3. Aren't =, != etc. used to compare strings and -eq, -ne, -gt etc. used to compare numbers? I see lot of numbers compared as strings, e.g.:
[ $? = 0 ]
[ $? != 0 ]
[ $exitcode = 0 ]
4. There are lot of "cat <<EOF" blocks without indentation. I understand that this is made because the shell expects "EOF" on the line start, but there is a special syntax designed on purpose for this use case, simply put a dash between << and the token, e.g. "cat <<-EOF".
In this case:
tomb_init() {
system="`uname -s`"
case "$system" in
FreeBSD)
cat <<-EOF
create=posix_create
format=posix_format
map=posix_map
mount=freebsd_mount
close=freebsd_close
EOF
;;
Linux)
thanks for your review! tho you are referring to the tomb-portable unfinished experiment which is about to be dismissed since cross-platform experiments with veracrypt show very bad performance.
you are welcome to share a review of the tomb script, but be warned in that we use a lot of zsh specific features. It is a script that works since 15+ years so it has a discrete amount of patchwork to avoid regressions.
1. What "exitcode" is set for:
2. I see a lot of "return $?". Why "$?" is returned if by default the shell returns the return value of the last command? Just ti name a few: 3. Aren't =, != etc. used to compare strings and -eq, -ne, -gt etc. used to compare numbers? I see lot of numbers compared as strings, e.g.: 4. There are lot of "cat <<EOF" blocks without indentation. I understand that this is made because the shell expects "EOF" on the line start, but there is a special syntax designed on purpose for this use case, simply put a dash between << and the token, e.g. "cat <<-EOF". In this case: 5. Aren't backtick deprecated in favor of $()?