aboutsummaryrefslogtreecommitdiffstats
path: root/src/construct
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--src/construct/definition.rs318
-rw-r--r--src/construct/mod.rs8
-rw-r--r--src/construct/partial_destination.rs154
-rw-r--r--src/construct/partial_label.rs100
-rw-r--r--src/construct/partial_title.rs136
5 files changed, 714 insertions, 2 deletions
diff --git a/src/construct/definition.rs b/src/construct/definition.rs
new file mode 100644
index 0000000..e540b44
--- /dev/null
+++ b/src/construct/definition.rs
@@ -0,0 +1,318 @@
+//! Definition is a construct that occurs in the [flow] content type.
+//!
+//! They’re formed with the following BNF:
+//!
+//! ```bnf
+//! definition ::= label ':' whitespace destination [ whitespace title ] [ space_or_tab ]
+//!
+//! ; Restriction: maximum `999` codes allowed between brackets.
+//! ; Restriction: no blank lines.
+//! ; Restriction: at least 1 non-space and non-eol code must exist.
+//! label ::= '[' *( label_text | label_escape ) ']'
+//! label_text ::= code - '[' - '\\' - ']'
+//! label_escape ::= '\\' [ '[' | '\\' | ']' ]
+//!
+//! destination ::= destination_enclosed | destination_raw
+//! destination_enclosed ::= '<' *( destination_enclosed_text | destination_enclosed_escape ) '>'
+//! destination_enclosed_text ::= code - '<' - '\\' - eol
+//! destination_enclosed_escape ::= '\\' [ '<' | '\\' | '>' ]
+//! destination_raw ::= 1*( destination_raw_text | destination_raw_escape )
+//! ; Restriction: unbalanced `)` characters are not allowed.
+//! destination_raw_text ::= code - '\\' - ascii_control - space_or_tab - eol
+//! destination_raw_escape ::= '\\' [ '(' | ')' | '\\' ]
+//!
+//! ; Restriction: no blank lines.
+//! ; Restriction: markers must match (in case of `(` with `)`).
+//! title ::= marker [ *( code - '\\' | '\\' [ marker ] ) ] marker
+//! marker ::= '"' | '\'' | '('
+//!
+//! whitespace ::= eol *whitespace | 1*space_or_tab [ eol *whitespace ]
+//! space_or_tab ::= ' ' | '\t'
+//! ```
+//!
+//! Definitions in markdown to not, on their own, relate to anything in HTML.
+//! When connected with a link (reference), they together relate to the `<a>`
+//! element in HTML.
+//! The definition forms its `href`, and optionally `title`, attributes.
+//! See [*§ 4.5.1 The `a` element*][html] in the HTML spec for more info.
+//!
+//! The `label`, `destination`, and `title` parts are interpreted as the
+//! [string][] content type.
+//! That means that character escapes and character reference are allowed.
+//!
+//! ## References
+//!
+//! * [`definition.js` in `micromark`](https://github.com/micromark/micromark/blob/main/packages/micromark-core-commonmark/dev/lib/definition.js)
+//! * [*§ 4.7 Link reference definitions* in `CommonMark`](https://spec.commonmark.org/0.30/#link-reference-definitions)
+//!
+//! [flow]: crate::content::flow
+//! [string]: crate::content::string
+//! [html]: https://html.spec.whatwg.org/multipage/text-level-semantics.html#the-a-element
+//!
+//! <!-- To do: link link (reference) -->
+//!
+//! <!-- To do: describe how references and definitions match -->
+
+use crate::construct::{
+ partial_destination::start as destination, partial_label::start as label,
+ partial_title::start as title, partial_whitespace::start as whitespace,
+};
+use crate::tokenizer::{Code, State, StateFnResult, TokenType, Tokenizer};
+
+/// At the start of a definition.
+///
+/// ```markdown
+/// |[a]: b "c"
+/// ```
+pub fn start(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::Char('[') => {
+ tokenizer.enter(TokenType::Definition);
+ tokenizer.go(label, label_after)(tokenizer, code)
+ }
+ _ => (State::Nok, None),
+ }
+}
+
+/// After the label of a definition.
