DataLoader ignores context manager for Data Sources

Grain's documentation notes that `DataLoader` will try using a data source as a context manager:

>Open file handles should be closed after use. Data sources typically open underlying files in order to read records from them. We recommend implementing data sources as context managers that close their open file handles within the __exit__ method. When opening a data source, the DataLoader will first attempt to use the data source as a context manager. If the data source doesn’t implement the context manager protocol, it will be used as-is, without a with statement.

However, from my testing, it seems that DataLoader doesn't do this. See the following code snippet:

```
import grain.python as pygrain
from grain.sources import RandomAccessDataSource

class ContextDataset(RandomAccessDataSource):
data: list

def __len__(self):
return len(self.data)

def __getitem__(self, idx):
return self.data[idx]

def __enter__(self):
print('entering')
self.data = [1,2,3,4,5,6]
return self

def __exit__(self, exc_type, exc_val, exc_tb):
print('exiting')
return False

if __name__ == '__main__':
index_sampler = pygrain.IndexSampler(
num_records=6,
num_epochs=1,
shard_options=pygrain.NoSharding(),
shuffle=False,
seed=0)
transformations = [pygrain.Batch(batch_size=1, drop_remainder=True)]
data_source = ContextDataset()
dataloader = pygrain.DataLoader(
data_source=data_source,
operations=transformations,
sampler=index_sampler,
worker_count=0)

for data in dataloader: # errors here due to data not being initialized
print(data)
```

In this example, should the datasource be accessed as a context manager, `data` will be populated. However, running the example raises `AttributeError: 'ContextDataset' object has no attribute 'data'` when calling `__getitem__` by iterating over the dataloader, and `entering` is never printed.

I tested with the latest stable version of grain (0.2.11) on Python 3.12.7.