My primary goal is to avoid frequent consultation of documentation.
In my work environment, there is a scenario like this: An Android development engineer may mainly work on developing the OS, but occasionally they need to use the Gerrit REST API to write some scripts. They are not frequent users, but they will use it a few times per quarter.
Since they cannot remember the API, they have to repeatedly experiment and consult the documentation every time they write a script, and they always need to write an example first to test if the API they are calling is correct.
The purpose of creating this open-source project is to help engineers easily write scripts based on simple impressions. For example, by thinking "The detailed information of 'change' should be 'detail', right? There might be a field called 'branch' in it?" they can quickly write the script.
So the main design is:
1. support calling method autocomplete (done)
2. support input entity fields autocomplete (ongoing)
3. support output entity fields (ongoing)
At least in my situation, the library you mentioned cannot solve the problem. User still needs to remember what the method is, and find the fields in their returned json