Bash – Issues with regex match using the =~ operator of the bash [[ command

bashdateosxregular expression

On OSX, I am building a function to validate date formats and then convert them to epoch times. The function should validate that the date is in one of the following formats, if not error: 01/01/1970 10:00PM or 10:00PM (%m/%d/%Y %I:%M%p or %I:%M%p)

FUNCTION

checkTIME () {
    local CONVERT_CHK_TIME="$1"
    if [[ "$CONVERT_CHK_TIME" =~ ^(0[0-9]|1[0-2]):[0-9][0-9](AM|PM)$ ]]; then
        CONVERT_TIME="$(date -j -f "%I:%M%p" "$CONVERT_CHK_TIME" "+%s")"
    elif [[ "$CONVERT_CHK_TIME" =~ (0[0-9]|1[0-2])\/([0-2][0-9]|3[0-1])\/\d{4}\s[0-9][0-9]:[0-9][0-9](AM|PM) ]]; then
        CONVERT_TIME="$(date -j -f "%m/%d/%Y %I:%M%p" "$CONVERT_CHK_TIME" "+%s")"
    else
        echo "ERROR!"
        exit 1
    fi
}

It currently works fine for 10:00PM but is failing to match when I try 01/10/2017 10:00PM

I'm calling it as follows:

./convert '01/10/2017 10:00PM'
...
...
+ [[ -n 01/10/2017 10:00PM ]]
+ checkTIME '01/10/2017 10:00PM'
+ local 'CONVERT_CHK_TIME=01/10/2017 10:00PM'
+ [[ 01/10/2017 10:00PM =~ ^(0[0-9]|1[0-2]):[0-9][0-9](AM|PM)$ ]]
+ [[ 01/10/2017 10:00PM =~ (0[0-9]|1[0-2])/([0-2][0-9]|3[0-1])/d{4}s[0-9][0-9]:[0-9][0-9](AM|PM) ]]
+ echo 'ERROR!'
ERROR!
+ exit 1

Thanks!

I've also tried the following regex:

(0[0-9]|1[0-2])\/([0-2][0-9]|3[0-1])\/\d{4}\ [0-9][0-9]:[0-9][0-9](AM|PM)

Best Answer

\d matches a decimal digit in some versions of regex (perl), but does not in the Extended Regular Expressions used for the =~ operator of the [[ command in bash.

Therefore, change the \d to [0-9] for a pattern that will match 4 decimal digits.

Similarly for \s. To match one literal space character, replace the \s with an escaped space (\). If you want to match 1 or more blanks (spaces or tabs) then replace the \s with [[:blank:]]+.

More importantly, to avoid these regex mix-ups:

man bash says that =~ regular expressions match according to the extended regular expression syntax, as documented in regex(3).
man 3 regex (POSIX regex functions) says SEE ALSO regex(7).
man 7 regex gives a description of the regular expression syntax, and says SEE ALSO POSIX.2, section 2.8 (Regular Expression Notation).

You can find the complete POSIX Extended Regular Expressions syntax described in The Open Group's Posix Regular Expressions documentation.

Related Solutions

Bash – Lazy regex using Bash

There is a reason why the internet is packed with alternative approaches. I can't really think of any situation where you would be forced to use bash for this. Why not use one of the tools designed for the job?

Anyway, as far as I know there is no way of doing non-greedy matches using the =~ operator. That's because it does not use bash's internal regex engine but your system's C one as defined in man 3 regex. This is explained in man bash:

   An additional binary operator, =~, is available, with the  same  prece‐
   dence  as  ==  and !=.  When it is used, the string to the right of the
   operator is considered  an  extended  regular  expression  and  matched
   accordingly  (as  in  regex(3)).

You can, however, do more or less what you want (bearing in mind that this is really not a good way of parsing HTML files) with a slightly different regex:

string='<span class="circle"> </span>foo</span></span>'
regex='<span class="circle"> </span>([^<]+)</span>'
[[ $string =~ $regex ]]; 
echo "${BASH_REMATCH[1]}"

The above will return foo as expected.

Bash Regex – How to Match Both Upper and Lower Case Letters

[OK] Will match either character within the brackets, the brackets do not tell it to be case insensitive.

You could do:

[[ "$status" =~ ^[Oo][Kk]$ ]]

or I would probably do the following:

[[ "${status,,}" == ok ]]

The ,, operator for parameter expansion will convert the entire variable to lowercase for the purpose of the comparison.

Best Answer

More importantly, to avoid these regex mix-ups:

Related Solutions

Bash – Lazy regex using Bash

Bash Regex – How to Match Both Upper and Lower Case Letters

Related Question