def update_content(self, text):
self.clear_widgets()
self.carousel.clear_widgets()
self.analyze(text)
self.update_title()
for block in self.blocks:
words = DbUtil().get_words_with_block(block, exclude=self.word)
self.carousel.add_widget(BlockColumn(block, words))
self.add_widget(self.title_label)
self.add_widget(self.carousel)
def __init__(self,
string='',
etymology=None,
meaning=None,
compute_etymology=False):
self.check_init_parameters(string, etymology, meaning)
self.string = _u(string) # e.g. user input string
self.db_util = DbUtil()
if etymology and meaning:
self.blocks = [[
Block(string[i], etymology=etymology[i])
for i in range(len(string))
]]
self.meanings = [meaning]
self.selected_meaning = 0 # the word is clearly defined
else:
self.compute_suffix()
self.blocks = self.compute_blocks(compute_etymology)
self.meanings = self.db_util.compute_meanings(
self.string_without_suffix) # Different meanings in English
self.selected_meaning = 0 # index of the selected meaning
def __init__(self, string='', ethym=None, meaning=None, compute_ethym=False):
self.string = _u(string) # e.g. user input string
self.language = 'Korean'
self.db_util = DbUtil()
if ethym and meaning:
assert(len(string) == len(ethym)) # to the best of my knowledge a
# Korean word and its hanja
# representation (when existing)
# have the same lengths
self.blocks = [[Block(string[i], ethym=ethym[i]) for i in range(len(string))]]
self.meanings = [meaning]
self.selected_meaning = 0 # the word is clearly defined
else:
self.compute_suffix()
self.blocks = self.compute_blocks(compute_ethym)
self.meanings = self.db_util.compute_meanings(self.string_without_suffix) # Different meanings in English
self.selected_meaning = 0 # index of the selected meaning
def test_get_hanja_name():
util = DbUtil()
assert u'클 대 / 큰 대' == util.get_hanja_name(u'大')
def test_compute_meanings():
util = DbUtil()
assert '(평안) (public) peace (안정) stability, well' in util.compute_meanings(u'안녕').values
def test_get_hanja_meaning():
util = DbUtil()
assert 'big' == util.get_hanja_meaning(u'大')
def test_get_hanja(self, hanja, expected):
assert expected == DbUtil().get_hanja(hanja)
def test_get_words_with_block_no_etymology(self):
block = Block(u'하세요')
assert [] == DbUtil().get_words_with_block(block)
def test_get_words_with_block(self, input_str, exclude, input_str_expected):
word = KoreanWord(input_str, compute_etymology=True)
block = word.get_blocks_for_selected_meaning()[0]
words = DbUtil().get_words_with_block(block, exclude=exclude)
assert len(words) > 0
assert input_str_expected == (input_str in [word[0] for word in words])
def test_compute_meanings(self):
util = DbUtil()
assert '(평안) (public) peace (안정) stability, well' in util.compute_meanings(u'안녕')
def test_get_hanja_meaning(self, hanja, expected):
util = DbUtil()
assert expected == util.get_hanja_meaning(hanja)
def test_get_hanja_name(self, hanja, expected):
util = DbUtil()
assert expected == util.get_hanja_name(hanja)
def test_no_db(self, monkeypatch):
monkeypatch.setattr('asian_word_analyzer.korean.db.connection_string', 'dummy')
with pytest.raises(FileNotFoundError):
DbUtil()
class KoreanWord(object):
""" This class is used to manipulate Korean words. """
def __init__(self, string='', ethym=None, meaning=None, compute_ethym=False):
self.string = _u(string) # e.g. user input string
self.language = 'Korean'
self.db_util = DbUtil()
if ethym and meaning:
assert(len(string) == len(ethym)) # to the best of my knowledge a
# Korean word and its hanja
# representation (when existing)
# have the same lengths
self.blocks = [[Block(string[i], ethym=ethym[i]) for i in range(len(string))]]
self.meanings = [meaning]
self.selected_meaning = 0 # the word is clearly defined
else:
self.compute_suffix()
self.blocks = self.compute_blocks(compute_ethym)
self.meanings = self.db_util.compute_meanings(self.string_without_suffix) # Different meanings in English
self.selected_meaning = 0 # index of the selected meaning
@property
def meaning(self):
""" Meaning getter """
return self.meanings[self.selected_meaning]
def get_blocks_for_selected_meaning(self):
""" Getter for the blocks corresponding to the selected meaning """
return self.blocks[self.selected_meaning]
@property
def ethym(self):
return ''.join([block.ethym for block in \
self.blocks[self.selected_meaning] if block.ethym])
#==========================================================================
# PRINT METHODS
#==========================================================================
def print_blocks_for_selected_meaning(self):
""" This methods prints the block strings for the selected meaning.
