Merge pull request matplotlib#31183 from Cemonix/fix-font-alt-family-names

ENH: Allow fonts to be addressed by any of their SFNT family names

Author: ksunden

doc/release/next_whats_new/font_alt_family_names.rst

@@ -0,0 +1,25 @@
+Fonts addressable by all their SFNT family names
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Fonts can now be selected by any of the family names they advertise in
+the OpenType name table, not just the one FreeType reports as the primary
+family name.
+
+Some fonts store different family names on different platforms or in
+different name-table entries.  For example, Ubuntu Light stores
+``"Ubuntu"`` in the Macintosh-platform Name ID 1 slot (which FreeType
+uses as the primary name) and ``"Ubuntu Light"`` in the Microsoft-platform
+Name ID 1 slot.  Previously only the FreeType-derived name was registered,
+requiring an obscure weight-based workaround::
+
+    # Previously required
+    matplotlib.rcParams['font.family'] = 'Ubuntu'
+    matplotlib.rcParams['font.weight'] = 300
+
+All name-table entries that describe a family — Name ID 1 on both
+platforms, the Typographic Family (Name ID 16), and the WWS Family
+(Name ID 21) — are now registered as separate entries in the
+`~matplotlib.font_manager.FontManager`, so any of those names can be
+used directly::
+
+    matplotlib.rcParams['font.family'] = 'Ubuntu Light'

lib/matplotlib/font_manager.py

@@ -532,6 +532,82 @@ def get_weight():  # From fontconfig's FcFreeTypeQueryFaceInternal.
                      style, variant, weight, stretch, size)
 
 
+def _get_font_alt_names(font, primary_name):
+    """
+    Return ``(name, weight)`` pairs for alternate family names of *font*.
+
+    A font file can advertise its family name in several places.  FreeType
+    exposes ``font.family_name``, which is typically derived from the
+    Macintosh-platform Name ID 1 entry.  However, other entries may carry
+    different (equally valid) names that users reasonably expect to work:
+
+    - **Name ID 1, other platform** — some fonts store a different family name
+      on the Microsoft platform than on the Macintosh platform.
+    - **Name ID 16** — "Typographic Family" (a.k.a. preferred family): groups
+      more than the traditional four styles under one name.
+    - **Name ID 21** — "WWS Family": an even narrower grouping used by some
+      fonts (weight/width/slope only).
+
+    Each name is paired with a weight derived from the corresponding subfamily
+    entry on the *same* platform.  This ensures that the weight of the alternate entry
+    reflects the font's role *within that named family* rather than its absolute
+    typographic weight.
+
+    Parameters
+    ----------
+    font : `.FT2Font`
+    primary_name : str
+        The family name already extracted from the font (``font.family_name``).
+
+    Returns
+    -------
+    list of (str, int)
+        ``(alternate_family_name, weight)`` pairs, not including *primary_name*.
+    """
+    try:
+        sfnt = font.get_sfnt()
+    except ValueError:
+        return []
+
+    mac_key = (1,  # platform: macintosh
+               0,  # id: roman
+               0)  # langid: english
+    ms_key = (3,   # platform: microsoft
+              1,   # id: unicode_cs
+              0x0409)  # langid: english_united_states
+
+    seen = {primary_name}
+    result = []
+
+    def _weight_from_subfam(subfam):
+        subfam = subfam.replace(" ", "")
+        for regex, weight in _weight_regexes:
+            if re.search(regex, subfam, re.I):
+                return weight
+        return 400  # "Regular" or unrecognised
+
+    def _try_add(name, subfam):
+        name = name.strip()
+        if not name or name in seen:
+            return
+        seen.add(name)
+        result.append((name, _weight_from_subfam(subfam.strip())))
+
+    # Each family-name ID is paired with its corresponding subfamily ID on the
+    # same platform: (family_id, subfamily_id).
+    for fam_id, subfam_id in ((1, 2), (16, 17), (21, 22)):
+        _try_add(
+            sfnt.get((*mac_key, fam_id), b'').decode('latin-1'),
+            sfnt.get((*mac_key, subfam_id), b'').decode('latin-1'),
+        )
+        _try_add(
+            sfnt.get((*ms_key, fam_id), b'').decode('utf-16-be'),
+            sfnt.get((*ms_key, subfam_id), b'').decode('utf-16-be'),
+        )
+
+    return result
+
+
 def afmFontProperty(fontpath, font):
     """
     Extract information from an AFM font file.
@@ -1131,7 +1207,7 @@ class FontManager:
     # Increment this version number whenever the font cache data
     # format or behavior has changed and requires an existing font
     # cache files to be rebuilt.
-    __version__ = '3.11.0a3'
+    __version__ = '3.11.0a4'
 
     def __init__(self, size=None, weight='normal'):
         self._version = self.__version__
@@ -1196,10 +1272,18 @@ def addfont(self, path):
             font = ft2font.FT2Font(path)
             prop = ttfFontProperty(font)
             self.ttflist.append(prop)
+            for alt_name, alt_weight in _get_font_alt_names(font, prop.name):
+                self.ttflist.append(
+                    dataclasses.replace(prop, name=alt_name, weight=alt_weight))
+
             for face_index in range(1, font.num_faces):
                 subfont = ft2font.FT2Font(path, face_index=face_index)
                 prop = ttfFontProperty(subfont)
                 self.ttflist.append(prop)
+                for alt_name, alt_weight in _get_font_alt_names(subfont, prop.name):
+                    self.ttflist.append(
+                        dataclasses.replace(prop, name=alt_name, weight=alt_weight))
+
         self._findfont_cached.cache_clear()
 
     @property

lib/matplotlib/font_manager.pyi

@@ -23,6 +23,9 @@ def get_fontext_synonyms(fontext: str) -> list[str]: ...
 def list_fonts(directory: str, extensions: Iterable[str]) -> list[str]: ...
 def win32FontDirectory() -> str: ...
 def _get_fontconfig_fonts() -> list[Path]: ...
+def _get_font_alt_names(
+    font: ft2font.FT2Font, primary_name: str
+) -> list[tuple[str, int]]: ...
 def findSystemFonts(
     fontpaths: Iterable[str | os.PathLike] | None = ..., fontext: str = ...
 ) -> list[str]: ...

