wow-dataset-registry-api/me.ahoo.wow.datasetregistry.api

Package-level declarations

Types

DatasetDrafted

data class DatasetDrafted(val table: String, val document: ObjectNode, val skipped: List<SkippedColumn> = emptyList())

DatasetPublished

data class DatasetPublished(val version: Int, val document: ObjectNode)

DatasetRegistryErrorCodes

object DatasetRegistryErrorCodes

The dataset registry's own error codes. The starter registers an HTTP status for each.

DatasetRetired

data class DatasetRetired(val version: Int)

DatasetRevised

data class DatasetRevised(val document: ObjectNode, val reviews: Map<String, FieldSensitivity>)

DraftDataset

@Order(value = 1)

@Summary(value = "Draft a dataset from a table")

@CreateAggregate

@CommandRoute(action = "draft", method = CommandRoute.Method.POST, appendIdPath = CommandRoute.AppendPath.ALWAYS)

data class DraftDataset(val id: String, val table: String, val source: String? = null)

Drafts the dataset id ({context}.{dataset}) from table: the server reads the table's columns and declares what they prove (types, nullability, dates, string arrays), suggests the scope and identity from conventional column names, and leaves sensitivity to the review.

FieldSensitivity

data class FieldSensitivity(val level: SensitivityLevel? = null, val keepPrefix: Int = 0, val keepSuffix: Int = 0)

How one field is protected, as a reviewer decided: level null is the decision that the field is not sensitive, which is a decision too.

PublishDataset

@Order(value = 4)

@Summary(value = "Publish a dataset")

@CommandRoute(action = "publish", method = CommandRoute.Method.POST, appendIdPath = CommandRoute.AppendPath.ALWAYS)

data class PublishDataset(val id: String)

Publishes the draft as the dataset's next version. Refused until the declaration names its scope, every string field's sensitivity is reviewed, and the declaration compiles against its table.

RetireDataset

@Order(value = 5)

@Summary(value = "Retire a dataset")

@CommandRoute(action = "retire", method = CommandRoute.Method.POST, appendIdPath = CommandRoute.AppendPath.ALWAYS)

data class RetireDataset(val id: String)

Stops serving the dataset. Its draft and history stay; publishing again serves it again.

ReviewSensitivity

@Order(value = 3)

@Summary(value = "Review the sensitivity of a dataset's fields")

@CommandRoute(action = "sensitivity", method = CommandRoute.Method.PUT, appendIdPath = CommandRoute.AppendPath.ALWAYS)

data class ReviewSensitivity(val id: String, val fields: Map<String, FieldSensitivity>)

Records the reviewer's decision on each of fields: sensitive at a level, or not sensitive.

ReviseDataset

@Order(value = 2)

@Summary(value = "Revise a dataset's draft")

@CommandRoute(action = "revise", method = CommandRoute.Method.PUT, appendIdPath = CommandRoute.AppendPath.ALWAYS)

data class ReviseDataset(val id: String, val document: ObjectNode)

Replaces the draft's declaration with document, in the format of a dataset file. Sensitivity is not part of it: it is reviewed field by field (ReviewSensitivity). A field whose declaration is unchanged keeps its review.

SensitivityReviewed

data class SensitivityReviewed(val fields: Map<String, FieldSensitivity>)

SkippedColumn

data class SkippedColumn(val name: String, val type: String)

A column the draft leaves out because a dataset cannot declare its type.