def load_bd(self, filename: str) -> None:
"""
Loads sentence view from big_dict structure file
Args:
filename: file name <str>
"""
with open(filename) as f:
data = json.load(f)
column_names = [
"word", "pos_start", "pos_end", "type", "tag", "class", "tone"
]
self.columns = column_names
for item in data:
text = item['constant']['text']
for i in range(0, len(text)):
sentence = text[i]
cur_sentence = graph.Node(self.graph, {})
cur_sentence["type"] = "sentence"
cur_sentence["words"] = graph.NodeList(self.graph)
for j in range(0, len(sentence)):
word = graph.Node(self.graph, {})
for c_name in column_names:
if c_name in ["class", "tone"]:
word[c_name] = sentence[j]["variable"][c_name][
0]
elif c_name in ["pos_start", "pos_end"]:
word["pos_start"] = str(sentence[j]["pos"][0])
word["pos_end"] = str(sentence[j]["pos"][1])
else:
word[c_name] = sentence[j][c_name]
cur_sentence["words"].append(word)
self.sentences.append(cur_sentence)
def get_text(self, text):
words = load_words_save_separators(text)
wsents = get_sentences(words, text)
self.columns = ["type", "pos_start", "pos_end", "word"]
for sent in wsents:
cur_sentence = graph.Node(self.graph, {})
cur_sentence["words"] = graph.NodeList(self.graph)
cur_sentence["type"] = "sentence"
for word_i in sent:
word = graph.Node(self.graph, {})
word["type"] = "word"
word["word"] = word_i["word"]
word["pos_start"] = word_i["pos_start"]
word["pos_end"] = word_i["pos_end"]
cur_sentence["words"].append(word)
self.sentences.append(cur_sentence)
def __init__(self, config=None):
BaseView.__init__(self, config=config)
self.texts = graph.NodeList(self.graph)
def load_column(self,
filename: str,
separator=" ",
maxlen=1000,
default_class=None,
expand=False,
maxwords=None) -> None:
'''loads sentence view from file with columns file'''
f = open(filename, "r")
column_names = split(f.readline().strip(), separator)
cur_sentence = graph.Node(self.graph, {})
cur_sentence["words"] = graph.NodeList(self.graph)
self.columns = column_names
line_num = 0
self.extended = expand
for line in f:
line = line.strip()
line = line.replace("\ufeff", "")
line_num = line_num + 1
if (maxwords is not None) and (line_num > maxwords):
return
if "<STOP>" in line or line == "" or len(
cur_sentence["words"]) > maxlen:
if len(cur_sentence["words"]) > 0:
self.sentences.append(cur_sentence)
cur_sentence = graph.Node(self.graph, {})
cur_sentence["words"] = graph.NodeList(self.graph)
cur_sentence["type"] = "sentence"
else:
word = graph.Node(self.graph, {})
words = split(line, separator)
# print (len(words))
# print (len(column_names))
#TODO: This code is horrible: need to simplify
#normal situation
if len(words) == len(column_names):
for i in range(0, len(column_names)):
word[column_names[i]] = words[i]
if expand == True:
nn = graph.Node(self.graph, {
"type": column_names[i],
"value": words[i]
})
word.Connect(nn)
#less words then colums - fill with default class if given
if len(words) < len(column_names):
if default_class is None:
print(
"Error: on col " + str(line_num) +
" not enough data and no default class provided, skipping"
)
else:
#fill specified columns
for i in range(0, len(words)):
word[column_names[i]] = words[i]
if expand is True:
nn = graph.Node(self.graph, {
"type": column_names[i],
"value": words[i]
})
word.Connect(nn)
#fill unspecified colums with default value
# print (range(len(words), len(self.columns)))
# print (words)
for i in range(len(words), len(self.columns)):
word[column_names[i]] = default_class
if expand is True:
nn = graph.Node(
self.graph, {
"type": column_names[i],
"value": default_class
})
word.Connect(nn)
if len(words) > len(column_names):
#we have too many classes
#fill normal range
for i in range(0, len(column_names)):
word[column_names[i]] = words[i]
if expand == True:
nn = graph.Node(self.graph, {
"type": column_names[i],
"value": words[i]
})
word.Connect(nn)
#if expand = true, fill load other classes into graph
for i in range(len(column_names), len(words)):
if expand == True:
nn = graph.Node(self.graph, {
"type": column_names[-1],
"value": words[i]
})
word.Connect(nn)
# print (word.children({"type":column_names[-1]}).Distinct("value"))
word["type"] = "word"
cur_sentence["words"].append(word)
if len(cur_sentence["words"]) > 0:
self.sentences.append(cur_sentence)
def __init__(self, config=None):
BaseView.__init__(self, config=config)
self.sentences = graph.NodeList(self.graph)
self.extended = False
self.clean_text = ""