Empty value is not actually empty?

Question

I get false checking "AA8187517" string by the regex:

if (/^[a-z0-9]*$/i.test(value))

What I suddenly see in the console?

> value.split('').filter(function(el){ return el != '' })
<  (11) ["", "A", "A", "8", "1", "8", "7", "5", "", "1", "7"]

What are these two values in the array?

I think you are missing CAPS checking in your regex ```console.log(/^[a-zA-Z0-9]*$/i.test('AA8187517'))``` — vaku, Apr 21 '19 at 15:50
What do you get from `value.split('').map(function (x) { return x.charCodeAt(0); })`? — melpomene, Apr 21 '19 at 15:51
Could you post enough of your code that we can see exactly what you're trying to do? The picture sort of clarifies it, but doesn't help people reproduce your problem and is essentially useless for future visitors that might be experiencing similar problems. — David Thomas, Apr 21 '19 at 15:52
How can I remove the Unicode Character 'RIGHT-TO-LEFT MARK' --> 8207 — Igor, Apr 21 '19 at 15:53
Without knowing the context of your problem, the accepted solution may not be a reliable solution. I have provided an answer with food for thought. — Anthony Rutledge, Apr 21 '19 at 17:49

melpomene · Accepted Answer · 2019-04-21T15:57:01.983

3

The two "empty" values in your array contain character 8207 (decimal), which is 200f (hex).

U+200F is RIGHT-TO-LEFT MARK in Unicode, an (invisible) marker that changes the direction text is displayed.

Here's a reproduction of your issue plus sample code to remove the character:

let value = "\u200FAA81875\u200F17";

console.log(value.split('').map(function (x) { return x.charCodeAt(0); }));

value = value.replace(/\u200F/g, '');

console.log(/^[a-z0-9]*$/i.test(value));

edited Apr 21 '19 at 15:57

answered Apr 21 '19 at 15:54

melpomene

84,125
8
85
148

Actually the problem was that I had the Unicode characters which were removed by value.replace(/[^\x00-\x7F]/g, "") – Igor Apr 21 '19 at 15:56
If the goal is filtering, then suggesting replacing as a solution is a half truth. One should be able to filter out unwanted values. – Anthony Rutledge Apr 21 '19 at 18:10

Anthony Rutledge · Answer 2 · 2019-04-21T18:02:33.247

Four thoughts.

1) I would use explicit regular expressions, where practical, instead of modifiers. Be wary of the * quantifier, as it may allow too much, including the absence of a value!!!

if (/^[A-Za-z0-9]*$/.test(value)) {

}

2) Examine closely the definition of Array.protoype.split(), and what happens when you use the empty string as the separator.

Attention: If an empty string ("") is used as the separator, the string is not split between each user-perceived character (grapheme cluster) or between each unicode character (codepoint) but between each UTF-16 codeunit. This destroys surrogate pairs. See also How do you get a string to a character array in JavaScript? on stackoverflow.

3) Could .trim() be of any use to you here?

value.trim().split('').filter(function(el){ return el != '' })

4) Consider changing your filter predicate (the callback method).

value.trim().split('').filter(function(element){ return /^[A-Za-z0-9]{1}$/.test(element) }

However, investigate the significance of splitting on the empty string and know the encoding of your source strings. Since you are filtering, you should not need to replace before filtering. Filtering alone should be sufficient. You want to whitelist wanted values, as blacklisting by replacing is bound to get you in trouble here.

Stackoverflow: How do you get a string to a character array in JavaScript?

score -2 · Answer 3 · answered Apr 21 '19 at 15:51

-2

No you just should save the result in a new variable:

let array = ["text", "text2", '', 'text3'];

let a = array.filter((currentValue) => currentValue);
console.log(a); //["text", "text2", 'text3']

answered Apr 21 '19 at 15:51

Mark Minerov

309
1
8

Empty value is not actually empty?

3 Answers3