From 58ba69452a25c3d4b2059c01cc6cd837159d2f90 Mon Sep 17 00:00:00 2001
From: Titus Wormer <tituswormer@gmail.com>
Date: Thu, 16 Jun 2022 11:34:35 +0200
Subject: Add support for hard break escape

---
 src/construct/character_escape.rs  | 11 ++++---
 src/construct/hard_break_escape.rs | 61 ++++++++++++++++++++++++++++++++++++++
 src/construct/mod.rs               |  3 +-
 3 files changed, 68 insertions(+), 7 deletions(-)
 create mode 100644 src/construct/hard_break_escape.rs

(limited to 'src/construct')

diff --git a/src/construct/character_escape.rs b/src/construct/character_escape.rs
index 7bab42d..baedd4b 100644
--- a/src/construct/character_escape.rs
+++ b/src/construct/character_escape.rs
@@ -11,11 +11,11 @@
 //! slash, or a slash followed by anything other than an ASCII punctuation
 //! character, is exactly that: just a slash.
 //! To escape (most) arbitrary characters, use a
-//! [character reference][] instead
+//! [character reference][character_reference] instead
 //! (as in, `&amp;`, `&#123;`, or say `&#x9;`).
 //! It is also possible to escape a line ending in text with a similar
-//! construct: a backslash followed by a line ending (that is part of the
-//! construct instead of ending it).
+//! construct: a [hard break escape][hard_break_escape] is a backslash followed
+//! by a line ending (that is part of the construct instead of ending it).
 //!
 //! ## References
 //!
@@ -24,9 +24,8 @@
 //!
 //! [string]: crate::content::string
 //! [text]: crate::content::text
-//! [character reference]: crate::construct::character_reference
-//!
-//! <!-- To do: link `hard_break_escape` -->
+//! [character_reference]: crate::construct::character_reference
+//! [hard_break_escape]: crate::construct::hard_break_escape
 
 use crate::tokenizer::{Code, State, StateFnResult, TokenType, Tokenizer};
 
diff --git a/src/construct/hard_break_escape.rs b/src/construct/hard_break_escape.rs
new file mode 100644
index 0000000..a7712d6
--- /dev/null
+++ b/src/construct/hard_break_escape.rs
@@ -0,0 +1,61 @@
+//! Hard break escapes are a construct that occurs in the  [text][] content
+//! type.
+//!
+//! They’re formed with the following BNF:
+//!
+//! ```bnf
+//! ; Restriction: followed by a line ending  (that is part of the construct
+//! ; instead of ending it).
+//! hard_break_escape ::= '\\'
+//! ```
+//! It is also possible to escape punctuation characters with a similar
+//! construct: a [character escape][character_escape] is a backslash followed
+//! by an ASCII punctuation character.
+//! Arbitrary characters can be escaped with
+//! [character reference][character_reference]s.
+//!
+//! ## References
+//!
+//! *   [`hard-break-escape.js` in `micromark`](https://github.com/micromark/micromark/blob/main/packages/micromark-core-commonmark/dev/lib/hard-break-escape.js)
+//! *   [*§ 6.7 Hard line breaks* in `CommonMark`](https://spec.commonmark.org/0.30/#hard-line-breaks)
+//!
+//! [text]: crate::content::text
+//! [character_escape]: crate::construct::character_escape
+//! [character_reference]: crate::construct::character_reference
+//!
+//! <!-- To do: link `hard_break_escape` -->
+
+use crate::tokenizer::{Code, State, StateFnResult, TokenType, Tokenizer};
+
+/// Start of a hard break escape.
+///
+/// ```markdown
+/// a|\
+/// ```
+pub fn start(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+    match code {
+        Code::Char('\\') => {
+            tokenizer.enter(TokenType::HardBreakEscape);
+            tokenizer.enter(TokenType::HardBreakEscapeMarker);
+            tokenizer.consume(code);
+            tokenizer.exit(TokenType::HardBreakEscapeMarker);
+            (State::Fn(Box::new(inside)), None)
+        }
+        _ => (State::Nok, None),
+    }
+}
+
+/// At the end of a hard break escape, after `\`.
+///
+/// ```markdown
+/// a\|
+/// ```
+fn inside(tokenizer: &mut Tokenizer, code: Code) -> StateFnResult {
+    match code {
+        Code::CarriageReturnLineFeed | Code::Char('\r' | '\n') => {
+            tokenizer.exit(TokenType::HardBreakEscape);
+            (State::Ok, Some(vec![code]))
+        }
+        _ => (State::Nok, None),
+    }
+}
diff --git a/src/construct/mod.rs b/src/construct/mod.rs
index 1fa57d5..27f4308 100644
--- a/src/construct/mod.rs
+++ b/src/construct/mod.rs
@@ -26,7 +26,7 @@
 //! *   [code (text)][code_text]
 //! *   content
 //! *   definition
-//! *   hard break escape
+//! *   [hard break escape][hard_break_escape]
 //! *   [heading (atx)][heading_atx]
 //! *   heading (setext)
 //! *   [html (flow)][html_flow]
@@ -60,6 +60,7 @@ pub mod character_reference;
 pub mod code_fenced;
 pub mod code_indented;
 pub mod code_text;
+pub mod hard_break_escape;
 pub mod heading_atx;
 pub mod html_flow;
 pub mod html_text;
-- 
cgit