Implement custom commands with 0 arguments

latex2mmlc/Cargo.toml

@@ -17,13 +17,10 @@ categories = ["science"]
 
 [dependencies]
 bumpalo = "3.17.0"
-no-panic = "0.1.33"
 phf = { version = "0.11.3", features = ["macros"] }
 strum_macros = "0.26.4"
 
 [dev-dependencies]
 insta = { version = "1.41.1", features = ["default", "ron"] }
 serde = { version = "1.0.217", features = ["derive"] }
-lazy_static = "1.5.0"
 regex = "1.11.1"
-similar = "2.7.0"

latex2mmlc/src/arena.rs

@@ -220,6 +220,9 @@ impl<'arena> NodeList<'arena> {
     }
 }
 
+// NodeList is sync, because we don't allow mutation through shared references.
+unsafe impl<'arena> Sync for NodeList<'arena> {}
+
 #[cfg(test)]
 impl<'arena> Serialize for NodeList<'arena> {
     fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>

latex2mmlc/src/ast.rs

@@ -97,6 +97,13 @@ pub enum Node<'arena> {
         tf: MathVariant,
         content: &'arena Node<'arena>,
     },
+    PredefinedNode(&'static Node<'static>),
+}
+
+impl PartialEq for &'static Node<'static> {
+    fn eq(&self, other: &&'static Node<'static>) -> bool {
+        std::ptr::eq(*self, *other)
+    }
 }
 
 const INDENT: &str = "    ";
@@ -178,6 +185,7 @@ impl MathMLEmitter {
                 | Node::ColumnSeparator
                 | Node::RowSeparator
                 | Node::TextTransform { .. }
+                | Node::PredefinedNode(_)
         ) {
             // Get the base indent out of the way.
             new_line_and_indent(&mut self.s, base_indent);
@@ -516,6 +524,7 @@ impl MathMLEmitter {
                 pushln!(&mut self.s, base_indent, "</mtable>");
             }
             Node::ColumnSeparator | Node::RowSeparator => (),
+            Node::PredefinedNode(node) => self.emit(node, base_indent),
         }
     }
 }

latex2mmlc/src/commands.rs

@@ -1,3 +1,4 @@
+use crate::ast::Node;
 use crate::attribute::{FracAttr, MathVariant, OpAttr, Size, Style, TextTransform};
 use crate::ops::{self, Op};
 use crate::token::Token;
@@ -68,6 +69,10 @@ static COMMANDS: phf::Map<&'static str, Token> = phf::phf_map! {
     "Pi" => Token::UprightLetter('Π'),
     "Pr" => Token::Function("Pr"),
     "Psi" => Token::UprightLetter('Ψ'),
+    "RR" => Token::PredefinedNode(&Node::TextTransform {
+        tf: MathVariant::Transform(TextTransform::DoubleStruck),
+        content: &Node::SingleLetterIdent('R', false),
+    }),
     "Re" => Token::Letter('ℜ'),
     "Rho" => Token::UprightLetter('Ρ'),
     "Rightarrow" => Token::Operator(ops::RIGHTWARDS_DOUBLE_ARROW),
@@ -146,7 +151,7 @@ static COMMANDS: phf::Map<&'static str, Token> = phf::phf_map! {
     "blacksquare" => Token::Letter(ops::BLACK_SQUARE),
     "bm" => Token::Transform(MathVariant::Transform(TextTransform::BoldItalic)),
     "boldsymbol" => Token::Transform(MathVariant::Transform(TextTransform::BoldItalic)),
-    "bot" => Token::Operator(ops::UP_TACK),
+    "bot" => Token::Letter(ops::UP_TACK),
     "botdoteq" => Token::Operator(ops::EQUALS_SIGN_WITH_DOT_BELOW),
     "boxbox" => Token::Operator(ops::SQUARED_SQUARE),
     "boxbslash" => Token::Operator(ops::SQUARED_FALLING_DIAGONAL_SLASH),
@@ -195,6 +200,10 @@ static COMMANDS: phf::Map<&'static str, Token> = phf::phf_map! {
     "curlywedge" => Token::Operator(ops::CURLY_LOGICAL_AND),
     "curvearrowleft" => Token::Operator(ops::ANTICLOCKWISE_TOP_SEMICIRCLE_ARROW),
     "curvearrowright" => Token::Operator(ops::CLOCKWISE_TOP_SEMICIRCLE_ARROW),
+    "d" => Token::PredefinedNode(&Node::TextTransform {
+        tf: MathVariant::Normal,
+        content: &Node::SingleLetterIdent('d', false),
+    }),
     "dag" => Token::Letter('†'),
     "dagger" => Token::Letter('†'),
     "daleth" => Token::Letter('ℸ'),
@@ -448,7 +457,7 @@ static COMMANDS: phf::Map<&'static str, Token> = phf::phf_map! {
     "overset" => Token::Overset,
     "parallel" => Token::Operator(ops::PARALLEL_TO),
     "partial" => Token::Letter(ops::PARTIAL_DIFFERENTIAL),
-    "perp" => Token::Operator(ops::UP_TACK),
+    "perp" => Token::Operator(ops::PERPENDICULAR),
     "phi" => Token::Letter('ϕ'),
     "pi" => Token::Letter('π'),
     "pm" => Token::Operator(ops::PLUS_MINUS_SIGN),
@@ -562,7 +571,7 @@ static COMMANDS: phf::Map<&'static str, Token> = phf::phf_map! {
     "tilde" => Token::OverUnder(ops::TILDE, true, Some(OpAttr::StretchyFalse)),
     "times" => Token::Operator(ops::MULTIPLICATION_SIGN),
     "to" => Token::Operator(ops::RIGHTWARDS_ARROW),
-    "top" => Token::Operator(ops::DOWN_TACK),
+    "top" => Token::Letter(ops::DOWN_TACK),
     "triangle" => Token::Letter('△'),
     "triangledown" => Token::Operator(ops::WHITE_DOWN_POINTING_TRIANGLE),
     "triangleleft" => Token::Operator(ops::WHITE_LEFT_POINTING_TRIANGLE),

latex2mmlc/src/lib.rs

@@ -57,6 +57,7 @@ pub(crate) mod lexer;
 pub(crate) mod ops;
 pub(crate) mod parse;
 pub mod token;
+
 pub use ast::MathMLEmitter;
 pub use error::{LatexErrKind, LatexError};
 
