def _standard_name(seq):
"""Use the standard emoji name, with some algorithmic modifications.
We want to ignore skin-tone modifiers (but of course if the sequence _is_
the skin-tone modifier itself we keep that). So we strip these so we can
start with the generic name ignoring skin tone.
Non-emoji that are turned into emoji using the emoji VS have '(emoji) '
prepended to them, so strip that.
Regional indicator symbol names are a bit long, so shorten them.
Regional sequences are assumed to be ok as-is in terms of capitalization and
punctuation, so no modifications are applied to them.
After title-casing we make some English articles/prepositions lower-case
again. We also replace '&' with 'and'; Unicode seems rather fond of
ampersand."""
if not unicode_data.is_skintone_modifier(seq[0]):
seq = tuple(
[cp for cp in seq if not unicode_data.is_skintone_modifier(cp)])
name = unicode_data.get_emoji_sequence_name(seq)
if name.startswith('(emoji) '):
name = name[8:]
if len(seq) == 1 and unicode_data.is_regional_indicator(seq[0]):
return 'Regional Symbol ' + unicode_data.regional_indicator_to_ascii(
seq[0])
if (unicode_data.is_regional_indicator_seq(seq)
or unicode_data.is_regional_tag_seq(seq)):
return name
name = name.title()
# Require space delimiting just in case...
name = re.sub(r'\s&\s', ' and ', name)
name = re.sub(
# not \b at start because we retain capital at start of phrase
r'(\s(:?A|And|From|In|Of|With|For))\b',
lambda s: s.group(1).lower(),
name)
return name
def _standard_name(seq):
"""Use the standard emoji name, with some algorithmic modifications.
We want to ignore skin-tone modifiers (but of course if the sequence _is_
the skin-tone modifier itself we keep that). So we strip these so we can
start with the generic name ignoring skin tone.
Non-emoji that are turned into emoji using the emoji VS have '(emoji) '
prepended to them, so strip that.
Regional indicator symbol names are a bit long, so shorten them.
Regional sequences are assumed to be ok as-is in terms of capitalization and
punctuation, so no modifications are applied to them.
After title-casing we make some English articles/prepositions lower-case
again. We also replace '&' with 'and'; Unicode seems rather fond of
ampersand."""
if not unicode_data.is_skintone_modifier(seq[0]):
seq = tuple([cp for cp in seq if not unicode_data.is_skintone_modifier(cp)])
name = unicode_data.get_emoji_sequence_name(seq)
if name.startswith('(emoji) '):
name = name[8:]
if len(seq) == 1 and unicode_data.is_regional_indicator(seq[0]):
return 'Regional Symbol ' + unicode_data.regional_indicator_to_ascii(seq[0])
if (unicode_data.is_regional_indicator_seq(seq) or
unicode_data.is_regional_tag_seq(seq)):
return name
name = name.title()
# Require space delimiting just in case...
name = re.sub(r'\s&\s', ' and ', name)
name = re.sub(
# not \b at start because we retain capital at start of phrase
r'(\s(:?A|And|From|In|Of|With|For))\b', lambda s: s.group(1).lower(),
name)
return name
def generate_names(src_dir,
dst_dir,
skip_limit=20,
omit_groups=None,
pretty_print=False,
verbose=False):
srcdir = tool_utils.resolve_path(src_dir)
if not path.isdir(srcdir):
print('%s is not a directory' % src_dir, file=sys.stderr)
return
if omit_groups:
unknown_groups = set(omit_groups) - set(
unicode_data.get_emoji_groups())
if unknown_groups:
print(
'did not recognize %d group%s: %s' %
(len(unknown_groups), '' if len(unknown_groups) == 1 else 's',
', '.join('"%s"' % g
for g in omit_groups if g in unknown_groups)),
file=sys.stderr)
print('valid groups are:\n %s' %
('\n '.join(g for g in unicode_data.get_emoji_groups())),
file=sys.stderr)
return
print('omitting %d group%s: %s' %
(len(omit_groups), '' if len(omit_groups) == 1 else 's',
', '.join('"%s"' % g for g in omit_groups)))
else:
# might be None
print('keeping all groups')
omit_groups = []
# make sure the destination exists
dstdir = tool_utils.ensure_dir_exists(tool_utils.resolve_path(dst_dir))
# _get_image_data returns canonical cp sequences
print('src dir:', srcdir)
seq_to_file = generate_emoji_html._get_image_data(srcdir, 'png', 'emoji_u')
print('seq to file has %d sequences' % len(seq_to_file))
