parse_date 0.3.3 → 0.4.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 42b7b5f7405ee4a79e7eece920b10fe2852a03c996b32da8c2b8edc58a6e3ede
-  data.tar.gz: eeba668fd1ee75994f743a010dbe37af4ffdcb8ca0c4143ca329c811739c4b28
+  metadata.gz: 60c444ed79ea6842f49bfb44d2adc5bd1a6790d483fe43e0bd4a0880001a6005
+  data.tar.gz: a62f906a4c908c4b774ef9fe1dae2a047deee79ee6569688d1c2db9796bb3b2e
 SHA512:
-  metadata.gz: a22f1e23899efc15e5fd13ab1aac724cddc39ddf9986d0295e2f4360b639483da116ef4540bb35fc0c03cf201c0d29f4e0ac6729339140663ff7622ed8a7a623
-  data.tar.gz: df518ca44ed63ea2925976fe49a11e28cd4ecc40b66752c7db11934d4bba6fe7ade795aca93402cb0f556816cf076e40006e08294162b8f01c527377a4603e51
+  metadata.gz: a949b307fb3259f1a20fa279aa2af395137b514b35b15dbce8c63701047462b06b5b6cb31723f25dbef2c8f1db249cf8e95914d92c113dd4439ed6737267878f
+  data.tar.gz: e1e3e154a6bc4851a7aafa199da3ee66e0d4b7d5d27537528cee36d2f0b2f80761fbbffffa1d2a544ca4a204f4fd7816f85a6d0b975824fff06cead4d64cf3f2

data/.github/pull_request_template.md

@@ -1,4 +1,12 @@
 ## Why was this change made?
 
 
-## Was the documentation (README, API, wiki, ...) updated?
+
+## How was this change tested?
+
+
+
+## Which documentation and/or configurations were updated?
+
+
+

data/.rubocop.yml

@@ -1,7 +1,7 @@
 inherit_from: .rubocop_todo.yml
 
-AllCops:
-  TargetRubyVersion: 2.6
+Gemspec/RequiredRubyVersion:
+  Enabled: false
 
 Layout/EmptyLinesAroundClassBody:
   Enabled: false
@@ -9,15 +9,27 @@ Layout/EmptyLinesAroundClassBody:
 Layout/EmptyLinesAroundModuleBody:
   Enabled: false
 
+Layout/LineLength:
+  Exclude:
+    - parse_date.gemspec
+
+Lint/MixedRegexpCaptureTypes:
+  Enabled: false
+
 Metrics/BlockLength:
   Exclude:
     - spec/**/*
 
-Metrics/LineLength:
-  Max: 120
-
 Metrics/MethodLength:
-  Max: 25
+  Max: 15
+
+Metrics/ModuleLength:
+  Exclude:
+    - lib/parse_date/int_from_string.rb
+
+Style/Documentation:
+  Exclude:
+    - lib/parse_date.rb
 
 Style/NumericLiterals:
   Enabled: false
@@ -35,3 +47,64 @@ Style/WordArray:
 
 Style/YodaCondition:
   Enabled: false
+
+Gemspec/DateAssignment: # (new in 1.10)
+  Enabled: true
+Layout/SpaceBeforeBrackets: # (new in 1.7)
+  Enabled: true
+Lint/AmbiguousAssignment: # (new in 1.7)
+  Enabled: true
+Lint/DeprecatedConstants: # (new in 1.8)
+  Enabled: true
+Lint/DuplicateBranch: # (new in 1.3)
+  Enabled: true
+Lint/DuplicateRegexpCharacterClassElement: # (new in 1.1)
+  Enabled: true
+Lint/EmptyBlock: # (new in 1.1)
+  Enabled: true
+Lint/EmptyClass: # (new in 1.3)
+  Enabled: true
+Lint/LambdaWithoutLiteralBlock: # (new in 1.8)
+  Enabled: true
+Lint/NoReturnInBeginEndBlocks: # (new in 1.2)
+  Enabled: true
+Lint/NumberedParameterAssignment: # (new in 1.9)
+  Enabled: true
+Lint/OrAssignmentToConstant: # (new in 1.9)
+  Enabled: true
+Lint/RedundantDirGlobSort: # (new in 1.8)
+  Enabled: true
+Lint/SymbolConversion: # (new in 1.9)
+  Enabled: true
+Lint/ToEnumArguments: # (new in 1.1)
+  Enabled: true
+Lint/TripleQuotes: # (new in 1.9)
+  Enabled: true
+Lint/UnexpectedBlockArity: # (new in 1.5)
+  Enabled: true
+Lint/UnmodifiedReduceAccumulator: # (new in 1.1)
+  Enabled: true
+Style/ArgumentsForwarding: # (new in 1.1)
+  Enabled: true
+Style/CollectionCompact: # (new in 1.2)
+  Enabled: true
+Style/DocumentDynamicEvalDefinition: # (new in 1.1)
+  Enabled: true
+Style/EndlessMethod: # (new in 1.8)
+  Enabled: true
+Style/HashConversion: # (new in 1.10)
+  Enabled: true
+Style/HashExcept: # (new in 1.7)
+  Enabled: true
+Style/IfWithBooleanLiteralBranches: # (new in 1.9)
+  Enabled: true
+Style/NegatedIfElseCondition: # (new in 1.2)
+  Enabled: true
+Style/NilLambda: # (new in 1.3)
+  Enabled: true
+Style/RedundantArgument: # (new in 1.4)
+  Enabled: true
+Style/StringChars: # (new in 1.12)
+  Enabled: true
+Style/SwapValues: # (new in 1.1)
+  Enabled: true

