import pyarrow as pa
import lance

data = pa.Table.from_pydict(
    {
        "id": list(range(2)),
        "tags": [
            ["puppy dinosaur", "puppy", "puppy horse"],
            ["horse dinosaur", "horse", "puppy dinosaur"],
        ],   
    }
)
ds = lance.write_dataset(data, "memory://") # maybe an old data storage version?

ds.create_scalar_index("tags", index_type="INVERTED")

ds.to_table(full_text_query="puppy").to_pylist()

[{'id': 0, 'tags': ['puppy dinosaur', 'puppy', 'puppy horse'], '_score': 0.5282782912254333}, 
 {'id': 1, 'tags': ['horse dinosaur', 'horse', 'puppy dinosaur'], '_score': 0.40841686725616455},
 {'id': 0, 'tags': ['puppy dinosaur', 'puppy', 'puppy horse'], '_score': 0.40841686725616455}, 
 {'id': 0, 'tags': ['puppy dinosaur', 'puppy', 'puppy horse'], '_score': 0.40841686725616455}]