+///
+/// ```markdown
+/// [a]|: b "c"
+/// ```
+pub fn label_after(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ // To do: get the identifier:
+ // identifier = normalizeIdentifier(
+ // self.sliceSerialize(self.events[self.events.length - 1][1]).slice(1, -1)
+ // )
+
+ match code {
+ Code::Char(':') => {
+ tokenizer.enter(TokenType::DefinitionMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionMarker);
+ (State::Fn(Box::new(marker_after)), None)
+ }
+ _ => (State::Nok, None),
+ }
+}
+
+/// After the marker of a definition.
+///
+/// ```markdown
+/// [a]:| b "c"
+///
+/// [a]:| ␊
+/// b "c"
+/// ```
+pub fn marker_after(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(marker_after_optional_whitespace),
+ )(tokenizer, code)
+}
+
+/// After the marker, after whitespace.
+///
+/// ```markdown
+/// [a]: |b "c"
+///
+/// [a]: |␊
+/// b "c"
+/// ```
+pub fn marker_after_optional_whitespace(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ tokenizer.enter(TokenType::LineEnding);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::LineEnding);
+ (State::Fn(Box::new(marker_after_optional_line_ending)), None)
+ }
+ _ => destination_before(tokenizer, code),
+ }
+}
+
+/// After the marker, after a line ending.
+///
+/// ```markdown
+/// [a]:
+/// | b "c"
+/// ```
+pub fn marker_after_optional_line_ending(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(destination_before),
+ )(tokenizer, code)
+}
+
+/// Before a destination.
+///
+/// ```markdown
+/// [a]: |b "c"
+///
+/// [a]:
+/// |b "c"
+/// ```
+pub fn destination_before(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ let event = tokenizer.events.last().unwrap();
+ // Blank line not ok.
+ let char_nok = matches!(
+ code,
+ Code::None | Code::CarriageReturnLineFeed | Code::Char('\r' | '\n')
+ );
+
+ if !char_nok
+ && (event.token_type == TokenType::LineEnding || event.token_type == TokenType::Whitespace)
+ {
+ tokenizer.go(destination, destination_after)(tokenizer, code)
+ } else {
+ (State::Nok, None)
+ }
+}
+
+/// After a destination.
+///
+/// ```markdown
+/// [a]: b| "c"
+///
+/// [a]: b| ␊
+/// "c"
+/// ```
+pub fn destination_after(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ tokenizer.attempt(title_before, |_ok| Box::new(after))(tokenizer, code)
+}
+
+/// After a definition.
+///
+/// ```markdown
+/// [a]: b|
+/// [a]: b "c"|
+/// ```
+pub fn after(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(after_whitespace),
+ )(tokenizer, code)
+}
+
+/// After a definition, after optional whitespace.
+///
+/// ```markdown
+/// [a]: b |
+/// [a]: b "c"|
+/// ```
+pub fn after_whitespace(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::None | Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ tokenizer.exit(TokenType::Definition);
+ (State::Ok, Some(vec![code]))
+ }
+ _ => (State::Nok, None),
+ }
+}
+
+/// After a destination, presumably before a title.
+///
+/// ```markdown
+/// [a]: b| "c"
+///
+/// [a]: b| ␊
+/// "c"
+/// ```
+pub fn title_before(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(title_before_after_optional_whitespace),
+ )(tokenizer, code)
+}
+
+/// Before a title, after optional whitespace.
+///
+/// ```markdown
+/// [a]: b |"c"
+///
+/// [a]: b |␊
+/// "c"
+/// ```
+pub fn title_before_after_optional_whitespace(
+ tokenizer: &mut Tokenizer,
+ code: Code,
+) -> StateFnResult {
+ match code {
+ Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ tokenizer.enter(TokenType::LineEnding);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::LineEnding);
+ (
+ State::Fn(Box::new(title_before_after_optional_line_ending)),
+ None,
+ )
+ }
+ _ => title_before_marker(tokenizer, code),
+ }
+}
+
+/// Before a title, after a line ending.
