Update ICU to 58.1

source/data/translit/Grek_Latn.txt

Index: source/data/translit/Grek_Latn.txt
diff --git a/source/data/translit/Greek_Latin.txt b/source/data/translit/Grek_Latn.txt
similarity index 75%
rename from source/data/translit/Greek_Latin.txt
rename to source/data/translit/Grek_Latn.txt
index 5118c6fe6e56948a533e97bd5872cb2572017806..a682aff6712b54f27e2c5a30ef698b9b7b762834 100644
--- a/source/data/translit/Greek_Latin.txt
+++ b/source/data/translit/Grek_Latn.txt
@@ -1,18 +1,30 @@
-# ***************************************************************************
-# *
-# *  Copyright (C) 2004-2015, International Business Machines
-# *  Corporation; Unicode, Inc.; and others.  All Rights Reserved.
-# *
-# ***************************************************************************
-# File: Greek_Latin.txt
-# Generated from CLDR 
+# © 2016 and later: Unicode, Inc. and others.
+# License & terms of use: http://www.unicode.org/copyright.html#License
 #
+# File: Grek_Latn.txt
+# Generated from CLDR
+#
+
+# Rules are predicated on running NFD first, and NFC afterwards
+# :: [\u0000-\u007F \u0370-Ͽ [:Greek:] [:nonspacing mark:]] ;
+# MINIMAL FILTER GENERATED FOR: Greek-Latin
 :: [;µ·ÄËÏÖÜäëïöüÿ-āĒ-ēĪ-īŌ-ōŪ-ūŸǕ-ǜǞ-ǣǬ-ǭȪ-ȭȰ-ȳ\u0304\u0308\u0313-\u0314\u0342-\u0345ͺ;Ά-ΊΌΎ-ΡΣ-ώϐ-ϗϛϝϟϡϣϥϧϩϫϭϯ-ϵϷ-\u07FBЁЇёїӒ-ӓӚ-ӟӢ-ӧӪ-ӱӴ-ӵӸ-ӹḔ-ḗḠ-ḡḦ-ḧḮ-ḯḸ-ḹṎ-ṓṜ-ṝṺ-ṻẄ-ẅẌ-ẍẗἀ-ἕἘ-Ἕἠ-ὅὈ-Ὅὐ-ὗὙὛὝὟ-ώᾀ-ᾴᾶ-ᾼι῁-ῄῆ-ῌ῏-ΐῖ-Ί῟-Ῥῲ-ῴῶ-ῼΩϹ] ;
 :: NFD (NFC) ;
+# TEST CASES
+# Ὀλίγοι ἔμφονες πολλῶν ἀφρόνων φοβερώτεροι — Πλάτωνος
+# ᾂ ᾒ ᾢ ᾃ ᾓ ᾣ
+# ᾳ ῃ ῳ ὃ ὄ
+# ὠς ὡς ὢς ὣς
+# Ὠς Ὡς Ὢς Ὣς
+# ὨΣ ὩΣ ὪΣ ὫΣ
+# Ạ, ạ, Ẹ, ẹ, Ọ, ọ
+# Useful variables
 $lower = [[:latin:][:greek:] & [:Ll:]];
 $glower = [[:greek:] & [:Ll:]];
 $upper = [[:latin:][:greek:] & [:Lu:]] ;
 $accent = [:M:] ;
+# NOTE: restrict to just the Greek & Latin accents that we care about
+# TODO: broaden out once interation is fixed
 $accentMinus = [ [\u0300-\u0345] & [:M:] - [\u0338]] ;
 $macron = \u0304 ;
 $ddot = \u0308 ;
@@ -37,18 +49,27 @@ $beforeLetter = [[:M:]\']* [:L:] ;
 $beforeLower = $accent * $lower ;
 $notLetter = [^[:L:][:M:]] ;
 $under = \u0331;
+# Fix punctuation
+# preserve original
 \: ↔ \: $under ;
 \? ↔ \? $under ;
 \; ↔ \? ;
 · ↔ \: ;
+# CIRCUMFLEX: convert greek circumflex to normal one. Could use tilde or inverted breve
 \u0342 ↔ \u0302 ;
+# IOTA: convert iota subscript to iota
+# first make previous alpha long!
