wip

src/block.rs

@@ -22,25 +22,19 @@ pub enum Block {
 #[derive(Debug, Clone, PartialEq, Eq)]
 pub enum Leaf {
     Paragraph,
-    Heading {
+    Heading { level: u8 },
-        level: usize,
-    },
     Attributes,
     Table,
-    ThematicBreak,
     LinkDefinition,
-    CodeBlock {
+    CodeBlock { fence_length: u8 },
-        fence_char: char,
-        fence_length: usize,
-    },
 }
 
 #[derive(Debug, Clone, PartialEq, Eq)]
 pub enum Container {
     Blockquote,
-    Div { fence_length: usize },
+    Div { fence_length: u8 },
-    ListItem { indent: usize },
+    ListItem { indent: u8 },
-    Footnote { indent: usize },
+    Footnote { indent: u8 },
 }
 
 #[derive(Debug, PartialEq, Eq)]
@@ -49,6 +43,8 @@ pub enum Atom {
     Inline,
     /// A line with no non-whitespace characters.
     Blankline,
+    /// Thematic break.
+    ThematicBreak,
 }
 
 struct Parser<'s> {
@@ -89,9 +85,9 @@ impl<'s> Parser<'s> {
             })
             .count();
         let lines = &mut lines[blanklines..];
-        Block::parse(lines.iter().map(|sp| (sp.of(self.src), sp.start()))).map_or(
+        Block::parse(lines.iter().map(|sp| sp.of(self.src))).map_or(0, |(kind, span, len)| {
-            0,
+            let start = lines.get(0).map(|sp| sp.start()).unwrap();
-            |(kind, span, len)| {
+            let span = span.translate(start);
             match &kind {
                 Block::Leaf(_) => {
                     self.tree.enter(kind, span);
@@ -116,7 +112,7 @@ impl<'s> Parser<'s> {
                             .chars()
                             .take_while(|c| c.is_whitespace())
                             .count()
-                                + skip_chars)
+                            + usize::from(skip_chars))
                         .min(sp.len());
                         *sp = sp.trim_start(skip);
                     });
@@ -130,27 +126,27 @@ impl<'s> Parser<'s> {
             }
             self.tree.exit();
             blanklines + len
-            },
+        })
-        )
     }
 }
 
 impl Block {
     /// Parse a single block. Return number of lines the block uses.
-    fn parse<'b, I: Iterator<Item = (&'b str, usize)>>(
+    fn parse<'b, I: Iterator<Item = &'b str>>(mut lines: I) -> Option<(Block, Span, usize)> {
-        mut lines: I,
+        lines.next().map(|l| {
-    ) -> Option<(Block, Span, usize)> {
-        if let Some((l, start)) = lines.next() {
             let (kind, sp) = Block::start(l);
-            let line_count = 1 + lines.take_while(|(l, _)| kind.continues(l)).count();
+            let line_count = 1
-            Some((kind, sp.translate(start), line_count))
+                + lines.take_while(|l| kind.continues(l)).count()
-        } else {
+                + usize::from(matches!(
-            None
+                    kind,
-        }
+                    Self::Leaf(CodeBlock { .. }) | Self::Container(Div { .. })
+                ));
+            (kind, sp, line_count)
+        })
     }
 
