project easier, I want to append a delimiter to each line beginning with a capital letter with a bash script. Could anyone help me out? I've tried reading the file line-by-line in tandem with the case command but it doesn't work I don't really understand what I'm doing. This is what i have so far: (I'm new to scripting, don't flame me)
#!/bin/bash
input=/home/user/file
while IFS= read -r var
do
case [A-Z]
sed 's/^/#/' file
done
You don't use shell loops to process text. Processing text is done by text processing tools which are the ones that process their input one line at a time:
#! /bin/sh -
input=/home/user/file
sed 's/^[[:upper:]]/#&/' < "$input"
Would insert a # before an uppercase letter found at the start (^) of each line of $input.
Here, uppercase letter is as determined by the locale. It will include the ABCDEFGHIJKLMNOPQRSTUVWXYZ English letters, and possibly things like Á, Ź, or Π if present.
If you wanted to restrict to ABCDEFGHIJKLMNOPQRSTUVWXYZ only, you'd write:
sed 's/^[ABCDEFGHIJKLMNOPQRSTUVWXYZ]/#&/' < "$input"
[A-Z] itself may match on about anything depending on the locale. You'll find systems where it matches on abcdefghijklmnopqrstuvwxy (not z) as well. On most systems, it would match on Á, but not Ź if present in the locale's charmap. It could even match on sequences of characters (multi-character collating elements like the Hungarian Ddzs).
sed open the file itself, for a number of reasons. See for instance Why is `sort < "$f1" ` preferred over `sort -- "$f1"`, and why is this preferred over `sort "$f1"`?
– Stéphane Chazelas
Jun 30 '20 at 07:22
sed "./$input" and then sed retains the ability to process several files. (If input is not a full path, which in your script it is. That is, it starts with a / and so it can not start with a -).
–
Jun 30 '20 at 07:25
sed '/^[A-Z]/s/^/#/' "$input". – Jun 30 '20 at 07:16