Split Unicode emoji into their own script

* admin/notes/unicode: Describe how to update emoji for new Unicode release.
* admin/unidata/Makefile.in: Pass emoji-data.txt to
blocks.awk script.
* admin/unidata/README: Add pointer to emoji-data.txt file.
* admin/unidata/blocks.awk: Parse emoji-data.txt, add emoji codepoints
to the 'emoji' script (except for the ASCII ones).
* admin/unidata/emoji-data.txt: New file.
* etc/NEWS: Describe new 'emoji' script.
* etc/TODO: Update item about 'emoji' script.
* lisp/international/fontset.el (script-representative-chars): Add
'emoji' script.
(setup-default-fontset): Add 'emoji' script.  Use "Noto Color Emoji"
as default font for it.

admin/notes/unicode

@@ -16,13 +16,14 @@ Emacs uses the following files from the Unicode Character Database
   . IVD_Sequences.txt
   . NormalizationTest.txt
   . SpecialCasing.txt
+  . emoji-data.txt
   . BidiCharacterTest.txt
 
-First, the first 7 files need to be copied into admin/unidata/, and
+First, the first 8 files need to be copied into admin/unidata/, and
 the file https://www.unicode.org/copyright.html should be copied over
-copyright.html in admin/unidata (that file might need trailing
-whitespace removed before it can be committed to the Emacs
-repository).
+copyright.html in admin/unidata (that file and emoji-data.txt might
+need trailing whitespace removed before they can be committed to the
+Emacs repository).
 
 Then Emacs should be rebuilt for them to take effect.  Rebuilding
 Emacs updates several derived files elsewhere in the Emacs source
@@ -85,8 +86,34 @@ modified to follow suit.  If there's trailing whitespace in
 BidiCharacterTest.txt, it should be removed before committing the new
 version.
 
-etc/NEWS should be updated to announce the support for the new Unicode
-version.
+Visit "emoji-data.txt" with the rebuilt Emacs, and check that an
+appropriate font is being used for the emoji (by default Emacs uses
+"Noto Color Emoji").  Running the following command in that buffer
+will give you an idea of which codepoints are not supported by
+whichever font Emacs is using.
+
+(defun check-emoji-coverage (font-name-regexp)
+"Display a buffer containing emoji codepoints for which FONT-NAME is not used.
+This must be run from a buffer in the format of emoji-data.txt.
+FONT-NAME-REGEXP is checked using `string-match'."
+(interactive "MFont Name: ")
+(save-excursion
+(goto-char (point-min))
+(let (res char name ifont)
+  (while (re-search-forward "; Emoji [^(]+(\\(.\\)[).\uFE0F]" nil t)
+    (setq char (aref (match-string 1) 0))
+    (setq ifont (car (internal-char-font nil char)))
+    (when ifont
+      (setq name (font-xlfd-name ifont)))
+    (if (or (not ifont) (not (string-match font-name-regexp name)))
+        (setq res (concat (string char) res))))
+  (when res
+    (with-output-to-temp-buffer "*Check-Emoji-Coverage*"
+      (princ (format "Font not matching '%s' was used for the following characters:\n%s"
+                     font-name-regexp (reverse res))))))))
+
+Finally, etc/NEWS should be updated to announce the support for the
+new Unicode version.
 
 Problems, fixmes and other unicode-related issues
 -------------------------------------------------------------

admin/unidata/Makefile.in

@@ -81,8 +81,10 @@ charscript.el: ${unidir}/charscript.el
 
 blocks = ${srcdir}/blocks.awk
 
-${unidir}/charscript.el: ${srcdir}/Blocks.txt ${blocks}
-	$(AM_V_GEN)$(AWK) -f ${blocks} < $< > $@
+${unidir}/charscript.el: ${blocks}
+
+${unidir}/charscript.el: ${srcdir}/Blocks.txt ${srcdir}/emoji-data.txt
+	$(AM_V_GEN)$(AWK) -f ${blocks} $^ > $@
 
 
 .PHONY: clean bootstrap-clean distclean maintainer-clean gen-clean

admin/unidata/README

@@ -32,3 +32,7 @@ http://www.unicode.org/Public/UNIDATA/NormalizationTest.txt
 SpecialCasing.txt
 http://unicode.org/Public/UNIDATA/SpecialCasing.txt
 2017-04-20
+
+emoji-data.txt
+https://www.unicode.org/Public/14.0.0/ucd/emoji/emoji-data.txt
+2021-08-26

admin/unidata/blocks.awk

@@ -131,7 +131,7 @@ function name2alias(name   , w, w2) {
     return name
 }
 
-/^[0-9A-F]/ {
+FILENAME == "Blocks.txt" && /^[0-9A-F]/ {
     sep = index($1, "..")
     len = length($1)
     s = substr($1,1,sep-1)
@@ -202,6 +202,29 @@ function name2alias(name   , w, w2) {
     }
 }
 
