"paranoid" switch

[psecio]

"paranoid" switch

[jeremyquinton]

Im a 12 hour flight tonight so thought this would be something cool to hack on if it still needs doing?

[enygma]

Yeah, that'd be great! Some of those checks may already exist but the ranges for them may not match up with the "paranoid" settings.

[jeremyquinton]

Having had a look at the source code there is a couple ways this could be implemented.

Approach 1
A real simple approach I took was to define a new rule with a context of paranoid. I also updated the context when doing a scan to use prod by default if context is not specified. I only defined memory limit as an example you can see the implementation here https://github.com/jeremyquinton/iniscan/compare/issue39

Approach 2
There is a another approach that could be used. Tests in the rules.json could instead be defined as a different structure where contexts supersede other contexts. My thinking was to perhaps modify the json and a rule could have multiple test operations. So for memory limit the test is updated from

"test": {
      "key": "memory_limit",
      "operation": "smaller",
      "value": "128M"
} 

to

"test": [{
            "key": "memory_limit",
        "operation": "smaller",
        "value": "128M"
       },
       {
        "key": "memory_limit",
        "operation": "smaller",
        "value": "32M",
        "context": ["paranoid"]
       }
  ] 

Which approach do you think is best or if you think there is a better way let me know and I could go about implementing it.

test operations can then override other operations based on context. The second item in that json would also need an operation of smallerthanorequalto.

For this feature to be complete the fix command needs to be updated to have contexts and apply approach 1 or approach 2.