greeklish_iso843 0.3.0 → 0.4.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3d4c26155ee5e66d06b9a44cc5a452bcac2ed78a9f5ed889c6ec9540a9fc5853
-  data.tar.gz: 22c733cff1a29127b39a36ca69caa9cd73d96b6e55009ebb017d0fa517484d01
+  metadata.gz: b5d2ef8c9b648af399cd8cbdae6371173a871ae5107031fc9f00d7cff26b81b1
+  data.tar.gz: 051d809e9efac4ed60176798ce46412f63bdfbb8e8b3c4d8ddb7ef721cdde690
 SHA512:
-  metadata.gz: 46c846c9b550c56a8530c232bc4db7d1eb561484d716758e62a23a4dd50956f63a29df91c99410ef7f9b4028f315660c42b21e4607b5200f3a9ea254dd2a2cc2
-  data.tar.gz: 802ac4eef5af569b492991de5316e7390cf2377acd05d04f3dbcacc8b0adba528d9835be250fd34b64231c9d7ef2917e9f6e614c22cb2289e9fae3b6f886417d
+  metadata.gz: 77fdf6af41076c632a35cb0bce2b1397c14cf50d6961959ea13b45eb6d351219a9b8a1be0f293f56b2d1a99735af8fbc2f5fa713704ac518fcb70fdc202bd050
+  data.tar.gz: 4bad1dea32c72a9014ff8a077cd052bd9ec2a0b520ad1a9733e93ce0999ddc875b7d2711150a053d4c14df5909ed0b5d2952fb5af15b317c229e43a7bab7901e

data/CHANGELOG.md

@@ -7,6 +7,44 @@ The format is based on [Keep a Changelog][] and this project adheres to
 
 ## [Unreleased][]
 
+## [0.4.3][] - 2022-01-21
+
+- [#3][]: Fix `ευς`, `αυς` and `ηυς` throwing `UnhandledCaseError`
+  ([@cherouvim][])
+
+[#3]: https://github.com/agorf/greeklish_iso843/issues/3
+[@cherouvim]: https://github.com/cherouvim
+
+## [0.4.2][] - 2021-05-29
+
+### Changed
+
+- Do not raise `UnhandledCaseError` if next character is a symbol when
+  converting a pair that ends with `υ` to `v` or `f`. For example, `άνευ` would
+  correctly transliterate to `anef` but `άνευ.` would raise due to `.` at the
+  end.
+- Fix case for leading `Θ`, `Χ` and `Ψ` when the word is capitalized. For
+  example, `ψάθα` would correctly transliterate to `psatha`, but `Ψάθα` would
+  incorrectly transliterate to `PSatha`. Now it is transliterated to `Psatha`.
+
+## [0.4.1][] - 2021-04-29
+
+### Changed
+
+- Do not raise `UnhandledCaseError` if next character is whitespace when
+  converting a pair that ends with `υ` to `v` or `f`. For example, `άνευ` would
+  correctly transliterate to `anef` but `άνευ λόγου` would raise due to the
+  whitespace after `ευ`.
+
+## [0.4.0][] - 2021-04-28
+
+### Changed
+
+- Return match as-is if it cannot be handled instead of crashing
+- Raise `UnhandledCaseError` in case of an unhandled case when converting pair
+  for υ or φ (should never happen)
+- Rewrite a significant part of the code to make it more readable
+
 ## [0.3.0][] - 2021-03-06
 
 ### Changed
@@ -34,4 +72,8 @@ Initial release.
 [Semantic Versioning]: http://semver.org/spec/v2.0.0.html
 [0.2.0]: https://github.com/agorf/greeklish_iso843/compare/0.1.0...0.2.0
 [0.3.0]: https://github.com/agorf/greeklish_iso843/compare/0.2.0...0.3.0
-[Unreleased]: https://github.com/agorf/greeklish_iso843/compare/0.3.0...HEAD
+[0.4.0]: https://github.com/agorf/greeklish_iso843/compare/0.3.0...0.4.0
+[0.4.1]: https://github.com/agorf/greeklish_iso843/compare/0.4.0...0.4.1
+[0.4.2]: https://github.com/agorf/greeklish_iso843/compare/0.4.1...0.4.2
+[0.4.3]: https://github.com/agorf/greeklish_iso843/compare/0.4.2...0.4.3
+[Unreleased]: https://github.com/agorf/greeklish_iso843/compare/0.4.3...HEAD

