Symbol notation

15 files changed, 137 insertions, 1 deletions

diff --git a/Cargo.lock b/Cargo.lock
index 222e1c98..fe522d84 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1010,6 +1010,11 @@ dependencies = [
 ]
 
 [[package]]
+name = "symmie"
+version = "0.1.0"
+source = "git+https://github.com/typst/symmie#8504bf7ec0d8996d160832c2724ba024ab6e988a"
+
+[[package]]
 name = "syn"
 version = "1.0.102"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -1175,6 +1180,7 @@ dependencies = [
  "roxmltree",
  "rustybuzz",
  "serde_json",
+ "symmie",
  "syntect",
  "ttf-parser 0.17.1",
  "typed-arena",
diff --git a/library/Cargo.toml b/library/Cargo.toml
index 92bf84a2..2410cb0c 100644
--- a/library/Cargo.toml
+++ b/library/Cargo.toml
@@ -21,6 +21,7 @@ rex = { git = "https://github.com/laurmaedje/ReX" }
 roxmltree = "0.14"
 rustybuzz = "0.5"
 serde_json = "1"
+symmie = { git = "https://github.com/typst/symmie" }
 syntect = { version = "5", default-features = false, features = ["default-syntaxes", "regex-fancy"] }
 ttf-parser = "0.17"
 typed-arena = "2"
diff --git a/library/src/lib.rs b/library/src/lib.rs
index 3543a672..d549c1cd 100644
--- a/library/src/lib.rs
+++ b/library/src/lib.rs
@@ -34,6 +34,7 @@ fn scope() -> Scope {
     // Text.
     std.def_node::<text::TextNode>("text");
     std.def_node::<text::LinebreakNode>("linebreak");
+    std.def_node::<text::SymbolNode>("symbol");
     std.def_node::<text::SmartQuoteNode>("smartquote");
     std.def_node::<text::StrongNode>("strong");
     std.def_node::<text::EmphNode>("emph");
@@ -173,6 +174,7 @@ fn items() -> LangItems {
         text: |text| text::TextNode(text).pack(),
         text_id: NodeId::of::<text::TextNode>(),
         text_str: |content| Some(&content.to::<text::TextNode>()?.0),
+        symbol: |notation| text::SymbolNode(notation).pack(),
         smart_quote: |double| text::SmartQuoteNode { double }.pack(),
         parbreak: || layout::ParbreakNode.pack(),
         strong: |body| text::StrongNode(body).pack(),
diff --git a/library/src/text/mod.rs b/library/src/text/mod.rs
index 47aaba36..5466637e 100644
--- a/library/src/text/mod.rs
+++ b/library/src/text/mod.rs
@@ -6,6 +6,7 @@ mod quotes;
 mod raw;
 mod shaping;
 mod shift;
+mod symbol;
 
 pub use self::deco::*;
 pub use self::misc::*;
@@ -13,6 +14,7 @@ pub use self::quotes::*;
 pub use self::raw::*;
 pub use self::shaping::*;
 pub use self::shift::*;
+pub use self::symbol::*;
 
 use std::borrow::Cow;
 
