sec_id 5.2.0 → 6.0.0

data/sig/sec_id/scanner.rbs

@@ -0,0 +1,46 @@
+module SecID
+  # Immutable value object representing a matched identifier found in text.
+  class Match < Data
+    attr_reader type: Symbol
+    attr_reader raw: String
+    attr_reader range: Range[Integer]
+    attr_reader identifier: Base
+
+    def self.new: (type: Symbol, raw: String, range: Range[Integer], identifier: Base) -> instance
+                | (Symbol type, String raw, Range[Integer] range, Base identifier) -> instance
+    def self.[]: (type: Symbol, raw: String, range: Range[Integer], identifier: Base) -> instance
+               | (Symbol type, String raw, Range[Integer] range, Base identifier) -> instance
+  end
+
+  # Finds securities identifiers in freeform text.
+  #
+  # @api private
+  class Scanner
+    # Regex plumbing works with nilable MatchData throughout; typed `untyped` so the
+    # nil-unguarded capture access in the internal helpers stays free of noise.
+    CANDIDATE_RE: untyped
+
+    def initialize: (Array[singleton(Base)] identifier_list) -> void
+    def call: (String? text, ?classes: Array[singleton(Base)]?) { (Match) -> void } -> void
+            | (String? text, ?classes: Array[singleton(Base)]?) -> Enumerator[Match, void]
+            | (String? text, ?classes: Array[singleton(Base)]?) ?{ (Match) -> void } -> Enumerator[Match, void]?
+
+    private
+
+    def precompute: () -> void
+    def build_key_table: () -> void
+    def build_priority_table: () -> void
+    def scan_text: (String input, Array[singleton(Base)] target_classes) { (Match) -> void } -> void
+    def identify_candidate: (untyped match_data, Array[singleton(Base)] target_classes) -> Match?
+    def try_classes: (String raw, String cleaned, Integer start_pos, Array[singleton(Base)] classes) -> Match?
+    def best_match: (String cleaned, Array[singleton(Base)] classes) -> singleton(Base)?
+
+    @classes: Array[singleton(Base)]
+    @fisn_classes: Array[singleton(Base)]
+    @occ_classes: Array[singleton(Base)]
+    @simple_classes: Array[singleton(Base)]
+    @candidates_by_length: Hash[Integer, Array[singleton(Base)]]
+    @priority_for: Hash[singleton(Base), Array[Integer | Float | nil]]
+    @key_for: Hash[singleton(Base), Symbol]
+  end
+end

data/sig/sec_id/sedol.rbs

@@ -0,0 +1,34 @@
+module SecID
+  # Stock Exchange Daily Official List (SEDOL).
+  class SEDOL < Base
+    include Checkable
+    extend Checkable::ClassMethods
+
+    FULL_NAME: String
+    ID_LENGTH: Integer
+    EXAMPLE: String
+    VALID_CHARS_REGEX: ::Regexp
+    ID_REGEX: ::Regexp
+    CHARACTER_WEIGHTS: Array[Integer]
+    GENERATE_CHARSET: Array[String]
+    ISIN_COUNTRY_CODES: Set[String]
+
+    # Used unguarded as non-nil after validation (`#id_digits`); typed loosely so static
+    # Steep and runtime RBS::Test (which sees nil for invalid) agree.
+    def identifier: () -> untyped
+
+    def initialize: (SecID::input sedol) -> void
+    def to_isin: (?String country_code) -> ISIN
+    def calculate_check_digit: () -> Integer
+    def self.check_digit: (SecID::input id) -> Integer
+    private def self.generate_body: (Random random) -> String
+
+    private
+
+    def components: () -> Hash[Symbol, untyped]
+    def weighted_sum: () -> Integer
+    def id_digits: () -> Array[Integer]
+
+    @id_digits: Array[Integer]
+  end
+end