Example:
--------
For the word '안녕', the printed blocks will be ['안', '녕']
"""
return [block.string for block in self.blocks[self.selected_meaning]]
#==========================================================================
# LANGUAGE METHODS
#==========================================================================
def compute_suffix(self):
""" This method computes:
self.suffix
self.suffix_meaning
self.string_without_suffix
"""
suffixes = {u'하다':u'하다 verb particule', \
u'합니다': u'formal 하다 ending', \
u'하세요': u'formal imperative form of 하다', \
u'요': u'politeness particle',\
u'님': u'honorific particle'}
# TODO: store the suffixes in the database instead of hardcoding them here
detected_suffix = ''
for suffix in suffixes.keys():
if self.string.endswith(suffix):
detected_suffix = suffix
continue
self.string_without_suffix = self.string[0:len(self.string)-len(detected_suffix)]
self.suffix = detected_suffix
self.suffix_meaning = suffixes.get(detected_suffix, None)
def compute_blocks(self, compute_ethym=False):
""" Compute the blocks given the input string.
Output:
Returns a list of lists of blocks, i.e.
[ [b11, ..., b1n1], [b21, ..., b2n2], ...], where each list of
blocks [bi1, ..., bini] corresponds to a possible meaning of the
input string.
Note:
In this implemenation, only one meaning is available.
"""
if DEBUG:
UI.render_info('compute_blocks(...) called for word ' + self.string)
if not compute_ethym:
blocks = [Block(self.string_without_suffix[i]) \
for i in range(len(self.string_without_suffix)) \
if self.string_without_suffix[i] != ' ']
else:
ethym = get_hanja(self.string_without_suffix)
if DEBUG:
UI.render_info(ethym)
blocks = [Block(self.string_without_suffix[i], ethym=ethym[i], \
meaning=self.db_util.get_hanja_meaning(ethym[i]), \
name=self.db_util.get_hanja_name(ethym[i])) \
for i in range(len(self.string_without_suffix)) \
if self.string_without_suffix[i] != ' ']
if self.suffix:
suffix_desc = 'Suffix: ' + self.suffix_meaning
blocks.append(Block(self.suffix, meaning=suffix_desc))
return [blocks]
class KoreanWord(AsianWord):
""" This class is used to manipulate Korean words. """
language = 'Korean'
def __init__(self,
string='',
etymology=None,
meaning=None,
compute_etymology=False):
self.check_init_parameters(string, etymology, meaning)
self.string = _u(string) # e.g. user input string
self.db_util = DbUtil()
if etymology and meaning:
self.blocks = [[
Block(string[i], etymology=etymology[i])
for i in range(len(string))
]]
self.meanings = [meaning]
self.selected_meaning = 0 # the word is clearly defined
else:
self.compute_suffix()
self.blocks = self.compute_blocks(compute_etymology)
self.meanings = self.db_util.compute_meanings(
self.string_without_suffix) # Different meanings in English
self.selected_meaning = 0 # index of the selected meaning
@staticmethod
def check_init_parameters(string, etymology, meaning):
if etymology is not None:
if len(string) != len(etymology):
# to the best of my knowledge a Korean word and its hanja
# representation (when existing) have the same lengths
raise ValueError(
'string and etymology must have the same lengths')
def compute_suffix(self):
""" This method computes:
self.suffix
self.suffix_meaning
self.string_without_suffix
"""
suffixes = {
u'하다': u'하다 verb particle',
u'합니다': u'formal 하다 ending',
u'하세요': u'formal imperative form of 하다',
u'요': u'politeness particle',
u'님': u'honorific particle'
}
# TODO: store the suffixes in the database instead of hardcoding them here
detected_suffix = ''
for suffix in suffixes.keys():
if self.string.endswith(suffix):
detected_suffix = suffix
continue
self.string_without_suffix = self.string[0:len(self.string) -
len(detected_suffix)]
self.suffix = detected_suffix
self.suffix_meaning = suffixes.get(detected_suffix, None)
def compute_blocks(self, compute_etymology=False):
""" Compute the blocks given the input string.
Output:
Returns a list of lists of blocks, i.e.
[ [b11, ..., b1n1], [b21, ..., b2n2], ...], where each list of
blocks [bi1, ..., bini] corresponds to a possible meaning of the
input string.
Note:
In this implementation, only one meaning is available.
"""
ui.render_debug('compute_blocks(...) called for word ' + self.string)
if compute_etymology:
etymology = self.db_util.get_hanja(self.string_without_suffix)
if etymology:
ui.render_debug('Found hanja={}'.format(etymology))
blocks = [
Block(self.string_without_suffix[i],
etymology=etymology[i],
meaning=self.db_util.get_hanja_meaning(etymology[i]),
name=self.db_util.get_hanja_name(etymology[i]))
for i in range(len(self.string_without_suffix))
if self.string_without_suffix[i] != ' '
]
else:
ui.render_error(
'Hanja not found for {}.'
'Please check the spelling or populate the `Korean` table '
'with more data.'.format(self.string_without_suffix))
blocks = []
else:
blocks = [
Block(self.string_without_suffix[i])
for i in range(len(self.string_without_suffix))
if self.string_without_suffix[i] != ' '
]
if self.suffix:
suffix_desc = 'Suffix: ' + self.suffix_meaning
blocks.append(Block(self.suffix, meaning=suffix_desc))
return [blocks]