data/README.md

@@ -1,12 +1,17 @@
 # greeklish_iso843 [![Gem](https://img.shields.io/gem/v/greeklish_iso843?color=blue)](https://rubygems.org/gems/greeklish_iso843/) [![RuboCop](https://github.com/agorf/greeklish_iso843/actions/workflows/rubocop.yml/badge.svg)](https://github.com/agorf/greeklish_iso843/actions/workflows/rubocop.yml) [![Tests](https://github.com/agorf/greeklish_iso843/actions/workflows/tests.yml/badge.svg)](https://github.com/agorf/greeklish_iso843/actions/workflows/tests.yml)
 
 A [Ruby][] library that converts Greek text to [Greeklish][], conforming to the
-[ISO 843][] (ELOT 743) standard used by the Greek State.
+[ISO 843][] standard (based on the Greek standard ELOT 743 or ΕΛΟΤ 743) used by
+the Greek state.
 
 [Ruby]: https://www.ruby-lang.org/en/
 [Greeklish]: https://en.wikipedia.org/wiki/Greeklish
 [ISO 843]: https://www.iso.org/standard/5215.html
 
+## Demo
+
+An online service that uses this Gem is available at <https://greeklish.xyz>
+
 ## Installation
 
 As a [Gem][]:
@@ -66,7 +71,8 @@ verbose output.
 
 ## Acknowledgements
 
-This library was based on [an official, buggy implementation in JavaScript][js].
+This library was initially based on an official, buggy [implementation][js] in
+JavaScript by the Hellenic Police.
 
 [js]: http://www.passport.gov.gr/passports/GrElotConverter/GrElotConverter.html
 

data/lib/greeklish_iso843/greek_text.rb

@@ -5,8 +5,12 @@ class GreeklishIso843::GreekText
 
   GREEK_VOWELS = 'αάεέηήιίϊΐοόυύϋΰωώ'.freeze
 
+  PAIRS_FOR_V_OR_F = %w[αυ αύ ευ εύ ηυ ηύ].freeze
+
   GREEK_LETTERS_AFTER_V = "#{GREEK_VOWELS}βγδζλμνρ".freeze
 
+  GREEK_LETTERS_AFTER_F = 'θκξπσςτφχψ'.freeze
+
   REPLACEMENTS = {
     'αι' => 'ai',
     'αί' => 'ai',
@@ -68,7 +72,11 @@ class GreeklishIso843::GreekText
     'ώ' => 'o'
   }.freeze
 
-  REPLACEMENTS_REGEXP = /#{REPLACEMENTS.keys.join('|')}/i.freeze
+  REPLACEMENT_KEYS_REGEXP = /#{REPLACEMENTS.keys.join('|')}/i.freeze
+
+  class Error < StandardError; end
+
+  class UnhandledCaseError < Error; end
 
   attr_reader :text
 
@@ -81,66 +89,115 @@ class GreeklishIso843::GreekText
   end
 
   def to_greeklish
-    text.gsub(REPLACEMENTS_REGEXP) do |match|
-      match_data = Regexp.last_match
-      next_char = match_data.post_match[0]&.downcase
-      greeklish = REPLACEMENTS[match.downcase]
-      greek = match
-
-      if greeklish
-        greek = match + next_char.to_s
-      else
-        greeklish = convert_to_greeklish(match, match_data, next_char)
-      end
-
-      fix_case(greeklish, greek)
+    text.gsub(REPLACEMENT_KEYS_REGEXP) do |match|
+      greeklish = REPLACEMENTS[match.downcase] ||
+        convert_pair(match, Regexp.last_match)
+
+      next match if greeklish.nil? # Unhandled case. Return as-is.
+
+      fix_case(greeklish, match, Regexp.last_match)
     end
   end
 
-  private def capitalized?(text)
-    GREEK_UPPER[text[0]]
+  private def convert_pair(match, match_data)
+    return 'ts' if match.casecmp?('τς')
+
+    next_char = match_data.post_match[0]&.downcase
+
+    if match.casecmp?('μπ')
+      prev_char = match_data.pre_match[-1]&.downcase
+      return convert_mp_or_b(prev_char, next_char)
+    end
+
+    if PAIRS_FOR_V_OR_F.any? { |pair| match.casecmp?(pair) }
+      return convert_pair_for_v_or_f(match, next_char)
+    end
+
+    match # Unhandled case. Return as-is.
   end
 