@@ -313,6 +314,9 @@ mod tests {
             ),
             ("middle_bracket", r"\left(\frac12\middle]\frac12\right)"),
             ("left_right_different_stretch", r"\left/\frac12\right)"),
+            ("d_command", r"\d"),
+            ("d_command_nested", r"\mathit{x\d x}"),
+            ("RR_command", r"\RR"),
         ];
 
         for (name, problem) in problems.into_iter() {

latex2mmlc/src/ops.rs

@@ -399,8 +399,8 @@ pub(crate) const SQUARED_TIMES: Op = Op('⊠');
 pub(crate) const SQUARED_DOT_OPERATOR: Op = Op('⊡');
 pub(crate) const RIGHT_TACK: Op = Op('⊢');
 pub(crate) const LEFT_TACK: Op = Op('⊣');
-pub(crate) const DOWN_TACK: Op = Op('⊤');
-pub(crate) const UP_TACK: Op = Op('⊥');
+pub(crate) const DOWN_TACK: char = '⊤';
+pub(crate) const UP_TACK: char = '⊥';
 // pub(crate) const ASSERTION: Op = Op('⊦');
 // pub(crate) const MODELS: Op = Op('⊧');
 pub(crate) const TRUE: Op = Op('⊨');
@@ -538,6 +538,7 @@ pub(crate) const BLACK_STAR: char = '★';
 //
 // Unicode Block: Miscellaneous Mathematical Symbols-A
 //
+pub(crate) const PERPENDICULAR: Op = Op('⟂');
 pub(crate) const MATHEMATICAL_LEFT_WHITE_SQUARE_BRACKET: &ParenOp =
     &ParenOp('⟦', false, Stretchy::Always);
 pub(crate) const MATHEMATICAL_RIGHT_WHITE_SQUARE_BRACKET: &ParenOp =

latex2mmlc/src/parse.rs

@@ -661,6 +661,7 @@ where
             Token::End | Token::Right | Token::GroupEnd => {
                 return Err(LatexError(loc, LatexErrKind::UnexpectedClose(cur_token)))
             }
+            Token::PredefinedNode(node) => Node::PredefinedNode(node),
         };
         Ok(self.commit(node))
     }

latex2mmlc/src/snapshots/latex2mmlc__tests__RR_command.snap

@@ -0,0 +1,7 @@
+---
+source: latex2mmlc/src/lib.rs
+expression: "\\RR"
+---
+<math>
+    <mi>ℝ</mi>
+</math>

latex2mmlc/src/snapshots/latex2mmlc__tests__d_command.snap

@@ -0,0 +1,7 @@
+---
+source: latex2mmlc/src/lib.rs
+expression: "\\d"
+---
+<math>
+    <mi mathvariant="normal">d</mi>
+</math>

latex2mmlc/src/snapshots/latex2mmlc__tests__d_command_nested.snap

@@ -0,0 +1,11 @@
+---
+source: latex2mmlc/src/lib.rs
+expression: "\\mathit{x\\d x}"
+---
+<math>
+    <mrow>
+        <mi>𝑥</mi>
+        <mi mathvariant="normal">d</mi>
+        <mi>𝑥</mi>
+    </mrow>
+</math>

latex2mmlc/src/token.rs

@@ -2,6 +2,7 @@ use std::mem::discriminant;
 
 use strum_macros::AsRefStr;
 
+use crate::ast::Node;
 use crate::attribute::{FracAttr, MathVariant, OpAttr, Size, Style, TextTransform};
 use crate::ops::{Op, ParenOp};
 
@@ -94,6 +95,7 @@ pub enum Token<'source> {
     #[strum(serialize = r"\mathstrut")]
     Mathstrut,
     Style(Style),
+    PredefinedNode(&'static Node<'static>),
     UnknownCommand(&'source str),
 }
 

latex2mmlc/tests/snapshots/wiki_test__wiki068.snap

@@ -3,7 +3,7 @@ source: latex2mmlc/tests/wiki_test.rs
 expression: "\\perp, \\angle, \\sphericalangle, \\measuredangle, 45^\\circ"
 ---
 <math>
-    <mo>⊥</mo>
+    <mo>⟂</mo>
     <mo>,</mo>
     <mi>∠</mi>
     <mo>,</mo>

latex2mmlc/tests/wiki_test.rs

@@ -1,5 +1,6 @@
+use std::sync::LazyLock;
+
 use insta::assert_snapshot;
-use lazy_static::lazy_static;
 use regex::Regex;
 // use similar::{ChangeTag, TextDiff};
 
@@ -268,9 +269,7 @@ fn wiki_test() {
 
 /// Prettify HTML input
 pub fn prettify(input: &str) -> String {
-    lazy_static! {
-        static ref OPEN_TAG: Regex = Regex::new("(?P<tag><[A-z])").unwrap();
-    }
+    static OPEN_TAG: LazyLock<Regex> = LazyLock::new(|| Regex::new("(?P<tag><[A-z])").unwrap());
 
     // First get all tags on their own lines
     let mut stage1 = input.to_string();