Query.any/all treat strings as sequences, causing false matches

Description

Query.any / Query.all (value-list form) use:

def is_sequence(obj):
    return hasattr(obj, '__iter__')

(tinydb/queries.py)

Strings (and bytes, dicts, etc.) are iterable, so a string field is treated like a list of characters. That yields false positives that the docs do not describe (docs talk about list fields).

Verified on current master (4aa5311, TinyDB 4.9.0).

Minimal reproduction

from tinydb.queries import Query

Q = Query()
print(Q.name.any(['a', 'b'])({'name': 'abc'}))  # True  (unexpected)
print(Q.name.all(['a', 'b'])({'name': 'abc'}))  # True  (unexpected)

Why: iterating 'abc' yields 'a', 'b', 'c', and 'a' in ['a','b'] / 'b' in ['a','b'].

Same class of surprise for bytes and for dict fields (iterates keys).

Expected behavior (for discussion)

Possible directions:

1. Treat only real sequences that are not str/bytes (e.g. list/tuple, or collections.abc.Sequence excluding text/bytes)
2. Keep current behavior but document that string fields are treated as character sequences
3. Something else you prefer

For typical document fields, (1) matches user expectations and the current docs.

Suggested tests

• Query().name.any(['a','b']) on {'name': 'abc'} → False (if option 1)
• Existing list-field any/all tests keep passing
• Optional: bytes / mapping field cases

Happy to open a PR once the preferred behavior is clear.

Environment

• TinyDB 4.9.0 / master @ 4aa5311