Commit: b78946ce53d23b0225ed0c09b64165b935f45b0d - luce (git)

A Markdown shard for the Crystal programming language

Commit MetaInfo

• Revision: b78946ce53d23b0225ed0c09b64165b935f45b0d (tree)
• Zeit: 2023-11-26 09:53:57
• Autor: supercell <stigma@disr...>
• Commiter: supercell

Log Message

Fix start of line detection in AutolinkExtension

Ändern Zusammenfassung

• modified: CHANGELOG.md
• modified: spec/extensions/autolink_extension.unit
• modified: src/luce/inline_syntaxes/autolink_extension_syntax.cr

Diff

--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -10,6 +10,7 @@ All notable changes to Luce will be documented in this file.
   * This will raise an exception should an infinite loop begin during `BlockSyntax#parse_lines`.
 * Fix an obscure issue with `HTMLBlockSyntax`.
 * Fix tables not rendering when there are trailing spaces.
+* Fix beginning of line detection in `AutolinkExtensionSyntax`.
 
 ## [0.4.0] - 2023-05-23
 

--- a/spec/extensions/autolink_extension.unit
+++ b/spec/extensions/autolink_extension.unit
@@ -1,4 +1,10 @@
 >>> not a link 
 mhttp://www.foo.com
 <<<
-<p>mhttp://www.foo.com</p>
\ No newline at end of file
+<p>mhttp://www.foo.com</p>
+>>> following a newline
+m
+http://www.foo.com
+<<<
+<p>m
+<a href="http://www.foo.com">http://www.foo.com</a></p>
\ No newline at end of file

--- a/src/luce/inline_syntaxes/autolink_extension_syntax.cr
+++ b/src/luce/inline_syntaxes/autolink_extension_syntax.cr
@@ -52,12 +52,12 @@ module Luce
         return false
       end
 
-      # When it is a link and it is not preceded by `*`, `_`, `~`, `(`, or `>`,
-      # it is invalid. See
-      # https://github.github.io/gfm/#extended-autolink-path-validation
+      # When it is a link and it is not at the beginning of a line, or preceded
+      # by whitespace, `*`, `_`, `~`, `(`, or `>`, it is invalid. See
+      # https://github.github.io/gfm/#autolinks-extension-.
       if start_match[1]? != nil && parser.pos > 0
         preceded_by = parser.char_at(parser.pos - 1).chr
-        valid_preceding_chars = [' ', '*', '_', '~', '(', '>']
+        valid_preceding_chars = ['\n', ' ', '*', '_', '~', '(', '>']
         return false unless valid_preceding_chars.includes? preceded_by
       end