def get_song_meta(song):
d = {}
search = set()
for k, v in song.meta.items():
if request.latin:
d[k] = v[(request.lc, "l")]
else:
d[k] = v[request.lc]
for k in ("title", "artist", "seenon", "album"):
if k in song.meta:
v = song.meta[k]
search.add(normalize(v[request.lc]))
search.add(normalize(v["k"]))
search.add(normalize(v["l"]))
search.add(
normalize(jaconv.kana2alphabet(jaconv.kata2hira(
v["k"]))).replace("ー", ""))
for k in ("genre", ):
if k in song.meta:
v = song.meta[k]
search.add(normalize(v[request.lc]))
d["search"] = list(search)
if request.latin:
d["sort"] = song.meta["title"][(request.lc, "l")]
if ord(d["sort"][0:1]) > 0x100:
# Try again with kana-to-romaji, might help manufacture some sensible sort order
d["sort"] = jaconv.kana2alphabet(
jaconv.kata2hira(song.meta["title"][(request.lc, "l", "k")]))
else:
d["sort"] = song.meta["title"][(request.lc, "k")]
return d
def get_song_meta(song):
d = {}
search = set()
for k, v in song.meta.items():
if request.latin:
d[k] = v[(request.lc, "l")]
else:
d[k] = v[request.lc]
for k in ("title", "artist", "seenon", "album"):
if k in song.meta:
v = song.meta[k]
search.add(normalize(v[request.lc]))
search.add(normalize(v["k"]))
search.add(normalize(v["l"]))
search.add(
normalize(jaconv.kana2alphabet(jaconv.kata2hira(
v["k"]))).replace("ー", ""))
for k in ("genre", ):
if k in song.meta:
v = song.meta[k]
search.add(normalize(v[request.lc]))
d["search"] = list(search)
if request.latin:
d["sort"] = song.meta["title"][(request.lc, "l")]
else:
d["sort"] = song.meta["title"][(request.lc, "k")]
return d
def test_kana2alphabet():
assert_equal(jaconv.kana2alphabet('まみさん'), 'mamisan')
assert_equal(jaconv.kana2alphabet('はっとり'), 'hattori')
assert_equal(jaconv.kana2alphabet('はっ'), 'haxtsu')
assert_equal(jaconv.kana2alphabet('ぽっ'), 'poxtsu')
assert_equal(jaconv.kana2alphabet('ふぁふぃふぇふぉ'), 'fafifefo')
assert_equal(jaconv.kana2alphabet('っって'), 'xtsutte')
def test_kana2alphabet():
assert_equal(jaconv.kana2alphabet('まみさん'), 'mamisan')
assert_equal(jaconv.kana2alphabet('はっとり'), 'hattori')
assert_equal(jaconv.kana2alphabet('はっ'), 'haxtsu')
assert_equal(jaconv.kana2alphabet('ぽっ'), 'poxtsu')
assert_equal(jaconv.kana2alphabet('ふぁふぃふぇふぉ'), 'fafifefo')
assert_equal(jaconv.kana2alphabet('っって'), 'xtsutte')
async def on_message(message):
message.content = message.content.lower()
if message.content.startswith("!hello"):
await message.channel.send("ビールとおっぱいがだいすきです！")
if message.content.startswith("!cotd"):
rand_num = random.randint(0, len(hiragana) - 1)
rand_hiragana_char = hiragana[rand_num]
translated_char = jaconv.kana2alphabet(rand_hiragana_char)
# TODO Let's try to make this look nicer!
await message.channel.send(
f'Character of the day: {rand_hiragana_char}\nTranslation: {translated_char}'
)
if message.content.startswith("!wotd"):
rand_num = random.randint(1, 3035)
word = query_db(rand_num)
await message.channel.send(
f"Hiragana: {word.hiragana}\nKatakana: {word.katakana}\nKanji: {word.kanji}\nMeaning: {word.meaning}\n"
)
if message.content.startswith("!potd"):
pass
if message.content.startswith("!help"):
e = discord.Embed(description='''
Commands
-------------
***!hello*** - Say hi to Gintoki Sensei
***!COTD*** (Character of the day!) - Get the character of the day
***!WOTD*** (Word of the day!) - Get the word of the day and link to the word pronounciation
***!POTD*** (Phrase of the day!) - Get the phrase of the day''',
colour=discord.Colour.teal())
await message.channel.send(embed=e)
def test_kana_to_hepburn():
assert_equal(jaconv.kana2alphabet('まみさん'), 'mamisan')
assert_equal(jaconv.kana2alphabet('はっとり'), 'hattori')
def _transliterate(token: str) -> str:
s = jaconv.kata2hira(token)
s = jaconv.kana2alphabet(s)
return s
def read_aozora_bunko_list(path: str, ndc_tr: Dict[str, str]) -> defaultdict:
'''
Reads in the list_person_all_extended_utf8.csv of Aozora Bunko and
constructs a nested dictionary keyed on author and title. This is
then used identify the correct path to the file as well as give
more metadata.
