Multi-threaded/multi-process support on top of SQLite. The intent is to ensure a SQL statement
will get executed, even if other threads or processes are trying to access the DB. Avoids
database is locked issues.
No additional package dependencies beyond regular Python.
Intended for relatively simple SQL statement execution. Locks the DB file on every access, with built-in retry mechanism.
Even though the DB is locked on every access, typically simple writes are much less than 1 second (more like 1 mS), so this latency is still usable for many use cases.
table_name = "example"
schema = {"id PRIMARY KEY": int, "name": str, "color": str, "year": int}
db_path = Path("temp", "example.sqlite")
db_path.parent.mkdir(exist_ok=True)
# Write and read data.
with MSQLite(db_path, table_name, schema) as db:
now = time.monotonic_ns() # some index value
# insert some data
db.execute(f"INSERT INTO {table_name} VALUES ({now}, 'plate', 'red', 2020), ({now + 1}, 'chair', 'green', 2019)")
# read the data back out
response = db.execute(f"SELECT * FROM {table_name}")
for row in response:
print(row)
# Read data out. No longer needs the schema.
with MSQLite(db_path, table_name) as db:
response = db.execute(f"SELECT * FROM {table_name}")
for row in response:
print(row)
