[develop 2.0] nickname generator fails with Path Parts

[fehguy]

When an operationId is not specified, the codegen logic makes invalid method names when path params are present

[maxlinc]

Quick question, because it relates to OAI/OpenAPI-Specification#29. I want to try and match the swagger-codegen behavior in other projects if possible. I tried to test it out but I'm having trouble figuring out how to run samples on the develop_2.0 branch.

It looks like you're generating method names based on the path + http method. So GET /foo/bar would become something like fooBar_get. What happens if the path is just GET /? Would the method name be _get?

[fehguy]

looks like we need special handling of this.

#279

[earth2marsh]

Could this have regressed, possibly? Importing a description with path params but no operationid is causing complaints about unsupported characters.

[fehguy]

@earth2marsh can you please share a sample?

[earth2marsh]

Yessir! https://gist.githubusercontent.com/earth2marsh/db8ab2ffcb4895e78d14/raw/3a4ebdd242fa4b4c9e0b0fcff5dd44ae8842b78f/apm.json

[webron]

@earth2marsh - that's because you have a typo there. You use operationid instead of operationId. Change that and you should be good.

[earth2marsh]

Sorry, it looks like I linked to the wrong revision of that spec where I had later added the operationid to test it. :facepalm

@webron try this: https://gist.githubusercontent.com/earth2marsh/db8ab2ffcb4895e78d14/raw/0241924f47061a5f6668874651926779a781df9f/apm.json

If you bring that into swagger-codegen, then you should see the error as it tries to use the brace char to create the operationId.

[webron]

@earth2marsh - do you get errors for a specific language? Just tried running it with the java client generator and didn't get any errors. However, I'm not sure that's the sample you wanted to share because in your original post you mentioned the path params seem to be causing the issue and the sample has no path parameters.

[earth2marsh]

@webron I'm not sure what happened here—I must have had some wires crossed somewhere. You're totally right that my sample doesn't actually cover the issue I described, and I can't seem to reproduce my original issue. My apologies for the wild goose chase! I'm going to go back to the start again and see if I can make this happen predictably.

[webron]

@earth2marsh - no worries, it's better to check on issues that can't be reproduced thanto not be aware of ones that can :)