Syntax: Refactor autolink contexts

This commit is a major overhaul of autolink contexts It addresses: - sublimehq/Packages#1930 - sublimehq/Packages#2542
SublimeText-Markdown · Sep 18, 2021 · 8d36e15 · 8d36e15
1 parent 16cd86f
commit 8d36e15
Show file tree

Hide file tree

Showing 3 changed files with 273 additions and 42 deletions.
diff --git a/messages/3.0.3.md b/messages/3.0.3.md
@@ -17,9 +17,11 @@ feedback you can use [GitHub issues][issues].
 * Remove obsolete keymap selectors (required due to recent syntax changes)
 * Add a macro to unbold bold italc text (required due to recent syntax changes)
 * Only strip whitespace separated trailing hashes of headings from symbol lists
+* Scope inet/email autolinks according to CommonMark 0.30.0
 
 ## New Features
 
+* Scope path separators and escapes in urls
 * Support fish fenced code (if supported syntax is installed)
 * Partially support xonsh fenced code (use Python syntax due to a lack of xonsh syntax support in ST)
 

diff --git a/syntaxes/Markdown.sublime-syntax b/syntaxes/Markdown.sublime-syntax
@@ -138,6 +138,11 @@ variables:
         )
         \s*$          # any amount of whitespace until EOL
       )
+
+    # https://spec.commonmark.org/0.30/#email-autolink
+    email_domain_commonmark: '[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?'
+    email_user_commonmark: '[a-zA-Z0-9.!#$%&''*+/=?^_`{|}~-]+'
+
     html_tag_open_commonmark: |-
       (?xi:
         <
@@ -733,8 +738,8 @@ contexts:
     - include: code-span
     - include: image-inline
     - include: link-inline
-    - include: autolink-inet
     - include: autolink-email
+    - include: autolink-inet
     - include: image-ref
     - include: link-ref-wiki
     - include: link-ref-footnote
@@ -2435,13 +2440,29 @@ contexts:
     - include: link-url-common
 
   link-url-common:
+    - include: link-url-path-separators
+    - include: link-url-scheme-separators
+    - include: link-url-escapes
+    - include: paragraph-end
+
+  link-url-escapes:
     - match: (%)\h{2}
       scope: constant.character.escape.url.markdown
       captures:
         1: punctuation.definition.escape.markdown
-    - match: '[/&?#]|://'
+
+  link-url-path-separators:
+    - match: '[/&?#]'
+      scope: punctuation.separator.path.markdown
+
+  link-url-scheme-separators:
+    - match: ':/{,2}'
       scope: punctuation.separator.path.markdown
-    - include: paragraph-end
+
+  link-url-scheme-separator:
+    - match: ':/{,2}'
+      scope: punctuation.separator.path.markdown
+      pop: true
 
 ###[ LINK/IMAGE/REFERENCE ATTRIBUTES ]########################################
 
@@ -2501,43 +2522,87 @@ contexts:
 ###[ AUTOLINKS ]##############################################################
 
   autolink-email:
-    - match: (<)((?:mailto:)?[-+.\w]+@[-a-z0-9]+(\.[-a-z0-9]+)*\.[a-z]+)(>)
-      scope: meta.link.email.lt-gt.markdown
+    # CommonMark
+    # https://spec.commonmark.org/0.30/#email-autolink
+    - match: |-
+        (?x)
+        (<)
+        (
+          (?:mailto(:))?
+          {{email_user_commonmark}}
+          (@)
+          {{email_domain_commonmark}}(?:\.{{email_domain_commonmark}})*
+        )
+        (>)
       captures:
+        0: meta.link.email.markdown
         1: punctuation.definition.link.begin.markdown
         2: markup.underline.link.markdown
-        4: punctuation.definition.link.end.markdown
-    - match: '[\w.+-]+@[\w-]+(\.((?![._-][\W])[\w_-])+)+(?![_-])'
-      scope: markup.underline.link.markdown
+        3: punctuation.separator.path.markdown
+        4: punctuation.separator.path.markdown
+        5: punctuation.definition.link.end.markdown
+    # Github Flavoured Markdown
+    - match: '[\w.+-]+(@)[\w-]+(?:\.(?:(?![._-][\W])[\w_-])+)+(?![_-])'
+      captures:
+        0: meta.link.email.markdown markup.underline.link.markdown
+        1: punctuation.separator.path.markdown
 
