feat(rules): enable/disable automated rules

[lkonno]

Fixes #773

http PATCH :8181/api/v2/rules/demo?enabled=false

When a rule created disabled is enabled, the rule is activated and it starts recording.

[andrewazores]

Looks mostly good from a reading. I haven't tested it out yet but judging by the test changes I think it would work as expected.

One change I would like to see is to use the request body for the PATCH rather than using a query parameter. This is a little more conventional and also matches how we stop or archive active recordings:

cryostat/src/main/java/io/cryostat/net/web/http/api/v1/TargetRecordingPatchHandler.java

Line 117 in d55889e

String mtd = ctx.getBodyAsString();

Here we get the request body as a simple String, which is either "save" (copy to archives) or "stop" (RUNNING active recording becomes STOPPED).

So rather than PATCH /api/v2/rules/:ruleName?enabled=state it would be simply PATCH /api/v2//rules/:ruleName with a request body content. I think the natural thing to do for the body content is to expect it as a JSON object like { "enabled": false }. After parsing the JSON the rest of the request would be handled the same way you're currently doing it.

[andrewazores]

It might be worth adding an integration test here. I think it can be pretty simple:

- create a rule definition with enabled=false
- start an external target container that matches the rule's matchExpression
- query the list of recordings on that target and assert that the list is empty
- patch the rule to enable it
- sleep a few seconds
- query the recording list again and assert that it has one recording with the expected name

[lkonno]

It might be worth adding an integration test here. I think it can be pretty simple:

- create a rule definition with enabled=false
- start an external target container that matches the rule's matchExpression
- query the list of recordings on that target and assert that the list is empty
- patch the rule to enable it
- sleep a few seconds
- query the recording list again and assert that it has one recording with the expected name

Created a new test in the AutoRulesIT.

[andrewazores]

Looks good, thanks!