savourylie / eyeota_hw Goto Github PK

The objective of this assignment is to simulate a typical development task. You will design and implement a component which encapsulates a data structure which will used to lookup reference data in a latency-sensitive data processing pipeline.

The data which needs to be ingested is included in the attached file: data.json. It is organized as follows:

• Outer Array of Organization Objects keyed by the orgKey and contain an array of ParameterNameObjects
  • ParameterNameObjects are keyed by parameterName and contain an array of 1...n SegmentConfig objects.
    • SegmentConfig objects are keyed by parameterValue which apply to that segment. These can be an empty string, a single value, or a newline-delimited string of values. Multiple segment objects in a ParameterNameObject can share the same key. Internally, Segment Objects have a segmentId field.

Functionally, the following interface must be supported:

type LookupCache interface {
	GetSegmentForOrgAndKey(orgKey string, paramKey string) []SegmentConfig
	GetSegmentForOrgAndKeyAndVal(orgKey string, paramKey string, paramVal string) []SegmentConfig
}

//SegmentConfig is a struct that holds an id for 1 segment
type SegmentConfig struct {
	Id string
}

func (cfg *SegmentConfig) GetId() string {
	return cfg.Id
}

The LookupCache interface specifies the public API your cache should expose.

The first function will return an array of SegmentConfig objects given an orgKey and a parameterName. This call should return all SegmentConfigs associated with that org and parameter, and with an empty parameterValue

The second function will return an array of SegmentConfig objects given an orgKey, a parameterName, and a parameterValue. This call should return all SegmentConfigs associated with that org, parameter, and parameter value.

Given the following JSON Structure:

[
    {
        "org1": [
            {
                "paramName1": [
                    {
                        "paramVal1": {
                            "segmentId": "seg_1234"
                        }
                    },
                    {
                        "paramVal2\nparamVal3\nparamVal4\nparamVal5": {
                            "segmentId": "intr.edu"
                        }
                    },
                    {
                        "paramVal6": {
                            "segmentId": "dem.infg.m"
                        }
                    },
                    {
                         "paramVal6": {
                             "segmentId": "intr.heal"
                         }
                    },
                    {
                          "paramVal6": {
                             "segmentId": "dem.infg.f"
                          }
                    }
                ]
            },
            {
                "testedu": [
                    {
                        "": {
                            "segmentId": "n277"
                        }
                    }
                ]
            },
            {
                "sid": [
                    {
                        "": {
                            "segmentId": "dem.life.expat"
                        }
                    }
                ]
            },
            {
                "gen": [
                    {
                        "Female": {
                            "segmentId": "dem.g.f"
                        }
                    },
                    {
                        "Male": {
                            "segmentId": "dem.g.m"
                        }
                    }
                ]
            }
        ]
    }
]

• The query getSegmentFor("org1", "paramName1") will return an empty SegmentConfig array.
• The query getSegmentFor("org1", "paramName1", "paramVal1") will return a 1-element SegmentConfigArray containing a SegmentConfig object for seg_1234
• The query getSegmentFor("org1", "paramName1", <<"paramVal2" OR "paramVal3" OR "paramVal4" OR "paramVal5">> ) will return a 1-element SegmentConfigArray containing a SegmentConfig object with id: "intr.edu". (Note that in the data file paramVals may be in a "\n" delimited string.)
• The query getSegmentFor("org1", "paramName1", "paramVal6" ) will return a 3-element SegmentConfigArray containing SegmentConfig objects with ids: dem.infg.m, intr.heal, dem.infg.f
• The query getSegmentFor("org1", "testedu") will return a 1-element SegmentConfigArray containing a SegmentConfig object with id "n277"
• The query getSegmentFor("org1", "testedu", "") will return a 1-element SegmentConfigArray containing a SegmentConfig object with id "n277"
• The query getSegmentFor("org1", "testedu", <>) will return an empty SegmentConfig array.
• The query getSegmentFor("org1", "gen", "Female") will return a 1-element SegmentConfigArray containing a SegmentConfig object with id "dem.g.f"
• The query getSegmentFor("org1", "gen", "Male") will return a 1-element SegmentConfigArray containing a SegmentConfig object with id "dem.g.m"
• The query getSegmentFor("org1", "gen", <<any value other than "Male" or "Female">>) will return an empty SegmentConfig array.

• The data structure should be as small as possible in memory
• HIGH IMPORTANCE: The design should be optimized for speed of lookup (i.e., the methods defined in the interface).
• HIGH IMPORTANCE: The design should be optimized to be as memory stable as possible for lookups; meaning high rates of lookup calls should not generate memory churn and significant gc activity.

• Please implement your solution in your own project. We have provided a basic library which should be imported into your solution as a dependency.
• The go tool should be used for all dependency sourcing, building, test runs, etc... Do not use any other toolchain.
• Do not change any of the supplied files.
• You can design your solution with the assumption that the data will fit in memory.
• All projects must build and should have runnable unit tests that demonstrate that the requirements have been met. only the standard go test framework should be used for tests.
• Solutions should target Go 1.9 and above. No other dependencies (including external libraries) should be added. This includes both application and test code.
• You may use any resources available to you in terms of research, however you need to work alone.

Please submit your solution (without binaries) in the form of a tarball (created with command: tar -czvf) or a zip file to your Eyeota recruiting contact within 7 days of receiving these instructions.

Good luck, and have fun!

The Eyeota Recruiting Team