regex positive negative decimal numbers, comma separated

Question

I am having trouble getting the regex right for the validation. Need multiple decimals, positive, negative numbers, and comma separated to be allowed and NOTHING else.

Should allow:

41.975481,-87.728222, 41.974141,-87.721569, 41.973247,-87.7135

Shouldn't allow:

41.975481,-87.728222, 41.97TEST4141,-87.721569, 41.973247,-87.7135, test

This is what I am using but its allowing for characters.

(\-?\d+(\.\d+))(,\s*(\-?\d+(\.\d+)))*

Am I missing something?

$('textarea[name="latlong"]').bind('focusout',function(){
        $(".polygon .geoerror").text("");
        var polygon = $(this).val();
        //remove trailing comma and extra space if its there
        polygon = polygon.replace(/,\s*$/, "");
        var checkpoly = /(\-?\d+(\.\d+))(,\s*(\-?\d+(\.\d+)))*/;
        var evenodd = polygon.split(',').length;

        if (checkpoly.test(polygon) && evenodd % 2 == 0) {
            $(".polygon .geoerror").text("Polygon string is valid." + polygon);
            $(".polygon .geoerror").css("color", "green");
        } else if (checkpoly.test(polygon) && evenodd % 2) {
            $(".polygon .geoerror").text("Mismatch of lat and long values." + polygon);
            $(".polygon .geoerror").css("color", "red");
        } else {
            $(".polygon .geoerror").text("Not valid." + polygon);
            $(".polygon .geoerror").css("color", "red");
        }
        return false;
    });

Based on the code if the user inputs - 41.975481,-87.728222, 41.974141,-87.721569, 41.973247,-87.7135, test - then it tests as valid. It needs to error instead.

Answer 1

This should do it:

/^(?:\-?\d+\.\d+[, ]*)+$/

https://regex101.com/r/nP9lE8/1

Answer 2

$('textarea[name="latlong"]').bind('focusout',function(){
    $(".polygon .geoerror").text("");
    var polygon = $(this).val();


    //remove extra space if its there
    polygon = polygon.replace(/\s*/g, "");

    var nums = polygon.split(',');
    var len = nums.length;
    var is_even = ( len % 2 == 0 );

    if (!is_even) {
        $(".polygon .geoerror").text("Mismatch of lat and long values." + polygon);
        $(".polygon .geoerror").css("color", "red");

        return false;
    }

    var validNums = true;

    for(var i = 0; i < len; i++) {
        if(isNaN(nums[i])) {
            validNums = false;
            break;
        }
    }

    if (validNums) {
        $(".polygon .geoerror").text("Polygon string is valid." + polygon);
        $(".polygon .geoerror").css("color", "green");
    } else {
        $(".polygon .geoerror").text("Not valid." + polygon);
        $(".polygon .geoerror").css("color", "red");
    }

    return false;
});

Answer 3

The core of your regex is absolutely okay. The point(s) you are missing were given in the answer of caeth , but without explanation.

Basically, he added ^ at the beginning and $ at the end, forcing the regex to match the whole word, not just parts of it. According to the description on RegExr (Reference -> Anchors) ^

[m]atches the beginning of the string, or the beginning of a line if the multiline flag (m) is enabled. This matches a position, not a character.

Similiarly $ (same source)

[m]atches the end of the string, or the end of a line if the multiline flag (m) is enabled. This matches a position, not a character.

So a corrected (and a little bit cleaner) version of your regex would be something like

/^(-?\d+\.\d+)(\s*,\s*-?\d+\.\d+)+$/