data/.rubocop_todo.yml

@@ -1,47 +1,7 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config`
-# on 2019-10-28 09:19:23 -0700 using RuboCop version 0.74.0.
+# on 2021-03-26 19:20:09 UTC using RuboCop version 1.12.0.
 # The point is for the user to remove these configuration records
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
 # versions of RuboCop, may require this file to be generated again.
-
-# Offense count: 5
-Metrics/AbcSize:
-  Max: 41
-
-# Offense count: 4
-Metrics/CyclomaticComplexity:
-  Max: 12
-
-# Offense count: 1
-# Configuration parameters: CountComments.
-Metrics/ModuleLength:
-  Max: 178
-
-# Offense count: 4
-Metrics/PerceivedComplexity:
-  Max: 12
-
-# Offense count: 2
-Style/Documentation:
-  Exclude:
-    - 'spec/**/*'
-    - 'test/**/*'
-    - 'lib/parse_date.rb'
-
-# Offense count: 1
-# Cop supports --auto-correct.
-# Configuration parameters: EnforcedOctalStyle.
-# SupportedOctalStyles: zero_with_o, zero_only
-Style/NumericLiteralPrefix:
-  Exclude:
-    - 'spec/parse_date/int_from_string_spec.rb'
-
-# Offense count: 1
-# Cop supports --auto-correct.
-# Configuration parameters: EnforcedStyle, AllowInnerSlashes.
-# SupportedStyles: slashes, percent_r, mixed
-Style/RegexpLiteral:
-  Exclude:
-    - 'lib/parse_date/int_from_string.rb'

data/README.md

@@ -1,5 +1,5 @@
 [![Gem Version](https://badge.fury.io/rb/parse_date.svg)](https://badge.fury.io/rb/parse_date)
-[![Build Status](https://travis-ci.org/sul-dlss/parse_date.svg?branch=master)](https://travis-ci.org/sul-dlss/parse_date)
+[![Build Status](https://travis-ci.com/sul-dlss/parse_date.svg?branch=main)](https://travis-ci.com/sul-dlss/parse_date)
 [![Maintainability](https://api.codeclimate.com/v1/badges/2d006b4ccb3100434f4a/maintainability)](https://codeclimate.com/github/sul-dlss/parse_date/maintainability)
 [![Test Coverage](https://api.codeclimate.com/v1/badges/2d006b4ccb3100434f4a/test_coverage)](https://codeclimate.com/github/sul-dlss/parse_date/test_coverage)
 
