Skip to main content

15 posts tagged with "Ruby"

View All Tags

· 2 min read

We have added a new CodeGen setting that allows you to specify a list of words that Code Generation will not split when converting identifiers from API specification to language-specific identifiers. This is useful for declaring brand names such as APIMatic.

E.g. if you provide the words apimatic and vmware in your the list, APIMaticandvmWare becomes ApimaticAndVmware or apimatic_and_vmware, depending on the place of use.

· One min read

There was a bug in the implementation of the Unindexed and Plain array serialization formats in Ruby SDKs. Arrays of models were not being properly encoded as form fields for the two mentioned formats.

Additionally, a bug in the initialization of inherited models caused duplication in fields with names longer than one word.

Both these bugs have now been fixed. Please regenerate your Ruby SDKs to get this fix.

· One min read

There was a bug in Ruby SDKs which were using OAuth2.

  • When updating the OAuth token, the compiler was trying to set the expiry time by adding an int and a string.

Besides this, there were a couple of edge cases that allowed Ruby SDKs to be generated, and even the gem to be built and installed, but the code would not execute because of the following reasons:

  • require statements in the base file were not being split into multiple lines correctly. This splitting is done to ensure compliance with The Ruby Style Guide
  • New line characters in the descriptions for models/enums/exceptions were not being dealt with, causing a comment to spill over into the next line.

These bugs have now been fixed. Please regenerate your Ruby SDKs to get the fixes.