hierogloss 0.0.5 → 0.0.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 013868a650a3464e07c17bcca8f37bcba3e52974
-  data.tar.gz: a02e8d67a85df5dcfedc9ae87447bf6604ed3587
+  metadata.gz: a5918c9d05821627262648470a2724043857a5a5
+  data.tar.gz: 22fabb89f839ddc8573645df993b639fa9006919
 SHA512:
-  metadata.gz: 9a362073542fa41bad251c4f0d38e9310b051fc617615036f9448d68e55a6855e0dd27e2e9c6d2489d7f0987245c07453775083c395817d23e8498d863b5801d
-  data.tar.gz: e17841bf61cb0b4b77680c498a6f93169a670246604b593b84b6b26a1ed3ec61b2edf5da8ca1aca437ebc7919cc138d088a8b0c892fa744b1ae23ed8ed36f52b
+  metadata.gz: ea4c5397bd99efae7e459555564fa4c31fb0c05c0f2dec7107fd193cae9430db28962743ef9ceeae20609f43b05214b21025e17764eb364ade718e900e8cf90c
+  data.tar.gz: 1471f3c7e4a6e9683d6610b4732a768c56736bbf5d3e05a71c4f385fe5a8824482656c27da22b268e3bbd555e17efab9be6308a6296de3ebacaf8697149378ec

data/lib/hierogloss/dictionary.rb

@@ -6,6 +6,18 @@ module Hierogloss
     DATA_DIR = File.join(File.dirname(__FILE__), '..', '..', 'data')
     MDC_MAPPING_PATH = File.join(DATA_DIR, "Unicode-MdC-Mapping-v1.utf8")
 
+    # Pattern identifying Gardiner codes (as opposed to phonetic ones.
+    GARDINER_REGEX = /\A[A-Z][0-9]+[A-Za-z]?([-:*\\].*)?\z/
+
+    # Convert variant signs to something that should work in JSesh.
+    def self.fix_gardiner(code)
+      if code =~ GARDINER_REGEX
+        code.upcase.sub(/\\R/, '\r')
+      else
+        code
+      end
+    end
+
     SIGN_TO_GARDINER = {}
     MDC_TO_SIGN = {}
     SIGN_TO_MDC = {}
@@ -15,11 +27,12 @@ module Hierogloss
         l.chomp!
         sign, hex, codes, remarks = l.split(/\t/, 4)
         for code in codes.split(/ /)
+          code = fix_gardiner(code)
           MDC_TO_SIGN[code] = sign
           # Unliterals.
           SIGN_TO_MDC[sign] = code if code.length == 1
           # Gardiner codes, and composite signs starting with Gardiner codes.
-          next unless code =~ /\A[A-Z][0-9]+([-:*].*)?\z/
+          next unless code =~ GARDINER_REGEX
           SIGN_TO_GARDINER[sign] = code
           SIGN_TO_MDC[sign] ||= code
         end

data/lib/hierogloss/version.rb

@@ -1,3 +1,3 @@
 module Hierogloss
-  VERSION = "0.0.5"
+  VERSION = "0.0.6"
 end

data/test/test_dictionary.rb

@@ -40,4 +40,13 @@ class TestDictionary < Minitest::Test
     assert_equal("A1", Hierogloss::Dictionary.sign_to_gardiner("𓀀"))
     assert_equal("D4", Hierogloss::Dictionary.sign_to_gardiner("𓁹"))
   end
+
+  def test_should_convert_variant_signs_to_gardiner
+    # These signs aren't in the Unicode table.
+    assert_equal("E8A", Hierogloss::Dictionary.sign_to_gardiner("𓃚"))
+    assert_equal("N33A", Hierogloss::Dictionary.sign_to_gardiner("𓈓"))
+    assert_equal("N35A", Hierogloss::Dictionary.sign_to_gardiner("𓈗"))
+    assert_equal("P1A", Hierogloss::Dictionary.sign_to_gardiner("𓊜"))
+    assert_equal("Y1\\r1", Hierogloss::Dictionary.sign_to_gardiner("𓏜"))
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hierogloss
 version: !ruby/object:Gem::Version
-  version: 0.0.5
+  version: 0.0.6
 platform: ruby
 authors:
 - Eric Kidd
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-01-25 00:00:00.000000000 Z
+date: 2014-01-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: kramdown