@@ -40,6 +40,14 @@ ParseDate.parse_range('195-')                           # (1950..1959).to_a
 ParseDate.parse_range('199u')                           # (1990..1999).to_a
 ParseDate.parse_range('197?')                           # (1970..1979).to_a
 ParseDate.parse_range('196x')                           # (1960..1969).to_a
+ParseDate.parse_range('1990s')                          # (1990..1999).to_a
+ParseDate.parse_range('1990s?')                         # (1990..1999).to_a
+ParseDate.parse_range('ca. 1930s')                      # (1930..1939).to_a
+ParseDate.parse_range('1928-1980s')                     # (1928..1989).to_a
+ParseDate.parse_range('1940s-1990')                     # (1940..1990).to_a
+ParseDate.parse_range('1980s-1990s')                    # (1980..1999).to_a
+ParseDate.parse_range('1675-7')                         # [1675, 1676, 1677]
+ParseDate.parse_range('1040–1 CE')                      # [1040, 1041]
 ParseDate.parse_range('18th century CE')                # (1700..1799).to_a
 ParseDate.parse_range('17uu')                           # (1700..1799).to_a
 ParseDate.parse_range('between 1694 and 1799')          # (1694..1799).to_a
@@ -52,19 +60,26 @@ ParseDate.parse_range('17--?-18--?')                    # (1700..1899).to_a
 ParseDate.parse_range('1835 or 1836')                   # [1835, 1836]
 ParseDate.parse_range('17-- or 18--?')                  # (1700..1899).to_a
 ParseDate.parse_range('-2 or 1?')                       # (-2..1).to_a
+ParseDate.parse_range('1500? to 1582')                  # (1500..1582).to_a
 ParseDate.parse_range('17th or 18th century?')          # (1600..1799).to_a
 ParseDate.parse_range('ca. 5th–6th century A.D.')       # (400..599).to_a
 ParseDate.parse_range('ca. 9th–8th century B.C.')       # (-999..-800).to_a
 ParseDate.parse_range('ca. 13th–12th century B.C.')     # (-1399..-1200).to_a
 ParseDate.parse_range('5th century B.C.')               # (-599..-500).to_a
 ParseDate.parse_range('502-504')                        # [502, 503, 504]
+ParseDate.parse_range('950-60')                         # (950..960).to_a
+ParseDate.parse_range('-0150 - -0100')                  # (-150..-100).to_a
 ParseDate.parse_range('-2100 - -2000')                  # (-2100..-2000).to_a
+ParseDate.parse_range('1230—1239 CE')                   # (1230..1239).to_a (alternate hyphen char)
+ParseDate.parse_range('996–1021 CE')                    # (996..1021).to_a (diff alternate hyphen char)
 ParseDate.parse_range('1975 - 1905')                    # last year > first year, raises error
 ParseDate.parse_range('-100 - -150')                    # last year > first year, raises error
 ParseDate.parse_range('1975 or 1905')                   # last year > first year, raises error
-ParseDate.parse_range('2050')                           # year later than current year + 1, raises error
-ParseDate.parse_range('random text')                    # can't parse years, raises error
-ParseDate.parse_range(nil)                              # can't parse years, raises error
+ParseDate.parse_range('1975 - 2050')                    # single invalid year endpoint, raises error
+ParseDate.parse_range('2050')                           # nil - only invalid year endpoints present
+ParseDate.parse_range('2045 - 2050')                    # nil - only invalid year endpoints present
+ParseDate.parse_range('random text')                    # nil - can't parse years
+ParseDate.parse_range(nil)                              # nil - can't parse years
 
 ParseDate.earliest_year('12/25/00')                     # 2000
 ParseDate.earliest_year('5-1-21')                       # 1921
@@ -82,7 +97,10 @@ ParseDate.earliest_year('17uu')                         # 1700
 ParseDate.earliest_year('between 1694 and 1799')        # 1694
 ParseDate.earliest_year('between 1 and 5')              # 1
 ParseDate.earliest_year('between 300 and 150 B.C.')     # -300
+ParseDate.earliest_year('1500? to 1582')                # 1500
 ParseDate.earliest_year('1496-1499')                    # 1496
+ParseDate.earliest_year('1230—1239 CE')                 # 1230 (alternate hyphen char)
+ParseDate.earliest_year('996–1021 CE')                  # 996 (diff alternate hyphen char)
 ParseDate.earliest_year('1750?-1867')                   # 1750
 ParseDate.earliest_year('17--?-18--?')                  # 1700
 ParseDate.earliest_year('1835 or 1836')                 # 1835
@@ -93,6 +111,8 @@ ParseDate.earliest_year('ca. 9th–8th century B.C.')     # -999
 ParseDate.earliest_year('ca. 13th–12th century B.C.')   # -1399
 ParseDate.earliest_year('5th century B.C.')             # -599
 ParseDate.earliest_year('502-504')                      # 502
+ParseDate.earliest_year('950-60')                       # 950
+ParseDate.earliest_year('-0150 - -0100')                # -150
 ParseDate.earliest_year('-2100 - -2000')                # -2100
 