     /// Determine what type of block a line can start.
-    fn start(line: &str) -> (Block, Span) {
+    fn start(line: &str) -> (Self, Span) {
         let start = line.chars().take_while(|c| c.is_whitespace()).count();
         let line = &line[start..];
         let mut chars = line.chars();
@@ -159,37 +155,41 @@ impl Block {
                 .find(|c| *c != '#')
                 .map_or(true, char::is_whitespace)
                 .then(|| {
-                    let span = Span::by_len(start, line.len() - chars.as_str().len() - 1);
+                    u8::try_from(line.len() - chars.as_str().len() - 1)
-                    (Self::Leaf(Heading { level: span.len() }), span)
+                        .ok()
-                }),
+                        .map(|level| {
+                            (
+                                Self::Leaf(Heading { level }),
+                                Span::by_len(start, level.into()),
+                            )
+                        })
+                })
+                .flatten(),
             '>' => chars.next().map_or(true, |c| c == ' ').then(|| {
                 (
                     Self::Container(Blockquote),
                     Span::by_len(start, line.len() - chars.as_str().len() - 1),
                 )
             }),
-            f @ ':' => {
+            f @ ('`' | ':') => {
                 let fence_length = chars.take_while(|c| *c == f).count() + 1;
-                (fence_length >= 3).then(|| {
+                (fence_length >= 3)
+                    .then(|| {
+                        u8::try_from(fence_length).ok().map(|fence_length| {
                             (
-                        Self::Container(Div { fence_length }),
+                                match f {
+                                    '`' => Self::Leaf(CodeBlock { fence_length }),
+                                    ':' => Self::Container(Div { fence_length }),
+                                    _ => unreachable!(),
+                                },
                                 Span::by_len(start, line.len()),
                             )
                         })
-            }
-            fence_char @ ('`' | '~') => {
-                let fence_length = chars.take_while(|c| *c == fence_char).count() + 1;
-                (fence_length >= 3).then(|| {
-                    (
-                        Self::Leaf(CodeBlock {
-                            fence_char,
-                            fence_length,
-                        }),
-                        Span::by_len(start, line.len()),
-                    )
                     })
+                    .flatten()
             }
             _ => {
+                /*
                 let thematic_break = || {
                     let mut without_whitespace = line.chars().filter(|c| !c.is_whitespace());
                     let length = without_whitespace.clone().count();
@@ -198,7 +198,9 @@ impl Block {
                             || without_whitespace.all(|c| c == '*')))
                     .then(|| (Self::Leaf(ThematicBreak), Span::by_len(start, line.len())))
                 };
-                thematic_break()
+                */
+                //thematic_break()
+                None
             }
         }
         .unwrap_or((Self::Leaf(Paragraph), Span::new(0, 0)))
@@ -210,18 +212,16 @@ impl Block {
             Self::Leaf(Paragraph | Heading { .. } | Table | LinkDefinition) => {
                 !line.trim().is_empty()
             }
-            Self::Leaf(Attributes | ThematicBreak) => false,
+            Self::Leaf(Attributes) => false,
-            Self::Leaf(CodeBlock {
-                fence_char,
-                fence_length,
-            }) => !line.chars().take(*fence_length).all(|c| c == *fence_char),
             Self::Container(Blockquote) => line.trim().starts_with('>'),
             Self::Container(Footnote { indent } | ListItem { indent }) => {
                 let spaces = line.chars().take_while(|c| c.is_whitespace()).count();
-                !line.trim().is_empty() && spaces >= *indent
+                !line.trim().is_empty() && spaces >= (*indent).into()
             }
-            Self::Container(Div { fence_length }) => {
+            Self::Container(Div { fence_length }) | Self::Leaf(CodeBlock { fence_length }) => {
-                line.chars().take(*fence_length).all(|c| c == ':')
+                let mut c = line.chars();
+                !((&mut c).take((*fence_length).into()).all(|c| c == ':')
+                    && c.next().map_or(false, char::is_whitespace))
             }
         }
     }
@@ -357,10 +357,30 @@ mod test {
         );
     }
 
+    #[test]
+    fn parse_code_block() {
+        let src = concat!(
+            "```lang\n",
+            "l0\n",
+            "l1\n",
+            "```", //
+        );
+
+        assert_eq!(
+            super::Parser::new(src).parse().iter().collect::<Vec<_>>(),
+            &[
+                Event::Enter(&Leaf(CodeBlock { fence_length: 3 }), Span::new(0, 8)),
+                Event::Element(&Inline, Span::new(8, 11)),
+                Event::Element(&Inline, Span::new(11, 14)),
+                Event::Exit
+            ]
+        );
+    }
+
     #[test]
     fn block_multiline() {
         let src = "# heading\n spanning two lines\n";
-        let lines = super::lines(src).map(|sp| (sp.of(src), sp.start()));
+        let lines = super::lines(src).map(|sp| sp.of(src));
         let (kind, sp, len) = Block::parse(lines).unwrap();
         assert_eq!(kind, Block::Leaf(Heading { level: 1 }));
         assert_eq!(sp.of(src), "#");
@@ -376,7 +396,7 @@ mod test {
         ">\n",
         "> c\n", //
     );
-        let lines = super::lines(src).map(|sp| (sp.of(src), sp.start()));
+        let lines = super::lines(src).map(|sp| sp.of(src));
         let (kind, sp, len) = Block::parse(lines).unwrap();
         assert_eq!(kind, Block::Container(Blockquote));
         assert_eq!(sp.of(src), ">");

src/inline.rs

@@ -35,22 +35,21 @@ pub struct Node {
 pub enum NodeKind {
     Str,
     // link
+    Url,
+    ImageSource,
+    LinkReference,
     FootnoteReference,
-    ReferenceLink,
-    Link,
-    Emoji,
     // verbatim
     Verbatim,
     RawFormat,
-    DisplayMath,
     InlineMath,
+    DisplayMath,
 }
 
 #[derive(Debug, Copy, Clone, PartialEq, Eq)]
 pub enum Container {
-    // attributes
-    Attributes,
     Span,
+    Attributes,
     // typesetting
     Subscript,
     Superscript,
@@ -58,17 +57,10 @@ pub enum Container {
     Delete,
     Emphasis,
     Strong,
-    Mark,
+    //Mark,
     // smart quoting
     SingleQuoted,
     DoubleQuoted,
-    // URLs
-    AutoUrl,
-    Url,
-    ImageText,
-    LinkText,
-    Reference,
-    Destination,
 }
 
