def split_glyph_list(self, glyph):
glyphs = []
if len(glyph) > 1:
# glyph names need to be converted to unicode
# we need to take into account, that there can be more than one first/second letter in the very same hkern element
# in this case they will be commas separated and each first letter needs to be combined with each next letter
# e.g. <hkern g1="A,Agrave,Aacute,Acircumflex,Atilde,Adieresis,Amacron,Abreve,Aogonek" g2="T,Tcaron" k="5" />
glyph_names = glyph.split(",")
for glyph_name in glyph_names:
# each glyph can have additional special markers, e.g. o.cmp
# toUnicode will not respect those and convert them to a simple letter
# this behaviour will generate a wrong spacing for this letter.
# Let's make sure to also transfer the separators and extensions to our json file
separators = [".", "_"]
used_separator = False
for separator in separators:
glyph_with_separator = glyph_name.split(separator)
if len(glyph_with_separator) == 2:
glyphs.append("%s%s%s" %
(toUnicode(glyph_with_separator[0]),
separator, glyph_with_separator[1]))
used_separator = True
continue
# there is no extra separator
if not used_separator:
glyphs.append(toUnicode(glyph_name))
else:
glyphs.append(glyph)
return glyphs
def test_spec_examples(self):
# https://github.com/adobe-type-tools/agl-specification#3-examples
self.assertEqual(agl.toUnicode("Lcommaaccent"), "Ļ")
self.assertEqual(agl.toUnicode("uni20AC0308"), "\u20AC\u0308")
self.assertEqual(agl.toUnicode("u1040C"), "\U0001040C")
self.assertEqual(agl.toUnicode("uniD801DC0C"), "")
self.assertEqual(agl.toUnicode("uni20ac"), "")
self.assertEqual(
agl.toUnicode("Lcommaaccent_uni20AC0308_u1040C.alternate"),
"\u013B\u20AC\u0308\U0001040C")
self.assertEqual(agl.toUnicode("Lcommaaccent_uni013B_u013B"), "ĻĻĻ")
self.assertEqual(agl.toUnicode("foo"), "")
self.assertEqual(agl.toUnicode(".notdef"), "")
def get_glyph(glyph_name, data=glyphdata_generated):
"""Return a named tuple (Glyph) containing information derived from a glyph
name akin to GSGlyphInfo.
The information is derived from an included copy of GlyphsData.xml,
going purely by the glyph name.
"""
# First, get the base name of the glyph. .notdef and .null are exceptions.
# Periods denote glyph variants as per the AGLFN convention, which should
# be in the same category as their base glyph.
if glyph_name in (".notdef", ".null"):
base_name = glyph_name
else:
base_name = glyph_name.split(".", 1)[0]
# Next, look up the glyph name in Glyph's name database to get a Unicode
# pseudoname, or "production name" as found in a font's post table
# (e.g. "A-cy" -> "uni0410") so that e.g. PDF readers can map from names
# to Unicode values. FontTool's agl module can turn this into the actual
# character.
production_name = _lookup_production_name(glyph_name)
# Some Glyphs files use production names instead of Glyph's "nice names".
# We catch this here, so that we can return the same properties as if
# the Glyphs file had been following the Glyphs naming conventions.
# https://github.com/googlei18n/glyphsLib/issues/232
if production_name is None:
rev_prodname = data.PRODUCTION_NAMES_REVERSED.get(base_name)
if rev_prodname is not None:
production_name = base_name
base_name = rev_prodname
# Finally, if we couldn't find a known production name one way or another,
# conclude that the glyph name doesn't carry any Unicode semantics. Use the
# bare name in that case.
if production_name is None:
production_name = glyph_name
# Next, derive the actual characters from the production name, e.g.
# "uni0414" -> "Д". Two caveats:
# 1. For some glyphs, Glyphs does not have a mapped character even when one
# could be derived.
# 2. For some others, Glyphs has a different idea than the agl module.
unicode_characters = None
if base_name not in data.MISSING_UNICODE_STRINGS: # 1.
unicode_characters = data.IRREGULAR_UNICODE_STRINGS.get(
base_name) # 2.
if unicode_characters is None:
unicode_characters = agl.toUnicode(production_name) or None
# Lastly, generate the category in the sense of Glyph's
# GSGlyphInfo.category and .subCategory.
category, sub_category = _get_category(base_name, unicode_characters, data)
return Glyph(glyph_name, production_name, unicode_characters, category,
sub_category)
def test_dingbats(self):
self.assertEqual(agl.toUnicode("a20", isZapfDingbats=True), "✔")
self.assertEqual(agl.toUnicode("a20.alt", isZapfDingbats=True), "✔")
self.assertEqual(agl.toUnicode("a206", isZapfDingbats=True), "❰")
self.assertEqual(agl.toUnicode("a20", isZapfDingbats=False), "")
self.assertEqual(agl.toUnicode("a0", isZapfDingbats=True), "")
self.assertEqual(agl.toUnicode("a207", isZapfDingbats=True), "")
self.assertEqual(agl.toUnicode("abcdef", isZapfDingbats=True), "")
def test_uniABCD(self):
self.assertEqual(agl.toUnicode("uni0041"), "A")
self.assertEqual(agl.toUnicode("uni0041_uni0042_uni0043"), "ABC")
self.assertEqual(agl.toUnicode("uni004100420043"), "ABC")
self.assertEqual(agl.toUnicode("uni"), "")
self.assertEqual(agl.toUnicode("uni41"), "")
self.assertEqual(agl.toUnicode("uni004101"), "")
self.assertEqual(agl.toUnicode("uniDC00"), "")
def test_dingbats(self):
self.assertEqual(agl.toUnicode("a20", isZapfDingbats=True), "✔")
self.assertEqual(agl.toUnicode("a20.alt", isZapfDingbats=True), "✔")
self.assertEqual(agl.toUnicode("a206", isZapfDingbats=True), "❰")
self.assertEqual(agl.toUnicode("a20", isZapfDingbats=False), "")
self.assertEqual(agl.toUnicode("a0", isZapfDingbats=True), "")
self.assertEqual(agl.toUnicode("a207", isZapfDingbats=True), "")
self.assertEqual(agl.toUnicode("abcdef", isZapfDingbats=True), "")
def test_uniABCD(self):
self.assertEqual(agl.toUnicode("uni0041"), "A")
self.assertEqual(agl.toUnicode("uni0041_uni0042_uni0043"), "ABC")
self.assertEqual(agl.toUnicode("uni004100420043"), "ABC")
self.assertEqual(agl.toUnicode("uni"), "")
self.assertEqual(agl.toUnicode("uni41"), "")
self.assertEqual(agl.toUnicode("uni004101"), "")
self.assertEqual(agl.toUnicode("uniDC00"), "")
def get_glyph(name, data=glyphdata_generated):
prodname = data.PRODUCTION_NAMES.get(name, name)
unistr = data.IRREGULAR_UNICODE_STRINGS.get(name)
if unistr is None:
unistr = agl.toUnicode(prodname)
if unistr != "" and name not in data.MISSING_UNICODE_STRINGS:
unistr_result = unistr
else:
unistr_result = None
category, subCategory = _get_category(name, unistr, data)
return Glyph(name, prodname, unistr_result, category, subCategory)
def test_spec_examples(self):
# https://github.com/adobe-type-tools/agl-specification#3-examples
#
# TODO: Currently, we only handle AGLFN instead of legacy AGL names.
# Therefore, the test cases below use Iogonek instead of Lcommaaccent.
# Change Iogonek to Lcommaaccent as soon as the implementation has
# been fixed to also support legacy AGL names.
# https://github.com/fonttools/fonttools/issues/775
self.assertEqual(agl.toUnicode("Iogonek"), "Į")
self.assertEqual(agl.toUnicode("uni20AC0308"), "\u20AC\u0308")
self.assertEqual(agl.toUnicode("u1040C"), "\U0001040C")
self.assertEqual(agl.toUnicode("uniD801DC0C"), "")
self.assertEqual(agl.toUnicode("uni20ac"), "")
self.assertEqual(agl.toUnicode("Iogonek_uni20AC0308_u1040C.alternate"),
"\u012E\u20AC\u0308\U0001040C")
self.assertEqual(agl.toUnicode("Iogonek_uni012E_u012E"), "ĮĮĮ")
self.assertEqual(agl.toUnicode("foo"), "")
self.assertEqual(agl.toUnicode(".notdef"), "")
def test_spec_examples(self):
# https://github.com/adobe-type-tools/agl-specification#3-examples
#
# TODO: Currently, we only handle AGLFN instead of legacy AGL names.
# Therefore, the test cases below use Iogonek instead of Lcommaaccent.
# Change Iogonek to Lcommaaccent as soon as the implementation has
# been fixed to also support legacy AGL names.
# https://github.com/fonttools/fonttools/issues/775
self.assertEqual(agl.toUnicode("Iogonek"), "Į")
self.assertEqual(agl.toUnicode("uni20AC0308"), "\u20AC\u0308")
self.assertEqual(agl.toUnicode("u1040C"), "\U0001040C")
self.assertEqual(agl.toUnicode("uniD801DC0C"), "")
self.assertEqual(agl.toUnicode("uni20ac"), "")
self.assertEqual(
agl.toUnicode("Iogonek_uni20AC0308_u1040C.alternate"),
"\u012E\u20AC\u0308\U0001040C")
self.assertEqual(agl.toUnicode("Iogonek_uni012E_u012E"), "ĮĮĮ")
self.assertEqual(agl.toUnicode("foo"), "")
self.assertEqual(agl.toUnicode(".notdef"), "")
def get_glyph(name, data=glyphdata_generated):
prodname = data.PRODUCTION_NAMES.get(name)
# Some Glyphs files use production names (instead of Glyphs names).
# We catch this here, so that we can return the same properties as if
# the Glyphs file had been following the Glyphs naming conventions.
# https://github.com/googlei18n/glyphsLib/issues/232
if prodname is None:
rev_prodname = data.PRODUCTION_NAMES_REVERSED.get(name)
if rev_prodname is not None:
prodname = name
name = rev_prodname
if prodname is None:
prodname = name
unistr = data.IRREGULAR_UNICODE_STRINGS.get(name)
if unistr is None:
unistr = agl.toUnicode(prodname)
if unistr != "" and name not in data.MISSING_UNICODE_STRINGS:
unistr_result = unistr
else:
unistr_result = None
category, subCategory = _get_category(name, unistr, data)
return Glyph(name, prodname, unistr_result, category, subCategory)
def _read_encoding_with_differences(self) -> None:
# check whether we've been here before
if len(self._unicode_lookup_to_character_identifier) > 0:
return
# figure out how many characters we'll need to calculate
assert "FirstChar" in self
assert isinstance(self["FirstChar"], pDecimal)
assert "LastChar" in self
assert isinstance(self["LastChar"], pDecimal)
first_char: int = int(self["FirstChar"])
last_char: int = int(self["LastChar"])
self._character_identifier_to_unicode_lookup = {}
# apply differences
i: int = 0
j: int = 0
while i < len(self["Encoding"]["Differences"]):
assert isinstance(self["Encoding"]["Differences"][i], pDecimal)
character_code: int = self["Encoding"]["Differences"][i]
j = i + 1
while j < len(self["Encoding"]["Differences"]) and not isinstance(
self["Encoding"]["Differences"][j], pDecimal
):
glyph_name: str = str(self["Encoding"]["Differences"][j])
self._character_identifier_to_unicode_lookup[
int(character_code)
] = toUnicode(glyph_name)
character_code += 1
j += 1
i = j
# build reverse map
self._unicode_lookup_to_character_identifier = {
v: k for k, v in self._character_identifier_to_unicode_lookup.items()
}
def test_aglfn(self):
self.assertEqual(agl.toUnicode("longs_t"), "ſt")
self.assertEqual(agl.toUnicode("f_f_i.alt123"), "ffi")
def test_uABCD(self):
self.assertEqual(agl.toUnicode("u0041"), "A")
self.assertEqual(agl.toUnicode("u00041"), "A")
self.assertEqual(agl.toUnicode("u000041"), "A")
self.assertEqual(agl.toUnicode("u0000041"), "")
self.assertEqual(agl.toUnicode("u0041_uni0041_A.alt"), "AAA")
def test_union(self):
# Interesting test case because "uni" is a prefix of "union".
self.assertEqual(agl.toUnicode("union"), "∪")
# U+222A U+FE00 is a Standardized Variant for UNION WITH SERIFS.
self.assertEqual(agl.toUnicode("union_uniFE00"), "\u222A\uFE00")
def test_uABCD(self):
self.assertEqual(agl.toUnicode("u0041"), "A")
self.assertEqual(agl.toUnicode("u00041"), "A")
self.assertEqual(agl.toUnicode("u000041"), "A")
self.assertEqual(agl.toUnicode("u0000041"), "")
self.assertEqual(agl.toUnicode("u0041_uni0041_A.alt"), "AAA")
def test_union(self):
# Interesting test case because "uni" is a prefix of "union".
self.assertEqual(agl.toUnicode("union"), "∪")
# U+222A U+FE00 is a Standardized Variant for UNION WITH SERIFS.
self.assertEqual(agl.toUnicode("union_uniFE00"), "\u222A\uFE00")
def test_aglfn(self):
self.assertEqual(agl.toUnicode("longs_t"), "ſt")
self.assertEqual(agl.toUnicode("f_f_i.alt123"), "ffi")