 ParseDate.latest_year('20000222')                       # 2000
@@ -100,12 +120,23 @@ ParseDate.latest_year('195-')                           # 1959
 ParseDate.latest_year('199u')                           # 1999
 ParseDate.latest_year('197?')                           # 1979
 ParseDate.latest_year('196x')                           # 1969
+ParseDate.latest_year('1990s')                          # 1999
+ParseDate.latest_year('1990s?')                         # 1999
+ParseDate.latest_year('ca. 1930s')                      # 1939
+ParseDate.latest_year('1928-1980s')                     # 1989
+ParseDate.latest_year('1940s-1990')                     # 1990
+ParseDate.latest_year('1980s-1990s')                    # 1999
+ParseDate.latest_year('1675-7')                         # 1677
+ParseDate.latest_year('1040–1 CE')                      # 1041
 ParseDate.latest_year('18th century CE')                # 1799
 ParseDate.latest_year('17uu')                           # 1799
 ParseDate.latest_year('between 1694 and 1799')          # 1799
 ParseDate.latest_year('between 1 and 5')                # 5
 ParseDate.latest_year('between 300 and 150 B.C.')       # -150
+ParseDate.latest_year('1500? to 1582')                  # 1582
 ParseDate.latest_year('1496-1499')                      # 1499
+ParseDate.latest_year('1230—1239 CE')                   # 1239 (alternate hyphen char)
+ParseDate.latest_year('996–1021 CE')                    # 1021 (diff alternate hyphen char)
 ParseDate.latest_year('1750?-1867')                     # 1867
 ParseDate.latest_year('17--?-18--?')                    # 1899
 ParseDate.latest_year('1757-58')                        # 1758
@@ -119,6 +150,8 @@ ParseDate.latest_year('ca. 13th–12th century B.C.')     # -1200
 ParseDate.latest_year('5th century B.C.')               # -500
 ParseDate.latest_year('-5 - 3')                         # 3
 ParseDate.latest_year('502-504')                        # 504
+ParseDate.latest_year('950-60')                         # 960
+ParseDate.latest_year('-0150 - -0100')                  # -100
 ParseDate.latest_year('-2100 - -2000')                  # -2000
 
 ParseDate.range_array('1993', '1995')                   # [1993, 1994, 1995]

data/lib/parse_date.rb

@@ -42,6 +42,7 @@ class ParseDate
   def self.parse_range(date_str)
     first = earliest_year(date_str)
     last = latest_year(date_str)
+    return nil unless first || last
     raise ParseDate::Error, "Unable to parse range from '#{date_str}'" unless year_range_valid?(first, last)
 
     range_array(first, last)
@@ -64,6 +65,8 @@ class ParseDate
   # @param [Integer, String] first_year, expecting integer or parseable string for .to_i
   # @param [Integer, String] last_year, expecting integer or parseable string for .to_i
   # @return [Array] array of Integer year values from first to last, inclusive
+  # rubocop:disable Metrics/CyclomaticComplexity
+  # rubocop:disable Metrics/PerceivedComplexity
   def self.range_array(first_year, last_year)
     first_year = first_year.to_i if first_year.is_a?(String) && first_year.match?(/^-?\d+$/)
     last_year = last_year.to_i if last_year.is_a?(String) && last_year.match?(/^-?\d+$/)
@@ -71,9 +74,11 @@ class ParseDate
     return [] unless last_year || first_year
     return [first_year] if last_year.nil? && first_year
     return [last_year] if first_year.nil? && last_year
-    raise(StandardError, "unable to create year range array from #{first_year}, #{last_year}") unless
+    raise(ParseDate::Error, "unable to create year range array from #{first_year}, #{last_year}") unless
       year_range_valid?(first_year, last_year)
 
     Range.new(first_year, last_year).to_a
   end
+  # rubocop:enable Metrics/CyclomaticComplexity
+  # rubocop:enable Metrics/PerceivedComplexity
 end

data/lib/parse_date/int_from_string.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require 'date' # so upstream callers don't have to require it
-
 class ParseDate
 
   # Parse (Year) Integers from Date Strings
@@ -20,23 +18,15 @@ class ParseDate
       return if date_str == '0000-00-00' # shpc collection has these useless dates
 
       # B.C. first (match longest string first)
-      return ParseDate.send(:earliest_century_bc, date_str) if date_str.match(YY_YY_CENTURY_BC_REGEX)
-      return ParseDate.send(:between_bc_earliest_year, date_str) if date_str.match(BETWEEN_Yn_AND_Yn_BC_REGEX)
-      return ParseDate.send(:year_int_for_bc, date_str) if date_str.match(YEAR_BC_REGEX)
+      bc_result = ParseDate.send(:earliest_year_bc_parsing, date_str)
+      return bc_result if bc_result
 
