# -*- coding: utf-8 -*-
import pycantonese as pc
import csv
corpus = pc.hkcancor()
hkcancorFreq = corpus.word_frequency()
with open('hkcancorFrequency.csv', encoding='utf-8', mode='w',
newline='') as csvfile:
writer = csv.writer(csvfile)
for key, value in hkcancorFreq.items():
writer.writerow([key, value])
sun = pc.read_chat('../00-source/sun_1924_tagged.cha')
sun.word_frequency()
sunFreq = sun.word_frequency()
with open('sunFrequency.csv', encoding='utf-8', mode='w',
newline='') as csvfile:
writer = csv.writer(csvfile)
for key, value in sunFreq.items():
writer.writerow([key, value])
import sys
import pytest
import pycantonese
HKCANCOR = pycantonese.hkcancor()
def almost_equal(x, y, tolerance):
# Don't bother to import numpy's assert_almost_equal just for testing
return abs(x - y) <= tolerance
def test_hkcancor_word_count():
assert almost_equal(len(HKCANCOR.words()), 149781, tolerance=3)
@pytest.mark.skipif(sys.version_info[0] == 2,
reason='character/unicode parsing not yet fixed '
'for python 2.7')
def test_hkcancor_character_count():
assert almost_equal(len(HKCANCOR.characters()), 186888, tolerance=3)
import sys
from collections import Counter
import json
from memoize import memoize
import pinyin
import jyutping
import pycantonese
corpus = pycantonese.hkcancor()
#from hanziconv import HanziConv
from opencc import OpenCC
s2hk = OpenCC('s2hk').convert
from mkdict import pinyin_to_zhuyin_real as pinyin_to_zhuyin
from mkdict import get_all_yue, get_merged_entries
def get_contents_in_dictionary(dictfile):
lines = open(dictfile).readlines()
output = []
is_started = False
for line in lines:
x = line.strip()
if x == '...':
is_started = True
continue
if not is_started:
continue
output.append(line)
def main():
corpus = hkcancor()
tagger = POSTagger(**_TAGGER_PARAMETERS)
tagger.train(_get_tagged_sents(corpus), save=_PICKLE_PATH)
#loading packages
import pycantonese as pc
import xlsxwriter
import collections
print("Please read the code and change directories accordingly (see line 9).")
#setting up workbook
workbook = xlsxwriter.Workbook('hkcancor_word.xlsx')
worksheet = workbook.add_worksheet('words')
#loading files
c = pc.hkcancor()
#getting words
tokens = c.search(tone='[1-6]')
#word frequency
word_count = collections.Counter(tokens)
#writing data
row = 1
worksheet.write(0, 0, 'word')
worksheet.write(0, 1, 'jyutping')
worksheet.write(0, 2, 'occured')
worksheet.write(0, 3, 'ratio')
word_sum = sum(word_count.values())
for i in word_count.keys():
worksheet.write(row, 0, i[0])
worksheet.write(row, 1, i[2])
def _get_tagged_sents():
return [
[(token.word, _FIX_HKCANCOR_TAGS.get(token.pos, token.pos)) for token in tokens]
for tokens in hkcancor().tokens(by_utterances=True)
]