Regex formatting

[PabloOQ]

So, I'm a bit lost on the why or why not the regex patterns have different ways of escaping characters depending on which part of the app they are on. So for example / needs to be escaped, but:

• On the URL Cleaner module, it is written like this \\\/, I'm assuming it is escaping both, \ and /, so it results on \\+\/
• On the Pattern checker module and in Automations, it is written like this \/ no need to escape both

But why? I vaguely remember there was some discussion about this, but I can't find it, maybe I'm just imagining it?

[TrianguloY]

I don't remember any discussion, but maybe an old one exists somewhere.

As for the actual question, the answer is: there is only a single way. Java regex doesn't require you to escape the forward slash '/' but some other languages like JavaScript need it, so probably the url cleaner catalog authors did it just in case. In a regex, writing foo/bar is the same as foo\/bar. Then, the json editor requires the regex to be a string inside, and even though you don't need to escape the slashes, and "foo/bar" is perfectly valid, you can do so if you want which means that "foo\/bar" and "foo\\\/bar" are also valid and represent the same.

So basically: all three ways / \/ and \\\/ represent the same / character and can be used interchangeably.

That was the short explanation for those of you who already know some details, now let me try to explain it step by step for those who want a more in depth explanation.

What you write with your keyboard in the json editor is actually a multiline string, let's say you write this:

{
"regex": "1\\\/2\/3/4"
}

This string contains the character '{', then a line break '\n', then '"', then 'r', 'e'...and so on. What you write is what the app gets.

This multiline string is then sent to the JSON decode function, which takes the string and returns a java JSON node with the parsed data. This node is of type object (because of the {}) and contains a single entry with key regex and value a string with the content 1\/2/3/4. First thing to notice is that the value was parsed (as a java string) so any backwards slash + something was replaced with that actual something. \\->\ and \/->/

Now, this string value is sent to the java regex engine to be used as a regex pattern, which means that 1\/2/3/4 is going to be interpreted as such. This pattern is expecting the number '1', then the escaped forward slash '\/' which was unnecessarily escaped, then '2' then '/' and so on.

One important thing to notice is that, even if you write simply "/", after you save, exit, and reopen it will be automatically converted to "\/". This is because what is actually being saved is the serialized JSON node (your input -> JSON parser -> JSON serializer -> stored) and the serialized for some reason prefers to serialize "/" as "\/" (which, as I said, represent the same).

[PabloOQ]

Understood! Thanks!