Revise coverage guidelines
Blind 100% unit coverage hinders developement and has no added benefits, especially when coverage can be achieved by other means. Change-Id: I216286fd55ea6a6a71a01f56c918bf04298ef5f6
This commit is contained in:
parent
645fd1807d
commit
e1bda4ea85
@ -10,9 +10,12 @@ When reviewing neutron-lib changes, please be aware:
|
|||||||
- Is all of the code shared? Don't move neutron-only code.
|
- Is all of the code shared? Don't move neutron-only code.
|
||||||
- Is the interface good, or does it need to be refactored?
|
- Is the interface good, or does it need to be refactored?
|
||||||
- Does it need new tests, specifically around the interface? We want
|
- Does it need new tests, specifically around the interface? We want
|
||||||
100% unit coverage on this library, so if neutron does not yet have
|
a global unit coverage greater than 90%, and a per-module coverage
|
||||||
a test, it needs to be added. Note that tests on things like constants
|
greater than 80%. If neutron does not yet have a test, it needs to
|
||||||
are uninteresting, but any code or interface should have a unit test.
|
be added. Note that tests on things like constants are uninteresting,
|
||||||
|
but any code or interface should have a unit test, if you cannot
|
||||||
|
tell for sure that it is not going to be traversed in some alternative
|
||||||
|
way (e.g. tempest/functional coverage).
|
||||||
- Is there a corresponding Depends-On review in neutron removing
|
- Is there a corresponding Depends-On review in neutron removing
|
||||||
this code, and adding backwards compatibility shims for Mitaka?
|
this code, and adding backwards compatibility shims for Mitaka?
|
||||||
- Do the public APIs have their parameters and return values documented
|
- Do the public APIs have their parameters and return values documented
|
||||||
|
Loading…
Reference in New Issue
Block a user