data/generate_constants.rb in unicode-emoji-3.8.0 vs data/generate_constants.rb in unicode-emoji-4.0.0

@@ -67,19 +67,27 @@
     character_class(rangify(ords))
   end
 end
 
 def compile(emoji_character:, emoji_modifier:, emoji_modifier_base:, emoji_component:, emoji_presentation:, text_presentation:, picto:, picto_no_emoji:)
+  visual_component = pack_and_join(VISUAL_COMPONENT)
+
   emoji_presentation_sequence = \
     join(
       text_presentation + pack(EMOJI_VARIATION_SELECTOR),
       emoji_presentation + "(?!" + pack(TEXT_VARIATION_SELECTOR) + ")" + pack(EMOJI_VARIATION_SELECTOR) + "?",
     )
 
   non_component_emoji_presentation_sequence = \
     "(?!" + emoji_component + ")" + emoji_presentation_sequence
 
+  basic_emoji = \
+    join(
+      non_component_emoji_presentation_sequence,
+      visual_component,
+    )
+
   text_keycap_sequence = \
     pack_and_join(EMOJI_KEYCAPS) + pack(EMOJI_KEYCAP_SUFFIX)
 
   text_presentation_sequence = \
     join(
@@ -167,69 +175,77 @@
     join(
       emoji_rgi_zwj_sequence,
       emoji_rgi_tag_sequence,
       emoji_valid_flag_sequence,
       emoji_core_sequence,
+      visual_component,
     )
 
   emoji_rgi_sequence_include_text = \
     join(
       emoji_rgi_zwj_sequence,
       emoji_rgi_tag_sequence,
       emoji_valid_flag_sequence,
       emoji_core_sequence,
+      visual_component,
       text_emoji,
     )
 
   emoji_rgi_include_mqe_sequence = \
     join(
       emoji_rgi_include_mqe_zwj_sequence,
       emoji_rgi_tag_sequence,
       emoji_valid_flag_sequence,
       emoji_core_sequence,
+      visual_component,
     )
 
   emoji_rgi_include_mqe_uqe_sequence = \
     join(
       emoji_rgi_include_mqe_uqe_zwj_sequence,
       text_emoji, # also uqe
       emoji_rgi_tag_sequence,
       emoji_valid_flag_sequence,
       emoji_core_sequence,
+      visual_component,
     )
 
   emoji_valid_sequence = \
     join(
       emoji_valid_zwj_sequence,
       emoji_valid_tag_sequence,
       emoji_valid_flag_sequence,
       emoji_core_sequence,
+      visual_component,
     )
 
   emoji_valid_sequence_include_text = \
     join(
       emoji_valid_zwj_sequence,
       emoji_valid_tag_sequence,
       emoji_valid_flag_sequence,
       emoji_core_sequence,
+      visual_component,
       text_emoji,
     )
 
   emoji_well_formed_sequence = \
     join(
       emoji_valid_zwj_sequence,
       emoji_well_formed_tag_sequence,
       emoji_well_formed_flag_sequence,
       emoji_core_sequence,
+      visual_component,
     )
 
   emoji_well_formed_sequence_include_text = \
     join(
       emoji_valid_zwj_sequence,
       emoji_well_formed_tag_sequence,
       emoji_well_formed_flag_sequence,
       emoji_core_sequence,
+      visual_component,
       text_emoji,
     )
 
   emoji_possible_modification = \
     join(
@@ -277,23 +293,30 @@
 
   # Quick test which might lead to false positves
   # See https://www.unicode.org/reports/tr51/#EBNF_and_Regex
   regexes[:REGEX_POSSIBLE] = Regexp.compile(emoji_possible)
 
-  # Matches only basic single, non-textual emoji, ignores "components" like modifiers or simple digits
-  regexes[:REGEX_BASIC] = Regexp.compile(non_component_emoji_presentation_sequence)
+  # Matches only basic single, non-textual emoji, ignores some components like simple digits
+  regexes[:REGEX_BASIC] = Regexp.compile(basic_emoji)
 
-  # Matches only basic single, textual emoji, ignores "components" like modifiers or simple digits
+  # Matches only basic single, textual emoji, ignores components like modifiers or simple digits
   regexes[:REGEX_TEXT] = Regexp.compile(text_emoji)
 
-  # Same as \p{Emoji} - to be removed or renamed
-  regexes[:REGEX_ANY] = Regexp.compile(emoji_character)
+  # Export regexes for Emoji properties so they can be used with newer Unicode than Ruby's
+  regexes[:REGEX_PROP_EMOJI] = Regexp.compile(emoji_character)
+  regexes[:REGEX_PROP_MODIFIER] = Regexp.compile(emoji_modifier)
+  regexes[:REGEX_PROP_MODIFIER_BASE] = Regexp.compile(emoji_modifier_base)
+  regexes[:REGEX_PROP_COMPONENT] = Regexp.compile(emoji_component)
+  regexes[:REGEX_PROP_PRESENTATION] = Regexp.compile(emoji_presentation)
 
+  # Same goes for ExtendedPictographic
   regexes[:REGEX_PICTO] = Regexp.compile(picto)
-
   regexes[:REGEX_PICTO_NO_EMOJI] = Regexp.compile(picto_no_emoji)
 
+  # Emoji keycaps
+  regexes[:REGEX_EMOJI_KEYCAP] = Regexp.compile(emoji_keycap_sequence)
+
   regexes
 end
 
 regexes = compile(
   emoji_character:      pack_and_join(EMOJI_CHAR),
@@ -311,10 +334,10 @@
   emoji_character:      "\\p{Emoji}",
   emoji_modifier:       "\\p{EMod}",
   emoji_modifier_base:  "\\p{EBase}",
   emoji_component:      "\\p{EComp}",
   emoji_presentation:   "\\p{EPres}",
-  text_presentation:    "\\p{Emoji}(?<!\\p{EPres})",
+  text_presentation:    "[\\p{Emoji}&&\\P{EPres}]",
   picto:                "\\p{ExtPict}",
-  picto_no_emoji:       "\\p{ExtPict}(?<!\\p{Emoji})"
+  picto_no_emoji:       "[\\p{ExtPict}&&\\P{Emoji}]"
 )
 write_regexes(native_regexes, File.expand_path("../lib/unicode/emoji/generated_native", __dir__))