-      result ||= ParseDate.send(:between_earliest_year, date_str)
-      result ||= ParseDate.send(:negative_first_four_digits, date_str)
-      result ||= ParseDate.send(:first_four_digits, date_str)
-      result ||= ParseDate.send(:year_from_mm_dd_yy, date_str)
-      result ||= ParseDate.send(:first_year_for_decade, date_str) # 198x or 201x
-      result ||= ParseDate.send(:first_year_for_century, date_str) # includes BC
-      result ||= ParseDate.send(:year_for_early_numeric, date_str)
-      unless result
-        # try removing brackets between digits in case we have 169[5] or [18]91
-        no_brackets = ParseDate.send(:remove_brackets, date_str)
-        return earliest_year(no_brackets) if no_brackets
-      end
-      result.to_i if result && year_int_valid?(result.to_i)
+      result = ParseDate.send(:earliest_year_parsing, date_str)
+      return result if result
+
+      # try removing brackets between digits in case we have 169[5] or [18]91
+      no_brackets = ParseDate.send(:remove_brackets, date_str)
+      earliest_year(no_brackets) if no_brackets
     end
 
     # latest year as Integer if we can parse one from date_str
@@ -52,30 +42,15 @@ class ParseDate
       return if date_str == '0000-00-00' # shpc collection has these useless dates
 
       # B.C. first (match longest string first)
-      return ParseDate.send(:last_year_mult_centuries_bc, date_str) if date_str.match(YY_YY_CENTURY_BC_REGEX)
-      return ParseDate.send(:between_bc_latest_year, date_str) if date_str.match(BETWEEN_Yn_AND_Yn_BC_REGEX)
-      return ParseDate.send(:last_year_for_bc_century, date_str) if date_str.match(BC_CENTURY_REGEX)
-      return ParseDate.send(:year_int_for_bc, date_str) if date_str.match(BC_REGEX)
+      bc_result = ParseDate.send(:latest_year_bc_parsing, date_str)
+      return bc_result if bc_result
 
-      result ||= ParseDate.send(:between_latest_year, date_str)
-      result ||= ParseDate.send(:hyphen_4digit_latest_year, date_str)
-      result ||= ParseDate.send(:hyphen_2digit_latest_year, date_str)
-      result ||= ParseDate.send(:yyuu_after_hyphen, date_str)
-      result ||= ParseDate.send(:year_after_or, date_str)
-      result ||= ParseDate.send(:negative_4digits_after_hyphen, date_str)
-      result ||= ParseDate.send(:negative_first_four_digits, date_str)
-      result ||= ParseDate.send(:first_four_digits, date_str)
-      result ||= ParseDate.send(:year_from_mm_dd_yy, date_str)
-      result ||= ParseDate.send(:last_year_for_decade, date_str) # 198x or 201x
-      result ||= ParseDate.send(:last_year_mult_centuries, date_str) # nth-nth century
-      result ||= ParseDate.send(:last_year_for_century, date_str)
-      result ||= ParseDate.send(:last_year_for_early_numeric, date_str)
-      unless result
-        # try removing brackets between digits in case we have 169[5] or [18]91
-        no_brackets = ParseDate.send(:remove_brackets, date_str)
-        return earliest_year(no_brackets) if no_brackets
-      end
-      result.to_i if result && year_int_valid?(result.to_i)
+      result = ParseDate.send(:latest_year_parsing, date_str)
+      return result if result
+
+      # try removing brackets between digits in case we have 169[5] or [18]91
+      no_brackets = ParseDate.send(:remove_brackets, date_str)
+      latest_year(no_brackets) if no_brackets
     end
 
