resolves asciidoctor#2485 convert string with non-ASCII characters to…

… NFD when applying smallcaps transformation

CHANGELOG.adoc

@@ -31,6 +31,7 @@ Enhancements::
 * add symbols for shift, command, option, and return keys to the fallback font
 * add support for `nowrap` and `nobreak` roles on formatted text
 * do not insert break opportunities into bare URL with `nobreak` role
+* convert string with non-ASCII characters to NFD when applying smallcaps transformation to support diacritics (#2485)
 
 Improvements::
 

docs/modules/theme/pages/text.adoc

@@ -144,44 +144,15 @@ It accepts the following keywords:
 capitalize:: Transforms the first letter of each word to a capital letter.
 lowercase:: Transforms all the text to lowercase letters.
 none:: Clears an inherited value and no case transformation is applied to the text.
-smallcaps:: Replaces lowercase ASCII letters (a-z) with their small capital variant.
-Lowercase letters outside the ASCII range are not transformed.
+smallcaps:: Normalizes text as NFD (decomposed normalized form) and replaces lowercase ASCII letters (a-z) with their small capital variant.
+The diacritic mark will be applied to the small capital letter using a combining character as a result of the text being normalized as NFD.
 uppercase:: Transforms all the text to capital letters.
 
 The `text-transform` key can't be set on the xref:base.adoc[base category].
 
-If you want the smallcaps transformation to support letters beyond the a-z range, you can do so by overridding the `smallcaps` method in an extended converter.
-
-[,ruby]
-----
-class MyPDFConverter < (Asciidoctor::Converter.for 'pdf')
-  register_for 'pdf'
-
-  def smallcaps string
-    string = super
-    string = string.gsub 'é', %(\u1d07\u0301)
-    string
-  end
-end
-----
-
-This transformation can be automated using `String#unicode_normalize` with the `:nfd` form.
-This method will rewrite all characters with diacritical marks so that the diacritical mark is added using a combining character (i.e., a two graphene form).
-
-[,ruby]
-----
-class MyPDFConverter < (Asciidoctor::Converter.for 'pdf')
-  register_for 'pdf'
-
-  def smallcaps string
-    string = string.unicode_normalize :nfd unless string.ascii_only?
-    super
-  end
-end
-----
-
-The smallcaps transformation for extended Latin characters (e.g., characters that include an accent) typically requires the addition of a combining character, such as the combining acute accent in the example above).
-Therefore, you must ensure that the font you're using supports these combining characters (meaning it provides the necessary glyphs).
+The smallcaps transformation for extended Latin characters (e.g., characters outside the a-z range that include an accent) require the use of a combining character, such as the combining acute accent.
+Therefore, if you want the smallcaps transformation to support letters beyond the a-z range, you need to ensure the font you're using provides the required https://en.wikipedia.org/wiki/Combining_character[combining characters^] to support diacritics (meaning it provides the necessary glyphs).
+Otherwise, these characters will appear as missing glyph boxes in the transformed text.
 
 [#border-background-color]
 == Border and background color 

lib/asciidoctor/pdf/text_transformer.rb

@@ -66,6 +66,7 @@ def smallcaps_pcdata string
       end
 
       def smallcaps string
+        string = string.unicode_normalize :nfd unless string.ascii_only?
         string.tr LowerAlphaChars, SmallCapsChars
       end
 

spec/formatted_text_formatter_spec.rb

@@ -1024,6 +1024,11 @@
       (expect pdf.lines).to eql ['HTML stands for HʏᴘᴇʀTᴇxᴛ Mᴀʀᴋᴜᴘ Lᴀɴɢᴜᴀɢᴇ']
     end
 
+    it 'should decompose non-ASCII characters when applying smallcaps text transform' do
+      pdf = to_pdf '== Références', pdf_theme: { heading_text_transform: 'smallcaps' }, analyze: true
+      (expect pdf.lines).to eql [%(R\u1d07\u0301ғ\u1d07\u0301ʀᴇɴᴄᴇs)]
+    end
+
     it 'should allow custom role to specify relative font size' do
       pdf_theme = {
         heading_h2_font_size: 24,