CLI insert tool now uses generators, closes #7

Also cleaned up the logic so we commit rows in batches too.
This commit is contained in:
Simon Willison 2019-01-27 22:26:45 -08:00
commit 50589f8523
2 changed files with 15 additions and 14 deletions

View file

@ -72,17 +72,17 @@ def optimize(path, no_vacuum):
@click.argument("json_file", type=click.File(), required=True)
@click.option("--pk", help="Column to use as the primary key, e.g. id")
@click.option("--nl", is_flag=True, help="Expect newline-delimited JSON")
def insert(path, table, json_file, pk, nl):
@click.option("--batch-size", type=int, default=100, help="Commit every X records")
def insert(path, table, json_file, pk, nl, batch_size):
"Insert records from JSON file into the table, create table if it is missing"
db = sqlite_utils.Database(path)
if nl:
# TODO: Use a generator once #7 is solved
docs = [json_std.loads(line) for line in json_file]
docs = (json_std.loads(line) for line in json_file)
else:
docs = json_std.load(json_file)
if isinstance(docs, dict):
docs = [docs]
db[table].insert_all(docs, pk=pk)
db[table].insert_all(docs, pk=pk, batch_size=batch_size)
@cli.command()

View file

@ -185,13 +185,14 @@ class Table:
def create(self, columns, pk=None, foreign_keys=None, column_order=None):
columns = {name: value for (name, value) in columns.items()}
self.db.create_table(
self.name,
columns,
pk=pk,
foreign_keys=foreign_keys,
column_order=column_order,
)
with self.db.conn:
self.db.create_table(
self.name,
columns,
pk=pk,
foreign_keys=foreign_keys,
column_order=column_order,
)
self.exists = True
return self
@ -384,9 +385,9 @@ class Table:
values.extend(
jsonify_if_needed(record.get(key, None)) for key in all_columns
)
result = self.db.conn.execute(sql, values)
self.db.conn.commit()
self.last_id = result.lastrowid
with self.db.conn:
result = self.db.conn.execute(sql, values)
self.last_id = result.lastrowid
return self
def upsert(self, record, pk=None, foreign_keys=None, column_order=None):