def test_parse_code_ranges(self):
"""Tests the _parse_code_ranges method."""
source = (
'0000..001F ; Common # Cc [32] <control-0000>..<control-001F>\n'
'0020 ; Common # Zs SPACE\n')
self.assertEqual([(0, 31, 'Common'), (32, 32, 'Common')],
unicode_data._parse_code_ranges(source))
def main(argv):
"""Subset the Noto Symbols font.
The first argument is the source file name, and the second argument is
the target file name.
"""
target_coverage = set()
# Add all characters in BLOCKS_TO_INCLUDE
for first, last, _ in unicode_data._parse_code_ranges(BLOCKS_TO_INCLUDE):
target_coverage.update(range(first, last + 1))
# Add one-off characters
target_coverage |= ONE_OFF_ADDITIONS
# Remove characters preferably coming from Roboto
target_coverage -= LETTERLIKE_CHARS_IN_ROBOTO
# Remove characters that are supposed to default to emoji
target_coverage -= BMP_DEFAULT_EMOJI | ANDROID_EMOJI
# Remove dentistry symbols, as their main use appears to be for CJK:
# http://www.unicode.org/L2/L2000/00098-n2195.pdf
target_coverage -= set(range(0x23BE, 0x23CC + 1))
# Remove COMBINING ENCLOSING KEYCAP. It's needed for Android's color emoji
# mechanism to work properly
target_coverage.remove(0x20E3)
source_file_name = argv[1]
target_file_name = argv[2]
subset.subset_font(source_file_name,
target_file_name,
include=target_coverage)
def test_parse_code_ranges(self):
"""Tests the _parse_code_ranges method."""
source = (
'0000..001F ; Common # Cc [32] <control-0000>..<control-001F>\n'
'0020 ; Common # Zs SPACE\n')
self.assertEqual(
[(0, 31, 'Common'), (32, 32, 'Common')],
unicode_data._parse_code_ranges(source))
def _symbol_set():
"""Returns set of characters that should be supported in Noto Symbols.
"""
global _SYMBOL_SET
if not _SYMBOL_SET:
ranges = unicode_data._parse_code_ranges(noto_data.SYMBOL_RANGES_TXT)
_SYMBOL_SET = code_range_to_set(ranges) & unicode_data.defined_characters()
return _SYMBOL_SET
def main(argv):
"""Subset the Noto Symbols font.
The first argument is the source file name, and the second argument is
the target file name.
"""
target_coverage = set()
# Add all characters in BLOCKS_TO_INCLUDE
for first, last, _ in unicode_data._parse_code_ranges(BLOCKS_TO_INCLUDE):
target_coverage.update(range(first, last+1))
# Add one-off characters
target_coverage |= ONE_OFF_ADDITIONS
# Remove characters preferably coming from Roboto
target_coverage -= LETTERLIKE_CHARS_IN_ROBOTO
# Remove characters that are supposed to default to emoji
android_emoji = get_android_emoji()
target_coverage -= DEFAULT_EMOJI | android_emoji
# Remove dentistry symbols, as their main use appears to be for CJK:
# http://www.unicode.org/L2/L2000/00098-n2195.pdf
target_coverage -= set(range(0x23BE, 0x23CC+1))
# Remove COMBINING ENCLOSING KEYCAP. It's needed for Android's color emoji
# mechanism to work properly.
target_coverage.remove(0x20E3)
source_file_name = argv[1]
target_file_name = argv[2]
subset.subset_font(
source_file_name,
target_file_name,
include=target_coverage)
second_subset_coverage = DEFAULT_EMOJI | android_emoji
second_subset_file_name = argv[3]
subset.subset_font(
source_file_name,
second_subset_file_name,
include=second_subset_coverage)
def subset_symbols(srcdir, dstdir):
"""Subset Noto Sans Symbols in a curated way.
Noto Sans Symbols is now subsetted in a curated way. Changes include:
* Currency symbols now included in Roboto are removed.
* All combining marks for symbols (except for combining keycap) are
added, to combine with other symbols if needed.
* Characters in symbol blocks that are also covered by Noto CJK fonts
are added, for better harmony with the rest of the fonts in non-CJK
settings. The dentistry characters at U+23BE..23CC are not added,
since they appear to be Japan-only and full-width.
* Characters that UTR #51 defines as default text are added, although
they may also exist in the color emoji font, to make sure they get
a default text style.
* Characters that UTR #51 defines as default emoji are removed, to
make sure they don't block the fallback to the color emoji font.
* A few math symbols that are currently included in Roboto are added,
to prepare for potentially removing them from Roboto when they are
lower-quality in Roboto.
Based on subset_noto_sans_symbols.py from AOSP external/noto-fonts."""
# TODO see if we need to change this subset based on Noto Serif coverage
# (so the serif fallback chain would support them)
target_coverage = set()
# Add all characters in BLOCKS_TO_INCLUDE
for first, last, _ in unicode_data._parse_code_ranges(BLOCKS_TO_INCLUDE):
target_coverage.update(range(first, last + 1))
# Add one-off characters
target_coverage |= ONE_OFF_ADDITIONS
# Remove characters preferably coming from Roboto
target_coverage -= LETTERLIKE_CHARS_IN_ROBOTO
# Remove default emoji presentation (including ones Android prefers default)
target_coverage -= EMOJI
# Remove COMBINING ENCLOSING KEYCAP. It's needed for Android's color emoji
# mechanism to work properly
target_coverage.remove(0x20E3)
# Remove dentistry symbols, as their main use appears to be for CJK:
# http://www.unicode.org/L2/L2000/00098-n2195.pdf
target_coverage -= set(range(0x23BE, 0x23CC + 1))
for font_file in glob.glob(path.join(srcdir, 'NotoSansSymbols-*.ttf')):
print 'main subset', font_file
out_file = path.join(dstdir,
path.basename(font_file)[:-4] + '-Subsetted.ttf')
subset.subset_font(font_file, out_file, include=target_coverage)
# The second subset will be a fallback after the color emoji, for
# explicit text presentation sequences.
target_coverage = EMOJI | unicode_data.get_unicode_emoji_variants()
for font_file in glob.glob(path.join(srcdir, 'NotoSansSymbols-*.ttf')):
print 'secondary subset', font_file
out_file = path.join(dstdir,
path.basename(font_file)[:-4] + '-Subsetted2.ttf')
subset.subset_font(font_file, out_file, include=target_coverage)
def subset_symbols(srcdir, dstdir):
"""Subset Noto Sans Symbols in a curated way.
Noto Sans Symbols is now subsetted in a curated way. Changes include:
* Currency symbols now included in Roboto are removed.
* All combining marks for symbols (except for combining keycap) are
added, to combine with other symbols if needed.
* Characters in symbol blocks that are also covered by Noto CJK fonts
are added, for better harmony with the rest of the fonts in non-CJK
settings. The dentistry characters at U+23BE..23CC are not added,
since they appear to be Japan-only and full-width.
* Characters that UTR #51 defines as default text are added, although
they may also exist in the color emoji font, to make sure they get
a default text style.
* Characters that UTR #51 defines as default emoji are removed, to
make sure they don't block the fallback to the color emoji font.
* A few math symbols that are currently included in Roboto are added,
to prepare for potentially removing them from Roboto when they are
lower-quality in Roboto.
Based on subset_noto_sans_symbols.py from AOSP external/noto-fonts."""
# TODO see if we need to change this subset based on Noto Serif coverage
# (so the serif fallback chain would support them)
target_coverage = set()
# Add all characters in BLOCKS_TO_INCLUDE
for first, last, _ in unicode_data._parse_code_ranges(BLOCKS_TO_INCLUDE):
target_coverage.update(range(first, last+1))
# Add one-off characters
target_coverage |= ONE_OFF_ADDITIONS
# Remove characters preferably coming from Roboto
target_coverage -= LETTERLIKE_CHARS_IN_ROBOTO
# Remove default emoji presentation (including ones Android prefers default)
target_coverage -= EMOJI
# Remove COMBINING ENCLOSING KEYCAP. It's needed for Android's color emoji
# mechanism to work properly
target_coverage.remove(0x20E3)
# Remove dentistry symbols, as their main use appears to be for CJK:
# http://www.unicode.org/L2/L2000/00098-n2195.pdf
target_coverage -= set(range(0x23BE, 0x23CC+1))
for font_file in glob.glob(path.join(srcdir, 'NotoSansSymbols-*.ttf')):
print 'main subset', font_file
out_file = path.join(
dstdir, path.basename(font_file)[:-4] + '-Subsetted.ttf')
subset.subset_font(font_file, out_file, include=target_coverage)
# The second subset will be a fallback after the color emoji, for
# explicit text presentation sequences.
target_coverage = EMOJI | unicode_data.get_unicode_emoji_variants()
for font_file in glob.glob(path.join(srcdir, 'NotoSansSymbols-*.ttf')):
print 'secondary subset', font_file
out_file = path.join(
dstdir, path.basename(font_file)[:-4] + '-Subsetted2.ttf')
subset.subset_font(font_file, out_file, include=target_coverage)
def _cjk_set(): """Returns set of characters that will be provided in CJK fonts.""" ranges = unicode_data._parse_code_ranges(noto_data.CJK_RANGES_TXT) return _code_range_to_set(ranges)
def _symbol_set(): """Returns set of characters that should be supported in Noto Symbols.""" ranges = unicode_data._parse_code_ranges(noto_data.SYMBOL_RANGES_TXT) return _code_range_to_set(ranges)
def _cjk_set(): """Returns set of characters that will be provided in CJK fonts.""" ranges = unicode_data._parse_code_ranges(noto_data.CJK_RANGES_TXT) return _code_range_to_set(ranges)
def _math_set():
"""Returns set of characters that should be supported in Noto Math."""
ranges = unicode_data._parse_code_ranges(noto_data.MATH_RANGES_TXT)
return _code_range_to_set(ranges)
def _symbol_set(): """Returns set of characters that should be supported in Noto Symbols.""" ranges = unicode_data._parse_code_ranges(noto_data.SYMBOL_RANGES_TXT) return _code_range_to_set(ranges)
def _math_set():
"""Returns set of characters that should be supported in Noto Math."""
ranges = unicode_data._parse_code_ranges(noto_data.MATH_RANGES_TXT)
return _code_range_to_set(ranges)