 $accent_minus = [[$accent]-[$iotasub$macron]];
 Α } $accent_minus * $iotasub → | Α $macron ;
 α } $accent_minus * $iotasub → | α $macron ;
+# now convert to uppercase if after uppercase, ow to lowercase
 $upper $accent * { $iotasub → I ;
 $iotasub → i ;
 | $1 $iotasub ← ($evowel $macron $accentMinus *) i ;
 | $1 $iotasub ← ($evowel $macron $accentMinus *) I ;
+# BREATHING
+# Convert rough breathing to h, and move before letters.
+# Make A ` x = → H a x
 Α ($macron?) $rough } $beforeLower → H | α $1;
 Ε $rough } $beforeLower → H | ε;
 Η $rough } $beforeLower → H | η ;
@@ -56,6 +77,7 @@ $iotasub → i ;
 Ο $rough } $beforeLower → H | ο ;
 Υ $rough } $beforeLower → H | υ ;
 Ω ($ddot?) $rough } $beforeLower → H | ω $1;
+# Make A x ` = → H a x
 Α ($glower $macron?) $rough → H | α $1 ;
 Ε ($glower) $rough → H | ε $1 ;
 Η ($glower) $rough → H | η $1 ;
@@ -63,14 +85,18 @@ $iotasub → i ;
 Ο ($glower) $rough → H | ο $1 ;
 Υ ($glower) $rough → H | υ $1 ;
 Ω ($glower  $ddot?) $rough → H | ω $1 ;
+#Otherwise, make x ` into h x and X ` into H X
 ($lcgvowel + $ddotmac? ) $rough → h | $1 ;
 ($gvowel + $ddotmac? ) $rough → H | $1 ;
+# Go backwards with H
 | $1 $rough ← h ($evowel $macron $ddot? $evowel2_i $macron?) ;
 | $1 $rough ← h ($evowel $ddot? $evowel2 $macron?) ;
 | $1 $rough ← h ($evowel $macron? $ddot?) ;
 | $1 $rough ← H ([AEIOUY] $macron $ddot? $evowel2_i $macron?) ;
 | $1 $rough ← H ([AEIOUY] $ddot? $evowel2 $macron?) ;
 | $1 $rough ← H ([AEIOUY] $macron? $ddot?) ;
+# titlecase, have to fix individually
+# in the future, we should add &uppercase() to make this easier
 | A $1 $rough ← H a ($macron  $ddot? $evowel2_i $macron?) ;
 | E $1 $rough ← H e ($macron  $ddot? $evowel2_i $macron?) ;
 | I $1 $rough ← H i ($macron  $ddot? $evowel2_i $macron?) ;
@@ -89,10 +115,18 @@ $iotasub → i ;
 | O $1 $rough ← H o ($macron? $ddot? ) ;
 | U $1 $rough ← H u ($macron? $ddot? ) ;
 | Y $1 $rough ← H y ($macron? $ddot? ) ;
+# Now do smooth
+#delete smooth breathing for Latin
 $smooth → ;
+# insert in Greek
+# the assumption is that all Marks are on letters.
