Add support for `mailto:`, `xmpp:` protocols

5 files changed, 167 insertions, 16 deletions

diff --git a/src/compiler.rs b/src/compiler.rs
index 0ea1638..b271768 100644
--- a/src/compiler.rs
+++ b/src/compiler.rs
@@ -430,9 +430,11 @@ fn exit(context: &mut CompileContext) {
         Name::DefinitionTitleString => on_exit_definition_title_string(context),
         Name::Emphasis => on_exit_emphasis(context),
         Name::Frontmatter => on_exit_frontmatter(context),
+        Name::GfmAutolinkLiteralEmail => on_exit_gfm_autolink_literal_email(context),
+        Name::GfmAutolinkLiteralMailto => on_exit_gfm_autolink_literal_mailto(context),
         Name::GfmAutolinkLiteralProtocol => on_exit_gfm_autolink_literal_protocol(context),
         Name::GfmAutolinkLiteralWww => on_exit_gfm_autolink_literal_www(context),
-        Name::GfmAutolinkLiteralEmail => on_exit_gfm_autolink_literal_email(context),
+        Name::GfmAutolinkLiteralXmpp => on_exit_gfm_autolink_literal_xmpp(context),
         Name::GfmFootnoteCall => on_exit_gfm_footnote_call(context),
         Name::GfmFootnoteDefinitionLabelString => {
             on_exit_gfm_footnote_definition_label_string(context);
@@ -1146,6 +1148,34 @@ fn on_exit_frontmatter(context: &mut CompileContext) {
     context.slurp_one_line_ending = true;
 }
 
+/// Handle [`Exit`][Kind::Exit]:[`GfmAutolinkLiteralEmail`][Name::GfmAutolinkLiteralEmail].
+fn on_exit_gfm_autolink_literal_email(context: &mut CompileContext) {
+    generate_autolink(
+        context,
+        Some("mailto:"),
+        Slice::from_position(
+            context.bytes,
+            &Position::from_exit_event(context.events, context.index),
+        )
+        .as_str(),
+        true,
+    );
+}
+
+/// Handle [`Exit`][Kind::Exit]:[`GfmAutolinkLiteralMailto`][Name::GfmAutolinkLiteralMailto].
+fn on_exit_gfm_autolink_literal_mailto(context: &mut CompileContext) {
+    generate_autolink(
+        context,
+        None,
+        Slice::from_position(
+            context.bytes,
+            &Position::from_exit_event(context.events, context.index),
+        )
+        .as_str(),
+        true,
+    );
+}
+
 /// Handle [`Exit`][Kind::Exit]:[`GfmAutolinkLiteralProtocol`][Name::GfmAutolinkLiteralProtocol].
 fn on_exit_gfm_autolink_literal_protocol(context: &mut CompileContext) {
     generate_autolink(
@@ -1174,11 +1204,11 @@ fn on_exit_gfm_autolink_literal_www(context: &mut CompileContext) {
     );
 }
 
-/// Handle [`Exit`][Kind::Exit]:[`GfmAutolinkLiteralEmail`][Name::GfmAutolinkLiteralEmail].
-fn on_exit_gfm_autolink_literal_email(context: &mut CompileContext) {
+/// Handle [`Exit`][Kind::Exit]:[`GfmAutolinkLiteralXmpp`][Name::GfmAutolinkLiteralXmpp].
+fn on_exit_gfm_autolink_literal_xmpp(context: &mut CompileContext) {
     generate_autolink(
         context,
-        Some("mailto:"),
+        None,
         Slice::from_position(
             context.bytes,
             &Position::from_exit_event(context.events, context.index),
diff --git a/src/construct/gfm_autolink_literal.rs b/src/construct/gfm_autolink_literal.rs
index 038330c..62f18ef 100644
--- a/src/construct/gfm_autolink_literal.rs
+++ b/src/construct/gfm_autolink_literal.rs
@@ -122,15 +122,20 @@
 //!
 //! ## Tokens
 //!
+//! *   [`GfmAutolinkLiteralEmail`][Name::GfmAutolinkLiteralEmail]
+//! *   [`GfmAutolinkLiteralMailto`][Name::GfmAutolinkLiteralMailto]
 //! *   [`GfmAutolinkLiteralProtocol`][Name::GfmAutolinkLiteralProtocol]
 //! *   [`GfmAutolinkLiteralWww`][Name::GfmAutolinkLiteralWww]
-//! *   [`GfmAutolinkLiteralEmail`][Name::GfmAutolinkLiteralEmail]
+//! *   [`GfmAutolinkLiteralXmpp`][Name::GfmAutolinkLiteralXmpp]
 //!
 //! ## References
 //!
 //! *   [`micromark-extension-gfm-autolink-literal`](https://github.com/micromark/micromark-extension-gfm-autolink-literal)
 //! *   [*§ 6.9 Autolinks (extension)* in `GFM`](https://github.github.com/gfm/#autolinks-extension-)
 //!
+//! > 👉 **Note**: `mailto:` and `xmpp:` protocols before email autolinks were
+//! > added in `cmark-gfm@0.29.0.gfm.5` and are as of yet undocumented.
+//!
 //! [text]: crate::construct::text
 //! [definition]: crate::construct::definition
 //! [attention]: crate::construct::attention
@@ -644,12 +649,17 @@ pub fn resolve(tokenizer: &mut Tokenizer) {
 
                 while byte_index < bytes.len() {
                     if bytes[byte_index] == b'@' {
-                        let mut range = (0, 0);
+                        let mut range = (0, 0, Name::GfmAutolinkLiteralEmail);
 
