Correct letter pattern

This previously copied the values from the Graphviz DOT lexer. That didn't actually work, since those are expressed in octal, and also don't refer to Unicode code points. According to a comment in the lexer, Graphviz will accept ASCII letters, underscore, and all non-ASCII characters for its "LETTER" class. (Non-ASCII characters in this case starts at 0200 or 0x80, non-breaking space.)
mdaines · Aug 1, 2023 · 3156202 · 3156202
1 parent 39099f6
commit 3156202
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 1 deletion.
diff --git a/packages/lang-dot/src/dot.grammar b/packages/lang-dot/src/dot.grammar
@@ -111,7 +111,7 @@ ConcatString {
   blockCommentRest { ![*] blockCommentRest | "*" blockCommentAfterStar }
   blockCommentAfterStar { "/" | "*" blockCommentAfterStar | ![/*] blockCommentRest }
 
-  letter { $[A-Za-z_\u{200}-\u{377}] }
+  letter { $[A-Za-z_$\u{80}-\u{10ffff}] }
 
   digit { $[0-9] }
 

diff --git a/packages/lang-dot/test/graph.txt b/packages/lang-dot/test/graph.txt
@@ -176,6 +176,27 @@ Graph(Header(...), Body(
 ))
 
 
+# Names
+
+graph {
+  a
+  A_1
+  café
+
+  図
+}
+
+==>
+
+Graph(Header(...), Body(
+  SimpleStatement(Node(Name(...))),
+  SimpleStatement(Node(Name(...))),
+  SimpleStatement(Node(Name(...))),
+  SimpleStatement(Node(Name(...))),
+  SimpleStatement(Node(Name(...)))
+))
+
+
 # Numbers
 
 graph 123 {