# Aliases add non-gendered versions using gendered images for the most part.
# But when we display the images, we don't distinguish genders in the
# naming, we rely on the images-- so these look redundant. So we
# intentionally don't generate images for these.
# However, the alias file also includes the flag aliases, which we do want,
# and it also fails to exclude the unknown flag pua (since it doesn't
# map to anything), so we need to adjust for this.
canonical_aliases = generate_emoji_html._get_canonical_aliases()
aliases = set([
cps for cps in canonical_aliases.keys()
if not unicode_data.is_regional_indicator_seq(cps)
])
aliases.add((0xfe82b, )) # unknown flag PUA
excluded = aliases | generate_emoji_html._get_canonical_excluded()
# The flag aliases have distinct names, so we _do_ want to show them
# multiple times.
to_add = {}
for seq in canonical_aliases:
if unicode_data.is_regional_indicator_seq(seq):
replace_seq = canonical_aliases[seq]
if seq in seq_to_file:
print('warning, alias %s has file %s' %
(unicode_data.regional_indicator_seq_to_string(seq),
seq_to_file[seq]))
continue
replace_file = seq_to_file.get(replace_seq)
if replace_file:
to_add[seq] = replace_file
seq_to_file.update(to_add)
data = []
last_skipped_group = None
skipcount = 0
for group in unicode_data.get_emoji_groups():
if group in omit_groups:
continue
name_data = []
for seq in unicode_data.get_emoji_in_group(group):
if seq in excluded:
continue
seq_file = seq_to_file.get(seq, None)
if seq_file is None:
skipcount += 1
if verbose:
if group != last_skipped_group:
print('group %s' % group)
last_skipped_group = group
print(' %s (%s)' %
(unicode_data.seq_to_string(seq), ', '.join(
unicode_data.name(cp, 'x') for cp in seq)))
if skip_limit >= 0 and skipcount > skip_limit:
raise Exception('skipped too many items')
else:
name_data.append(_name_data(seq, seq_file))
data.append({'category': group, 'emojis': name_data})
outfile = path.join(dstdir, 'data.json')
with open(outfile, 'w') as f:
indent = 2 if pretty_print else None
separators = None if pretty_print else (',', ':')
json.dump(data, f, indent=indent, separators=separators)
print('wrote %s' % outfile)
def generate_names(
src_dir, dst_dir, skip_limit=20, omit_groups=None, pretty_print=False,
verbose=False):
srcdir = tool_utils.resolve_path(src_dir)
if not path.isdir(srcdir):
print('%s is not a directory' % src_dir, file=sys.stderr)
return
if omit_groups:
unknown_groups = set(omit_groups) - set(unicode_data.get_emoji_groups())
if unknown_groups:
print('did not recognize %d group%s: %s' % (
len(unknown_groups), '' if len(unknown_groups) == 1 else 's',
', '.join('"%s"' % g for g in omit_groups if g in unknown_groups)), file=sys.stderr)
print('valid groups are:\n %s' % (
'\n '.join(g for g in unicode_data.get_emoji_groups())), file=sys.stderr)
return
print('omitting %d group%s: %s' % (
len(omit_groups), '' if len(omit_groups) == 1 else 's',
', '.join('"%s"' % g for g in omit_groups)))
else:
# might be None
print('keeping all groups')
omit_groups = []
# make sure the destination exists
dstdir = tool_utils.ensure_dir_exists(
tool_utils.resolve_path(dst_dir))
# _get_image_data returns canonical cp sequences
print('src dir:', srcdir)
seq_to_file = generate_emoji_html._get_image_data(srcdir, 'png', 'emoji_u')
print('seq to file has %d sequences' % len(seq_to_file))
# Aliases add non-gendered versions using gendered images for the most part.
# But when we display the images, we don't distinguish genders in the
# naming, we rely on the images-- so these look redundant. So we
# intentionally don't generate images for these.
# However, the alias file also includes the flag aliases, which we do want,
# and it also fails to exclude the unknown flag pua (since it doesn't
# map to anything), so we need to adjust for this.
canonical_aliases = generate_emoji_html._get_canonical_aliases()
aliases = set([
cps for cps in canonical_aliases.keys()
if not unicode_data.is_regional_indicator_seq(cps)])
aliases.add((0xfe82b,)) # unknown flag PUA
excluded = aliases | generate_emoji_html._get_canonical_excluded()
# The flag aliases have distinct names, so we _do_ want to show them
# multiple times.
to_add = {}
for seq in canonical_aliases:
if unicode_data.is_regional_indicator_seq(seq):
replace_seq = canonical_aliases[seq]
if seq in seq_to_file:
print('warning, alias %s has file %s' % (
unicode_data.regional_indicator_seq_to_string(seq),
seq_to_file[seq]))
continue
replace_file = seq_to_file.get(replace_seq)
if replace_file:
to_add[seq] = replace_file
seq_to_file.update(to_add)
data = []
last_skipped_group = None
skipcount = 0
for group in unicode_data.get_emoji_groups():
if group in omit_groups:
continue
name_data = []
for seq in unicode_data.get_emoji_in_group(group):
if seq in excluded:
continue
seq_file = seq_to_file.get(seq, None)
if seq_file is None:
skipcount += 1
if verbose:
if group != last_skipped_group:
print('group %s' % group)
last_skipped_group = group
print(' %s (%s)' % (
unicode_data.seq_to_string(seq),
', '.join(unicode_data.name(cp, 'x') for cp in seq)))
if skip_limit >= 0 and skipcount > skip_limit:
raise Exception('skipped too many items')
else:
name_data.append(_name_data(seq, seq_file))
data.append({'category': group, 'emojis': name_data})
outfile = path.join(dstdir, 'data.json')
with open(outfile, 'w') as f:
indent = 2 if pretty_print else None
separators = None if pretty_print else (',', ':')
json.dump(data, f, indent=indent, separators=separators)
print('wrote %s' % outfile)