feat: support line breaks

src/syntax/line_break.rs

@@ -0,0 +1,82 @@
+use nom::{
+    branch::alt,
+    character::complete::{line_ending, space0},
+    combinator::{eof, map},
+    sequence::tuple,
+    IResult,
+};
+
+use crate::{
+    syntax::combinator::{backslash_token, node},
+    SyntaxKind,
+};
+
+use super::{combinator::GreenElement, input::Input};
+
+pub fn line_break_node(input: Input) -> IResult<Input, GreenElement, ()> {
+    debug_assert!(input.s.starts_with('\\'));
+    let mut parser = map(
+        tuple((
+            backslash_token,
+            backslash_token,
+            space0,
+            alt((line_ending, eof)),
+        )),
+        |(b1, b2, ws, nl)| {
+            node(
+                SyntaxKind::LINE_BREAK,
+                [b1, b2, ws.ws_token(), nl.nl_token()],
+            )
+        },
+    );
+    crate::lossless_parser!(parser, input)
+}
+
+#[test]
+fn parse() {
+    use crate::ast::LineBreak;
+    use crate::tests::to_ast;
+
+    let to_line_break = to_ast::<LineBreak>(line_break_node);
+
+    insta::assert_debug_snapshot!(
+        to_line_break("\\\\\n").syntax,
+        @r###"
+    LINE_BREAK@0..3
+      BACKSLASH@0..1 "\\"
+      BACKSLASH@1..2 "\\"
+      WHITESPACE@2..2 ""
+      NEW_LINE@2..3 "\n"
+    "###
+    );
+    insta::assert_debug_snapshot!(
+        to_line_break("\\\\   \n").syntax,
+        @r###"
+    LINE_BREAK@0..6
+      BACKSLASH@0..1 "\\"
+      BACKSLASH@1..2 "\\"
+      WHITESPACE@2..5 "   "
+      NEW_LINE@5..6 "\n"
+    "###
+    );
+    insta::assert_debug_snapshot!(
+        to_line_break("\\\\\r\n").syntax,
+        @r###"
+    LINE_BREAK@0..4
+      BACKSLASH@0..1 "\\"
+      BACKSLASH@1..2 "\\"
+      WHITESPACE@2..2 ""
+      NEW_LINE@2..4 "\r\n"
+    "###
+    );
+    insta::assert_debug_snapshot!(
+        to_line_break("\\\\    ").syntax,
+        @r###"
+    LINE_BREAK@0..6
+      BACKSLASH@0..1 "\\"
+      BACKSLASH@1..2 "\\"
+      WHITESPACE@2..6 "    "
+      NEW_LINE@6..6 ""
+    "###
+    );
+}

src/syntax/mod.rs

@@ -21,6 +21,7 @@ pub mod input;
 pub mod keyword;
 pub mod latex_environment;
 pub mod latex_fragment;
+pub mod line_break;
 pub mod link;
 pub mod list;
 pub mod macros;
@@ -184,6 +185,7 @@ pub enum SyntaxKind {
     INLINE_SRC,
     LINK,
     LINK_PATH,
+    LINE_BREAK,
     COOKIE,
     RADIO_TARGET,
     FN_REF,
@@ -234,6 +236,7 @@ impl SyntaxKind {
                 | SyntaxKind::FN_REF
                 | SyntaxKind::INLINE_CALL
                 | SyntaxKind::INLINE_SRC
+                | SyntaxKind::LINE_BREAK
                 | SyntaxKind::LINK
                 | SyntaxKind::MACROS
                 | SyntaxKind::RADIO_TARGET

src/syntax/object.rs

@@ -10,6 +10,7 @@ use super::{
     inline_src::inline_src_node,
     input::Input,
     latex_fragment::latex_fragment_node,
+    line_break::line_break_node,
     link::link_node,
     macros::macros_node,
     radio_target::radio_target_node,
@@ -110,10 +111,10 @@ impl<'a> Iterator for ObjectPositions<'a> {
 /// - Statistics Cookies
 /// - Timestamps
 /// - Text Markup (bold code strike verbatim underline italic)
+/// - Line Breaks
 ///
 /// // todo:
 /// - Citations
-/// - Line Breaks
 /// - Subscript and Superscript
 pub fn object_nodes(input: Input) -> Vec<GreenElement> {
     // TODO:
@@ -124,11 +125,6 @@ pub fn object_nodes(input: Input) -> Vec<GreenElement> {
 
     'l: while !i.is_empty() {
         for (input, head) in ObjectPositions::standard(i) {
-            debug_assert!(
-                input.s.len() >= 2,
-                "object must have at least two characters: {:?}",
-                input.s
-            );
             if let Ok((input, node)) = standard_object_node(input) {
                 if !head.is_empty() {
                     nodes.push(head.text_token())
@@ -170,11 +166,6 @@ pub fn minimal_object_nodes(input: Input) -> Vec<GreenElement> {
 
     'l: while !i.is_empty() {
         for (input, head) in ObjectPositions::minimal(i) {
-            debug_assert!(
-                input.s.len() >= 2,
-                "object must have at least two characters: {:?}",
-                input.s
-            );
             if let Ok((input, node)) = minimal_object_node(input) {
                 if !head.is_empty() {
                     nodes.push(head.text_token())
@@ -205,6 +196,12 @@ pub fn minimal_object_nodes(input: Input) -> Vec<GreenElement> {
 
 /// parse an object from standard sets
 fn standard_object_node(i: Input) -> IResult<Input, GreenElement, ()> {
+    debug_assert!(
+        i.s.len() >= 2,
+        "object must have at least two characters: {:?}",
+        i.s
+    );
+
     match &i.as_bytes()[0] {
         b'*' => bold_node(i),
         b'+' => strike_node(i),
@@ -225,7 +222,13 @@ fn standard_object_node(i: Input) -> IResult<Input, GreenElement, ()> {
         b'c' => inline_call_node(i),
         b's' => inline_src_node(i),
         b'$' => latex_fragment_node(i),
-        b'\\' => entity_node(i).or_else(|_| latex_fragment_node(i)),
+        b'\\' => {
+            if i.as_bytes()[1] == b'\\' {
+                line_break_node(i)
+            } else {
+                entity_node(i).or_else(|_| latex_fragment_node(i))
+            }
+        }
         _ => Err(nom::Err::Error(())),
     }
 }