I saw in bash regex match string that I should compare regexes with =~.
Tried the following:
if [[ "____[9 / 101] Linking" =~ "[0-9]*" ]]; then echo "YES"; fi
And nothing is printed...
Tried without the quotes:
if [[ "____[9 / 101] Linking" =~ [0-9]* ]]; then echo "YES"; fi
And it works fine. But what to do if my regex contains white spaces (quotes required)?
Put your regex in a variable. You are free to use quotes when defining the variable:
$ re="[0-9]*" ; [[ "____[9 / 101] Linking" =~ $re ]] && echo "YES"
YES
$ re="9 /" ; [[ "____[9 / 101] Linking" =~ $re ]] && echo "YES"
YES
Since the reference to $re inside [[...]] is unquoted, the value of $re is treated as a regex. Anything on the right-side of =~ that is quoted, however, will be treated as a literal string.
In regular expressions, as opposed to globs, * means zero or more of the preceding. Thus [0-9]* is considered a match even if zero characters are matching:
$ re="[0-9]*" ; [[ "____[a / bcd] Linking" =~ $re ]] && echo "YES"
YES
$ re="[0-9]" ; [[ "____[a / bcd] Linking" =~ $re ]] && echo "YES"
$
If you want to match one or more digits, use [0-9]+.
Precede the whitespace with a \:
if [[ "____[9 / 101] Linking" =~ [0-9]*\ /\ [0-9]* ]]; then echo "YES"; fi
[0-9]*matches zero or more occurrences so it will always be true. Use[0-9][0-9]*to match one or more digits.=~in quotation marks,=~works like==.