Collapse Runs of Spaces

Source: Kernighan & Ritchie

Version: Phuoc Hung OJ Extended

Problem Statement

Read the entire UTF-8 stream to EOF; replace each contiguous run of ordinary spaces (U+0020) by exactly one space. Preserve leading/trailing spaces, tabs, line breaks and all other content. Empty input gives empty output. This PHOJ edition excludes CR and caps length at 100000 bytes.

Input

A UTF-8 stream of at most 100000 bytes, no CR, possibly empty or unterminated.

Output

Output transformed text exactly, without inserting a final newline.

Subtasks

• Subtask 1 (20%): UTF-8 stream length at most 100 bytes, no CR.

• Subtask 2 (30%): UTF-8 stream length at most 2000 bytes, no CR.

• Subtask 3 (50%): UTF-8 stream length at most 100000 bytes, no CR.

Examples

Example 1

Input:

A   B	  C

Output:

A B	 C

Explanation:

Collapse adjacent ordinary spaces only; tab and newline delimit separate runs. Actual escaped input: "A B\t C\n"; escaped output: "A B\t C\n".

Example 2

Input:

   A  B   

Output:

 A B 

Explanation:

Collapse adjacent ordinary spaces only; tab and newline delimit separate runs. Actual escaped input: " A B "; escaped output: " A B ".

Example 3

Input:

Output:

Explanation:

Collapse adjacent ordinary spaces only; tab and newline delimit separate runs. Actual escaped input: ""; escaped output: "".

Example 4

Input:

A	  B

Output:

A	 B

Explanation:

Collapse adjacent ordinary spaces only; tab and newline delimit separate runs. Actual escaped input: "A\t B"; escaped output: "A\t B".

The quoted forms below use \t for a literal tab, \n for an LF, and \ for a literal backslash; they describe the actual byte stream, not additional characters in input.