     # true if the year is between -9999 and (current year + 1), inclusive
@@ -88,24 +63,95 @@ class ParseDate
 
     protected
 
+    def earliest_year_bc_parsing(date_str)
+      return ParseDate.send(:earliest_century_bc, date_str) if date_str.match(YY_YY_CENTURY_BC_REGEX)
+      return ParseDate.send(:between_bc_earliest_year, date_str) if date_str.match(BETWEEN_Yn_AND_Yn_BC_REGEX)
+      return ParseDate.send(:year_int_for_bc, date_str) if date_str.match(YEAR_BC_REGEX)
+    end
+
+    def earliest_year_parsing(date_str)
+      [
+        # longest string first, more or less
+        :between_earliest_year,
+        :hyphen_4digit_earliest_year,
+        :negative_first_four_digits,
+        :first_four_digits,
+        :year_from_mm_dd_yy,
+        :first_year_for_decade, # 198x or 201x
+        :first_year_for_century, # includes some BC
+        :year_for_early_numeric
+      ].each do |method_name|
+        result = ParseDate.send(method_name, date_str)
+        return result.to_i if result && year_int_valid?(result.to_i)
+      end
+      nil
+    end
+
+    def latest_year_bc_parsing(date_str)
+      return ParseDate.send(:last_year_mult_centuries_bc, date_str) if date_str.match(YY_YY_CENTURY_BC_REGEX)
+      return ParseDate.send(:between_bc_latest_year, date_str) if date_str.match(BETWEEN_Yn_AND_Yn_BC_REGEX)
+      return ParseDate.send(:last_year_for_bc_century, date_str) if date_str.match(BC_CENTURY_REGEX)
+      return ParseDate.send(:year_int_for_bc, date_str) if date_str.match(BC_REGEX)
+    end
+
+    # rubocop:disable Metrics/MethodLength
+    def latest_year_parsing(date_str)
+      result = nil
+      [
+        # longest string first, more or less
+        :between_latest_year,
+        :hyphen_4digit_latest_year,
+        :hyphen_2digit_latest_year,
+        :hyphen_1digit_latest_year,
+        :yyuu_after_hyphen,
+        :year_after_or,
+        :negative_4digits_after_hyphen,
+        :negative_first_four_digits,
+        :last_year_for_0s_decade,
+        :first_four_digits,
+        :year_from_mm_dd_yy,
+        :last_year_for_decade, # 198x or 201x
+        :last_year_mult_centuries, # nth-nth century
+        :last_year_for_century,
+        :last_year_for_early_numeric
+      ].each do |method|
+        result ||= ParseDate.send(method, date_str)
+        return result.to_i if result && year_int_valid?(result.to_i)
+      end
+      nil
+    end
+    # rubocop:enable Metrics/MethodLength
+
     REGEX_OPTS = Regexp::IGNORECASE | Regexp::MULTILINE
     BC_REGEX = Regexp.new(/\s*B\.?\s*C\.?/im)
-    BRACKETS_BETWEEN_DIGITS_REGEX = Regexp.new('\d[' + Regexp.escape('[]') + ']\d')
+    BRACKETS_BETWEEN_DIGITS_REGEX = Regexp.new("\\d[#{Regexp.escape('[]')}]\\d")
 
     # removes brackets between digits such as 169[5] or [18]91
     def remove_brackets(date_str)
       date_str.delete('[]') if date_str.match(BRACKETS_BETWEEN_DIGITS_REGEX)
     end
 
-    YYYY_HYPHEN_YYYY_REGEX = Regexp.new(/(?<first>\d{4})\??\s*[-—]\s*(?<last>\d{4})\??/m)
+    YYYY_HYPHEN_YYYY_REGEX = Regexp.new(/(?<first>\d{3,4})s?\??\s*(-|—|–|to)\s*(?<last>\d{4}s?)\??/m)
+
+    # Integer value for latest year if we have "yyyy-yyyy" pattern
+    # @return [Integer, nil] yyyy if date_str matches pattern; nil otherwise
+    def hyphen_4digit_earliest_year(date_str)
+      Regexp.last_match(:first).to_i if date_str.match(YYYY_HYPHEN_YYYY_REGEX)
+    end
 
     # Integer value for latest year if we have "yyyy-yyyy" pattern
     # @return [Integer, nil] yyyy if date_str matches pattern; nil otherwise
     def hyphen_4digit_latest_year(date_str)
-      Regexp.last_match(:last).to_i if date_str.match(YYYY_HYPHEN_YYYY_REGEX)
+      latest = Regexp.last_match(:last) if date_str.match(YYYY_HYPHEN_YYYY_REGEX)
+      if ParseDate.year_int_valid?(latest.to_i)
+        ParseDate.latest_year(latest) # accommodates '1980s - 1990s'
+      else
+        # return the bad value;  parse_range might need to complain about it
+        latest
+      end
     end
 
-    YYYY_HYPHEN_YY_REGEX = Regexp.new(/(?<first>\d{4})\??\s*[-—]\s*(?<last>\d{2})\??([^-0-9].*)?$/)
+    YYYY_HYPHEN_YY_REGEX = Regexp.new(/(?<first>\d{3,4})\??\s*(-|—|–|to)\s*(?<last>\d{2})\??([^-0-9].*)?$/)
 