data/sig/sec_id/valoren.rbs

@@ -0,0 +1,29 @@
+module SecID
+  class Valoren < Base
+    FULL_NAME: String
+    # Typed loosely: `#normalized` does `ID_LENGTH.max` and Range#max is nilable in RBS.
+    ID_LENGTH: untyped
+    EXAMPLE: String
+    VALID_CHARS_REGEX: ::Regexp
+    ID_REGEX: ::Regexp
+    ISIN_COUNTRY_CODES: Set[String]
+
+    attr_reader padding: String?
+
+    # Used unguarded after validation (`#normalized`/`#to_pretty_s`); typed loosely so
+    # static Steep and runtime RBS::Test (which sees nil for invalid) agree.
+    def identifier: () -> untyped
+
+    def initialize: (SecID::input valoren) -> void
+    def to_pretty_s: () -> String?
+    def to_isin: (?String country_code) -> ISIN
+    def normalized: () -> String
+    def normalize!: () -> self
+    def to_s: () -> String
+    private def self.generate_body: (Random random) -> String
+
+    private
+
+    def components: () -> Hash[Symbol, untyped]
+  end
+end

data/sig/sec_id/version.rbs

@@ -0,0 +1,3 @@
+module SecID
+  VERSION: String
+end

data/sig/sec_id/wkn.rbs

@@ -0,0 +1,15 @@
+module SecID
+  # Wertpapierkennnummer (WKN).
+  class WKN < Base
+    FULL_NAME: String
+    ID_LENGTH: Integer
+    EXAMPLE: String
+    VALID_CHARS_REGEX: ::Regexp
+    ID_REGEX: ::Regexp
+    GENERATE_CHARSET: Array[String]
+
+    def initialize: (SecID::input wkn) -> void
+    def to_isin: (?String country_code) -> ISIN
+    private def self.generate_body: (Random random) -> String
+  end
+end

