Exercise 3 solution

    import re

    user_re = re.compile(r"[A-Za-z][A-Za-z0-9_]{2,15}")

    def check(s):
        return "ok" if user_re.fullmatch(s) else "invalid"

    for s in ["bob", "ab", "9lives", "good_name_1", "name with space", "a" * 17, "ok\n"]:
        print(repr(s[:12]), check(s))
    # 'bob' ok
    # 'ab' invalid
    # '9lives' invalid
    # 'good_name_1' ok
    # 'name with sp' invalid
    # 'aaaaaaaaaaaa' invalid
    # 'ok\n' invalid

    print(bool(re.match(r"[A-Za-z]\w{2,15}$", "bob\n")))
    # True

(All lines were run and printed exactly what is shown.)

Explanation: fullmatch rejects the trailing newline, but the "match with $"
version accepts it, because $ also matches just before a final newline.
That is why fullmatch is the safer choice for validation.