Generating fake data with Faker¶

drongo.seed fills the mocked services with realistic test data, so you don't have to hand-write fixtures. It uses Faker and drives the real google-cloud clients inside a mock_gcp scope, so the data lands in the in-memory backends exactly like your code's own writes.

Install¶

Faker is an optional dependency:

pip install "drongo[faker]"

Calling a seeder without Faker installed raises a clear ImportError telling you to install the extra.

API¶

All seeders take a client you already built inside the mock scope, so they use the same project and mock backends as your code.

`seed.bigquery_rows(client, table, count=10, *, overrides=None)`¶

Generate count rows that match a table's schema and stream-insert them (insert_rows_json). Returns the generated rows.

from drongo import mock_gcp, seed


@mock_gcp
def test_orders():
    from google.cloud import bigquery

    client = bigquery.Client(project="shop")
    client.create_dataset("shop.analytics")
    client.create_table(
        bigquery.Table(
            "shop.analytics.orders",
            schema=[
                bigquery.SchemaField("order_id", "STRING"),
                bigquery.SchemaField("amount", "FLOAT"),
                bigquery.SchemaField("quantity", "INTEGER"),
                bigquery.SchemaField("paid", "BOOLEAN"),
            ],
        )
    )

    rows = seed.bigquery_rows(client, "shop.analytics.orders", count=500)

    assert len(rows) == 500
    assert len(list(client.list_rows("shop.analytics.orders"))) == 500

Values are generated per BigQuery column type:

BigQuery type	Generated value
`STRING`	a word
`INTEGER` / `INT64`	int in `[0, 100000]`
`FLOAT` / `NUMERIC`	float in `[0, 10000]`, 2 dp
`BOOLEAN`	bool
`TIMESTAMP` / `DATETIME`	ISO-8601 datetime
`DATE`	date
`BYTES`	base64 of 16 random bytes
anything else	a word

Pin specific columns with overrides:

seed.bigquery_rows(client, "shop.analytics.orders", count=10, overrides={"paid": True})

`seed.storage_blobs(client, bucket, count=10, *, prefix="file", suffix=".txt")`¶

Upload count fake text objects to a bucket. Returns the blob names.

from google.cloud import storage

gcs = storage.Client(project="shop")
gcs.create_bucket("dropzone")
names = seed.storage_blobs(gcs, "dropzone", count=25, prefix="invoice", suffix=".txt")

`seed.pubsub_messages(publisher, topic, count=10)`¶

Publish count fake messages to a topic. Returns the message ids.

from google.cloud import pubsub_v1

publisher = pubsub_v1.PublisherClient()
publisher.create_topic(request={"name": "projects/shop/topics/events"})
seed.pubsub_messages(publisher, "projects/shop/topics/events", count=50)

Reproducible runs¶

Seed Faker for deterministic data:

from drongo import seed

seed.seed(1234)  # same data every run

Notes¶

Seeders are ordinary Python; use them to build any starting state, then run the code under test against it.
bigquery_rows fills every column by type. If your code reads specific values, set them with overrides so assertions are stable.