def test_even_tree_for_string_2():
# aaaaabbabb
# 1111122122
inputstr = farach.str2int('1111122122')
t_odd = farach.T_odd(inputstr)
t_even = farach.T_even(t_odd, inputstr)
root = utils.Node(aId="root")
inner = utils.Node(1, "inner")
inner2 = utils.Node(2, "inner")
inner.add_child(inner2)
inner.add_child(utils.Node(3, 8))
inner2.add_child(utils.Node(9, 2))
inner2.add_child(utils.Node(7, 4))
root.add_child(inner)
root.add_child(utils.Node(5, 6))
root.add_child(utils.Node(1, 10))
root.update_leaf_list()
# print("t_even:")
# print(t_even.fancyprint(inputstr))
# print("root:")
# print(root.fancyprint(inputstr))
assert (t_even.fancyprint(inputstr) == root.fancyprint(inputstr))
def test_even_tree_for_string_1():
inputstr = utils.str2int("11112122")
t_odd = farach.T_odd(inputstr)
t_even = farach.T_even(t_odd, inputstr)
root = utils.Node(aId="root")
inner = utils.Node(1, "inner")
root.add_child(inner)
root.add_child(utils.Node(1, 8))
inner.add_child(utils.Node(7, 2))
inner2 = utils.Node(2, "inner")
inner.add_child(inner2)
inner2.add_child(utils.Node(5, 4))
inner2.add_child(utils.Node(3, 6))
root.update_leaf_list()
assert t_even.fancyprint(inputstr) == root.fancyprint(inputstr)
def conll_to_action(self, tgt, num_, permutation):
cnt = 0
if 'dev' in tgt:
oracle_fname = './data/processed/dev.oracle.en'
txt_fname = './data/processed/dev.en'
elif 'test' in tgt:
oracle_fname = './data/processed/test.oracle.en'
txt_fname = './data/processed/test.en'
elif 'train' in tgt:
oracle_fname = './data/processed/train.oracle.en'
txt_fname = './data/processed/train.en'
else:
print('Error: invalid file name of ' + tgt)
exit(1)
oracle_f = open(oracle_fname, 'w', encoding='utf-8')
plain_f = open(txt_fname, 'w', encoding='utf-8')
tagged_file = open(tgt, 'r', encoding='utf-8')
bulk = tagged_file.read()
blocks = re.compile(r"\n{2,}").split(bulk)
blocks = list(filter(None, blocks))
for i in permutation:
block = blocks[i]
tokens = []
buffer = []
child_to_head_dict = {}
for line in block.splitlines():
attr_list = line.split('\t')
if attr_list[1] == '.' or attr_list[1] == '?':
continue
tokens.append(attr_list[1])
num = int(attr_list[0])
head = int(attr_list[6])
label = attr_list[7]
node = utils.Node(num, head, label)
child_to_head_dict[num] = head
buffer.append(node)
arcs = utils.write_oracle(buffer, child_to_head_dict)
for i, token in enumerate(tokens):
token_lowered = token.lower()
if i == 0:
plain_f.write(token_lowered)
else:
plain_f.write(' ')
plain_f.write(token_lowered)
plain_f.write('\n')
for arc in arcs:
oracle_f.write(arc + '\n')
oracle_f.write('\n')
cnt += 1
if cnt == num_:
break
tagged_file.close()
oracle_f.close()
plain_f.close()
return txt_fname, oracle_fname
def insert(head, val):
'''
Insert val in to linked list
@params: val: int
@params: head: Node
@return: head: Node
'''
if not head or head.val >= val:
new_head = utils.Node(val)
new_head.next = head
return new_head
cur = head
while cur.next and cur.next.val < val:
cur = cur.next
if not cur.next:
# val is the last node
cur.next = utils.Node(val)
else:
node = utils.Node(val)
node.next = cur.next
cur.next = node
return head
def remove(head, target):
'''
Remove node if val is target
@params: head: Node
@params: target: int
@return: head: Node
'''
cur = dummy = utils.Node(-1)
dummy.next = head
while cur.next:
if cur.next.val == target:
cur.next = cur.next.next
else:
cur = cur.next
return dummy.next
