Closed xaka closed 8 years ago
I think that is rather obsolete as the spec clearly says that the value of a protocol node is case-insensitive. So the example above could be valid.
But that, of course, depends on if you use the same example in your case and the raml2json returns you these two different outputs?
@sichvoge It also depends on whether the JSON we receive is considered normalized. Considering that all the metadata is being added to populate things, it's already being normalized - so I'd expect that little things like this could be handled too.
Agree! @KonstantinSviridov we should return a consistent representation.
Guys, I do not quite understand, what you've finally decided. Do we convert protocols values to upper case or not?
I don't care either way, it really comes down to @xaka and if (or how much) we want to improve the representation output for the new parser development. I've noticed this with some other issues opened on diffs between 0.8 and 1.0 parser output and I feel we could improve the output slightly over both while understanding the changes.
However, for consistency, let's use 0.8 output here.
However, for consistency, let's use 0.8 output here.
LGTM :+1:
Ok. In the intermediate parser it's already implemented in 0.8 output style.
RAML 0.8 (uppercase):
RAML 1.0: