A programmer who is great at writing bug-ridden programs decided to build a program that finds bugs instead.
The program takes source code as input and, whenever it finds a bug, comments out that part.
But the bug-finding program had a bug of its own: instead of commenting out the part it found, it replaces that spot with the string BUG.
Given the source code after the bug-finding program has processed it, write a program that removes every BUG from the code.
The input consists of several lines of source code that have already been processed by the bug-finding program described above. Each line is at most 100 characters long, and the input ends at end of file (EOF). The number of lines is unlimited.
Print the source code with every BUG removed. The printed source code must not contain any BUG.
Note that after deleting a BUG, the remaining characters join together and may form a new BUG, which must also be removed. For example, ABUBUGGB becomes ABUGB after removing the middle BUG, and removing BUG again finally yields AB. In other words, keep removing until no BUG remains.
The line structure is preserved; only the BUG occurrences inside each line are removed, and if every character of a line is removed, that line stays as an empty line.