Match abbreviations across line breaks and other whitespace

gettalong · Nov 17, 2024 · 9ac9402 · 9ac9402
1 parent 6f53282
commit 9ac9402
Show file tree

Hide file tree

Showing 5 changed files with 24 additions and 1 deletion.
diff --git a/lib/kramdown/parser/base.rb b/lib/kramdown/parser/base.rb
@@ -54,6 +54,12 @@ def initialize(source, options)
         @options = Kramdown::Options.merge(options)
         @root = Element.new(:root, nil, nil, encoding: (source.encoding rescue nil), location: 1,
                             options: {}, abbrev_defs: {}, abbrev_attr: {})
+
+        @root.options[:abbrev_defs].default_proc = @root.options[:abbrev_attr].default_proc =
+          lambda do |h, k|
+            k_mod = k.gsub(/[\s\p{Z}]+/, " ")
+            k != k_mod ? h[k_mod] : nil
+          end
         @warnings = []
         @text_type = :text
       end

diff --git a/lib/kramdown/parser/kramdown/abbreviation.rb b/lib/kramdown/parser/kramdown/abbreviation.rb
@@ -42,7 +42,9 @@ def replace_abbreviations(el, regexps = nil)
         return if @root.options[:abbrev_defs].empty?
         unless regexps
           sorted_abbrevs = @root.options[:abbrev_defs].keys.sort {|a, b| b.length <=> a.length }
-          regexps = [Regexp.union(*sorted_abbrevs.map {|k| /#{Regexp.escape(k)}/ })]
+          regexps = [Regexp.union(*sorted_abbrevs.map do |k|
+            /#{Regexp.escape(k).gsub(/\\\s/, "[\\s\\p{Z}]+").force_encoding(Encoding::UTF_8)}/
+          end)]
           regexps << /(?=(?:\W|^)#{regexps.first}(?!\w))/ # regexp should only match on word boundaries
         end
         el.children.map! do |child|

diff --git a/test/test_files.rb b/test/test_files.rb
@@ -282,7 +282,11 @@ def tidy_output(out)
       define_method("test_whether_#{conv_class}_modifies_tree_with_file_#{text_file.tr('.', '_')}") do
         doc = Kramdown::Document.new(File.read(text_file), options)
         options_before = Marshal.load(Marshal.dump(doc.options))
+        abbrev_proc = doc.root.options[:abbrev_defs].default_proc
+        doc.root.options[:abbrev_defs].default_proc = doc.root.options[:abbrev_attr].default_proc = nil
         tree_before = Marshal.load(Marshal.dump(doc.root))
+        doc.root.options[:abbrev_defs].default_proc = doc.root.options[:abbrev_attr].default_proc =
+          abbrev_proc
         Kramdown::Converter.const_get(conv_class).convert(doc.root, doc.options)
         assert_equal(options_before, doc.options)
         assert_tree_not_changed(tree_before, doc.root)

diff --git a/test/testcases/span/abbreviations/abbrev.html b/test/testcases/span/abbreviations/abbrev.html
@@ -19,3 +19,7 @@
 
 <p>This is <abbr class="testit test" title="Some text here">awesome</abbr>.</p>
 
+<p>hello <abbr title="baz">foo bar</abbr> babble.</p>
+
+<p>hello <abbr title="baz">foo
+bar</abbr> babble.</p>
diff --git a/test/testcases/span/abbreviations/abbrev.text b/test/testcases/span/abbreviations/abbrev.text
@@ -32,3 +32,10 @@ This is awesome.
 {:.testit}
 *[awesome]: Some text here
 {:.test}
+
+*[foo bar]: baz
+
+hello foo bar babble.
+
+hello foo
+bar babble.