switch from NFKC to NFC normalization form in the lettering tool (#3828)

* switch from NFKC to NFC

* Additional normalization.

---------

Co-authored-by: CapellanCitizen <>

4 files changed, 11 insertions, 6 deletions

diff --git a/lib/gui/lettering/main_panel.py b/lib/gui/lettering/main_panel.py
index 64312b5a..f0532560 100644
--- a/lib/gui/lettering/main_panel.py
+++ b/lib/gui/lettering/main_panel.py
@@ -5,6 +5,7 @@
 
 import json
 from base64 import b64decode
+import unicodedata
 
 import inkex
 import wx
@@ -126,11 +127,11 @@ class LetteringPanel(wx.Panel):
         filter_glyph = self.options_panel.font_glyph_filter.GetValue()
         filter_category = self.options_panel.font_category_filter.GetSelection() - 1
 
-        # glyph filter string without spaces
-        glyphs = [*self.options_panel.text_editor.GetValue().replace(" ", "").replace("\n", "")]
+        # Set of all glyphs in input string (except whitespace characters), normalized in the same way that we normalize font glyphs
+        glyphs = set(*unicodedata.normalize("NFC", self.options_panel.text_editor.GetValue().replace(r"\s", "")))
 
         for font in self.font_list:
-            if filter_glyph and glyphs and not set(glyphs).issubset(font.available_glyphs):
+            if filter_glyph and glyphs and not glyphs.issubset(font.available_glyphs):
                 continue
 
             if filter_category != -1:
diff --git a/lib/lettering/font.py b/lib/lettering/font.py
index d1ed6154..5ae6743a 100644
--- a/lib/lettering/font.py
+++ b/lib/lettering/font.py
@@ -8,6 +8,7 @@ import os
 from collections import defaultdict
 from copy import deepcopy
 from random import randint
+import unicodedata
 
 import inkex
 
@@ -217,6 +218,9 @@ class Font(object):
         """Render text into an SVG group element."""
         self._load_variants()
 
+        # Normalize the text in the same way that glyph names are normalized (NFC)
+        text = unicodedata.normalize('NFC', text)
+
         if variant is None:
             variant = self.default_variant
 
diff --git a/lib/lettering/font_variant.py b/lib/lettering/font_variant.py
index 4862b00b..80d9d3e6 100644
--- a/lib/lettering/font_variant.py
+++ b/lib/lettering/font_variant.py
@@ -74,7 +74,7 @@ class FontVariant(object):
             for layer in glyph_layers:
                 self._clean_group(layer)
                 layer.attrib[INKSCAPE_LABEL] = layer.attrib[INKSCAPE_LABEL].replace("GlyphLayer-", "", 1)
-                glyph_name = normalize('NFKC', layer.attrib[INKSCAPE_LABEL])
+                glyph_name = normalize('NFC', layer.attrib[INKSCAPE_LABEL])
                 try:
                     self.glyphs[glyph_name] = Glyph(layer)
                 except (AttributeError, ValueError):
@@ -144,7 +144,7 @@ class FontVariant(object):
         # binding glyph only have  two shapes, isol and fina
 
         non_binding_char = ['ا', 'أ', 'ﺇ', 'آ', 'ٱ', 'د', 'ذ', 'ر', 'ز', 'و', 'ؤ']
-        normalized_non_binding_char = [normalize('NFKC', letter) for letter in non_binding_char]
+        normalized_non_binding_char = [normalize('NFC', letter) for letter in non_binding_char]
         return not (character in normalized_non_binding_char)
 
     def is_mark(self, character):
diff --git a/lib/lettering/glyph.py b/lib/lettering/glyph.py
index fd1e1985..98f69ecb 100644
--- a/lib/lettering/glyph.py
+++ b/lib/lettering/glyph.py
@@ -39,7 +39,7 @@ class Glyph(object):
 
         self.name = group.label
         if len(self.name) > 11:
-            self.name = normalize('NFKC', self.name[11:])
+            self.name = normalize('NFC', self.name[11:])
         self._process_baseline(group.getroottree().getroot())
         self.clips = self._process_clips(group)
         self.node = self._process_group(group)