import re

test_str = u"""
10
00:00:33,900 --> 00:00:34,767
Get to Rabaul.
Yeah.

11
00:00:34,767 --> 00:00:36,033
[Ground rumbling]
 
12
00:00:36,033 --> 00:00:37,533
Earthquake. Whoa.
[Children screaming]
 
13
00:00:37,533 --> 00:00:39,200
Holy [bleep]
 
14
00:00:40,133 --> 00:00:44,333
We heard that your tribe has
found wreckage in the jungles.
 
70
00:03:12,800 --> 00:03:14,767
[Airplane engine roars]
 
71
00:03:17,200 --> 00:03:20,767
In the last 75 years, there
have been countless dead ends,
"""

pattern = re.compile(r"(?=\d+:\d+)(.*)(\s|\s.*\s)(?=.*\[).*\[(.*)\]", re.MULTILINE | re.UNICODE)
matches = re.findall(pattern, test_str)

for item in matches:
    print item[0] + ' ' + item[2]