+///
+/// ```markdown
+/// [a]: b␊
+/// | "c"
+/// ```
+pub fn title_before_after_optional_line_ending(
+ tokenizer: &mut Tokenizer,
+ code: Code,
+) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(title_before_marker),
+ )(tokenizer, code)
+}
+
+/// Before a title, after a line ending.
+///
+/// ```markdown
+/// [a]: b␊
+/// | "c"
+/// ```
+pub fn title_before_marker(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ let event = tokenizer.events.last().unwrap();
+
+ if event.token_type == TokenType::LineEnding || event.token_type == TokenType::Whitespace {
+ tokenizer.go(title, title_after)(tokenizer, code)
+ } else {
+ (State::Nok, None)
+ }
+}
+
+/// After a title.
+///
+/// ```markdown
+/// [a]: b "c"|
+///
+/// [a]: b␊
+/// "c"|
+/// ```
+pub fn title_after(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(title_after_after_optional_whitespace),
+ )(tokenizer, code)
+}
+
+/// After a title, after optional whitespace.
+///
+/// ```markdown
+/// [a]: b "c"|
+///
+/// [a]: b "c" |
+/// ```
+pub fn title_after_after_optional_whitespace(
+ _tokenizer: &mut Tokenizer,
+ code: Code,
+) -> StateFnResult {
+ match code {
+ Code::None | Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ (State::Ok, Some(vec![code]))
+ }
+ _ => (State::Nok, None),
+ }
+}
diff --git a/src/construct/mod.rs b/src/construct/mod.rs
index ca1149f..fb79f68 100644
--- a/src/construct/mod.rs
+++ b/src/construct/mod.rs
@@ -16,7 +16,7 @@
//! The following constructs are found in markdown:
//!
//! * attention (strong, emphasis) (text)
-//! * [autolink][autolink]
+//! * [autolink][]
//! * [blank line][blank_line]
//! * block quote
//! * [character escape][character_escape]
@@ -25,7 +25,7 @@
//! * [code (indented)][code_indented]
//! * [code (text)][code_text]
//! * content
-//! * definition
+//! * [definition][]
//! * [hard break (escape)][hard_break_escape]
//! * [hard break (trailing)][hard_break_trailing]
//! * [heading (atx)][heading_atx]
@@ -61,11 +61,15 @@ pub mod character_reference;
pub mod code_fenced;
pub mod code_indented;
pub mod code_text;
+pub mod definition;
pub mod hard_break_escape;
pub mod hard_break_trailing;
pub mod heading_atx;
pub mod heading_setext;
pub mod html_flow;
pub mod html_text;
+pub mod partial_destination;
+pub mod partial_label;
+pub mod partial_title;
pub mod partial_whitespace;
pub mod thematic_break;
diff --git a/src/construct/partial_destination.rs b/src/construct/partial_destination.rs
new file mode 100644
index 0000000..8cf5b77
--- /dev/null
+++ b/src/construct/partial_destination.rs
@@ -0,0 +1,154 @@
+// To do: pass token types in.
+
+use crate::tokenizer::{Code, State, StateFnResult, TokenType, Tokenizer};
+
+pub fn start(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::Char('<') => {
+ tokenizer.enter(TokenType::DefinitionDestination);
+ tokenizer.enter(TokenType::DefinitionDestinationLiteral);
+ tokenizer.enter(TokenType::DefinitionDestinationLiteralMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionDestinationLiteralMarker);
+ (State::Fn(Box::new(enclosed_before)), None)
+ }
+ Code::None | Code::CarriageReturnLineFeed | Code::VirtualSpace | Code::Char(')') => {
+ (State::Nok, None)
+ }
+ Code::Char(char) if char.is_ascii_control() => (State::Nok, None),
+ Code::Char(_) => {
+ tokenizer.enter(TokenType::DefinitionDestination);
+ tokenizer.enter(TokenType::DefinitionDestinationRaw);
+ tokenizer.enter(TokenType::DefinitionDestinationString);
+ // To do: link.
