feat: add datafusion adbc table provider notebook#12
Conversation
|
Here's a screenshot of the recipe: 
|
|
I’d suggest adding a paragraph explaining that although this demo uses DuckDB for convenience, the real power of this ADBC table provider is that it gives DataFusion access to more than a dozen different data systems that don’t have a dedicated DataFusion table provider, especially cloud data platforms like Snowflake. The second paragraph says some of this already but I think it should explicitly acknowledge that DuckDB is not really the greatest system to demonstrate this with. |
| ".unwrap()" | ||
| ] | ||
| }, | ||
| { |
There was a problem hiding this comment.
The output of this cell is just None so it'd probably be better to hide it.
| "\n", | ||
| "[ADBC](https://arrow.apache.org/adbc/) (Arrow Database Connectivity) is a database-agnostic API that retrieves query results directly in Arrow format, making it an efficient bridge between DataFusion and external databases. This demo uses DuckDB for simplicity, but ADBC's real value is providing DataFusion with access to more than a dozen different data systems that lack dedicated table providers.\n", | ||
| "\n", | ||
| "In this notebook, you will:\n", |
There was a problem hiding this comment.
| "In this notebook, you will:\n", | |
| "In this Rust-language notebook, you will:\n", |
| ], | ||
| "source": [ | ||
| "let df = ctx\n", | ||
| " .sql(\"SELECT * FROM datafusion.public.penguins LIMIT 10\")\n", |
There was a problem hiding this comment.
What dialect of SQL is expected here? DataFusion or DuckDB? Can you add a note to clarify this?
2 participants
Closes #8