I just wondered what exactly the difference between
[[ $STRING != foo ]]
[ $STRING != foo ]
is, apart from that the latter is posix-compliant, found in sh and the former is an extension found in bash.
There are several differences. In my opinion, a few of the most important are:
[is a builtin in Bash and many other modern shells. The builtin
[is similar to
testwith the additional requirement of a closing
]. The builtins
testimitate the functionality
/bin/testalong with their limitations so that scripts would be backwards compatible. The original executables still exist mostly for POSIX compliance and backwards compatibility. Running the command
type [in Bash indicates that
[is interpreted as a builtin by default. (Note:
which [only looks for executables on the PATH and is equivalent to
type -p [)
[[is not as compatible, it won't necessarily work with whatever
/bin/shpoints to. So
[[is the more modern Bash / Zsh / Ksh option.
[[is built into the shell and does not have legacy requirements, you don't need to worry about word splitting based on the IFS variable to mess up on variables that evaluate to a string with spaces. Therefore, you don't really need to put the variable in double quotes.
For the most part, the rest is just some nicer syntax. To see more differences, I recommend this link to an FAQ answer: What is the difference between test, [ and [[ ?. In fact, if you are serious about bash scripting, I recommend reading the entire wiki, including the FAQ, Pitfalls, and Guide. The test section from the guide section explains these differences as well, and why the author(s) think
[[ is a better choice if you don't need to worry about being as portable. The main reasons are:
< >with backslashes in order for them not to get evaluated as input redirection, which can really mess some stuff up by overwriting files. This again goes back to
[[being a builtin. If [ (test) is an external program the shell would have to make an exception in the way it evaluates
/bin/testis being called, which wouldn't really make sense.
[ is a bash Builtin
[[ ]] are bash Keywords
Keywords: Keywords are quite like builtins, but the main difference is that special parsing rules apply to them. For example, [ is a bash builtin, while [[ is a bash keyword. They are both used for testing stuff, but since [[ is a keyword rather than a builtin, it benefits from a few special parsing rules which make it a lot easier:
$ [ a < b ] -bash: b: No such file or directory $ [[ a < b ]]
The first example returns an error because bash tries to redirect the file b to the command [ a ]. The second example actually does what you expect it to. The character < no longer has its special meaning of File Redirection operator.
Tested in Bash 4.3.11:
POSIX vs Bash extension:
[[is a Bash extension
regular command vs magic
[ is just a regular command with a weird name.
] is just an argument of
[ that prevents further arguments from being used.
Ubuntu 16.04 actually has an executable for it at
/usr/bin/[ provided by coreutils, but the bash built-in version takes precedence.
Nothing is altered in the way that Bash parses the command.
< is redirection,
|| concatenate multiple commands,
( ) generates subshells unless escaped by
\, and word expansion happens as usual.
[[ X ]] is a single construct that makes
X be parsed magically.
() are treated specially, and word splitting rules are different.
There are also further differences like
[ is a built-in command, and
[[ is a keyword: https://askubuntu.com/questions/445749/whats-the-difference-between-shell-builtin-and-shell-keyword
[[ a < b ]]: lexicographical comparison
[ a \< b ]: Same as above.
\required or else does redirection like for any other command. Bash extension.
[[ a = a && b = b ]]: true, logical and
[ a = a && b = b ]: syntax error,
&&parsed as an AND command separator
cmd1 && cmd2
[ a = a -a b = b ]: equivalent, but deprecated by POSIX
[ a = a ] && [ b = b ]: POSIX recommendation
[[ (a = a || a = b) && a = b ]]: false
[ ( a = a ) ]: syntax error,
()is interpreted as a subshell
[ \( a = a -o a = b \) -a a = b ]: equivalent, but
()is deprecated by POSIX
([ a = a ] || [ a = b ]) && [ a = b ]POSIX recommendation
x='a b'; [[ $x = 'a b' ]]: true, quotes not needed
x='a b'; [ $x = 'a b' ]: syntax error, expands to
[ a b = 'a b' ]
x='a b'; [ "$x" = 'a b' ]: equivalent
[[ ab = a? ]]: true, because it does pattern matching (
* ? [are magic). Does not glob expand to files in current directory.
[ ab = a? ]:
a?glob expands. So may be true or false depending on the files in the current directory.
[ ab = a\? ]: false, not glob expansion
==are the same in both
==is a Bash extension.
printf 'ab' | grep -Eq 'a.': POSIX ERE equivalent
[[ ab =~ 'ab?' ]]: false, loses magic with
[[ ab? =~ 'ab?' ]]: true
[[ ab =~ ab? ]]: true, POSIX extended regular expression match,
?does not glob expand
[ a =~ a ]: syntax error
printf 'ab' | grep -Eq 'ab?': POSIX equivalent
I prefer to always use
There are POSIX equivalents for every
[[ ]] construct I've seen.
If you use
[[ ]] you:
[is just a regular command with a weird name, no special semantics are involved.
Single Bracket i.e.
 is POSIX shell compliant to to enclose a conditional expression.
Double Brackets i.e.
[] is an enhanced (or extension) version of standard POSIX version, this is supported by bash and other shells(zsh,ksh).
In bash, for numeric comparison we use
gt, with double brackets for comparison we can use
[is a synonym for test command. Even if it is built in to the shell it creates a new process.
[[is a new improved version of it, which is a keyword, not a program.
[ var1 lt var2] #works [ var1 < var2] #error: var2 No such file or directory [ var1 \< var2] #works with escape [[ var1 < var2]] #works