+ tokenizer.enter(TokenType::ChunkString);
+ raw(tokenizer, code, 0)
+ }
+ }
+}
+
+/// To do.
+fn enclosed_before(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ if let Code::Char('>') = code {
+ tokenizer.enter(TokenType::DefinitionDestinationLiteralMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionDestinationLiteralMarker);
+ tokenizer.exit(TokenType::DefinitionDestinationLiteral);
+ tokenizer.exit(TokenType::DefinitionDestination);
+ (State::Ok, None)
+ } else {
+ tokenizer.enter(TokenType::DefinitionDestinationString);
+ // To do: link.
+ tokenizer.enter(TokenType::ChunkString);
+ enclosed(tokenizer, code)
+ }
+}
+
+/// To do.
+fn enclosed(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::Char('>') => {
+ tokenizer.exit(TokenType::ChunkString);
+ tokenizer.exit(TokenType::DefinitionDestinationString);
+ enclosed_before(tokenizer, code)
+ }
+ Code::None | Code::CarriageReturnLineFeed | Code::Char('\r' | '\n' | '<') => {
+ (State::Nok, None)
+ }
+ Code::Char('\\') => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(enclosed_escape)), None)
+ }
+ _ => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(enclosed)), None)
+ }
+ }
+}
+
+/// To do.
+fn enclosed_escape(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::Char('<' | '>' | '\\') => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(enclosed)), None)
+ }
+ _ => enclosed(tokenizer, code),
+ }
+}
+
+/// To do.
+// To do: these arms can be improved?
+fn raw(tokenizer: &mut Tokenizer, code: Code, balance: usize) -> StateFnResult {
+ // To do: configurable.
+ let limit = usize::MAX;
+
+ match code {
+ Code::Char('(') if balance >= limit => (State::Nok, None),
+ Code::Char('(') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| raw(t, c, balance + 1))),
+ None,
+ )
+ }
+ Code::Char(')') if balance == 0 => {
+ tokenizer.exit(TokenType::ChunkString);
+ tokenizer.exit(TokenType::DefinitionDestinationString);
+ tokenizer.exit(TokenType::DefinitionDestinationRaw);
+ tokenizer.exit(TokenType::DefinitionDestination);
+ (State::Ok, Some(vec![code]))
+ }
+ Code::Char(')') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| raw(t, c, balance - 1))),
+ None,
+ )
+ }
+ Code::None
+ | Code::CarriageReturnLineFeed
+ | Code::VirtualSpace
+ | Code::Char('\t' | '\r' | '\n' | ' ')
+ if balance > 0 =>
+ {
+ (State::Nok, None)
+ }
+ Code::None
+ | Code::CarriageReturnLineFeed
+ | Code::VirtualSpace
+ | Code::Char('\t' | '\r' | '\n' | ' ') => {
+ tokenizer.exit(TokenType::ChunkString);
+ tokenizer.exit(TokenType::DefinitionDestinationString);
+ tokenizer.exit(TokenType::DefinitionDestinationRaw);
+ tokenizer.exit(TokenType::DefinitionDestination);
+ (State::Ok, Some(vec![code]))
+ }
+ Code::Char(char) if char.is_ascii_control() => (State::Nok, None),
+ Code::Char('\\') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| raw_escape(t, c, balance))),
+ None,
+ )
+ }
+ Code::Char(_) => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(move |t, c| raw(t, c, balance))), None)
+ }
+ }
+}
+
+/// To do.
+fn raw_escape(tokenizer: &mut Tokenizer, code: Code, balance: usize) -> StateFnResult {
+ match code {
+ Code::Char('(' | ')' | '\\') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| raw(t, c, balance + 1))),
+ None,
+ )
+ }
+ _ => raw(tokenizer, code, balance),
+ }
+}
diff --git a/src/construct/partial_label.rs b/src/construct/partial_label.rs
new file mode 100644
index 0000000..c772c56
--- /dev/null
+++ b/src/construct/partial_label.rs
@@ -0,0 +1,100 @@
+// To do: pass token types in.