   autolink-inet:
-    - match: (<)((?:https?|ftp)://.*?)(>)
-      scope: meta.link.inet.markdown
+    # CommonMark
+    # https://spec.commonmark.org/0.30/#autolinks
+    - match: <(?=[[:alpha:]][[:alnum:].+-]+:)
+      scope: punctuation.definition.link.begin.markdown
+      push:
+        - autolink-inet-angled-content
+        - link-url-scheme-separator
+    # Github Flavoured Markdown
+    # After a valid domain, zero or more non-space non-< characters may follow
+    - match: (?:(?:https|http|ftp)(://)|www\.)[\w-]+
       captures:
-        1: punctuation.definition.link.begin.markdown
-        2: markup.underline.link.markdown
-        3: punctuation.definition.link.end.markdown
-    - match: (((https|http|ftp)://)|www\.)[\w-]+(\.[\w-]+)+
-      scope: markup.underline.link.markdown-gfm
-      push: autolink-inet-body
+        1: punctuation.separator.path.markdown
+      push: autolink-inet-unquoted-content
 
-  autolink-inet-body:
-    # After a valid domain, zero or more non-space non-< characters may follow
-    - match: (?=[?!.,:*_~]*(?:[\s<]|$)) # Trailing punctuation (specifically, ?, !, ., ,, :, *, _, and ~) will not be considered part of the autolink, though they may be included in the interior of the link
+  autolink-inet-angled-content:
+    - meta_scope: meta.link.inet.markdown
+    - meta_content_scope: markup.underline.link.markdown
+    - match: \>
+      scope: punctuation.definition.link.end.markdown
       pop: true
-    - match: (?={{html_entity}}[?!.,:*_~]*[\s<]) # If an autolink ends in a semicolon (;), we check to see if it appears to resemble an entity reference; if the preceding text is & followed by one or more alphanumeric characters. If so, it is excluded from the autolink
+    # Spaces are not allowed in autolinks
+    - match: (?=\s)
       pop: true
-    - match: \( # When an autolink ends in ), we scan the entire autolink for the total number of parentheses. If there is a greater number of closing parentheses than opening ones, we don’t consider the last character part of the autolink, in order to facilitate including an autolink inside a parenthesis
-      push:
-        - meta_scope: markup.underline.link.markdown-gfm
-        - match: (?=[?!.,:*_~]*[\s<])
-          pop: true
-        - match: \)
-          pop: true
+    - include: autolink-inet-common
+
+  autolink-inet-unquoted-content:
+    - meta_scope: meta.link.inet.markdown markup.underline.link.markdown-gfm
+    # Trailing punctuation (specifically, ?, !, ., ,, :, *, _, and ~) will not be
+    # considered part of the autolink, though they may be included in the interior
+    # of the link
+    - match: (?=[?!.,:*_~]*[\s<])
+      pop: true
+    # If an autolink ends in a semicolon (;), we check to see if it appears to
+    # resemble an entity reference; if the preceding text is & followed by one
+    # or more alphanumeric characters. If so, it is excluded from the autolink
+    - match: (?={{html_entity}})
+      pop: true
+    # When an autolink ends in ), we scan the entire autolink for the total number
+    # of parentheses. If there is a greater number of closing parentheses than
+    # opening ones, we don’t consider the last character part of the autolink,
+    # in order to facilitate including an autolink inside a parenthesis
     - match: (?=\)[?!.,:*_~]*[\s<])
       pop: true
-    - match: '[^?!.,:*_~\s<&()]+|\S'
-      scope: markup.underline.link.markdown-gfm
+    - match: \(
+      push: autolink-inet-group
+    - include: autolink-inet-common
+
+  autolink-inet-group:
+    - match: \)
+      pop: true
+    - include: autolink-inet-unquoted-content
+
+  autolink-inet-common:
+    - include: link-url-path-separators
+    - include: link-url-escapes
 
 ###[ TABLE ]##################################################################