Jun 2, 2022 · Jun 12, 2022 · Jun 12, 2022
diff --git a/lib/matplotlib/_mathtext.py b/lib/matplotlib/_mathtext.py
        if bunch is not None:
            return bunch

        font, num, slanted = self._get_glyph(
        font, num, slanted, substituted_glyph = self._get_glyph(
            fontname, font_class, sym, fontsize)

        font.set_size(fontsize, dpi)
            ymax    = ymax+offset,
            # iceberg is the equivalent of TeX's "height"
            iceberg = glyph.horiBearingY/64.0 + offset,
            slanted = slanted
            slanted = slanted,
            substituted_glyph = substituted_glyph
            )

        result = self.glyphd[key] = types.SimpleNamespace(
            if font is not None:
                num = ord(sym)
        if font is not None and font.get_char_index(num) != 0:
            return font, num, slanted
            return font, num, slanted, False
        else:
            return self._stix_fallback._get_glyph(
                fontname, font_class, sym, fontsize)
            found_symbol = False
            _log.warning("No TeX to Unicode mapping for {!a}.".format(sym))

        substituted_glyph = False

        fontname, uniindex = self._map_virtual_font(
            fontname, font_class, uniindex)

                font = self._get_font('rm')
                uniindex = 0xA4  # currency char, for lack of anything better
                slanted = False
                substituted_glyph = True

        return font, uniindex, slanted
        return font, uniindex, slanted, substituted_glyph

    def get_sized_alternatives_for_symbol(self, fontname, sym):
        if self._fallback_font:
    def is_slanted(self):
        return self._metrics.slanted

    def is_substituted(self):
        return self._metrics.substituted_glyph

    def get_kerning(self, next):
        """
        Return the amount of kerning between this and the given character.
        raise ParseFatalException(s, loc, f"Unknown symbol: {toks['name']}")

    _accent_map = {
        r'hat':            r'\circumflexaccent',
        r'hat':            r'\combiningcircumflexaccent',
        r'breve':          r'\combiningbreve',
        r'bar':            r'\combiningoverline',
        r'grave':          r'\combininggraveaccent',
        r"'":              r'\combiningacuteaccent',
        r'~':              r'\combiningtilde',
        r'.':              r'\combiningdotabove',
        r'^':              r'\circumflexaccent',
        r'overrightarrow': r'\rightarrow',
        r'overleftarrow':  r'\leftarrow',
        r'mathring':       r'\circ',
        r'^':              r'\combiningcircumflexaccent',
        r'overrightarrow': r'\combiningrightarrowabove',
        r'overleftarrow':  r'\combiningleftarrowabove',
        r'mathring':       r'\combiningringabove',
        r'=':              r'\combiningmacron',
        r'H':              r'\combiningdoubleacuteaccent',
        r'check':          r'\combiningcaron',
    }

    _wide_accents = set(r"widehat widetilde widebar".split())
            accent_box = AutoWidthChar(
                '\\' + accent, sym.width, state, char_class=Accent)
        else:
            # Check if accent and character can be combined
            a = get_unicode_index(self._accent_map[accent])
            if isinstance(sym, Char):
                c = sym.c
            else:
                c = sym.children[0].c
            if c == chr(305):  # Dotless i, but normal i combines
                c = 'i'
            comb = unicodedata.normalize('NFC', c + chr(a))
            if len(comb) == 1:  # Check that they did combine
                newsym = Char(comb, state)
                # Check that glyph exists
                if not newsym.is_substituted():
                    return newsym
            if c == 'i':  # Turn i into dotless i
                sym = Char(chr(305), state)
                if sym.is_substituted():
                    # Dotless i does not exist
                    sym = Char('i', state)
            # Cannot be combined
            accent_box = Accent(self._accent_map[accent], state)
        if accent == 'mathring':
            accent_box.shrink()
            accent_box.shrink()
        centered = HCentered([Hbox(sym.width / 4.0), accent_box])
        centered.hpack(sym.width, 'exactly')
        return Vlist([
            return nucleus.is_slanted()
        return False

    def is_between_brackets(self, s, loc):
        return False

    def subsuper(self, s, loc, toks):
        nucleus = toks.get("nucleus", Hbox(0))
        subsuper = toks.get("subsuper", [])
diff --git a/lib/matplotlib/_mathtext_data.py b/lib/matplotlib/_mathtext_data.py
    '\\Upsilon'                  : ('cmr10', 0xa8),
    '\\Xi'                       : ('cmr10', 0xa5),
    '\\circumflexaccent'         : ('cmr10', 0x5e),
    '\\combiningcircumflexaccent' : ('cmr10', 0x5e),
    '\\combiningacuteaccent'     : ('cmr10', 0xb6),
    '\\combiningbreve'           : ('cmr10', 0xb8),
    '\\combiningdiaeresis'       : ('cmr10', 0xc4),
    '\\combiningdotabove'        : ('cmr10', 0x5f),
    '\\combininggraveaccent'     : ('cmr10', 0xb5),
    '\\combiningoverline'        : ('cmr10', 0xb9),
    '\\combiningmacron'          : ('cmr10', 0x16),
    '\\combiningringabove'       : ('cmr10', 0x15),
    '\\combiningcaron'           : ('cmr10', 0x14),
    '\\combiningtilde'           : ('cmr10', 0x7e),
    '\\combiningdoubleacuteaccent' : ('cmr10', 0x7d),
    '\\leftbracket'              : ('cmr10', 0x5b),
    '\\leftparen'                : ('cmr10', 0x28),
    '\\rightbracket'             : ('cmr10', 0x5d),
    'circumflexaccent'         : 770,
    'combiningbreve'           : 774,
    'combiningoverline'        : 772,
    'combiningmacron'          : 773,
    'combininggraveaccent'     : 768,
    'combiningacuteaccent'     : 769,
    'combiningdiaeresis'       : 776,
    'combiningtilde'           : 771,
    'combiningrightarrowabove' : 8407,
    'combiningleftarrowabove'  : 8406,
    'combiningdotabove'        : 775,
    'combiningringabove'       : 778,
    'combiningthreedotsabove'  : 8411,
    'combiningfourdotsabove'   : 8412,
    'combiningdoubleacuteaccent' : 779,
    'combiningcaron'           : 780,
    'combiningcircumflexaccent' : 770,
    'to'                       : 8594,
    'succeq'                   : 8829,
    'emptyset'                 : 8709,
diff --git a/lib/matplotlib/tests/baseline_images/test_mathtext/mathtext_cm_83.pdf b/lib/matplotlib/tests/baseline_images/test_mathtext/mathtext_cm_83.pdf
diff --git a/lib/matplotlib/tests/baseline_images/test_mathtext/mathtext_cm_83.png b/lib/matplotlib/tests/baseline_images/test_mathtext/mathtext_cm_83.png
Original file line number	Diff line number	Diff line change
Expand Up		@@ -220,7 +220,7 @@ def _get_info(self, fontname, font_class, sym, fontsize, dpi):
		if bunch is not None:
		return bunch

		font, num, slanted = self._get_glyph(
		font, num, slanted, substituted_glyph = self._get_glyph(
		fontname, font_class, sym, fontsize)

		font.set_size(fontsize, dpi)
Expand All		@@ -239,7 +239,8 @@ def _get_info(self, fontname, font_class, sym, fontsize, dpi):
		ymax = ymax+offset,
		# iceberg is the equivalent of TeX's "height"
		iceberg = glyph.horiBearingY/64.0 + offset,
		slanted = slanted
		slanted = slanted,
		substituted_glyph = substituted_glyph
		)

		result = self.glyphd[key] = types.SimpleNamespace(
Expand DownExpand Up		@@ -323,7 +324,7 @@ def _get_glyph(self, fontname, font_class, sym, fontsize):
		if font is not None:
		num = ord(sym)
		if font is not None and font.get_char_index(num) != 0:
		return font, num, slanted
		return font, num, slanted, False
		else:
		return self._stix_fallback._get_glyph(
		fontname, font_class, sym, fontsize)
Expand DownExpand Up		@@ -448,6 +449,8 @@ def _get_glyph(self, fontname, font_class, sym, fontsize):
		found_symbol = False
		_log.warning("No TeX to Unicode mapping for {!a}.".format(sym))

		substituted_glyph = False

		fontname, uniindex = self._map_virtual_font(
		fontname, font_class, uniindex)

Expand DownExpand Up		@@ -499,8 +502,9 @@ def _get_glyph(self, fontname, font_class, sym, fontsize):
		font = self._get_font('rm')
		uniindex = 0xA4 # currency char, for lack of anything better
		slanted = False
		substituted_glyph = True

		return font, uniindex, slanted
		return font, uniindex, slanted, substituted_glyph

		def get_sized_alternatives_for_symbol(self, fontname, sym):
		if self._fallback_font:
Expand DownExpand Up		@@ -935,6 +939,9 @@ def _update_metrics(self):
		def is_slanted(self):
		return self._metrics.slanted

		def is_substituted(self):
		return self._metrics.substituted_glyph

		def get_kerning(self, next):
		"""
		Return the amount of kerning between this and the given character.
Expand DownExpand Up		@@ -2011,7 +2018,7 @@ def unknown_symbol(self, s, loc, toks):
		raise ParseFatalException(s, loc, f"Unknown symbol: {toks['name']}")

		_accent_map = {
		r'hat': r'\circumflexaccent',
		r'hat': r'\combiningcircumflexaccent',
		r'breve': r'\combiningbreve',
		r'bar': r'\combiningoverline',
		r'grave': r'\combininggraveaccent',
Expand All		@@ -2027,10 +2034,13 @@ def unknown_symbol(self, s, loc, toks):
		r"'": r'\combiningacuteaccent',
		r'~': r'\combiningtilde',
		r'.': r'\combiningdotabove',
		r'^': r'\circumflexaccent',
		r'overrightarrow': r'\rightarrow',
		r'overleftarrow': r'\leftarrow',
		r'mathring': r'\circ',
		r'^': r'\combiningcircumflexaccent',
		r'overrightarrow': r'\combiningrightarrowabove',
		r'overleftarrow': r'\combiningleftarrowabove',
		r'mathring': r'\combiningringabove',
		r'=': r'\combiningmacron',
		r'H': r'\combiningdoubleacuteaccent',
		r'check': r'\combiningcaron',
		}

		_wide_accents = set(r"widehat widetilde widebar".split())
Expand All		@@ -2050,10 +2060,27 @@ def accent(self, s, loc, toks):
		accent_box = AutoWidthChar(
		'\\' + accent, sym.width, state, char_class=Accent)
		else:
		# Check if accent and character can be combined
Copy link MemberAuthor oscargusJun 12, 2022 Choose a reason for hiding this comment The reason will be displayed to describe this comment to others.Learn more. One can possibly consider splitting the accents into those that may have precomposed characters and those that may not. https://en.wikipedia.org/wiki/List_of_precomposed_Latin_characters_in_Unicode Possibly one should check that the character is one of the standard latin characters as well, although that may lead to that those precomposed with two accents may not work (which should be checked if they even do to start with...).
		a = get_unicode_index(self._accent_map[accent])
		if isinstance(sym, Char):
		c = sym.c
		else:
		c = sym.children[0].c
		if c == chr(305): # Dotless i, but normal i combines
		c = 'i'
		comb = unicodedata.normalize('NFC', c + chr(a))
		if len(comb) == 1: # Check that they did combine
		newsym = Char(comb, state)
		# Check that glyph exists
		if not newsym.is_substituted():
		return newsym
		if c == 'i': # Turn i into dotless i
		sym = Char(chr(305), state)
		if sym.is_substituted():
		# Dotless i does not exist
		sym = Char('i', state)
		# Cannot be combined
		accent_box = Accent(self._accent_map[accent], state)
		if accent == 'mathring':
		accent_box.shrink()
		accent_box.shrink()
		centered = HCentered([Hbox(sym.width / 4.0), accent_box])
		centered.hpack(sym.width, 'exactly')
		return Vlist([
Expand DownExpand Up		@@ -2137,9 +2164,6 @@ def is_slanted(self, nucleus):
		return nucleus.is_slanted()
		return False

		def is_between_brackets(self, s, loc):
		return False

		def subsuper(self, s, loc, toks):
		nucleus = toks.get("nucleus", Hbox(0))
		subsuper = toks.get("subsuper", [])
Expand Down
Original file line number	Diff line number	Diff line change
Expand Up		@@ -117,13 +117,18 @@
		'\\Upsilon' : ('cmr10', 0xa8),
		'\\Xi' : ('cmr10', 0xa5),
		'\\circumflexaccent' : ('cmr10', 0x5e),
		'\\combiningcircumflexaccent' : ('cmr10', 0x5e),
		'\\combiningacuteaccent' : ('cmr10', 0xb6),
		'\\combiningbreve' : ('cmr10', 0xb8),
		'\\combiningdiaeresis' : ('cmr10', 0xc4),
		'\\combiningdotabove' : ('cmr10', 0x5f),
		'\\combininggraveaccent' : ('cmr10', 0xb5),
		'\\combiningoverline' : ('cmr10', 0xb9),
		'\\combiningmacron' : ('cmr10', 0x16),
		'\\combiningringabove' : ('cmr10', 0x15),
		'\\combiningcaron' : ('cmr10', 0x14),
		'\\combiningtilde' : ('cmr10', 0x7e),
		'\\combiningdoubleacuteaccent' : ('cmr10', 0x7d),
		'\\leftbracket' : ('cmr10', 0x5b),
		'\\leftparen' : ('cmr10', 0x28),
		'\\rightbracket' : ('cmr10', 0x5d),
Expand DownExpand Up		@@ -994,14 +999,20 @@
		'circumflexaccent' : 770,
		'combiningbreve' : 774,
		'combiningoverline' : 772,
		'combiningmacron' : 773,
		'combininggraveaccent' : 768,
		'combiningacuteaccent' : 769,
		'combiningdiaeresis' : 776,
		'combiningtilde' : 771,
		'combiningrightarrowabove' : 8407,
		'combiningleftarrowabove' : 8406,
		'combiningdotabove' : 775,
		'combiningringabove' : 778,
		'combiningthreedotsabove' : 8411,
		'combiningfourdotsabove' : 8412,
		'combiningdoubleacuteaccent' : 779,
		'combiningcaron' : 780,
		'combiningcircumflexaccent' : 770,
		'to' : 8594,
		'succeq' : 8829,
		'emptyset' : 8709,
Expand Down