+
+use crate::constant::LINK_REFERENCE_SIZE_MAX;
+use crate::tokenizer::{Code, State, StateFnResult, TokenType, Tokenizer};
+
+/// To do.
+pub fn start(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ match code {
+ Code::Char('[') => {
+ tokenizer.enter(TokenType::DefinitionLabel);
+ tokenizer.enter(TokenType::DefinitionLabelMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionLabelMarker);
+ tokenizer.enter(TokenType::DefinitionLabelData);
+ (State::Fn(Box::new(|t, c| at_break(t, c, false, 0))), None)
+ }
+ // To do: allow?
+ _ => unreachable!("expected `[` at start of label"),
+ }
+}
+
+/// To do.
+fn at_break(tokenizer: &mut Tokenizer, code: Code, data: bool, size: usize) -> StateFnResult {
+ match code {
+ Code::None | Code::Char('[') => (State::Nok, None),
+ Code::Char(']') if !data => (State::Nok, None),
+ _ if size > LINK_REFERENCE_SIZE_MAX => (State::Nok, None),
+ Code::Char(']') => {
+ tokenizer.exit(TokenType::DefinitionLabelData);
+ tokenizer.enter(TokenType::DefinitionLabelMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionLabelMarker);
+ tokenizer.exit(TokenType::DefinitionLabel);
+ (State::Ok, None)
+ }
+ Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ tokenizer.enter(TokenType::LineEnding);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::LineEnding);
+ (
+ State::Fn(Box::new(move |t, c| at_break(t, c, data, size))),
+ None,
+ )
+ }
+ _ => {
+ tokenizer.enter(TokenType::ChunkString);
+ // To do: link.
+ label(tokenizer, code, data, size)
+ }
+ }
+}
+
+/// To do.
+fn label(tokenizer: &mut Tokenizer, code: Code, data: bool, size: usize) -> StateFnResult {
+ match code {
+ Code::None | Code::CarriageReturnLineFeed | Code::Char('\r' | '\n' | '[' | ']') => {
+ tokenizer.exit(TokenType::ChunkString);
+ at_break(tokenizer, code, data, size)
+ }
+ _ if size > LINK_REFERENCE_SIZE_MAX => {
+ tokenizer.exit(TokenType::ChunkString);
+ at_break(tokenizer, code, data, size)
+ }
+ Code::VirtualSpace | Code::Char('\t' | ' ') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| label(t, c, data, size + 1))),
+ None,
+ )
+ }
+ Code::Char('/') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| escape(t, c, true, size + 1))),
+ None,
+ )
+ }
+ Code::Char(_) => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| label(t, c, true, size + 1))),
+ None,
+ )
+ }
+ }
+}
+
+/// To do.
+fn escape(tokenizer: &mut Tokenizer, code: Code, data: bool, size: usize) -> StateFnResult {
+ match code {
+ Code::Char('[' | '\\' | ']') => {
+ tokenizer.consume(code);
+ (
+ State::Fn(Box::new(move |t, c| label(t, c, true, size + 1))),
+ None,
+ )
+ }
+ _ => label(tokenizer, code, data, size),
+ }
+}
diff --git a/src/construct/partial_title.rs b/src/construct/partial_title.rs
new file mode 100644
index 0000000..4c7b527
--- /dev/null
+++ b/src/construct/partial_title.rs
@@ -0,0 +1,136 @@
+// To do: pass token types in.
+
+use crate::construct::partial_whitespace::start as whitespace;
+use crate::tokenizer::{Code, State, StateFnResult, TokenType, Tokenizer};
+
+/// Type of quote, if we’re in an attribure, in complete (condition 7).
+#[derive(Debug, Clone, PartialEq)]
+enum TitleKind {
+ /// In a parenthesised (`(` and `)`) title.
+ Paren,
+ /// In a double quoted (`"`) title.
+ Double,
+ /// In a single quoted (`"`) title.
