import sys # detect numbers. separate numbers by spaces def solve(lines): res = [] j = 0 for l in lines: numbers = [] for i in range(len(l)): c = l[i] if j > 0 and not c.isdigit(): numbers.append(l[i-j:i]) j = j+1 if (c.isdigit() and c.isascii()) else 0 if numbers: res.append(' '.join(numbers)) return '\n'.join(res) if __name__ == "__main__": lines = [] fp = sys.argv[1] with open(fp, encoding="utf-8") as f: lines = f.readlines() sol = solve(lines) with open("./polish_wiki_excerpt.out", 'w', encoding="utf-8") as f: f.writelines(sol) f.write('\n') # it expects for file to end with newline :)