class Capsium::Package::Dataset

A loaded dataset. “source”/“schemaFile”/“databaseFile” are package-relative paths (ARCHITECTURE.md section 5), resolved against the package directory.

Attributes

config [R] DatasetConfig

data [R] untyped

name [R] String

package_path [R] String

Public Class Methods

new (name:, config:, package_path:)

(name: String name, config: DatasetConfig config, package_path: String package_path) → void

Source

# File lib/capsium/package/dataset.rb, line 22
def initialize(name:, config:, package_path:)
  @name = name
  @config = config
  @package_path = package_path
  @data = load_data
end

Public Instance Methods

json_schema ()

() → Hash[String, untyped]?

Source

# File lib/capsium/package/dataset.rb, line 72
def json_schema
  return nil unless @config.schema_file && @config.schema_type == "json-schema"
  return nil unless File.file?(schema_path)

  load_schema
end

The parsed JSON schema for this dataset, or nil when none is declared or the file is unreadable.

load_data ()

() → untyped

Source

# File lib/capsium/package/dataset.rb, line 39
def load_data
  return nil unless File.file?(source_path)

  case @config.format
  when "yaml" then YAML.load_file(source_path)
  when "json" then JSON.parse(File.read(source_path))
  when "csv" then CSV.read(source_path, headers: true)
  when "tsv" then CSV.read(source_path, col_sep: "\t", headers: true)
  when "sqlite" then load_sqlite_data
  else
    raise Error, "Unsupported data file type: #{@config.format}"
  end
end

schema_errors_for (data)

(untyped data) → Array[String]

Source

# File lib/capsium/package/dataset.rb, line 83
def schema_errors_for(data)
  return [] unless @config.schema_file && @config.schema_type == "json-schema"
  return [] unless File.file?(schema_path)

  JSON::Validator.fully_validate(load_schema, JSON.parse(JSON.generate(data)))
end

Schema validation of an arbitrary candidate document (used by the reactor’s writable overlay before persisting a mutation): human-readable schema problems, empty when valid or when the dataset has no JSON schema.

schema_path ()

() → String?

Source

# File lib/capsium/package/dataset.rb, line 33
def schema_path
  return unless @config.schema_file

  File.join(@package_path, @config.schema_file)
end

source_path ()

() → String

Source

# File lib/capsium/package/dataset.rb, line 29
def source_path
  File.join(@package_path, @config.backing_file)
end

validate ()

() → bool

Source

# File lib/capsium/package/dataset.rb, line 53
def validate
  return true unless @config.schema_file && @config.schema_type == "json-schema"

  JSON::Validator.validate!(load_schema, @data.to_json)
end

validation_errors ()

() → Array[String]

Source

# File lib/capsium/package/dataset.rb, line 61
def validation_errors
  problems = []
  unless File.file?(source_path)
    problems << "dataset source missing on disk: #{@config.backing_file}"
  end
  problems.concat(schema_validation_errors)
  problems
end

File-existence and schema validations for this dataset. Returns a list of human-readable problems; empty when valid.