From 3d760ba9e47d356cabc548c46c99a5d25865cb03 Mon Sep 17 00:00:00 2001
From: Markus Lehtonen <markus.lehtonen@intel.com>
Date: Wed, 6 Sep 2023 23:58:58 +0300
Subject: [PATCH] MD027: handle anchor elements correctly (#463)

Fixex false positives of MD027 multiple-spaces-after-blockquote-symbol
if quote lines begin with a link, e.g.

  > This is a
  > [link](link) to somewhere

Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
---
 lib/mdl/doc.rb                       | 25 +++++++++++++++++++++++++
 lib/mdl/rules.rb                     |  6 +++++-
 test/rule_tests/blockquote_spaces.md |  2 ++
 3 files changed, 32 insertions(+), 1 deletion(-)

diff --git a/lib/mdl/doc.rb b/lib/mdl/doc.rb
index 844f4741..6c35678b 100644
--- a/lib/mdl/doc.rb
+++ b/lib/mdl/doc.rb
@@ -268,6 +268,31 @@ def extract_text(element, prefix = '', restore_whitespace = true)
       lines
     end
 
+    ##
+    # Returns the element as plaintext
+
+    def extract_as_text(element)
+      quotes = {
+        :rdquo => '"',
+        :ldquo => '"',
+        :lsquo => "'",
+        :rsquo => "'",
+      }
+      # If anything goes amiss here, e.g. unknown type, then nil will be
+      # returned and we'll just not catch that part of the text, which seems
+      # like a sensible failure mode.
+      lines = element.children.map do |e|
+        if e.type == :text or e.type == :codespan
+          e.value
+        elsif %i{strong em p  a}.include?(e.type)
+          extract_as_text(e).join("\n")
+        elsif e.type == :smart_quote
+          quotes[e.value]
+        end
+      end.join.split("\n")
+      lines
+    end
+
     private
 
     ##
diff --git a/lib/mdl/rules.rb b/lib/mdl/rules.rb
index 4544611f..cc1dfff5 100644
--- a/lib/mdl/rules.rb
+++ b/lib/mdl/rules.rb
@@ -458,7 +458,11 @@
     errors = []
     doc.find_type_elements(:blockquote).each do |e|
       linenum = doc.element_linenumber(e)
-      lines = doc.extract_text(e, /^\s*> /)
+      lines = doc.extract_as_text(e)
+      # Handle first line specially as whitespace is stripped from the text element
+      if doc.element_line(e).match(/^\s*>  /)
+        errors << linenum
+      end
       lines.each do |line|
         errors << linenum if line.start_with?(' ')
         linenum += 1
diff --git a/test/rule_tests/blockquote_spaces.md b/test/rule_tests/blockquote_spaces.md
index 214a3a46..a54cf66c 100644
--- a/test/rule_tests/blockquote_spaces.md
+++ b/test/rule_tests/blockquote_spaces.md
@@ -17,6 +17,8 @@ This tests other things embedded in the blockquote:
 > **bar** more text
 > 'baz' more text
 > `qux` more text
+> [link](example.com) to site
+>  [link](#link) {MD027}
 >
 > - foo