Problem

At Dropbox, you are asked to implement a lightweight string compressor using run-length encoding. Given a string s, compress consecutive repeated characters so that each group becomes the character followed by its count, then return whichever is shorter: the compressed string or the original string.

Formal Specification

• Input: a string s
• Output: a string
• Consecutive repeated characters are encoded as character + count
• If the compressed result is not strictly shorter than the original string, return the original string

Examples

Example 1

• Input: s = "aaabb"
• Output: "a3b2"
• Explanation: The groups are aaa and bb, so the compressed form is a3b2, which is shorter than aaabb.

Example 2

• Input: s = "abc"
• Output: "abc"
• Explanation: The compressed form would be a1b1c1, which is longer than the original string.

Constraints

• 0 <= len(s) <= 10^5
• s may contain uppercase letters, lowercase letters, digits, spaces, or symbols
• Compression is based only on consecutive characters
• You must preserve character order exactly