 #[derive(Debug, PartialEq, Eq)]
@@ -165,7 +157,14 @@ impl<'s> Parser<'s> {
                             len,
                         }) = self.peek()
                         {
-                            Some((DisplayMath, first.len))
+                            Some((
+                                if first.len == 2 {
+                                    DisplayMath
+                                } else {
+                                    InlineMath
+                                },
+                                *len,
+                            ))
                         } else {
                             None
                         }
@@ -201,14 +200,14 @@ impl<'s> Parser<'s> {
             lex::Kind::Sym(Symbol::Tilde) => Some((Subscript, Dir::Both)),
             lex::Kind::Sym(Symbol::Quote1) => Some((SingleQuoted, Dir::Both)),
             lex::Kind::Sym(Symbol::Quote2) => Some((DoubleQuoted, Dir::Both)),
-            lex::Kind::Open(Delimiter::Bracket) => Some((LinkText, Dir::Open)),
+            lex::Kind::Open(Delimiter::Bracket) => Some((Span, Dir::Open)),
-            lex::Kind::Close(Delimiter::Bracket) => Some((LinkText, Dir::Close)),
+            lex::Kind::Close(Delimiter::Bracket) => Some((Span, Dir::Close)),
             lex::Kind::Open(Delimiter::BraceAsterisk) => Some((Strong, Dir::Open)),
             lex::Kind::Close(Delimiter::BraceAsterisk) => Some((Strong, Dir::Close)),
             lex::Kind::Open(Delimiter::BraceCaret) => Some((Superscript, Dir::Open)),
             lex::Kind::Close(Delimiter::BraceCaret) => Some((Superscript, Dir::Close)),
-            lex::Kind::Open(Delimiter::BraceEqual) => Some((Mark, Dir::Open)),
+            //lex::Kind::Open(Delimiter::BraceEqual) => Some((Mark, Dir::Open)),
-            lex::Kind::Close(Delimiter::BraceEqual) => Some((Mark, Dir::Close)),
+            //lex::Kind::Close(Delimiter::BraceEqual) => Some((Mark, Dir::Close)),
             lex::Kind::Open(Delimiter::BraceHyphen) => Some((Delete, Dir::Open)),
             lex::Kind::Close(Delimiter::BraceHyphen) => Some((Delete, Dir::Close)),
             lex::Kind::Open(Delimiter::BracePlus) => Some((Insert, Dir::Open)),
@@ -308,6 +307,12 @@ mod test {
         test_parse!("abc `def`", Node(Str.span(0, 4)), Node(Verbatim.span(5, 8)));
     }
 
+    #[test]
+    fn math() {
+        test_parse!("$`abc`", Node(InlineMath.span(2, 5)));
+        test_parse!("$$```abc", Node(DisplayMath.span(5, 8)));
+    }
+
     #[test]
     fn container_basic() {
         test_parse!(

src/lex.rs

@@ -30,7 +30,7 @@ pub enum Delimiter {
     Brace,
     BraceAsterisk,
     BraceCaret,
-    BraceEqual,
+    //BraceEqual,
     BraceHyphen,
     BracePlus,
     BraceTilde,
@@ -149,7 +149,7 @@ impl<'s> Lexer<'s> {
                 let explicit = match self.peek() {
                     '*' => Some(Open(BraceAsterisk)),
                     '^' => Some(Open(BraceCaret)),
-                    '=' => Some(Open(BraceEqual)),
+                    //'=' => Some(Open(BraceEqual)),
                     '-' => Some(Open(BraceHyphen)),
                     '+' => Some(Open(BracePlus)),
                     '~' => Some(Open(BraceTilde)),
@@ -165,7 +165,7 @@ impl<'s> Lexer<'s> {
             }
             '*' => self.maybe_eat_close_brace(Asterisk, BraceAsterisk),
             '^' => self.maybe_eat_close_brace(Caret, BraceCaret),
-            '=' => self.maybe_eat_close_brace(Equal, BraceEqual),
+            //'=' => self.maybe_eat_close_brace(Equal, BraceEqual),
             '+' => self.maybe_eat_close_brace(Plus, BracePlus),
             '~' => self.maybe_eat_close_brace(Tilde, BraceTilde),
             '_' => self.maybe_eat_close_brace(Underscore, BraceUnderscore),
@@ -178,6 +178,7 @@ impl<'s> Lexer<'s> {
                 }
             }
 
+            '=' => Sym(Equal),
             '!' => Sym(Exclaim),
             '%' => Sym(Percentage),
             '<' => Sym(Lt),

src/lib.rs

@@ -35,6 +35,29 @@ impl<'s> Parser<'s> {
     }
 }
 
+pub enum ListType {
+    Unordered,
+    Ordered,
+}
+
+pub enum TagKind<'s> {
+    Paragraph,
+    Heading { level: u8 },
+    Table,
+    TableRow,
+    TableCell,
+    RawBlock { format: &'s str },
+    CodeBlock { language: &'s str },
+    Blockquote,
+    Div,
+    UnorderedList,
+    OrderedList { start: usize },
+    ListItem,
+    DescriptionList,
+    DescriptionItem,
+    Footnote { tag: &'s str },
+}
+
 #[derive(Debug, PartialEq, Eq)]
 pub enum Event {
     Start(block::Block),