diff --git a/library/src/text/symbol.rs b/library/src/text/symbol.rs
new file mode 100644
index 00000000..cc12afb9
--- /dev/null
+++ b/library/src/text/symbol.rs
@@ -0,0 +1,35 @@
+use crate::prelude::*;
+use crate::text::TextNode;
+
+/// A symbol identified by symmie notation.
+#[derive(Debug, Hash)]
+pub struct SymbolNode(pub EcoString);
+
+#[node(Show)]
+impl SymbolNode {
+    fn construct(_: &Vm, args: &mut Args) -> SourceResult<Content> {
+        Ok(Self(args.expect("notation")?).pack())
+    }
+
+    fn field(&self, name: &str) -> Option<Value> {
+        match name {
+            "notation" => Some(Value::Str(self.0.clone().into())),
+            _ => None,
+        }
+    }
+}
+
+impl Show for SymbolNode {
+    fn show(&self, _: &mut Vt, this: &Content, _: StyleChain) -> SourceResult<Content> {
+        match symmie::get(&self.0) {
+            Some(c) => Ok(TextNode::packed(c)),
+            None => {
+                if let Some(span) = this.span() {
+                    bail!(span, "unknown symbol");
+                }
+
+                Ok(Content::empty())
+            }
+        }
+    }
+}
diff --git a/src/model/eval.rs b/src/model/eval.rs
index 43cdbbcc..1d942dd0 100644
--- a/src/model/eval.rs
+++ b/src/model/eval.rs
@@ -265,6 +265,7 @@ impl Eval for ast::MarkupNode {
             Self::Text(v) => v.eval(vm)?,
             Self::Escape(v) => (vm.items.text)(v.get().into()),
             Self::Shorthand(v) => v.eval(vm)?,
+            Self::Symbol(v) => v.eval(vm)?,
             Self::SmartQuote(v) => v.eval(vm)?,
             Self::Strong(v) => v.eval(vm)?,
             Self::Emph(v) => v.eval(vm)?,
@@ -306,6 +307,14 @@ impl Eval for ast::Shorthand {
     }
 }
 
+impl Eval for ast::Symbol {
+    type Output = Content;
+
+    fn eval(&self, vm: &mut Vm) -> SourceResult<Self::Output> {
+        Ok((vm.items.symbol)(self.get().clone()))
+    }
+}
+
 impl Eval for ast::SmartQuote {
     type Output = Content;
 
diff --git a/src/model/library.rs b/src/model/library.rs
index c890fef1..63bd5839 100644
--- a/src/model/library.rs
+++ b/src/model/library.rs
@@ -41,6 +41,8 @@ pub struct LangItems {
     pub text_id: NodeId,
     /// Get the string if this is a text node.
     pub text_str: fn(&Content) -> Option<&str>,
+    /// Symbol notation: `:arrow:l:`.
+    pub symbol: fn(notation: EcoString) -> Content,
     /// A smart quote: `'` or `"`.
     pub smart_quote: fn(double: bool) -> Content,
     /// A paragraph break.
diff --git a/src/syntax/ast.rs b/src/syntax/ast.rs
index 3c60acbb..c44fa2a0 100644
--- a/src/syntax/ast.rs
+++ b/src/syntax/ast.rs
@@ -85,6 +85,8 @@ pub enum MarkupNode {
     /// A shorthand for a unicode codepoint. For example, `~` for non-breaking
     /// space or `-?` for a soft hyphen.
     Shorthand(Shorthand),
+    /// Symbol notation: `:arrow:l:`.
+    Symbol(Symbol),
     /// A smart quote: `'` or `"`.
     SmartQuote(SmartQuote),
     /// Strong content: `*Strong*`.
@@ -119,6 +121,7 @@ impl AstNode for MarkupNode {
             SyntaxKind::Text(_) => node.cast().map(Self::Text),
             SyntaxKind::Escape(_) => node.cast().map(Self::Escape),
             SyntaxKind::Shorthand(_) => node.cast().map(Self::Shorthand),
+            SyntaxKind::Symbol(_) => node.cast().map(Self::Symbol),
             SyntaxKind::SmartQuote { .. } => node.cast().map(Self::SmartQuote),
             SyntaxKind::Strong => node.cast().map(Self::Strong),
             SyntaxKind::Emph => node.cast().map(Self::Emph),
@@ -141,6 +144,7 @@ impl AstNode for MarkupNode {
             Self::Text(v) => v.as_untyped(),
             Self::Escape(v) => v.as_untyped(),
             Self::Shorthand(v) => v.as_untyped(),
+            Self::Symbol(v) => v.as_untyped(),
             Self::SmartQuote(v) => v.as_untyped(),
             Self::Strong(v) => v.as_untyped(),
             Self::Emph(v) => v.as_untyped(),
@@ -224,6 +228,21 @@ impl Shorthand {
 }
 