'''
d = defaultdict(dict)
url_rx = re.compile(r'https://www\.aozora\.gr\.jp/cards/(\d+)/(.+)')
with ZipFile(path) as z:
with z.open('list_person_all_extended_utf8.csv', 'r') as f:
for row in csv.DictReader(TextIOWrapper(f)):
# Some works have versions in both new- and old-style
# kana. As we are only interested in the new-style
# version, we skip the old one while keeping only
# old-style works.
if row['文字遣い種別'] != '新字新仮名':
log.warning(f'Skipping processing of old-syle kana work: {row}')
continue
# Use the lower value from 底本初版発行年1 and 初出:
year = ''
year_rx = re.compile(r'(\d{4})（.+）年\s?(\d{1,2})月((\d{1,2})日)?')
year_matches = year_rx.match(row['底本初版発行年1'])
if year_matches and year_matches.groups():
year = year_matches.groups()[0]
year_alternate_matches = year_rx.search(row['初出'])
if year_alternate_matches and year_alternate_matches.groups():
alt_year = year_alternate_matches.groups()[0]
if year == '':
year = alt_year
elif int(alt_year) < int(year):
year = alt_year
# Sanity check for year:
year_death = re.search(r'\d{4}', row['没年月日'])
if year_death and year_death.groups() and int(year_death.groups()[0]) < int(year):
year = '<' + year_death # Specify upper bound as last resort.
author_ja = row['姓'] + row['名']
author_en = row['名ローマ字'] + ' ' + row['姓ローマ字']
title = row['作品名']
title_ja = title
title_en = jaconv.kana2alphabet(jaconv.kata2hira(row['作品名読み'])).title()
subtitle = row['副題']
if subtitle != '':
title_ja += ': ' + subtitle
title_en += ': ' + romanize(row['副題読み']).title()
try:
match = url_rx.match(row['XHTML/HTMLファイルURL'])
id = match.group(1)
file_path = match.group(2)
except AttributeError:
log.debug('Missing XHTML/HTML file for record {}, skipping...'.format(row))
continue
ndc = row['分類番号'].replace('NDC ', '').replace('K', '')
if len(ndc) > 3:
ndcs = ndc.split()
ndc = '/'.join(ndc_tr[n] for n in ndcs)
elif not ndc:
ndc = ''
else:
ndc = ndc_tr[ndc]
if 'K' in row['分類番号']:
ndc += ' (児童書)'
if title in d[author_ja]:
# Remove translations.
d[author_ja].pop(title, None)
if len(d[author_ja]) == 0:
d.pop(author_ja, None)
else:
d[author_ja][title] = {
'author_ja': author_ja,
'author': author_en,
'author_year': f'{row["生年月日"]}--{row["没年月日"]}',
'title_ja': title_ja,
'title': title_en,
'year': year,
'ndc': ndc,
'file_path': 'aozorabunko/cards/{}/{}'.format(id, file_path),
'file_name': '{}_{}_{}'.format( # TODO Do we need to shorthen these?
row['姓ローマ字'],
row['名ローマ字'][0:1],
romanize(row['作品名読み'][0:7]).title()
)
}
return d
def romanize(s: str) -> str:
return re.sub(r'_+',
'_',
re.sub(r'[^a-zA-Z]',
'_',
jaconv.kana2alphabet(jaconv.kata2hira(s.replace('ゔ', 'v')))))
def test_kana_to_hepburn():
assert_equal(jaconv.kana2alphabet('まみさん'), 'mamisan')
assert_equal(jaconv.kana2alphabet('はっとり'), 'hattori')