it seems like test#4 includes big latin letters, but it's impossible according problem statement "все слова записаны в нижнем регистре, слово может содержать только латинские буквы от 'a' до 'z' (то есть «one-seventh» — это два слова — «one» и «seventh»)."
admin, please check

Requires to split the line into sequence of words, where word could be alphabet word or non-alphabet word. e.g.,

input:
it ockupies ebout one-seventh of the earth's surfase.

output:
it[ ]ockupies[ ]ebout[ ]one[-]seventh[ ]of[ ]the[ ]earth[']s[ ]surfase[.]

each bracketed word is non-alphabet and others are alphabet words.

Edited by author 24.07.2013 00:23

Edited by author 24.07.2013 00:23