Create test cases for complex unicode grapheme clusters

[christianparpart]

The following list of checks would be nice to be added.

Requirements

The test cases must always run the same, regardless of execution environment. No randomly generated data.

Checklist

• 2-byte UTF-8 sequence
• 3-byte UTF-8 sequence
• 4-byte UTF-8 sequence
• invalid 1/2/3/4-byte UTF-8 sequence
• a 2-codepoint grapheme cluster
• a 6-codepoint grapheme cluster (e.g. \u270c\U0001F926\U0001F3FC\u200D\u2642\uFE0F 🤦🏼‍♂️ Face Palm)
• a wide graapheme cluster with VS15 to make it narrow (remains wide, as per terminal-unicode-core)
• a narrow grapheme cluster with VS16 to make it wide