                         if let Some(start) = peek_bytes_atext(bytes, byte_index) {
-                            if let Some(end) = peek_bytes_email_domain(bytes, byte_index + 1) {
-                                let end = peek_bytes_truncate(bytes, start, end);
-                                range = (start, end);
+                            let (start, kind) = peek_protocol(bytes, start);
+
+                            if let Some(end) = peek_bytes_email_domain(
+                                bytes,
+                                byte_index + 1,
+                                kind == Name::GfmAutolinkLiteralXmpp,
+                            ) {
+                                range = (start, peek_bytes_truncate(bytes, start, end), kind);
                             }
                         }
 
@@ -678,7 +688,7 @@ pub fn resolve(tokenizer: &mut Tokenizer) {
                             // Add the link.
                             replace.push(Event {
                                 kind: Kind::Enter,
-                                name: Name::GfmAutolinkLiteralEmail,
+                                name: range.2.clone(),
                                 point: point.clone(),
                                 link: None,
                             });
@@ -686,7 +696,7 @@ pub fn resolve(tokenizer: &mut Tokenizer) {
                                 point.shift_to(tokenizer.parse_state.bytes, start_index + range.1);
                             replace.push(Event {
                                 kind: Kind::Exit,
-                                name: Name::GfmAutolinkLiteralEmail,
+                                name: range.2.clone(),
                                 point: point.clone(),
                                 link: None,
                             });
@@ -728,8 +738,6 @@ pub fn resolve(tokenizer: &mut Tokenizer) {
     }
 }
 
-// To do: add `xmpp`, `mailto` support.
-
 /// Move back past atext.
 ///
 /// Moving back is only used when post processing text: so for the email address
@@ -763,6 +771,40 @@ fn peek_bytes_atext(bytes: &[u8], end: usize) -> Option<usize> {
     }
 }
 
+/// Move back past a `mailto:` or `xmpp:` protocol.
+///
+/// Moving back is only used when post processing text: so for the email address
+/// algorithm.
+///
+/// ```markdown
+/// > | a mailto:contact@example.org b
+///              ^-- from
+///       ^-- to
+/// ```
+fn peek_protocol(bytes: &[u8], end: usize) -> (usize, Name) {
+    let mut index = end;
+
+    if index > 0 && bytes[index - 1] == b':' {
+        index -= 1;
+
+        // Take alphanumerical.
+        while index > 0 && matches!(bytes[index - 1], b'0'..=b'9' | b'A'..=b'Z' | b'a'..=b'z') {
+            index -= 1;
+        }
+
+        let slice = Slice::from_indices(bytes, index, end - 1);
+        let name = slice.as_str().to_ascii_lowercase();
+
+        if name == "xmpp" {
+            return (index, Name::GfmAutolinkLiteralXmpp);
+        } else if name == "mailto" {
+            return (index, Name::GfmAutolinkLiteralMailto);
+        }
+    }
+
+    (end, Name::GfmAutolinkLiteralEmail)
+}
+
 /// Move past email domain.
 ///
 /// Peeking like this only used when post processing text: so for the email
@@ -773,7 +815,7 @@ fn peek_bytes_atext(bytes: &[u8], end: usize) -> Option<usize> {
 ///               ^-- from
 ///                         ^-- to
 /// ```
-fn peek_bytes_email_domain(bytes: &[u8], start: usize) -> Option<usize> {
+fn peek_bytes_email_domain(bytes: &[u8], start: usize, xmpp: bool) -> Option<usize> {
     let mut index = start;
     let mut dot = false;
 
