Fix non striped coffe blender

[ngouy]

Example: due to how coffee blend names are built, and the content of coffee.name_2, some blend names are not stripped, this creates some flaky test on feature test because HTML is removing trailing spaces, so when you look for a match where the reference contains a tailing space, this leads to a faulty test

Here you can see that when name_2 is randomly the empty string, then #{name_1} #{name_2} leads to a non stripped string

No-Story

[vbrazo]

I wonder why we need those empty spaces in the .yml file?

If there are no real reasons, I'd prefer to remove the spaces instead of doing a .strip.

[ngouy]

I think @vbrazo the idea is the following: if you add "", you multiply chances of a unique coffee blender name by 2 (ish, u make single named coffee blender a thing)

I'm completely ok with what you said, it's just it will roughly divide number of unique possibilities by 2. If you are comfortable with that, I am

edit: This comment is so wrong x)
it's not n + n it's n * n so if we remove one word, impact is really low

[vbrazo]

@ngouy Thanks!