-  private def lowercase_char?(char)
-    GREEK_LOWER[char]
+  private def fix_case(greeklish, match, match_data)
+    if !uppercase?(match[0])
+      return greeklish
+    end
+
+    if match.size == 1
+      return fix_case_single_letter_match(greeklish, match_data)
+    end
+
+    if match.size == 2
+      return fix_case_two_letter_match(greeklish, match)
+    end
+
+    raise UnhandledCaseError
   end
 
-  private def fix_case(greeklish, greek)
-    return greeklish if !capitalized?(greek[0])
+  private def fix_case_single_letter_match(greeklish, match_data)
+    if greeklish.size == 1
+      return greeklish.upcase
+    end
+
+    if greeklish.size == 2 # match is one of Θ, Χ, Ψ
+      return fix_case_th_ch_ps(greeklish, match_data)
+    end
+
+    raise UnhandledCaseError
+  end
+
+  private def fix_case_two_letter_match(greeklish, match)
+    if uppercase?(match[1])
+      return greeklish.upcase
+    end
+
+    greeklish[0].upcase + greeklish[1].to_s
+  end
 
-    if greek.length == 1 || capitalized?(greek[1])
-      greeklish.upcase
-    else
-      greeklish[0].upcase + greeklish[1..-1]
+  private def fix_case_th_ch_ps(greeklish, match_data)
+    next_char = match_data.post_match[0]
+
+    if next_char.nil? ||
+        next_char !~ REPLACEMENT_KEYS_REGEXP ||
+        !uppercase?(next_char)
+      return greeklish[0].upcase + greeklish[1].to_s
     end
+
+    greeklish.upcase
   end
 
   private def convert_mp_or_b(prev_char, next_char)
-    if prev_char && lowercase_char?(prev_char) && # *μπ
-        next_char && lowercase_char?(next_char) # and μπ*
+    if prev_char && lowercase?(prev_char) && # *μπ
+        next_char && lowercase?(next_char) # and μπ*
       'mp'
     else # μπ* or *μπ
       'b'
     end
   end
 
-  private def convert_v_or_f(next_char)
-    if next_char && GREEK_LETTERS_AFTER_V[next_char]
-      'v'
-    else
-      'f'
+  private def convert_pair_for_v_or_f(match, next_char)
+    v_or_f = convert_v_or_f(next_char)
+
+    if v_or_f.nil?
+      raise UnhandledCaseError # Should never happen
     end
+
+    REPLACEMENTS[match[0].downcase] + v_or_f
   end
 
-  private def convert_to_greeklish(match, match_data, next_char)
-    return 'ts' if match.casecmp?('τς')
+  private def uppercase?(char)
+    GREEK_UPPER[char]
+  end
 
-    if match.casecmp?('μπ')
-      prev_char = match_data.pre_match[-1]&.downcase
-      return convert_mp_or_b(prev_char, next_char)
+  private def lowercase?(char)
+    GREEK_LOWER[char]
+  end
+
+  private def convert_v_or_f(next_char)
+    if next_char.nil? ||
+        GREEK_LETTERS_AFTER_F[next_char] ||
+        next_char !~ REPLACEMENT_KEYS_REGEXP
+      return 'f'
     end
 
-    # αυ αύ ευ εύ ηυ ηύ
-    REPLACEMENTS[match[0].downcase] + convert_v_or_f(next_char)
+    'v' if GREEK_LETTERS_AFTER_V[next_char]
   end
 end

data/lib/greeklish_iso843/version.rb

@@ -1,3 +1,3 @@
 module GreeklishIso843
-  VERSION = '0.3.0'.freeze
+  VERSION = '0.4.3'.freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: greeklish_iso843
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.3
 platform: ruby
 authors:
 - Angelos Orfanakos
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-03-06 00:00:00.000000000 Z
+date: 2022-01-21 00:00:00.000000000 Z
 dependencies: []
 description: 
 email: rubygems@angelos.dev
@@ -28,7 +28,8 @@ files:
 homepage: https://github.com/agorf/greeklish_iso843
 licenses:
 - MIT
-metadata: {}
+metadata:
+  rubygems_mfa_required: 'true'
 post_install_message: 
 rdoc_options: []
 require_paths: