def main(argv):
"""Fix all the fonts given in the command line.
If they are Lao fonts, make sure they have ZWSP and dotted circle. If they
are Khmer fonts, make sure they have ZWSP, joiners, and dotted circle."""
for font_name in argv[1:]:
if 'Khmer' in font_name:
script = 'Khmr'
elif 'Lao' in font_name:
script = 'Laoo'
needed_chars = set(opentype_data.SPECIAL_CHARACTERS_NEEDED[script])
lgc_font_name = (
os.path.basename(font_name).replace('Khmer', '').replace('Lao', ''))
lgc_font_name = os.path.join(_UNHINTED_FONTS_DIR, lgc_font_name)
font_charset = coverage.character_set(font_name)
missing_chars = needed_chars - font_charset
if missing_chars:
merge_chars_from_bank(
font_name,
lgc_font_name,
os.path.dirname(font_name)+'/new/'+os.path.basename(font_name),
missing_chars)
def main(argv):
"""Fix all the fonts given in the command line.
If they are Lao fonts, make sure they have ZWSP and dotted circle. If they
are Khmer fonts, make sure they have ZWSP, joiners, and dotted circle."""
for font_name in argv[1:]:
if 'Khmer' in font_name:
script = 'Khmr'
elif 'Lao' in font_name:
script = 'Laoo'
needed_chars = set(opentype_data.SPECIAL_CHARACTERS_NEEDED[script])
lgc_font_name = (os.path.basename(font_name).replace('Khmer',
'').replace(
'Lao', ''))
lgc_font_name = os.path.join(_UNHINTED_FONTS_DIR, lgc_font_name)
font_charset = coverage.character_set(font_name)
missing_chars = needed_chars - font_charset
if missing_chars:
merge_chars_from_bank(
font_name, lgc_font_name,
os.path.dirname(font_name) + '/new/' +
os.path.basename(font_name), missing_chars)
def test_rendering(data,
font_file_name,
min_allowed,
max_allowed,
language=None):
"""Test the rendering of the input data in a given font.
The input data is first filtered for sequences supported in the font.
"""
font_characters = coverage.character_set(font_file_name)
# Hack to add ASCII digits, even if the font doesn't have them,
# to keep potential frequency info in the input intact
font_characters |= set(range(ord('0'), ord('9') + 1))
supported_chars_regex = _regular_expression_from_set(font_characters)
harfbuzz_input = []
for match in supported_chars_regex.finditer(data):
harfbuzz_input.append(match.group(0))
harfbuzz_input = '\n'.join(harfbuzz_input)
return render.test_text_vertical_extents(harfbuzz_input, font_file_name,
min_allowed, max_allowed,
language)
def test_sanity(self):
"""Test basic sanity of the method."""
font_file_name = path.join(path.dirname(__file__), os.pardir, 'fonts',
'individual', 'unhinted',
'NotoSansAvestan-Regular.ttf')
charset = coverage.character_set(font_file_name)
self.assertTrue(ord(' ') in charset)
self.assertTrue(0x10B00 in charset)
self.assertFalse(ord('A') in charset)
def test_sanity(self):
"""Test basic sanity of the method."""
font_file_name = path.join(
path.dirname(__file__), os.pardir,
'fonts', 'individual', 'unhinted', 'NotoSansAvestan-Regular.ttf')
charset = coverage.character_set(font_file_name)
self.assertTrue(ord(' ') in charset)
self.assertTrue(0x10B00 in charset)
self.assertFalse(ord('A') in charset)
def subset_font(source_file,
target_file,
include=None,
exclude=None,
options=None):
"""Subsets a font file.
Subsets a font file based on a specified character set. If only include is
specified, only characters from that set would be included in the output
font. If only exclude is specified, all characters except those in that
set will be included. If neither is specified, the character set will
remain the same, but inaccessible glyphs will be removed.
Args:
source_file: Input file name.
target_file: Output file name
include: The list of characters to include from the source font.
exclude: The list of characters to exclude from the source font.
options: A dictionary listing which options should be different from the
default.
Raises:
NotImplementedError: Both include and exclude were specified.
"""
opt = subset.Options()
opt.name_IDs = ['*']
opt.name_legacy = True
opt.name_languages = ['*']
opt.layout_features = ['*']
opt.notdef_outline = True
opt.recalc_bounds = True
opt.recalc_timestamp = True
opt.canonical_order = True
if options is not None:
for name, value in options.iteritems():
setattr(opt, name, value)
if include is not None:
if exclude is not None:
raise NotImplementedError(
'Subset cannot include and exclude a set at the same time.')
target_charset = include
else:
if exclude is None:
exclude = []
source_charset = coverage.character_set(source_file)
target_charset = source_charset - set(exclude)
font = subset.load_font(source_file, opt)
subsetter = subset.Subsetter(options=opt)
subsetter.populate(unicodes=target_charset)
subsetter.subset(font)
subset.save_font(font, target_file, opt)
def subset_font(source_file, target_file,
include=None, exclude=None, options=None):
"""Subsets a font file.
Subsets a font file based on a specified character set. If only include is
specified, only characters from that set would be included in the output
font. If only exclude is specified, all characters except those in that
set will be included. If neither is specified, the character set will
remain the same, but inaccessible glyphs will be removed.
Args:
source_file: Input file name.
target_file: Output file name
include: The list of characters to include from the source font.
exclude: The list of characters to exclude from the source font.
options: A dictionary listing which options should be different from the
default.
Raises:
NotImplementedError: Both include and exclude were specified.
"""
opt = subset.Options()
opt.name_IDs = ['*']
opt.name_legacy = True
opt.name_languages = ['*']
opt.layout_features = ['*']
opt.notdef_outline = True
opt.recalc_bounds = True
opt.recalc_timestamp = True
opt.canonical_order = True
opt.drop_tables = ['+TTFA']
if options is not None:
for name, value in options.iteritems():
setattr(opt, name, value)
if include is not None:
if exclude is not None:
raise NotImplementedError(
'Subset cannot include and exclude a set at the same time.')
target_charset = include
else:
if exclude is None:
exclude = []
source_charset = coverage.character_set(source_file)
target_charset = source_charset - set(exclude)
font = subset.load_font(source_file, opt)
subsetter = subset.Subsetter(options=opt)
subsetter.populate(unicodes=target_charset)
subsetter.subset(font)
subset.save_font(font, target_file, opt)
def test_all_combinations(
max_len, font_file_name, min_allowed, max_allowed, language=None):
"""Tests the rendering of all combinations up to certain length."""
font_characters = coverage.character_set(font_file_name)
font_characters -= set(range(0x00, 0x20)) # Remove ASCII controls
font_characters = [unichr(code) for code in font_characters]
font_characters = sorted(font_characters)
all_strings = []
for length in range(1, max_len+1):
all_combinations = itertools.product(font_characters, repeat=length)
all_strings += [''.join(comb) for comb in all_combinations]
test_data = '\n'.join(all_strings)
return test_rendering(
test_data, font_file_name, min_allowed, max_allowed, language)
def test_rendering(
data, font_file_name, min_allowed, max_allowed, language=None):
"""Test the rendering of the input data in a given font.
The input data is first filtered for sequences supported in the font.
"""
font_characters = coverage.character_set(font_file_name)
# Hack to add ASCII digits, even if the font doesn't have them,
# to keep potential frequency info in the input intact
font_characters |= set(range(ord('0'), ord('9')+1))
supported_chars_regex = _regular_expression_from_set(font_characters)
harfbuzz_input = []
for match in supported_chars_regex.finditer(data):
harfbuzz_input.append(match.group(0))
harfbuzz_input = '\n'.join(harfbuzz_input)
return render.test_text_vertical_extents(
harfbuzz_input, font_file_name, min_allowed, max_allowed, language)
def find_fonts():
font_name_regexp = re.compile(
'(NotoSans|NotoSerif|NotoNaskh|NotoKufi|Arimo|Cousine|Tinos)'
'(.*?)'
'(UI|Eastern|Estrangela|Western)?'
'-'
'(|Black|Bold|DemiLight|Light|Medium|Regular|Thin)'
'(Italic)?'
'(-Windows)?'
'.[ot]t[cf]')
unicode_data.load_data()
for directory in [
path.join(FONT_DIR, 'hinted'),
path.join(FONT_DIR, 'unhinted'), CJK_DIR
]:
for filename in os.listdir(directory):
match = font_name_regexp.match(filename)
if not match:
assert (filename == 'NotoSansCJK.ttc'
or # All-comprenehsive CJK
filename.endswith('.ttx') or
filename.startswith('README.') or filename
in ['COPYING', 'LICENSE', 'NEWS'])
continue
family, script, variant, weight, style, platform = match.groups()
if family in {'Arimo', 'Cousine', 'Tinos'}:
continue # Skip these three for the website
if family.startswith('Noto'):
family = family.replace('Noto', 'Noto ')
if weight == '':
weight = 'Regular'
assert platform is None
if script == '': # LGC
supported_scripts.update({'Latn', 'Grek', 'Cyrl'})
elif script in {'JP', 'KR', 'SC', 'TC', 'CJK'}:
continue # Skip unified or old CJK fonts
else:
script = convert_to_four_letter(script)
supported_scripts.add(script)
file_path = path.join(directory, filename)
if filename.endswith('.ttf') or filename.endswith('.otf'):
charset = coverage.character_set(file_path)
else:
charset = NotImplemented
if directory == CJK_DIR:
hint_status = 'hinted'
else:
hint_status = path.basename(directory)
assert hint_status in ['hinted', 'unhinted']
key = family.replace(' ', '-')
if script:
key += '-' + script
if variant not in {None, 'UI'}:
key += '-' + variant
key = key.lower()
font = Font(file_path, hint_status, key, family, script, variant,
weight, style, platform, charset)
all_fonts.append(font)
def find_fonts():
font_name_regexp = re.compile(
'(NotoSans|NotoSerif|NotoNaskh|NotoKufi|Arimo|Cousine|Tinos)'
'(.*?)'
'(UI|Eastern|Estrangela|Western)?'
'-'
'(|Black|Bold|DemiLight|Light|Medium|Regular|Thin)'
'(Italic)?'
'(-Windows)?'
'.[ot]t[cf]')
unicode_data.load_data()
for directory in [path.join(FONT_DIR, 'hinted'),
path.join(FONT_DIR, 'unhinted'),
CJK_DIR]:
for filename in os.listdir(directory):
match = font_name_regexp.match(filename)
if not match:
assert (
filename == 'NotoSansCJK.ttc' or # All-comprenehsive CJK
filename.endswith('.ttx') or
filename.startswith('README.') or
filename in ['COPYING', 'LICENSE', 'NEWS'])
continue
family, script, variant, weight, style, platform = match.groups()
if family in {'Arimo', 'Cousine', 'Tinos'}:
continue # Skip these three for the website
if family.startswith('Noto'):
family = family.replace('Noto', 'Noto ')
if weight == '':
weight = 'Regular'
assert platform is None
if script == '': # LGC
supported_scripts.update({'Latn', 'Grek', 'Cyrl'})
elif script in {'JP', 'KR', 'SC', 'TC', 'CJK'}:
continue # Skip unified or old CJK fonts
else:
script = convert_to_four_letter(script)
supported_scripts.add(script)
file_path = path.join(directory, filename)
if filename.endswith('.ttf') or filename.endswith('.otf'):
charset = coverage.character_set(file_path)
else:
charset = NotImplemented
if directory == CJK_DIR:
hint_status = 'hinted'
else:
hint_status = path.basename(directory)
assert hint_status in ['hinted', 'unhinted']
key = family.replace(' ', '-')
if script:
key += '-' + script
if variant not in {None, 'UI'}:
key += '-' + variant
key = key.lower()
font = Font(file_path, hint_status, key,
family, script, variant, weight, style, platform,
charset)
all_fonts.append(font)