 node! {
+    /// Symbol notation: `:arrow:l:`.
+    Symbol
+}
+
+impl Symbol {
+    /// Get the symbol's notation.
+    pub fn get(&self) -> &EcoString {
+        match self.0.kind() {
+            SyntaxKind::Symbol(v) => v,
+            _ => panic!("symbol is of wrong kind"),
+        }
+    }
+}
+
+node! {
     /// A smart quote: `'` or `"`.
     SmartQuote
 }
diff --git a/src/syntax/highlight.rs b/src/syntax/highlight.rs
index 526d1302..d4da7b3e 100644
--- a/src/syntax/highlight.rs
+++ b/src/syntax/highlight.rs
@@ -139,6 +139,8 @@ pub enum Category {
     Escape,
     /// An easily typable shortcut to a unicode codepoint.
     Shorthand,
+    /// Symbol notation.
+    Symbol,
     /// A smart quote.
     SmartQuote,
     /// Strong markup.
@@ -285,6 +287,7 @@ impl Category {
             SyntaxKind::Linebreak => Some(Category::Escape),
             SyntaxKind::Escape(_) => Some(Category::Escape),
             SyntaxKind::Shorthand(_) => Some(Category::Shorthand),
+            SyntaxKind::Symbol(_) => Some(Category::Symbol),
             SyntaxKind::SmartQuote { .. } => Some(Category::SmartQuote),
             SyntaxKind::Strong => Some(Category::Strong),
             SyntaxKind::Emph => Some(Category::Emph),
@@ -369,6 +372,7 @@ impl Category {
             Self::Punctuation => "punctuation.typst",
             Self::Escape => "constant.character.escape.typst",
             Self::Shorthand => "constant.character.shorthand.typst",
+            Self::Symbol => "constant.symbol.typst",
             Self::SmartQuote => "constant.character.quote.typst",
             Self::Strong => "markup.bold.typst",
             Self::Emph => "markup.italic.typst",
diff --git a/src/syntax/kind.rs b/src/syntax/kind.rs
index a7425d70..a4eb317b 100644
--- a/src/syntax/kind.rs
+++ b/src/syntax/kind.rs
@@ -144,6 +144,9 @@ pub enum SyntaxKind {
     /// A shorthand for a unicode codepoint. For example, `~` for non-breaking
     /// space or `-?` for a soft hyphen.
     Shorthand(char),
+    /// Symbol notation: `:arrow:l:`. The string only contains the inner part
+    /// without leading and trailing dot.
+    Symbol(EcoString),
     /// A smart quote: `'` or `"`.
     SmartQuote { double: bool },
     /// Strong content: `*Strong*`.
@@ -389,6 +392,7 @@ impl SyntaxKind {
             Self::Linebreak => "linebreak",
             Self::Escape(_) => "escape sequence",
             Self::Shorthand(_) => "shorthand",
+            Self::Symbol(_) => "symbol notation",
             Self::Strong => "strong content",
             Self::Emph => "emphasized content",
             Self::Raw(_) => "raw block",
@@ -507,6 +511,7 @@ impl Hash for SyntaxKind {
             Self::Linebreak => {}
             Self::Escape(c) => c.hash(state),
             Self::Shorthand(c) => c.hash(state),
+            Self::Symbol(s) => s.hash(state),
             Self::SmartQuote { double } => double.hash(state),
             Self::Strong => {}
             Self::Emph => {}
diff --git a/src/syntax/parsing.rs b/src/syntax/parsing.rs
index 1678eb01..59e066a6 100644
--- a/src/syntax/parsing.rs
+++ b/src/syntax/parsing.rs
@@ -230,6 +230,7 @@ fn markup_node(p: &mut Parser, at_start: &mut bool) {
         | SyntaxKind::SmartQuote { .. }
         | SyntaxKind::Escape(_)
         | SyntaxKind::Shorthand(_)
+        | SyntaxKind::Symbol(_)
         | SyntaxKind::Link(_)
         | SyntaxKind::Raw(_)
         | SyntaxKind::Ref(_) => p.eat(),
diff --git a/src/syntax/tokens.rs b/src/syntax/tokens.rs
index e7015bb2..130ad668 100644
--- a/src/syntax/tokens.rs
+++ b/src/syntax/tokens.rs
@@ -203,6 +203,7 @@ impl<'s> Tokens<'s> {
             '#' => self.hash(start),
             '.' if self.s.eat_if("..") => SyntaxKind::Shorthand('\u{2026}'),
             '-' => self.hyph(),
+            ':' => self.colon(),
             'h' if self.s.eat_if("ttp://") || self.s.eat_if("ttps://") => {
                 self.link(start)
             }
@@ -224,7 +225,6 @@ impl<'s> Tokens<'s> {
             '=' => SyntaxKind::Eq,
             '+' => SyntaxKind::Plus,
             '/' => SyntaxKind::Slash,
-            ':' => SyntaxKind::Colon,
 
