feat(resolver): support generic constraint using CEL #2506
Conversation
openshift-ci
bot
added
the
do-not-merge/work-in-progress
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: dinhxuanvu The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
openshift-ci
bot
added
the
approved
dinhxuanvu
force-pushed
the
generic-con
branch
4 times, most recently
from
efce4b9 to
f379574
Compare
dinhxuanvu
changed the title
openshift-ci
bot
removed
the
do-not-merge/work-in-progress
dinhxuanvu
added
kind/feature
dinhxuanvu
force-pushed
the
generic-con
branch
from
8f90b14 to
07fd34e
Compare
openshift-ci
bot
added
the
approved
|
Due to the automatic approved label added by the bot, I place a hold on here to avoid accident merging. |
openshift-ci
bot
added
the
do-not-merge/hold
dinhxuanvu
force-pushed
the
generic-con
branch
6 times, most recently
from
b798482 to
574293b
Compare
| evaluator constraints.Evaluator | ||
| rule string | ||
| message string |
There was a problem hiding this comment.
Maybe I'm being naive, but I'm a little surprised to see constraint-specific logic on the OLM side of this. I was imagining that we'd define a generic Constraint interface in o-f/api that Cel, All, Any, None, etc. would all implement.
That way, all OLM would need to do would be unmarshal the constraints and then use that generic interface to evaluate them. What am I missing?
There was a problem hiding this comment.
This is predicate information. The rule and message are there to surface those information to the resolver during failure with .String() function. It is not the same as Constraint struct in api repo.
dinhxuanvu
force-pushed
the
generic-con
branch
from
9c5730a to
1c0510a
Compare
openshift-ci
bot
added
the
needs-rebase
dinhxuanvu
force-pushed
the
generic-con
branch
from
1c0510a to
63f39d2
Compare
openshift-ci
bot
removed
the
needs-rebase
dinhxuanvu
force-pushed
the
generic-con
branch
from
63f39d2 to
a311f52
Compare
|
It would be nice to refactor some of this into a more generic setup like in #2418 but I can always rebase my PR if it goes in after this one, or do a follow-up if this goes in last. /lgtm |
Introduce the new type of constraint that uses CEL (Common Expression Language) as an expression language to define the constraint in a generic way. This type of constraint will allow operator authors to specify a dependency on any arbitrary properties in the bundle. This constraint also supports logical operator such as AND and OR in the CEL expression. Signed-off-by: Vu Dinh <[email protected]>
Add unit test cases for generic constraint using CEL to ensure it works properly in the resolver. Add some context for the custom library for semver comparison in CEL. Signed-off-by: Vu Dinh <[email protected]>
Signed-off-by: Vu Dinh <[email protected]>
Signed-off-by: Vu Dinh <[email protected]>
dinhxuanvu
force-pushed
the
generic-con
branch
from
a311f52 to
2b5313b
Compare
|
/hold cancel |
openshift-ci
bot
removed
the
do-not-merge/hold
Signed-off-by: Vu Dinh [email protected]
Description of the change:
Introduce the new type of constraint that uses CEL (Common Expression Language)
as an expression language to define the constraint in a generic way.
This type of constraint will allow operator authors to specify a dependency
on any arbitrary properties in the bundle. This constraint also supports
logical operator such as AND and OR in the CEL expression.
Motivation for the change:
https://github.com/operator-framework/enhancements/blob/master/enhancements/generic-constraints.md
Reviewer Checklist
/doc