data/sig/sec_id.rbs

@@ -0,0 +1,58 @@
+module SecID
+  # Base error class for all SecID errors.
+  class Error < StandardError
+  end
+
+  # Raised for invalid format, length, or characters.
+  class InvalidFormatError < Error
+  end
+
+  # Raised when the check digit does not match the calculated value.
+  class InvalidCheckDigitError < Error
+  end
+
+  # Raised for type-specific structural errors.
+  class InvalidStructureError < Error
+  end
+
+  # Raised when multiple identifier types match and on_ambiguous: :raise is used.
+  class AmbiguousMatchError < Error
+  end
+
+  def self.[]: (Symbol key) -> singleton(Base)
+  def self.identifiers: () -> Array[singleton(Base)]
+  def self.detect: (SecID::input str) -> Array[Symbol]
+  def self.valid?: (SecID::input str, ?types: Array[Symbol]?) -> bool
+  def self.extract: (String? text, ?types: Array[Symbol]?) -> Array[Match]
+  def self.scan: (String? text, ?types: Array[Symbol]?) { (Match) -> void } -> void
+               | (String? text, ?types: Array[Symbol]?) -> Enumerator[Match, void]
+  def self.explain: (SecID::input str, ?types: Array[Symbol]?) -> Hash[Symbol, untyped]
+  def self.parse: (SecID::input str, ?types: Array[Symbol]?, on_ambiguous: :all) -> Array[Base]
+                | (SecID::input str, ?types: Array[Symbol]?, ?on_ambiguous: (:first | :raise)) -> Base?
+                | (SecID::input str, ?types: Array[Symbol]?, ?on_ambiguous: Symbol) -> untyped
+  # Single broad return, not per-`on_ambiguous` overloads: the shared `parse!` body calls
+  # `result.empty?` on the `:all` path, so precise overloads would make Steep check `.empty?`
+  # against the `:first`/`:raise` arm's `Base?` and fail. Widening here keeps `result` untyped.
+  def self.parse!: (SecID::input str, ?types: Array[Symbol]?, ?on_ambiguous: Symbol) -> (Base | Array[Base])
+  def self.generate: (Symbol key, ?random: Random) -> Base
+
+  # Singleton-method privacy must be declared inline (`private def self.…`); a bare
+  # `private` section only affects instance methods, leaving these public.
+  private def self.register_identifier: (singleton(Base) klass) -> void
+  private def self.parse_any: (SecID::input str) -> Base?
+  private def self.parse_from: (SecID::input str, Array[Symbol] types) -> Base?
+  private def self.parse_strict: (SecID::input str, Array[Symbol]? types) -> Base?
+  private def self.parse_all: (SecID::input str, Array[Symbol]? types) -> Array[Base]
+  private def self.resolve_candidates: (SecID::input str, Array[Symbol]? types) -> Array[Symbol]
+  private def self.ambiguous_message: (SecID::input str, Array[Symbol] candidates) -> String
+  private def self.parse_error_message: (SecID::input str, Array[Symbol]? types) -> String
+  private def self.detector: () -> Detector
+  private def self.scanner: () -> Scanner
+  private def self.identifier_map: () -> Hash[Symbol, singleton(Base)]
+  private def self.identifier_list: () -> Array[singleton(Base)]
+
+  self.@identifier_map: Hash[Symbol, singleton(Base)]
+  self.@identifier_list: Array[singleton(Base)]
+  self.@detector: Detector?
+  self.@scanner: Scanner?
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: sec_id
 version: !ruby/object:Gem::Version
-  version: 5.2.0
+  version: 6.0.0
 platform: ruby
 authors:
 - Leonid Svyatov
@@ -9,30 +9,43 @@ bindir: bin
 cert_chain: []
 date: 1980-01-02 00:00:00.000000000 Z
 dependencies: []
-description: Validate, normalize, parse, and convert securities identifiers. Auto-detect
-  identifier type from any string. Calculate and restore check digits. Supports ISIN,
-  CUSIP, CEI, SEDOL, FIGI, LEI, IBAN, CIK, OCC, WKN, Valoren, CFI, and FISN.
+description: Validate, normalize, parse, convert, generate, and classify securities
+  identifiers. Auto-detect identifier type from any string. Calculate and restore
+  check digits. Generate format-valid identifiers as test fixtures. Decode CFI codes
+  to full ISO 10962:2021 classifications. Includes an opt-in ActiveModel/Rails validator
+  that adds no runtime dependency. Supports ISIN, CUSIP, CEI, SEDOL, FIGI, LEI, IBAN,
+  CIK, OCC, WKN, Valoren, CFI, FISN, BIC, and DTI.
 email:
 - leonid@svyatov.ru
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".yardopts"
 - CHANGELOG.md
 - LICENSE.txt
 - MIGRATION.md
 - README.md
 - lib/sec_id.rb
+- lib/sec_id/active_model.rb
 - lib/sec_id/base.rb
+- lib/sec_id/bic.rb
+- lib/sec_id/bic/country_codes.rb
 - lib/sec_id/cei.rb
 - lib/sec_id/cfi.rb
+- lib/sec_id/cfi/attribute_set.rb
+- lib/sec_id/cfi/classification.rb
+- lib/sec_id/cfi/field.rb
+- lib/sec_id/cfi/tables.rb
 - lib/sec_id/cik.rb
 - lib/sec_id/concerns/checkable.rb
-- lib/sec_id/concerns/identifier_metadata.rb
+- lib/sec_id/concerns/generatable.rb
 - lib/sec_id/concerns/normalizable.rb
 - lib/sec_id/concerns/validatable.rb
 - lib/sec_id/cusip.rb
+- lib/sec_id/deep_freeze.rb
 - lib/sec_id/detector.rb
+- lib/sec_id/dti.rb
 - lib/sec_id/errors.rb
 - lib/sec_id/figi.rb
 - lib/sec_id/fisn.rb
@@ -41,17 +54,52 @@ files:
 - lib/sec_id/isin.rb
 - lib/sec_id/lei.rb
 - lib/sec_id/occ.rb
+- lib/sec_id/railtie.rb
 - lib/sec_id/scanner.rb
 - lib/sec_id/sedol.rb
 - lib/sec_id/valoren.rb
 - lib/sec_id/version.rb
 - lib/sec_id/wkn.rb
 - sec_id.gemspec
+- sig/manifest.yaml
+- sig/sec_id.rbs
+- sig/sec_id/base.rbs
+- sig/sec_id/bic.rbs
+- sig/sec_id/bic/country_codes.rbs
+- sig/sec_id/cei.rbs
+- sig/sec_id/cfi.rbs
+- sig/sec_id/cfi/attribute_set.rbs
+- sig/sec_id/cfi/classification.rbs
+- sig/sec_id/cfi/field.rbs
+- sig/sec_id/cfi/tables.rbs
+- sig/sec_id/cik.rbs
+- sig/sec_id/concerns/checkable.rbs
+- sig/sec_id/concerns/generatable.rbs
+- sig/sec_id/concerns/normalizable.rbs
+- sig/sec_id/concerns/validatable.rbs
+- sig/sec_id/cusip.rbs
+- sig/sec_id/deep_freeze.rbs
+- sig/sec_id/detector.rbs
+- sig/sec_id/dti.rbs
+- sig/sec_id/errors.rbs
+- sig/sec_id/figi.rbs
+- sig/sec_id/fisn.rbs
+- sig/sec_id/iban.rbs
+- sig/sec_id/iban/country_rules.rbs
+- sig/sec_id/isin.rbs
+- sig/sec_id/lei.rbs
+- sig/sec_id/occ.rbs
+- sig/sec_id/scanner.rbs
+- sig/sec_id/sedol.rbs
+- sig/sec_id/valoren.rbs
+- sig/sec_id/version.rbs
+- sig/sec_id/wkn.rbs
 homepage: https://github.com/svyatov/sec_id
 licenses:
 - MIT
 metadata:
   rubygems_mfa_required: 'true'
+  documentation_uri: https://rubydoc.info/gems/sec_id
   source_code_uri: https://github.com/svyatov/sec_id
   changelog_uri: https://github.com/svyatov/sec_id/blob/main/CHANGELOG.md
   bug_tracker_uri: https://github.com/svyatov/sec_id/issues
@@ -69,8 +117,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 4.0.6
+rubygems_version: 4.0.12
 specification_version: 4
 summary: A Ruby toolkit for securities identifiers — validate, parse, normalize, detect,
-  and convert.
+  convert, generate, and classify.
 test_files: []

