Regex doesn't match. Online generator does

Question

I've want to check with a regex this kind of string:

2020_2021_01_01

I've putted it in a variable, say $session

so i do:

if  [[  "$session" =~ \d{4}[_]\d{4}[_]\d{2}[_]\d{2} ]]; then
    stuff
fi

you see...it doesn't work... but I don't know why....

any help? THANKS!

Answer 1

The bash manual rather tersely explains that when the =~ operator "is used, the string to the right of the operator is considered an extended regular expression and matched accordingly (as in regex(3))".

Here, regex(3) is a reference to man 3 regex, which might explain what an "extended regular expression" is. A longer description would be "Posix standard extended regular expressions", and you can find the documentation for those in the Posix document. If you're using an online regular expression tester, make sure you select "Posix regular expressions".

In short, they don't include Perlisms like \d. You can write [[:digit:]] or (if you are using the C locale) [0-9].

So your regex could have been written:

([[:digit:]]{4}_){2}[[:digit:]]{2}_[[:digit:]]{2}

(there is no need to quote _). However, be aware that the =~ operator looks for a substring which matches the pattern, rather than testing whether the left-hand operator precisely matches the pattern. So you quite possibly actually wanted an anchored match:

^([[:digit:]]{4}_){2}[[:digit:]]{2}_[[:digit:]]{2}$