import re, os
from util.model.baseline_classifier import copy_folder
from util.d_fill_link_attrs import generate_qslinks, generate_olinks,\
generate_movelinks
from util.c_fill_tag_attrs import generate_attributes
if __name__ == "__main__":
with open('config.txt') as fo:
text = fo.read()
train_path = re.findall('TRAINING_PATH *= *(.*)', text)[0]
gen_path = re.findall('CONFIG_2_GEN_PATH *= *(.*)', text)[0]
parent_path = os.path.dirname(gen_path)
hyp_a = os.path.join(parent_path, '2a')
hyp_b = os.path.join(parent_path, '2b')
hyp_c = os.path.join(parent_path, '2c')
# 2a
copy_folder(gen_path, hyp_a)
generate_attributes(train_path, hyp_a, hyp_a)
# 2b + c
copy_folder(hyp_a, hyp_b)
generate_qslinks(train_path, hyp_b, hyp_b)
generate_olinks(train_path, hyp_b, hyp_b)
generate_movelinks(train_path, hyp_b, hyp_b)
copy_folder(hyp_b, hyp_c)
# pre-1a
with open('config.txt') as fo:
text = fo.read()
train_path = re.findall('TRAINING_PATH *= *(.*)', text)[0]
gen_path = re.findall('CONFIG_1_GEN_PATH *= *(.*)', text)[0]
parent_path = os.path.dirname(gen_path)
hyp_1 = os.path.join(parent_path, '1')
hyp_a = os.path.join(parent_path, '1a')
hyp_b = os.path.join(parent_path, '1b')
hyp_c = os.path.join(parent_path, '1c')
hyp_d = os.path.join(parent_path, '1d')
hyp_e = os.path.join(parent_path, '1e')
# 1a + 1b
generate_elements(train_path, gen_path, hyp_1)
generate_tags(train_path, hyp_1, gen_path, hyp_a)
copy_folder(hyp_a, hyp_b)
# 1c
copy_folder(hyp_b, hyp_c)
generate_attributes(train_path, hyp_c, hyp_c)
#1d + e
copy_folder(hyp_c, hyp_d)
generate_qslinks(train_path, hyp_d, hyp_d)
generate_olinks(train_path, hyp_d, hyp_d)
generate_movelinks(train_path, hyp_d, hyp_d)
copy_folder(hyp_d, hyp_e)
'id': '{}{}'.format(type_keys[type_name][0], id_number),
}
tag.update(type_fields)
curr_doc.insert_tag(tag)
id_number += 1
# switch to established files
curr_doc.save_xml(os.path.join(out_path, doc_name))
clean_corpus = HypotheticalCorpus(out_path)
clean_data = list(clean_corpus.documents())
#===============================================================================
if __name__ == "__main__":
# DEMO
train_path = './data/training'
test_path = './data/baseline/test.configuration1/'
clean_path = './data/baseline/test.configuration1/'
hyp_a = './data/final/test/configuration1/a'
hyp_b = './data/final/test/configuration1/b'
hyp_c = './data/final/test/configuration1/c'
generate_tags(train_path, test_path, clean_path, hyp_a)
# copy to next stages
copy_folder(hyp_a, hyp_b)
copy_folder(hyp_a, hyp_c)
d = TypesClassifier('PATH', train_path, test_path, gold_path = '')
d.run_demo()
tags = extent.document.query_extents(['SPATIAL_SIGNAL'], extent.lex[0].begin, extent.lex[-1].end)
if tags:
tag = tags[0]
tag.attrs['semantic_type'] = signal_type_labels[offsets]
curr_doc.save_xml(os.path.join(out_path, doc_name))
def generate_attributes(train_path, test_path, out_path):
# make outpath
if not os.path.exists(out_path):
os.makedirs(out_path)
# attributes for 7 tag types
generate_motion_attr(train_path, test_path, out_path)
generate_motion_signal_attr(train_path, test_path, out_path)
generate_event_attr(train_path, test_path, out_path)
generate_path_attr(train_path, test_path, out_path)
generate_place_attr(train_path, test_path, out_path)
generate_entity_attr(train_path, test_path, out_path)
generate_signal_attr(train_path, test_path, out_path)
if __name__ == "__main__":
# DEMO
training_path = './data/training'
hyp_c = './data/final/test/configuration1/c'
hyp_d = './data/final/test/configuration1/d'
generate_attributes(training_path, hyp_c, hyp_c)
copy_folder(hyp_c, hyp_d)
