def main():
options, args = parse_options()
# File paths of input IDL files are passed in a file, which is generated at
# GN time. It is OK because the target IDL files are static.
idl_files = read_file_to_list(options.idl_files_list)
# Output IDL files (to generate) are passed at the command line, since
# these are in the build directory, which is determined at build time, not
# GN time.
# These are passed as pairs of GlobalObjectName, global_object.idl
interface_name_idl_filename = [(args[i], args[i + 1])
for i in range(0, len(args), 2)]
interface_name_to_global_names.update(read_pickle_file(options.global_objects_file))
for idl_filename in idl_files:
record_global_constructors(idl_filename)
# Check for [Exposed] / [Global] mismatch.
known_global_names = EXPOSED_EXECUTION_CONTEXT_METHOD.keys()
exposed_global_names = frozenset(global_name_to_constructors)
if not exposed_global_names.issubset(known_global_names):
unknown_global_names = exposed_global_names.difference(known_global_names)
raise ValueError('The following global names were used in '
'[Exposed=xxx] but do not match any global names: %s'
% list(unknown_global_names))
# Write partial interfaces containing constructor attributes for each
# global interface.
for interface_name, idl_filename in interface_name_idl_filename:
constructors = interface_name_to_constructors(interface_name)
write_global_constructors_partial_interface(
interface_name, idl_filename, constructors)
def main():
options, args = parse_options()
# Input IDL files are passed in a file, due to OS command line length
# limits. This is generated at GYP time, which is ok b/c files are static.
idl_files = read_file_to_list(options.idl_files_list)
# Output IDL files (to generate) are passed at the command line, since
# these are in the build directory, which is determined at build time, not
# GYP time.
# These are passed as pairs of GlobalObjectName, GlobalObject.idl
interface_name_idl_filename = [(args[i], args[i + 1])
for i in range(0, len(args), 2)]
interface_name_to_global_names.update(read_pickle_file(options.global_objects_file))
for idl_filename in idl_files:
record_global_constructors(idl_filename)
# Check for [Exposed] / [Global] mismatch.
known_global_names = EXPOSED_EXECUTION_CONTEXT_METHOD.keys()
exposed_global_names = frozenset(global_name_to_constructors)
if not exposed_global_names.issubset(known_global_names):
unknown_global_names = exposed_global_names.difference(known_global_names)
raise ValueError('The following global names were used in '
'[Exposed=xxx] but do not match any [Global] / '
'[PrimaryGlobal] interface: %s'
% list(unknown_global_names))
# Write partial interfaces containing constructor attributes for each
# global interface.
for interface_name, idl_filename in interface_name_idl_filename:
constructors = interface_name_to_constructors(interface_name)
write_global_constructors_partial_interface(
interface_name, idl_filename, constructors)
def main():
options, _ = parse_options()
# IDL files are passed in a file, due to OS command line length limits
idl_files = read_idl_files_list_from_file(options.idl_files_list)
# Compute information for individual files
# Information is stored in global variables interfaces_info and
# partial_interface_files.
info_collector = InterfaceInfoCollector(options.cache_directory)
for idl_filename in idl_files:
info_collector.collect_info(idl_filename)
write_pickle_file(options.interfaces_info_file,
info_collector.get_info_as_dict())
runtime_enabled_features = read_pickle_file(
options.runtime_enabled_features_file)
write_pickle_file(
options.component_info_file,
info_collector.get_component_info_as_dict(runtime_enabled_features))
def main():
options, args = parse_options()
# Input IDL files are passed in a file, due to OS command line length
# limits. This is generated at GYP time, which is ok b/c files are static.
idl_files = read_file_to_list(options.idl_files_list)
# Output IDL files (to generate) are passed at the command line, since
# these are in the build directory, which is determined at build time, not
# GYP time.
# These are passed as pairs of GlobalObjectName, GlobalObject.idl
interface_name_idl_filename = [(args[i], args[i + 1])
for i in range(0, len(args), 2)]
interface_name_to_global_names.update(read_pickle_file(options.global_objects_file))
for idl_filename in idl_files:
record_global_constructors(idl_filename)
# Check for [Exposed] / [Global] mismatch.
known_global_names = EXPOSED_EXECUTION_CONTEXT_METHOD.keys()
exposed_global_names = frozenset(global_name_to_constructors)
if not exposed_global_names.issubset(known_global_names):
unknown_global_names = exposed_global_names.difference(known_global_names)
raise ValueError('The following global names were used in '
'[Exposed=xxx] but do not match any [Global] / '
'[PrimaryGlobal] interface: %s'
% list(unknown_global_names))
# Write partial interfaces containing constructor attributes for each
# global interface.
for interface_name, idl_filename in interface_name_idl_filename:
# Work around gyp's path relativization for this parameter that is not
# a path, but gets interpreted as such.
interface_name = os.path.basename(interface_name)
constructors = interface_name_to_constructors(interface_name)
write_global_constructors_partial_interface(
interface_name, idl_filename, constructors)
from underthesea import pos_tag, word_tokenize
from utilities import filter_stopword, read_pickle_file, BRANDS
BRAND = 'brand'
GENDER = 'gender'
COLOR = 'color'
SIZE = 'size'
MEN = 'nam'
WOMEN = 'nữ'
MONGODB_URI = os.getenv('MONGODB_URI')
client = MongoClient(MONGODB_URI)
collection = client['nlp']['test']
intents = read_pickle_file('data/intents.pkl')
# Load trained model
model = load_model('saved_models/model_1')
# Load bag of words and tags
bag_of_words = read_pickle_file('data/words.pkl')
tags = read_pickle_file('data/tags.pkl')
# Initialize LabelBinarizer
data_lb = LabelBinarizer()
data_lb.fit(bag_of_words)
label_lb = LabelBinarizer()
label_lb.fit(tags)
def classify_question(tokens):
temp = [
from underthesea import word_tokenize
import time
from functools import reduce
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelBinarizer
from tensorflow.keras.layers import Dense, Flatten, Dropout
from tensorflow.keras.losses import CategoricalCrossentropy
from tensorflow.keras.models import Sequential
from utilities import read_pickle_file
if __name__ == "__main__":
data = []
labels = read_pickle_file('data/labels.pkl')
documents = read_pickle_file('data/documents.pkl')
bag_of_words = read_pickle_file('data/words.pkl')
tags = read_pickle_file('data/tags.pkl')
data_lb = LabelBinarizer()
data_lb.fit(bag_of_words)
for doc in documents:
temp = data_lb.transform(doc)
temp = reduce(np.add, list(temp))
data.append(temp)
data = np.array(data)
labels = np.array(labels)
(train_data, test_data, train_labels,
test_labels) = train_test_split(data, labels, test_size=0.2)
# train_data = data