+ Single,
+}
+
+fn kind_to_marker(kind: &TitleKind) -> char {
+ match kind {
+ TitleKind::Double => '"',
+ TitleKind::Single => '\'',
+ TitleKind::Paren => ')',
+ }
+}
+
+pub fn start(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+ let kind = match code {
+ Code::Char('"') => Some(TitleKind::Double),
+ Code::Char('\'') => Some(TitleKind::Single),
+ Code::Char('(') => Some(TitleKind::Paren),
+ _ => None,
+ };
+
+ if let Some(kind) = kind {
+ tokenizer.enter(TokenType::DefinitionTitle);
+ tokenizer.enter(TokenType::DefinitionTitleMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionTitleMarker);
+ (State::Fn(Box::new(|t, c| at_first_break(t, c, kind))), None)
+ } else {
+ (State::Nok, None)
+ }
+}
+
+/// To do.
+fn at_first_break(tokenizer: &mut Tokenizer, code: Code, kind: TitleKind) -> StateFnResult {
+ match code {
+ Code::Char(char) if char == kind_to_marker(&kind) => {
+ tokenizer.enter(TokenType::DefinitionTitleMarker);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::DefinitionTitleMarker);
+ tokenizer.exit(TokenType::DefinitionTitle);
+ (State::Ok, None)
+ }
+ _ => {
+ tokenizer.enter(TokenType::DefinitionTitleString);
+ at_break(tokenizer, code, kind)
+ }
+ }
+}
+
+/// To do.
+fn at_break(tokenizer: &mut Tokenizer, code: Code, kind: TitleKind) -> StateFnResult {
+ match code {
+ Code::Char(char) if char == kind_to_marker(&kind) => {
+ tokenizer.exit(TokenType::DefinitionTitleString);
+ at_first_break(tokenizer, code, kind)
+ }
+ Code::None => (State::Nok, None),
+ Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ tokenizer.enter(TokenType::LineEnding);
+ tokenizer.consume(code);
+ tokenizer.exit(TokenType::LineEnding);
+ (
+ State::Fn(Box::new(|t, c| at_break_line_start(t, c, kind))),
+ None,
+ )
+ }
+ _ => {
+ // To do: link.
+ tokenizer.enter(TokenType::ChunkString);
+ title(tokenizer, code, kind)
+ }
+ }
+}
+
+fn at_break_line_start(tokenizer: &mut Tokenizer, code: Code, kind: TitleKind) -> StateFnResult {
+ tokenizer.attempt(
+ |t, c| whitespace(t, c, TokenType::Whitespace),
+ |_ok| Box::new(|t, c| at_break_line_begin(t, c, kind)),
+ )(tokenizer, code)
+}
+
+fn at_break_line_begin(tokenizer: &mut Tokenizer, code: Code, kind: TitleKind) -> StateFnResult {
+ match code {
+ // Blank line not allowed.
+ Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => (State::Nok, None),
+ _ => at_break(tokenizer, code, kind),
+ }
+}
+
+/// To do.
+fn title(tokenizer: &mut Tokenizer, code: Code, kind: TitleKind) -> StateFnResult {
+ match code {
+ Code::Char(char) if char == kind_to_marker(&kind) => {
+ tokenizer.exit(TokenType::ChunkString);
+ at_break(tokenizer, code, kind)
+ }
+ Code::None | Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+ tokenizer.exit(TokenType::ChunkString);
+ at_break(tokenizer, code, kind)
+ }
+ Code::Char('\\') => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(|t, c| escape(t, c, kind))), None)
+ }
+ _ => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(|t, c| title(t, c, kind))), None)
+ }
+ }
+}
+
+/// To do.
+fn escape(tokenizer: &mut Tokenizer, code: Code, kind: TitleKind) -> StateFnResult {
+ match code {
+ Code::Char(char) if char == kind_to_marker(&kind) => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(move |t, c| title(t, c, kind))), None)
+ }
+ Code::Char('\\') => {
+ tokenizer.consume(code);
+ (State::Fn(Box::new(move |t, c| title(t, c, kind))), None)
+ }
+ _ => title(tokenizer, code, kind),
+ }
+}