     # Integer value for latest year if we have "yyyy-yy" pattern
     # @return [Integer, nil] yyyy if date_str matches pattern; nil otherwise
@@ -114,14 +160,28 @@ class ParseDate
       return unless matches
 
       first = Regexp.last_match(:first)
-      century = first[0, 2]
+      century = first[0..-3] # whatever is before the last 2 digits
       last = "#{century}#{Regexp.last_match(:last)}"
       last.to_i if ParseDate.year_range_valid?(first.to_i, last.to_i)
     end
 
+    YYYY_HYPHEN_Y_REGEX = Regexp.new(/(?<first>\d{3,4})\??\s*(-|—|–|to)\s*(?<last>\d{1})\??([^-0-9].*)?$/)
+
+    # Integer value for latest year if we have "yyyy-y" pattern
+    # @return [Integer, nil] yyyy if date_str matches pattern; nil otherwise
+    def hyphen_1digit_latest_year(date_str)
+      matches = date_str.match(YYYY_HYPHEN_Y_REGEX)
+      return unless matches
+
+      first = Regexp.last_match(:first)
+      decade = first[0..-2] # whatever is before the last digit
+      last = "#{decade}#{Regexp.last_match(:last)}"
+      last.to_i if ParseDate.year_range_valid?(first.to_i, last.to_i)
+    end
+
     YYUU = '\\d{1,2}[u\\-]{2}'
     YYuu_HYPHEN_YYuu_REGEX =
-      Regexp.new("(?<first>#{YYUU})\\??\\s*[-—]\\s*(?<last>#{YYUU})\\??([^u\\-]|$)??", REGEX_OPTS)
+      Regexp.new("(?<first>#{YYUU})\\??\\s*(-|—|–|to)\\s*(?<last>#{YYUU})\\??([^u\\-]|$)??", REGEX_OPTS)
 
     # Integer value for latest year if we have "yyuu-yyuu" pattern
     # @return [Integer, nil] yyyy if date_str matches pattern; nil otherwise
@@ -139,7 +199,8 @@ class ParseDate
     end
 
     # NOTE: some actual data seemed to have a diff hyphen char. (slightly longer)
-    YY_YY_CENTURY_REGEX = Regexp.new(/(?<first>\d{1,2})[a-z]{2}?\s*(-|–|or)\s*(?<last>\d{1,2})[a-z]{2}?\s+centur.*/im)
+    YY_YY_CENTURY_REGEX =
+      Regexp.new(/(?<first>\d{1,2})[a-z]{2}?\s*(-|–|–|or|to)\s*(?<last>\d{1,2})[a-z]{2}?\s+centur.*/im)
 
     # Integer value for latest year if we have nth-nth century pattern
     # @return [Integer, nil] yy99 if date_str matches pattern; nil otherwise
@@ -176,13 +237,13 @@ class ParseDate
     # looks for -yyyy at beginning of date_str and returns if found
     # @return [String, nil] negative 4 digit year (e.g. -1865) if date_str has -yyyy, nil otherwise
     def negative_first_four_digits(date_str)
-      Regexp.last_match(1) if date_str.match(/^(\-\d{4})/)
+      Regexp.last_match(1) if date_str.match(/^(-\d{4})/)
     end
 
     # looks for -yyyy after hyphen and returns if found
     # @return [String, nil] negative 4 digit year (e.g. -1865) if date_str has -yyyy - -yyyy, nil otherwise
     def negative_4digits_after_hyphen(date_str)
-      Regexp.last_match(1) if date_str.match(/\-\d{4}\s*\-\s*(\-\d{4})/)
+      Regexp.last_match(1) if date_str.match(/-\d{4}\s*(?:-|–|–|or|to)\s*(-\d{4})/)
     end
 
     # looks for 4 consecutive digits in date_str and returns first occurrence if found
@@ -198,7 +259,7 @@ class ParseDate
     #   1/1/27  ->  1927
     # @return [String, nil] 4 digit year (e.g. 1865, 0950) if date_str matches pattern, nil otherwise
     def year_from_mm_dd_yy(date_str)
-      slash_matches = date_str.match(/\d{1,2}\/\d{1,2}\/\d{2}/)
+      slash_matches = date_str.match(%r{\d{1,2}/\d{1,2}/\d{2}})
       if slash_matches
         date_obj = Date.strptime(date_str, '%m/%d/%y')
       else
@@ -211,6 +272,16 @@ class ParseDate
       nil # explicitly want nil if date won't parse
     end
 
