Add kerning to Type 3 font, ASCII strings in PDFs

sauerburger · sauerburger · commit f4115cb9ce8b · 2021-03-01T17:02:41.000+01:00
diff --git a/lib/matplotlib/backends/backend_pdf.py b/lib/matplotlib/backends/backend_pdf.py
@@ -2253,6 +2253,40 @@ def encode_string(self, s, fonttype):
             return s.encode('cp1252', 'replace')
         return s.encode('utf-16be', 'replace')
 
+    def build_kern_sequence(self, string, font, fontsize,
+                            kern_mode=KERNING_UNFITTED):
+        """
+        Return an ordered sequence of contiguous substrings interspersed with
+        integer kerning offsets.
+
+        The return value is compatible with the PDF TJ operator.
+        """
+        kern_sequence = [[]]
+        prev_glyph_idx = None
+        for char in string:
+            glyph_idx = font.get_char_index(ord(char))
+
+            # Absolute kerning in pt for actual font size
+            kern = (font.get_kerning(prev_glyph_idx, glyph_idx, kern_mode)
+                    if prev_glyph_idx is not None else 0) / 64
+
+            # Transform to thousandths of a unit of PDF text space
+            kern *= 1000 / fontsize
+
+            if kern:
+                # Position of char must be adjusted
+                kern_sequence[-1] = "".join(kern_sequence[-1])
+                kern_sequence.append(-kern)  # TJ operands are 'subtracted'
+                kern_sequence.append([char])  # Start new sequence
+            else:
+                # No kerning required, append to running sequence
+                kern_sequence[-1].append(char)
+
+            prev_glyph_idx = glyph_idx
+
+        return ["".join(subs) if isinstance(subs, list) else subs
+                for subs in kern_sequence]
+
     def draw_text(self, gc, x, y, s, prop, angle, ismath=False, mtext=None):
         # docstring inherited
 
@@ -2287,14 +2321,26 @@ def draw_text(self, gc, x, y, s, prop, angle, ismath=False, mtext=None):
             }
             self.file._annotations[-1][1].append(link_annotation)
 
-        # If fonttype != 3 or there are no multibyte characters, emit the whole
-        # string at once.
-        if fonttype != 3 or all(ord(char) <= 255 for char in s):
+        # If fonttype != 3 emit the whole string at once.
+        if fonttype != 3:
+            self.file.output(Op.begin_text,
+                             self.file.fontName(prop), fontsize, Op.selectfont)
+            self._setup_textpos(x, y, angle)
+            self.file.output(self.encode_string(s, fonttype),
+                             Op.show, Op.end_text)
+
+        # If fonttype == 3 and there are no multibyte characters, emit the
+        # whole string with interspersed kerning info.
+        elif all(ord(char) <= 255 for char in s):
+            kern_sequence = self.build_kern_sequence(s, font, fontsize)
+            kern_s = [self.encode_string(subs, fonttype)
+                      if isinstance(subs, str) else subs
+                      for subs in kern_sequence]
+
             self.file.output(Op.begin_text,
                              self.file.fontName(prop), fontsize, Op.selectfont)
             self._setup_textpos(x, y, angle)
-            self.file.output(self.encode_string(s, fonttype), Op.show,
-                             Op.end_text)
+            self.file.output(kern_s, Op.showkern, Op.end_text)
 
         # There is no way to access multibyte characters of Type 3 fonts, as
         # they cannot have a CIDMap.  Therefore, in this case we break the
