'have.text' and 'have.html' ignore Special characters

[noudadrichem]

tl;dr;
Is there any other way to get the 'raw' text/html of a node beside have.html ?

I'm stuck on an assertion that is in a scenario that I have to implement. We're adding hyphenation to text where the HTML will end up looking like this:

The problem now is that when using an assertion like this:

cy.get('summary').should('contain', 'this text will hy­­phen­­ate');

Will fail because of this:

It removes the ­. So I investigated a bit further. But this also happens with & and €.
I'd expect have.html to get the 'raw' values of a node. Any flags on this function that I'm missing?

I'm kinda stuck here.. Thanks in advance!