+    DECADE_0S_REGEX = Regexp.new('(^|\D)\d{3}0\'?s($|\D)', REGEX_OPTS)
+
+    # last year of decade (as String) if we have:  yyy0s flavor pattern
+    # @return [String, nil] 4 digit year (e.g. 1869, 1959) if date_str matches pattern, nil otherwise
+    def last_year_for_0s_decade(date_str)
+      decade_matches = date_str.match(DECADE_0S_REGEX)
+      changed_to_nine = decade_matches.to_s.sub(/0'?s/, '9') if decade_matches
+      ParseDate.first_four_digits(changed_to_nine) if changed_to_nine
+    end
+
     DECADE_4CHAR_REGEX = Regexp.new('(^|\D)\d{3}[u\-?x]($|\D)', REGEX_OPTS)
 
     # first year of decade (as String) if we have:  yyyu, yyy-, yyy? or yyyx pattern
@@ -237,12 +308,14 @@ class ParseDate
 
     # first year of century if we have:  yyuu, yy--, yy--? or xxth century pattern; handles B.C.
     # @return [Integer, nil] yy00 if date_str matches pattern, nil otherwise
+    # rubocop:disable Metrics/AbcSize
     def first_year_for_century(date_str)
       return Regexp.last_match(1).to_i * -100 - 99 if date_str.match(BC_CENTURY_REGEX)
       return Regexp.last_match(1).to_i * 100 if date_str.match(CENTURY_4CHAR_REGEX)
       return (Regexp.last_match(:first).to_i - 1) * 100 if date_str.match(YY_YY_CENTURY_REGEX)
       return (Regexp.last_match(1).to_i - 1) * 100 if date_str.match(CENTURY_WORD_REGEX)
     end
+    # rubocop:enable Metrics/AbcSize
 
     # last year of century if we have:  yyuu, yy--, yy--? or xxth century pattern
     # @return [Integer, nil] yy99 if date_str matches pattern, nil otherwise; also nil if B.C. in pattern
@@ -304,7 +377,7 @@ class ParseDate
     end
 
     FIRST_LAST_EARLY_NUMERIC_REGEX =
-      Regexp.new(/^(?<first>\-?\d{1,3})\??\s*(-|–|or)\s*(?<last>\-?\d{1,4})\??([^\du\-\[]|$)/im)
+      Regexp.new(/^(?<first>-?\d{1,3})\??\s*(-|–|–|or|to)\s*(?<last>-?\d{1,4})\??([^\du\-\[]|$)/im)
 
     # Integer value for latest year if we have early numeric year range or single early numeric year
     # @return [Integer, nil] year if date_str matches pattern; nil otherwise

data/lib/parse_date/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 class ParseDate
-  VERSION = '0.3.3'
+  VERSION = '0.4.3'
 end

data/parse_date.gemspec

@@ -32,8 +32,9 @@ Gem::Specification.new do |spec|
 
   spec.add_development_dependency 'bundler', '~> 2.0'
   spec.add_development_dependency 'pry-byebug'
-  spec.add_development_dependency 'rake', '~> 10.0'
+  spec.add_development_dependency 'rake', '~> 13.0.3'
   spec.add_development_dependency 'rspec', '~> 3.0'
-  spec.add_development_dependency 'rubocop', '~> 0.74.0'
-  spec.add_development_dependency 'simplecov'
+  spec.add_development_dependency 'rubocop', '~> 1.12'
+  spec.add_development_dependency 'rubocop-rspec'
+  spec.add_development_dependency 'simplecov', '~> 0.17.1' # CodeClimate cannot use SimpleCov >= 0.18.0 for generating test coverage
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: parse_date
 version: !ruby/object:Gem::Version
-  version: 0.3.3
+  version: 0.4.3
 platform: ruby
 authors:
 - Naomi Dushay
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2019-10-28 00:00:00.000000000 Z
+date: 2021-03-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: zeitwerk
@@ -58,14 +58,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: 13.0.3
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: 13.0.3
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -86,16 +86,16 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.74.0
+        version: '1.12'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.74.0
+        version: '1.12'
 - !ruby/object:Gem::Dependency
-  name: simplecov
+  name: rubocop-rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -108,6 +108,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: simplecov
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.17.1
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.17.1
 description: Get normalized date values for searching, faceting and display (e.g.
   in Solr search engine)
 email:
@@ -138,7 +152,7 @@ metadata:
   allowed_push_host: https://rubygems.org/
   homepage_uri: https://github.com/sul-dlss/parse_date
   source_code_uri: https://github.com/sul-dlss/parse_date
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -153,8 +167,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.0.3
-signing_key: 
+rubygems_version: 3.2.3
+signing_key:
 specification_version: 4
 summary: parse date values out of strings and normalize them
 test_files: []