@@ -784,6 +826,7 @@ fn peek_bytes_email_domain(bytes: &[u8], start: usize) -> Option<usize> {
         match bytes[index] {
             // Alphanumerical, `-`, and `_`.
             b'-' | b'0'..=b'9' | b'A'..=b'Z' | b'_' | b'a'..=b'z' => {}
+            b'/' if xmpp => {}
             // Dot followed by alphanumerical (not `-` or `_`).
             b'.' if index + 1 < bytes.len()
                 && matches!(bytes[index + 1], b'0'..=b'9' | b'A'..=b'Z' | b'a'..=b'z') =>
diff --git a/src/construct/raw_flow.rs b/src/construct/raw_flow.rs
index 395d0ae..105a031 100644
--- a/src/construct/raw_flow.rs
+++ b/src/construct/raw_flow.rs
@@ -252,8 +252,7 @@ pub fn sequence_open(tokenizer: &mut Tokenizer) -> State {
         tokenizer.tokenize_state.size += 1;
         tokenizer.consume();
         State::Next(StateName::RawFlowSequenceOpen)
-    }
-    else if tokenizer.tokenize_state.size
+    } else if tokenizer.tokenize_state.size
         < (if tokenizer.tokenize_state.marker == b'$' {
             MATH_FLOW_SEQUENCE_SIZE_MIN
         } else {
diff --git a/src/event.rs b/src/event.rs
index 458bd19..fad2c64 100644
--- a/src/event.rs
+++ b/src/event.rs
@@ -986,6 +986,24 @@ pub enum Name {
     ///     ^^^^^^^^^^^^^^^^^^^
     /// ```
     GfmAutolinkLiteralEmail,
+    /// GFM extension: email autolink w/ explicit `mailto`.
+    ///
+    /// ## Info
+    ///
+    /// *   **Context**:
+    ///     [text content][crate::construct::text]
+    /// *   **Content model**:
+    ///     void.
+    /// *   **Construct**:
+    ///     [`gfm_autolink_literal`][crate::construct::gfm_autolink_literal]
+    ///
+    /// ## Example
+    ///
+    /// ```markdown
+    /// > | mailto:context@example.com
+    ///     ^^^^^^^^^^^^^^^^^^^^^^^^^^
+    /// ```
+    GfmAutolinkLiteralMailto,
     /// GFM extension: autolink w/ protocol.
     ///
     /// ## Info
@@ -1022,6 +1040,24 @@ pub enum Name {
     ///     ^^^^^^^^^^^^^^^
     /// ```
     GfmAutolinkLiteralWww,
+    /// GFM extension: email autolink w/ explicit `xmpp`.
+    ///
+    /// ## Info
+    ///
+    /// *   **Context**:
+    ///     [text content][crate::construct::text]
+    /// *   **Content model**:
+    ///     void.
+    /// *   **Construct**:
+    ///     [`gfm_autolink_literal`][crate::construct::gfm_autolink_literal]
+    ///
+    /// ## Example
+    ///
+    /// ```markdown
+    /// > | mailto:a@b.c/d
+    ///     ^^^^^^^^^^^^^^
+    /// ```
+    GfmAutolinkLiteralXmpp,
     /// GFM extension: whole footnote call.
     ///
     /// ## Info
diff --git a/tests/gfm_autolink_literal.rs b/tests/gfm_autolink_literal.rs
index 2e84e6d..0c646b6 100644
--- a/tests/gfm_autolink_literal.rs
+++ b/tests/gfm_autolink_literal.rs
@@ -228,6 +228,49 @@ fn gfm_autolink_literal() {
     assert_eq!(
         micromark_with_options(
             r###"
+mailto:scyther@pokemon.com
+
+This is a mailto:scyther@pokemon.com
+
+mailto:scyther@pokemon.com.
+
+mmmmailto:scyther@pokemon.com
+
+mailto:scyther@pokemon.com/
+
+mailto:scyther@pokemon.com/message
+
+mailto:scyther@pokemon.com/mailto:beedrill@pokemon.com
+
+xmpp:scyther@pokemon.com
+
+xmpp:scyther@pokemon.com.
+
+xmpp:scyther@pokemon.com/message
+
+xmpp:scyther@pokemon.com/message.
+
+Email me at:scyther@pokemon.com"###,
+            &gfm
+        ),
+        r###"<p><a href="mailto:scyther@pokemon.com">mailto:scyther@pokemon.com</a></p>
+<p>This is a <a href="mailto:scyther@pokemon.com">mailto:scyther@pokemon.com</a></p>
+<p><a href="mailto:scyther@pokemon.com">mailto:scyther@pokemon.com</a>.</p>
+<p>mmmmailto:<a href="mailto:scyther@pokemon.com">scyther@pokemon.com</a></p>
+<p><a href="mailto:scyther@pokemon.com">mailto:scyther@pokemon.com</a>/</p>
+<p><a href="mailto:scyther@pokemon.com">mailto:scyther@pokemon.com</a>/message</p>
+<p><a href="mailto:scyther@pokemon.com">mailto:scyther@pokemon.com</a>/<a href="mailto:beedrill@pokemon.com">mailto:beedrill@pokemon.com</a></p>
+<p><a href="xmpp:scyther@pokemon.com">xmpp:scyther@pokemon.com</a></p>
+<p><a href="xmpp:scyther@pokemon.com">xmpp:scyther@pokemon.com</a>.</p>
+<p><a href="xmpp:scyther@pokemon.com/message">xmpp:scyther@pokemon.com/message</a></p>
+<p><a href="xmpp:scyther@pokemon.com/message">xmpp:scyther@pokemon.com/message</a>.</p>
+<p>Email me at:<a href="mailto:scyther@pokemon.com">scyther@pokemon.com</a></p>"###,
+        "should support `mailto:` and `xmpp:` protocols"
+    );
+
+    assert_eq!(
+        micromark_with_options(
+            r###"
 a www.example.com&xxx;b c
 
 a www.example.com&xxx;. b