[gtksourceview] css.lang: Support Unicode and character escapes for IDs and classes

[Sébastien Wilmet <swilmet src gnome org>]

commit e6162a12fbbd39b300f64a219663f947b8b74f6b
Author: Jeffery To <jeffery to gmail com>
Date:   Thu May 10 04:23:53 2018 +0800

    css.lang: Support Unicode and character escapes for IDs and classes
    
    This adds an identifier regex that closely follows the rules for CSS
    identifiers, and uses this regex wherever CSS identifiers are used
    (ID/class/tag selectors, function names), so that Unicode characters and
    character escapes are correctly highlighted.
    
    This also removes the unicode-character-reference context, as Unicode
    character escapes do not appear anywhere else (except for inside
    strings, but then they are highlighted as part of the string).
    
    https://bugzilla.gnome.org/show_bug.cgi?id=796130

 data/language-specs/css.lang       |   63 +++++++++++++++++++++++++++++++----
 tests/syntax-highlighting/file.css |   11 ++++++
 2 files changed, 66 insertions(+), 8 deletions(-)
---
diff --git a/data/language-specs/css.lang b/data/language-specs/css.lang
index dbb4dd0..d60ea31 100644
--- a/data/language-specs/css.lang
+++ b/data/language-specs/css.lang
@@ -7,6 +7,7 @@
  Copyright (C) 2004 Scott Martin <scott coffeeblack org>
  Copyright (C) 2010 Patryk Zawadzki <patrys pld-linux org>
  Copyright (C) 2016 Tobias Schönberg <tobias47n9e gmail com>
+ Copyright (C) 2018 Jeffery To <jeffery to gmail com>
 
  GtkSourceView is free software; you can redistribute it and/or
  modify it under the terms of the GNU Lesser General Public
@@ -64,9 +65,56 @@
 
   <definitions>
 
-    <context id="unicode-character-reference" style-ref="others-2">
-      <match>\\([a-f0-9]{1,5}[ \t]|[a-f0-9]{6})</match>
-    </context>
+    <!-- https://drafts.csswg.org/css-syntax/#escaping -->
+    <define-regex id="escape-regex" extended="true">
+      (?:
+        \\                   # backslash
+        (?:
+          [^\n\r\f0-9a-f] |  # not newline or hex digit; or
+          [0-9a-f]{1,6} \s?  # 1-6 hex digits, trailing whitespace (not necessary in some cases)
+        )
+      )
+    </define-regex>
+
+    <!-- https://drafts.csswg.org/css-syntax/#non-ascii-code-point -->
+    <define-regex id="non-ascii-regex">[\x{80}-\x{10ffff}]</define-regex>
+
+    <!--
+    a combination of name-start code point and escape
+    https://drafts.csswg.org/css-syntax/#name-start-code-point
+    https://www.w3.org/TR/selectors-3/#lex
+    -->
+    <define-regex id="identifier-start-char-regex" extended="true">
+      (?:
+        [a-z_] |
+        \%{non-ascii-regex} |
+        \%{escape-regex}
+      )
+    </define-regex>
+
+    <!--
+    a combination of name code point and escape
+    https://drafts.csswg.org/css-syntax/#name-code-point
+    https://www.w3.org/TR/selectors-3/#lex
+    -->
+    <define-regex id="identifier-char-regex" extended="true">
+      (?:
+        [a-z0-9_-] |
+        \%{non-ascii-regex} |
+        \%{escape-regex}
+      )
+    </define-regex>
+
+    <!-- https://drafts.csswg.org/css-syntax/#ident-token-diagram -->
+    <define-regex id="identifier-regex" extended="true">
+      (?:
+        (?:
+          -- |
+          -? \%{identifier-start-char-regex}
+        )
+        \%{identifier-char-regex}*
+      )
+    </define-regex>
 
     <context id="selector-pseudo-elements-one-colon" style-ref="function">
       <prefix>::?</prefix>
@@ -132,15 +180,15 @@
     </context>
 
     <context id="selector-id" style-ref="selector-id">
-      <match>#[a-z][a-z0-9_-]*\b</match>
+      <match>#\%{identifier-regex}</match>
     </context>
 
     <context id="selector-class" style-ref="selector-class">
-      <match>\.[a-z][a-z0-9_-]*\b</match>
+      <match>\.\%{identifier-regex}</match>
     </context>
 
     <context id="selector-tagname" style-ref="selector-tagname">
-      <match>\b[a-z][a-z0-9_-]*\b</match>
+      <match>\%{identifier-regex}</match>
     </context>
 
     <context id="hexadecimal-color" style-ref="color">
@@ -1263,7 +1311,7 @@
     </context>
 
     <context id="function">
-      <start>[a-z][a-z0-9-]+\(</start>
+      <start>\%{identifier-regex}\(</start>
       <end>\)</end>
       <include>
         <context sub-pattern="0" where="start" style-ref="function"/>
@@ -1289,7 +1337,6 @@
         <context ref="def:single-quoted-string"/>
         <context ref="def:c-like-comment-multiline"/>
         <context ref="def:c-like-close-comment-outside-comment"/>
-        <context ref="unicode-character-reference"/>
         <context ref="selector-pseudo-elements-one-colon"/>
         <context ref="selector-pseudo-elements"/>
         <context ref="selector-pseudo-classes"/>
diff --git a/tests/syntax-highlighting/file.css b/tests/syntax-highlighting/file.css
index 7b52609..783a45c 100644
--- a/tests/syntax-highlighting/file.css
+++ b/tests/syntax-highlighting/file.css
@@ -36,6 +36,17 @@ INPUT::PLACEHOLDER {
     COLOR: REBECCAPURPLE;
 }
 
+#-- .--hello-world, /* two hyphens: https://stackoverflow.com/a/30822662 */
+.-❤#-español.♫#你好, /* leading hyphen, non-ASCII */
+span.\. + .\+\ space\@\>, /* special character escape */
+div#\E9 dition .\0000E9dition .motion_\e9motion, /* Unicode character escape */
+.\e33 div, /* trailing space terminates Unicode character escape */
+.\e33  div, /* need extra space to denote separate tokens */
+#-notvendor-specific,
+.\31 23 {
+    background: \u\72\l("unicode.png");
+}
+
 @import URL("fineprint.css");
 
 @media all {