[fpdf2] Make fpdf.fonts.Glyph a dataclass (#13645)

Author: srittau

stubs/fpdf2/@tests/stubtest_allowlist.txt

@@ -10,3 +10,7 @@ fpdf.fonts.HarfBuzzFont
 
 # Stubtest wants us to use Literals, but that is unreasonable.
 fpdf.unicode_script.UNICODE_RANGE_TO_SCRIPT
+
+# Ignore stubtest weirdness "fpdf.fonts.Glyph._DT is not present at runtime"
+# https://github.com/python/mypy/issues/18811
+fpdf.fonts.Glyph._DT

stubs/fpdf2/fpdf/fonts.pyi

@@ -118,19 +118,13 @@ class PDFFontDescriptor(PDFObject):
     font_name: Incomplete
     def __init__(self, ascent, descent, cap_height, flags, font_b_box, italic_angle, stem_v, missing_width) -> None: ...
 
+@dataclass(order=True)
 class Glyph:
     glyph_id: int
     unicode: tuple[Incomplete, ...]
     glyph_name: str
     glyph_width: int
-    def __hash__(self): ...
-    def __init__(self, glyph_id, unicode, glyph_name, glyph_width) -> None: ...
-    def __lt__(self, other): ...
-    def __gt__(self, other): ...
-    def __le__(self, other): ...
-    def __ge__(self, other): ...
-
-    __match_args__ = ("glyph_id", "unicode", "glyph_name", "glyph_width")
+    def __hash__(self) -> int: ...
 
 class SubsetMap:
     font: TTFFont