From 8cd1b1ba8018507e4f66b64fa9f81279cb5b71a8 Mon Sep 17 00:00:00 2001 From: Roozbeh Pournader Date: Mon, 25 Jul 2016 14:04:34 -0700 Subject: [PATCH] Re-enable disabled emoji tests [DO NOT MERGE] Since I2e508ced90515409ff6882b3c98d8911094b9b86 updates the emoji data to version 4.0 beta, we can now re-enable the disabled emoji tests. Also fix some style issues, update data file parsing method to handle the new emoji format, add UN to unsupported flags, and do a little refactoring. Bug: 30379358 Bug: 26187231 Change-Id: I441fdfed68381d08e5e0b3af91e94bfae742eef6 --- tools/fonts/fontchain_lint.py | 61 ++++++++++++++----------------------------- 1 file changed, 20 insertions(+), 41 deletions(-) diff --git a/tools/fonts/fontchain_lint.py b/tools/fonts/fontchain_lint.py index 2956d87f247c..bbac83c6efbf 100755 --- a/tools/fonts/fontchain_lint.py +++ b/tools/fonts/fontchain_lint.py @@ -256,8 +256,8 @@ def parse_fonts_xml(fonts_xml_path): def check_emoji_coverage(all_emoji, equivalent_emoji): - emoji_font = get_emoji_font() - check_emoji_font_coverage(emoji_font, all_emoji, equivalent_emoji) + emoji_font = get_emoji_font() + check_emoji_font_coverage(emoji_font, all_emoji, equivalent_emoji) def get_emoji_font(): @@ -274,15 +274,12 @@ def check_emoji_font_coverage(emoji_font, all_emoji, equivalent_emoji): assert sequence in coverage, ( '%s is not supported in the emoji font.' % printable(sequence)) - # disable temporarily - we cover more than this - """ for sequence in coverage: if sequence in {0x0000, 0x000D, 0x0020}: # The font needs to support a few extra characters, which is OK continue assert sequence in all_emoji, ( 'Emoji font should not support %s.' % printable(sequence)) - """ for first, second in sorted(equivalent_emoji.items()): assert coverage[first] == coverage[second], ( @@ -290,8 +287,6 @@ def check_emoji_font_coverage(emoji_font, all_emoji, equivalent_emoji): printable(first), printable(second))) - # disable temporarily - some equivalent sequences we don't even know about - """ for glyph in set(coverage.values()): maps_to_glyph = [seq for seq in coverage if coverage[seq] == glyph] if len(maps_to_glyph) > 1: @@ -307,7 +302,7 @@ def check_emoji_font_coverage(emoji_font, all_emoji, equivalent_emoji): 'The sequences %s should not result in the same glyph %s' % ( printable(equivalent_seqs), glyph)) - """ + def check_emoji_defaults(default_emoji): missing_text_chars = _emoji_properties['Emoji'] - default_emoji @@ -342,7 +337,8 @@ def check_emoji_defaults(default_emoji): 0x2764, # HEAVY BLACK HEART } assert missing_text_chars == set(), ( - 'Text style version of some emoji characters are missing: ' + repr(missing_text_chars)) + 'Text style version of some emoji characters are missing: ' + + repr(missing_text_chars)) # Setting reverse to true returns a dictionary that maps the values to sets of @@ -362,7 +358,7 @@ def parse_unicode_datafile(file_path, reverse=False): if not line: continue - chars, prop = line.split(';') + chars, prop = line.split(';')[:2] chars = chars.strip() prop = prop.strip() @@ -423,26 +419,6 @@ def parse_ucd(ucd_path): _emoji_zwj_sequences = parse_unicode_datafile( path.join(ucd_path, 'emoji-zwj-sequences.txt')) - # filter modern pentathlon, as it seems likely to be removed from final spec - # also filter rifle - def is_excluded(n): - return n in [0x1f93b, 0x1f946] - - def contains_excluded(t): - if type(t) == int: - return is_excluded(t) - return any(is_excluded(cp) for cp in t) - - # filter modern pentathlon, as it seems likely to be removed from final spec - _emoji_properties['Emoji'] = set( - t for t in _emoji_properties['Emoji'] if not contains_excluded(t)) - _emoji_sequences = dict( - (t, v) for (t, v) in _emoji_sequences.items() if not contains_excluded(t)) - - # add in UN flag - UN_seq = flag_sequence('UN') - _emoji_sequences[UN_seq] = 'Emoji_Flag_Sequence' - def flag_sequence(territory_code): return tuple(0x1F1E6 + ord(ch) - ord('A') for ch in territory_code) @@ -454,7 +430,8 @@ UNSUPPORTED_FLAGS = frozenset({ flag_sequence('GF'), flag_sequence('GP'), flag_sequence('GS'), flag_sequence('MF'), flag_sequence('MQ'), flag_sequence('NC'), flag_sequence('PM'), flag_sequence('RE'), flag_sequence('TF'), - flag_sequence('WF'), flag_sequence('XK'), flag_sequence('YT'), + flag_sequence('UN'), flag_sequence('WF'), flag_sequence('XK'), + flag_sequence('YT'), }) EQUIVALENT_FLAGS = { @@ -502,7 +479,17 @@ ZWJ_IDENTICALS = { def is_fitzpatrick_modifier(cp): - return 0x1f3fb <= cp <= 0x1f3ff + return 0x1F3FB <= cp <= 0x1F3FF + + +def reverse_emoji(seq): + rev = list(reversed(seq)) + # if there are fitzpatrick modifiers in the sequence, keep them after + # the emoji they modify + for i in xrange(1, len(rev)): + if is_fitzpatrick_modifier(rev[i-1]): + rev[i], rev[i-1] = rev[i-1], rev[i] + return tuple(rev) def compute_expected_emoji(): @@ -522,15 +509,7 @@ def compute_expected_emoji(): sequence_pieces.update(sequence) # Add reverse of all emoji ZWJ sequences, which are added to the fonts # as a workaround to get the sequences work in RTL text. - reversed_seq = list(reversed(sequence)) - # if there are fitzpatrick modifiers in the sequence, keep them after - # the emoji they modify - for i in xrange(1, len(reversed_seq)): - if is_fitzpatrick_modifier(reversed_seq[i - 1]): - tmp = reversed_seq[i] - reversed_seq[i] = reversed_seq[i-1] - reversed_seq[i-1] = tmp - reversed_seq = tuple(reversed_seq) + reversed_seq = reverse_emoji(sequence) all_sequences.add(reversed_seq) equivalent_emoji[reversed_seq] = sequence -- 2.11.0