Bug: different number of minimum characters for a string generators, if field meta in SQLAlchemyFactory have constraints

### Description

Different number of minimum characters for a string, if field meta in SQLAlchemyFactory have constraints

### URL to code causing the issue

https://github.com/litestar-org/polyfactory/blob/main/polyfactory/value_generators/primitives.py

### MCVE

```python

```

### Steps to reproduce

```bash
If we get `field_meta` for a `String` column with non-empty constraints and without `min_lengths`, then in `process_kwargs` in `get_field_value`, we will call the `get_constrained_field_value` function, which will eventually call the `create_random_bytes` function with `min_length=None`, and this function will set `min_length=0`. Although `faker.pystr()`, which we would use, generates strings of at least 4 characters.

Because of this, it often happened to me that my unique fields with a limit on the maximum number of characters generated empty lines for me and caused an error limiting the integrity of unique elements for a column.

It was hard to detect this behavior. I think it would be better to have `min_length` default to 4 when `None` is passed to `create_random_bytes`, or to catch `min_length=None` in `create_random_string` and set it to 4.
```

### Screenshots

"In the format of: ``"

### Logs

```bash

```

### Release Version

3.3.0

### Platform

- [x] Linux
- [ ] Mac
- [ ] Windows
- [ ] Other (Please specify in the description above)

Start in polyfactory/value_generators/primitives.py and trace create_random_bytes, create_random_string, and get_constrained_field_value when min_length is None. Reproduce the SQLAlchemyFactory case with a constrained String field and inspect the generated lengths. Done means the minimum length is consistent with the intended behavior and regression coverage verifies it.

Written by the indexing model from the issue text.