+# The space after 'Emoji' is significant in the next two rules.
+# This purposely and deliberately excludes codepoints <= 00FF
+FILENAME == "emoji-data.txt" && /^00[0-9A-F]{2}.*; Emoji / {
+    next
+}
+FILENAME == "emoji-data.txt" && /^[0-9A-F].*; Emoji / {
+    sep = index($1, "..")
+    len = length($1)
+    if (sep > 0)  {
+        s = substr($1,1,sep-1)
+        e = substr($1,sep+2,len-sep-1)
+    } else {
+        s = $1
+        e = $1
+    }
+    $1 = ""
+    i++
+    start[i] = s
+    end[i] = e
+    alt[i] = "emoji"
+    name[i] = "Autogenerated emoji"
+}
+
 END {
     print ";;; charscript.el --- character script table  -*- lexical-binding:t -*-"
     print ";;; Automatically generated from admin/unidata/Blocks.txt"
@@ -223,6 +246,6 @@ END {
     print "    (or (memq (nth 2 elt) script-list)"
     print "	(setq script-list (cons (nth 2 elt) script-list))))"
     print "  (set-char-table-extra-slot char-script-table 0 (nreverse script-list)))"
-    print ""
+    print "\n"
     print "(provide 'charscript)"
 }

etc/NEWS

@@ -131,6 +131,19 @@ of files visited via 'C-x C-f' and other commands.
 ---
 ** Emacs now supports Unicode Standard version 14.0.
 
++++
+** New character script 'emoji' has been created.
+Various blocks of codepoints have been split out of the 'symbol'
+script into their own 'emoji' script to allow easier specification of
+their treatment.  Which codepoints are treated as emoji is derived
+from the Unicode specifications.  Also, Emacs will now use "Noto Color
+Emoji" by default for that script.  Use:
+
+(set-fontset-font t 'emoji
+                  '("My New Emoji Font" . "iso10646-1") nil 'prepend)
+
+to change the font used.
+
 +++
 ** New command 'execute-extended-command-for-buffer'.
 This new command, bound to 'M-S-x', works like
@@ -379,6 +392,9 @@ When customized to nil, it uses 'minibuffer-restore-windows' in
 'minibuffer-exit-hook' to remove only the window showing the
 "*Completions*" buffer.
 
+
+* Editing Changes in Emacs 28.1
+
 ---
 *** New variable 'redisplay-adhoc-scroll-in-resize-mini-windows'.
 Customizing it to nil will disable the ad-hoc auto-scrolling of

etc/TODO

@@ -405,7 +405,8 @@ punctuation characters, disregarding the fontsets, should be modified
 to exempt Emoji from this rule (since Emoji characters belong to the
 'symbol' script in Emacs), so that use-default-font-for-symbols would
 not have to be tweaked to have Emoji display by default with a capable
-font.
+font. (This has now been implemented, but only one font is currently
+considered, please augment the list).
 
 *** Consider changing the default display of Variation Selectors
 Emacs by default displays the Variation Selector (VS) codepoints not

lisp/international/fontset.el

@@ -278,7 +278,8 @@
 	(indic-siyaq-number #x1ec71)
 	(ottoman-siyaq-number #x1ed01)
 	(mahjong-tile #x1F000)
-	(domino-tile #x1F030)))
+	(domino-tile #x1F030)
+        (emoji #x1F300 #x1F600 #xFE0F)))
 
 (defvar otf-script-alist)
 
@@ -781,7 +782,8 @@
                     toto
 		    adlam
 		    mahjong-tile
-		    domino-tile))
+		    domino-tile
+                    emoji))
     (set-fontset-font "fontset-default"
 		      script (font-spec :registry "iso10646-1" :script script)
 		      nil 'append))
@@ -894,6 +896,9 @@
              (#x1FA00 . #x1FA6F)))	;; Chess Symbols
     (set-fontset-font "fontset-default" symbol-subgroup
                       '("Symbola" . "iso10646-1") nil 'prepend))
+  ;; This sets up the Emoji codepoints to use prettier fonts.
+  (set-fontset-font "fontset-default" 'emoji
+                    '("Noto Color Emoji" . "iso10646-1") nil 'prepend)
   ;; Box Drawing and Block Elements
   (set-fontset-font "fontset-default" '(#x2500 . #x259F)
                     '("FreeMono" . "iso10646-1") nil 'prepend)