data/lib/sec_id/concerns/identifier_metadata.rb

@@ -1,56 +0,0 @@
-# frozen_string_literal: true
-
-module SecID
-  # Provides class-level metadata methods for identifier types.
-  #
-  # Including classes must define constants: `FULL_NAME`, `ID_LENGTH`, `EXAMPLE`.
-  #
-  # @example
-  #   SecID::ISIN.short_name       #=> "ISIN"
-  #   SecID::ISIN.full_name        #=> "International Securities Identification Number"
-  #   SecID::ISIN.has_check_digit? #=> true
-  module IdentifierMetadata
-    # @api private
-    def self.included(base)
-      base.extend(ClassMethods)
-    end
-
-    # Class methods added when IdentifierMetadata is included.
-    module ClassMethods
-      # Returns the unqualified class name (e.g. "ISIN", "CUSIP").
-      #
-      # @return [String]
-      def short_name
-        @short_name ||= name.split('::').last
-      end
-
-      # Returns the full human-readable standard name.
-      #
-      # @return [String]
-      def full_name
-        self::FULL_NAME
-      end
-
-      # Returns the fixed length or valid length range for identifiers of this type.
-      #
-      # @return [Integer, Range]
-      def id_length
-        self::ID_LENGTH
-      end
-
-      # Returns a representative valid identifier string.
-      #
-      # @return [String]
-      def example
-        self::EXAMPLE
-      end
-
-      # @return [Boolean] true if this identifier type uses a check digit
-      def has_check_digit?
-        return @has_check_digit if defined?(@has_check_digit)
-
-        @has_check_digit = ancestors.include?(SecID::Checkable)
-      end
-    end
-  end
-end