Add ability to "inject"options from the results of other `Evaluatable`s

There are times when you want to dynamically generate the value of an option using a dataset or other `Evaluatable` object. This is most often useful when using a `dataset` that someone else wrote.

Currently, a similar ability exists in the [`Map`](https://8451.github.io/labrea/usage.html#map) type, but only when iterating over the result of an `Evaluatable` that returns an interable.

The most natural way to do this would probably be to extend the `WithOptions` type (which powers the `Dataset.with_options` method) so that you can provide dynamic options as well as static ones. For example

```python
from labrea import dataset, Option

@dataset
def foo(bar: str = Option("BAR")) -> str:
return bar

@dataset
def bar(msg: str = Option("MSG")) -> str:
return msg

foo.with_options({"BAR": bar}).evaluate({"MSG": "Hello World!"})
# 'Hello World!'

Look at the `WithOptions` type in the codebase, likely in a file like `with_options.py` or `evaluatable.py`. Understand how static options are currently handled. The goal is to allow an option's value to be the result of another `Evaluatable` (like a dataset). Check the `Map` type's implementation for iterable-based dynamic options as a reference. The change involves modifying the option resolution logic to evaluate provided `Evaluatable` objects and inject their results. Run existing tests for `WithOptions` and `Map` to ensure the new feature works and doesn't break existing functionality.

Scritto dal modello di indicizzazione a partire dal testo della issue.