 | $1 $smooth ← $notLetter { ([rR]) } [^hH$smooth$rough] ;
 | $1 $smooth ← $notLetter { ($evowel $macron? $evowel2 $macron?) } [^$smooth$rough] ;
 | $1 $smooth ← $notLetter { ($evowel $macron?) } [^$evowel2$smooth$rough] ;
+# TODO: preserve smooth/rough breathing if not
+# on initial vowel sequence
+# need to have these up here so the rules don't mask
+# remove now superfluous macron when returning
 Α ← A $macron ;
 α ← a $macron ;
 η ↔ e $macron ;
@@ -105,6 +139,7 @@ $smooth → ;
 ψ ↔ ps ;
 ω ↔ o $macron ;
 Ω ↔  O $macron;
+# NORMAL
 α ↔ a ;
 Α ↔ A ;
 β ↔ b ;
@@ -145,17 +180,24 @@ $smooth → ;
 Ρ $rough ↔ RH ;
 ρ ↔ r ;
 Ρ ↔ R ;
+# insert separator before things that turn into s
 [Pp] { } [ςσΣϷϸϺϻ] → \' ;
+# special S variants
 Ϸ ↔ S\u030C ; # Ϸ GREEK CAPITAL LETTER SHO Uppercase_Letter Grek - L
 ϸ ↔ s\u030C ; #ϸ GREEK SMALL LETTER SHO Lowercase_Letter Grek - L
 Ϻ ↔ S\u0302 ; # Ϻ GREEK CAPITAL LETTER SAN Uppercase_Letter Grek - L
 ϻ ↔ s\u0302 ; # ϻ GREEK SMALL LETTER SAN Lowercase_Letter Grek - L
+# underbar means exception
+# before a letter, initial
 ς } $beforeLetter ↔ s $underbar } $beforeLetter;
 σ } $beforeLetter ↔ s } $beforeLetter;
+# otherwise, after a letter = final
 $afterLetter { σ ↔ $afterLetter { s $underbar;
 $afterLetter { ς ↔ $afterLetter { s ;
+# otherwise (isolated) = initial
 ς ↔ s $underbar;
 σ ↔ s ;
+# [Pp] { Σ ↔ \'S ;
 Σ ↔ S ;
 τ ↔ t ;
 Τ ↔ T ;
@@ -166,6 +208,7 @@ $vowel { Υ ↔ U ;
 χ ↔ ch ;
 Χ } $beforeLower ↔ Ch ;
 Χ ↔ CH ;
+# Completeness for ASCII
 $ignore = [[:Mark:]''] * ;
 | k  ← c ;
 | ph ← f ;
@@ -187,6 +230,7 @@ $rough } $ignore [:UppercaseLetter:] → H ;
 $ignore [:UppercaseLetter:] { $rough → H ;
 $rough ← H ;
 $rough ↔ h ;
+# Completeness for Greek
 ϐ → | β ;
 ϑ → | θ ;
 ϒ → | Υ ;
@@ -201,7 +245,12 @@ $rough ↔ h ;
 ϵ → | ε ;
 µ → | μ ;
 ͺ → i;
+# delete any trailing ' marks used for roundtripping
 ← [Ππ] { \' } [Ss] ;
 ← [Νν] { \' } $egammaLike ;
 ::NFC (NFD) ;
+# ([\u0000-\u007F [:Latin:] [:Greek:] [:nonspacing mark:]]) ;
+# ([\u0000-\u007F · [:Latin:] [:nonspacing mark:]]) ;
+# MINIMAL FILTER GENERATED FOR: Latin-Greek BACKWARD
 :: ( [':?A-Za-zÀ-ÅÇ-ÏÑ-ÖÙ-Ýà-åç-ïñ-öù-ýÿ-ďĒ-ĥĨ-İĴ-ķĹ-ľŃ-ňŌ-őŔ-ťŨ-žƠ-ơƯ-ưǍ-ǜǞ-ǣǦ-ǰǴ-ǵǸ-țȞ-ȟȦ-ȳ\u0300-\u0337\u0339-\u0345΅-ΆΈ-ΊΌΎ-ΐΪ-ΰϊ-ώϓ-ϔЀ-ЁЃЇЌ-ЎЙйѐ-ёѓїќ-ўѶ-ѷӁ-ӂӐ-ӓӖ-ӗӚ-ӟӢ-ӧӪ-ӵӸ-ӹḀ-ẙẛẠ-ỹἀ-ἕἘ-Ἕἠ-ὅὈ-Ὅὐ-ὗὙὛὝὟ-ώᾀ-ᾴᾶ-ᾼ῁-ῄῆ-ΐῖ-Ί῝-΅ῲ-ῴῶ-ῼK-Å] ) ;
+