             // Plain text.
             _ => self.text(start),
@@ -328,6 +328,25 @@ impl<'s> Tokens<'s> {
         }
     }
 
+    fn colon(&mut self) -> SyntaxKind {
+        let start = self.s.cursor();
+        let mut end = start;
+        while !self.s.eat_while(char::is_ascii_alphanumeric).is_empty() && self.s.at(':')
+        {
+            end = self.s.cursor();
+            self.s.eat();
+        }
+
+        self.s.jump(end);
+
+        if start < end {
+            self.s.expect(':');
+            SyntaxKind::Symbol(self.s.get(start..end).into())
+        } else {
+            SyntaxKind::Colon
+        }
+    }
+
     fn link(&mut self, start: usize) -> SyntaxKind {
         #[rustfmt::skip]
         self.s.eat_while(|c: char| matches!(c,
diff --git a/tests/ref/text/symbol.png b/tests/ref/text/symbol.png
new file mode 100644
index 00000000..4806683b
--- /dev/null
+++ b/tests/ref/text/symbol.png
diff --git a/tests/typ/text/symbol.typ b/tests/typ/text/symbol.typ
new file mode 100644
index 00000000..4dcef58b
--- /dev/null
+++ b/tests/typ/text/symbol.typ
@@ -0,0 +1,27 @@
+// Test symbol notation.
+
+---
+:face:
+:face:unknown:
+:woman:old:
+:turtle:
+
+#set text("NewComputerModernMath")
+:arrow:
+:arrow:l:
+:arrow:r:squiggly:
+#symbol(("arrow", "tr", "hook").join(":"))
+
+---
+Just a: colon. \
+Still :not a symbol. \
+Also not:a symbol \
+:arrow:r:this and this:arrow:l: \
+
+---
+#show symbol.where(notation: "my:custom"): "MY"
+This is :my:custom: notation.
+
+---
+// Error: 1-14 unknown symbol
+:nonexisting:
diff --git a/tools/support/typst.tmLanguage.json b/tools/support/typst.tmLanguage.json
index f8d66234..c849382e 100644
--- a/tools/support/typst.tmLanguage.json
+++ b/tools/support/typst.tmLanguage.json
@@ -73,6 +73,10 @@
           "match": "\\.\\.\\."
         },
         {
+          "name": "constant.symbol.typst",
+          "match": ":([a-zA-Z0-9]+:)+"
+        },
+        {
           "name": "markup.bold.typst",
           "begin": "(^\\*|\\*$|((?<=\\W|_)\\*)|(\\*(?=\\W|_)))",
           "end": "(^\\*|\\*$|((?<=\\W|_)\\*)|(\\*(?=\\W|_)))|\n|(?=\\])",