I tested it and simple connection works fine via python more advanced usage
with nim needs consideration of how to pass strings from nim to duckdb so that
thet appear single quoted after being passed through nimpy.
pip install duckdb==0.2.3 nimble install nimpy
nimduck.nim
import nimpy
let duckdb = pyImport("duckdb")
let os = pyImport("os")
let conn = duckdb.connect()
let cursor = conn.cursor()
discard conn.execute("CREATE TABLE test_table (i INTEGER, j STRING)")
# add some data
discard conn.execute("""INSERT INTO test_table VALUES (1, 'one')""")
discard conn.execute("""INSERT INTO test_table VALUES (1, 'two')""")
discard conn.execute("""INSERT INTO test_table VALUES (1, 'three')""")
discard conn.execute("""INSERT INTO test_table VALUES (1, 'four')""")
# as pandas df
echo (conn.execute("SELECT * FROM test_table").fetchdf())
echo()
echo()
# as numpy array
echo (conn.execute("SELECT * FROM test_table").fetchnumpy())
Run
I have done a similar approach with Firebird3.0 server and their latest python
firebird driver , which now works nicely for simple queries. Firebird can also
work as embedded server and packs much more power than sqlite3.
<https://github.com/qqtop/nimfdb>
