How $regex works in mongoose with empty value

user_6178324 0
Viewed 470

I'm using $regex in mongoose, a however a bit confused with its behavior.

Here's my data, I have one document which has one attribute "owner" as below.

{ "owner" : "lorem ipsum" }

now when I do search with {owner: {$regex:"",$options:"i"}} in MongoDB compass it returns that object, and when I do {owner: ""}, it returns nothing.

So can somebody explain to me why this behavior is like this, I do not have an empty string as an owner though while doing $regex it returns that document.

Even I tried to remove space between lorem and "epsum" as "loremepsum" but it had the same behavior. Shouldn't it return same as {owner: ""} ?

mongoose mongodb javascript regex
2 Answers

I'm not a Mongoose expert by any means but it sounds like it's behaving correctly.

Basically, regexing for "" results in a compiled regex that looks like // which loosely translates to "find me nothing". This "nothing" exists anywhere that something does not exist so "nothing" exists between chars such as l and o of lorem.

This can be best visualized using this code snippet:

echo preg_replace( '//', '8', 'lorem ipsum' );

which results in:

8l8o8r8e8m8 8i8p8s8u8m8

So Mongoose would be doing something like:

var_dump( preg_match( '//', 'lorem ipsum' ) );

which results in successfully finding "nothing"

In contrast {owner: ""} would be looking for an absence of owner data.

user_21915720 
{owner: {$regex: ""}}

this means find me nothing nothing, well there is nothing in your string which is in between letters

while this

{owner: ""}

means find me owner that's equal to this ""

i've tested your query in Robo 3T and i got the same results so i think it has to do with the regex version that mongo uses which is (PCRE-8.42)

user_137417870
Related