lib/matplotlib/tests/test_font_manager.py

@@ -11,11 +11,14 @@
 import numpy as np
 import pytest
 
+from unittest.mock import MagicMock, patch
+
 import matplotlib as mpl
+import matplotlib.font_manager as fm_mod
 from matplotlib.font_manager import (
     findfont, findSystemFonts, FontEntry, FontPath, FontProperties, fontManager,
     json_dump, json_load, get_font, is_opentype_cff_font,
-    MSUserFontDirectories, ttfFontProperty,
+    MSUserFontDirectories, ttfFontProperty, _get_font_alt_names,
     _get_fontconfig_fonts, _normalize_weight)
 from matplotlib import cbook, ft2font, pyplot as plt, rc_context, figure as mfigure
 from matplotlib.testing import subprocess_run_helper, subprocess_run_for_testing
@@ -400,23 +403,145 @@ def test_get_font_names():
     paths_mpl = [cbook._get_data_path('fonts', subdir) for subdir in ['ttf']]
     fonts_mpl = findSystemFonts(paths_mpl, fontext='ttf')
     fonts_system = findSystemFonts(fontext='ttf')
-    ttf_fonts = []
+    ttf_fonts = set()
     for path in fonts_mpl + fonts_system:
         try:
             font = ft2font.FT2Font(path)
             prop = ttfFontProperty(font)
-            ttf_fonts.append(prop.name)
+            ttf_fonts.add(prop.name)
             for face_index in range(1, font.num_faces):
                 font = ft2font.FT2Font(path, face_index=face_index)
                 prop = ttfFontProperty(font)
-                ttf_fonts.append(prop.name)
+                ttf_fonts.add(prop.name)
         except Exception:
             pass
-    available_fonts = sorted(list(set(ttf_fonts)))
-    mpl_font_names = sorted(fontManager.get_font_names())
-    assert set(available_fonts) == set(mpl_font_names)
-    assert len(available_fonts) == len(mpl_font_names)
-    assert available_fonts == mpl_font_names
+    # fontManager may contain additional entries for alternative family names
+    # (e.g. typographic family, platform-specific Name ID 1) registered by
+    # addfont(), so primary names must be a subset of the manager's names.
+    assert ttf_fonts <= set(fontManager.get_font_names())
+
+
+def test_addfont_alternative_names(tmp_path):
+    """
+    Fonts that advertise different family names across platforms or name IDs
+    should be registered under all of those names so users can address the font
+    by any of them.
+
+    Two real-world patterns are covered:
+
+    - **MS platform ID 1 differs from Mac platform ID 1** (e.g. Ubuntu Light):
+      FreeType returns the Mac ID 1 value as ``family_name``; the MS ID 1
+      value ("Ubuntu Light") is an equally valid name that users expect to work.
+    - **Name ID 16 (Typographic Family) differs from ID 1** (older fonts):
+      some fonts store a broader family name in ID 16.
+    """
+    mac_key = (1, 0, 0)
+    ms_key = (3, 1, 0x0409)
+
+    # Case 1: MS ID1 differs from Mac ID1 (Ubuntu Light pattern)
+    # Mac ID1="Test Family" → FreeType family_name (primary)
+    # MS  ID1="Test Family Light" → alternate name users expect to work
+    ubuntu_style_sfnt = {
+        (*mac_key, 1): "Test Family".encode("latin-1"),
+        (*ms_key,  1): "Test Family Light".encode("utf-16-be"),
+        (*mac_key, 2): "Light".encode("latin-1"),
+        (*ms_key,  2): "Regular".encode("utf-16-be"),
+    }
+    fake_font = MagicMock()
+    fake_font.get_sfnt.return_value = ubuntu_style_sfnt
+
+    assert _get_font_alt_names(fake_font, "Test Family") == [("Test Family Light", 400)]
+    assert _get_font_alt_names(fake_font, "Test Family Light") == [
+        ("Test Family", 300)]
+
+    # Case 2: ID 16 differs from ID 1 (older typographic-family pattern)
+    # ID 17 (typographic subfamily) is absent → defaults to weight 400
+    id16_sfnt = {
+        (*mac_key, 1):  "Test Family".encode("latin-1"),
+        (*ms_key,  1):  "Test Family".encode("utf-16-be"),
+        (*ms_key,  16): "Test Family Light".encode("utf-16-be"),
+    }
+    fake_font_id16 = MagicMock()
+    fake_font_id16.get_sfnt.return_value = id16_sfnt
+
+    assert _get_font_alt_names(
+        fake_font_id16, "Test Family"
+    ) == [("Test Family Light", 400)]
+
+    # Case 3: all entries agree → no alternates
+    same_sfnt = {
+        (*mac_key, 1): "Test Family".encode("latin-1"),
+        (*ms_key,  1): "Test Family".encode("utf-16-be"),
+    }
+    fake_font_same = MagicMock()
+    fake_font_same.get_sfnt.return_value = same_sfnt
+    assert _get_font_alt_names(fake_font_same, "Test Family") == []
+
+    # Case 4: get_sfnt() raises ValueError (e.g. non-SFNT font) → empty list
+    fake_font_no_sfnt = MagicMock()
+    fake_font_no_sfnt.get_sfnt.side_effect = ValueError
+    assert _get_font_alt_names(fake_font_no_sfnt, "Test Family") == []
+
+    fake_path = str(tmp_path / "fake.ttf")
+    primary_entry = FontEntry(fname=fake_path, name="Test Family",
+                              style="normal", variant="normal",
+                              weight=300, stretch="normal", size="scalable")
+
+    with patch("matplotlib.font_manager.ft2font.FT2Font",
+               return_value=fake_font), \
+         patch("matplotlib.font_manager.ttfFontProperty",
+               return_value=primary_entry):
+        fm_instance = fm_mod.FontManager.__new__(fm_mod.FontManager)
+        fm_instance.ttflist = []
+        fm_instance.afmlist = []
+        fm_instance._findfont_cached = MagicMock()
+        fm_instance._findfont_cached.cache_clear = MagicMock()
+        fm_instance.addfont(fake_path)
+
+    names = [e.name for e in fm_instance.ttflist]
+    assert names == ["Test Family", "Test Family Light"]
+    alt_entry = fm_instance.ttflist[1]
+    assert alt_entry.weight == 400
+    assert alt_entry.style == primary_entry.style
+    assert alt_entry.fname == primary_entry.fname
+
+
+@pytest.mark.parametrize("subfam,expected", [
+    ("Thin",        100),
+    ("ExtraLight",  200),
+    ("UltraLight",  200),
+    ("DemiLight",   350),
+    ("SemiLight",   350),
+    ("Light",       300),
+    ("Book",        380),
+    ("Regular",     400),
+    ("Normal",      400),
+    ("Medium",      500),
+    ("DemiBold",    600),
+    ("Demi",        600),
+    ("SemiBold",    600),
+    ("ExtraBold",   800),
+    ("SuperBold",   800),
+    ("UltraBold",   800),
+    ("Bold",        700),
+    ("UltraBlack", 1000),
+    ("SuperBlack", 1000),
+    ("ExtraBlack", 1000),
+    ("Ultra",      1000),
+    ("Black",       900),
+    ("Heavy",       900),
+    ("",            400),  # fallback: unrecognised → regular
+])
+def test_alt_name_weight_from_subfamily(subfam, expected):
+    """_get_font_alt_names derives weight from the paired subfamily string."""
+    ms_key = (3, 1, 0x0409)
+    fake_font = MagicMock()
+    fake_font.get_sfnt.return_value = {
+        (*ms_key, 1): "Family Alt".encode("utf-16-be"),
+        (*ms_key, 2): subfam.encode("utf-16-be"),
+    }
+    result = _get_font_alt_names(fake_font, "Family")
+    assert result == [("Family Alt", expected)]
 
 
 def test_donot_cache_tracebacks():