Binary is available for Linux, macOS, and Windows. You can download the latest release from GitHub Releases.

Install:

$ docker pull ghcr.io/shibukawa/dbtestify:latest 

Execute: :

# Execute CLI command
$ docker run --rm -it -e "DBTESTIFY_CONN=(url)" ghcr.io/shibukawa/dbtestify:latest dbtestify [command] [args...]

# Publish API Server
$ docker run --rm -it -p 8080:8000-e "DBTESTIFY_CONN=(url)" ghcr.io/shibukawa/dbtestify:latest dbtestify http -p 8000

Install:

$ go get -tool github.com/shibukawa/dbtestify/cmd/dbtestify

Execute:

$ go tool dbtestify [command] [args...]

It supports PostgreSQL, MySQL, SQLite. Set the following use the following connection string to DBTESTIFY_CONN environment variable (or --db flag in CLI).

• PostgreSQL: detail
  • postgres://user:pass@localhost:5432/dbname?sslmode=disable
• MySQL: detail
  • mysql://root:pass@tcp(localhost:3306)/foo?tls=skip-verify
• SQLite: detail
  • sqlite://file:dbfilename.db

Then prepare the data set file in YAML format. This file is used for seeding the database and asserting the data in the database.

member:
- { id: 1, name: Frank, email: frank@example.com }
- { id: 2, name: Grace, email: grace@example.com }
- { id: 3, name: Heidi, email: heidi@example.com }
- { id: 4, name: Ivan }

belonging:
- { member_id: 1, group_id: 1 }
- { member_id: 2, group_id: 1 }
- { member_id: 3, group_id: 2 }
- { member_id: 4, group_id: 2 }

group:
- { id: 1, name: "Group A" }
- { id: 2, name: "Group B" }

You can seed/assert the data set from CLI, HTTP API, Go library

Use the following command to seed the data set into the database:

$ dbtestify seed testdata/users.yaml
$ dbtestify assert testdata/users.yaml

http subcommand launches a HTTP server.

$ dbtestify http -p 8000 ../testdata
dbtestify API server

        GET  http://localhost:8000/api/list                    : Show data set file list
        POST http://localhost:8000/api/seed/{data set path}    : Seed database content with the specified data set
        GET  http://localhost:8000/api/assert/{data set path}  : Assert database content with the specified data set
        start receiving at :8000

$ curl -X POST http://localhost:8000/api/seed/users.yaml
$ curl http://localhost:8000/api/assert/users.yaml

const DBTESTIFY_URL = 'http://localhost:8000/api';

test.beforeEach(async ({ request }) => {
    await request.post(`${DBTESTIFY_URL}/api/seed/users.yaml`)
});

test("add user", async ({ page, request }) => {
    await page.getByRole("button", { name: "Add User" }).click();
    await request.get(`${DBTESTIFY_URL}/api/assert/users.yaml`);
});

github.com/shibukawa/dbtestify/assertdb packages provides a helper for Go unit tests. Just calling assertdb.SeedDataSet and assertdb.AssertDB functions in your test code.

import (
	"embed"
	"os"
	"testing"

	"github.com/shibukawa/dbtestify/assertdb"
)

//go:embed dataset/*
var dataSet embed.FS

const dbtestifyConn = "sqlite://file:database.db"

func TestUsage(t *testing.T) {
    assertdb.SeedDataSet(t, dbtestifyConn, dataSet, "initial.yaml", nil)

    // some logic that modifies the database

    assertdb.AssertDB(t, dbtestifyConn, dataSet, "expect.yaml", nil)

Data set definition is a key feature of dbtestify. Data set is defined in YAML format. Basic structure is like this:

member:
- { id: 1, name: Frank, email: frank@example.com }
- { id: 2, name: Grace, email: grace@example.com }
- { id: 3, name: Heidi, email: heidi@example.com }
- { id: 4, name: Ivan }

belonging:
- { member_id: 1, group_id: 1 }
- { member_id: 2, group_id: 1 }
- { member_id: 3, group_id: 2 }
- { member_id: 4, group_id: 2 }

group:
- { id: 1, name: "Group A" }
- { id: 2, name: "Group B" }

Data set is used in two purposes. There are special options:

• For seeding of test data in unit tests This data set is imported into the database before running tests.
• For assertion the data in database This data set is compared with the data in the actual database.

There are several options for seeding

• clear-insert(default): Truncate the table then insert data.
• insert: Just insert data into the table.
• upsert: Insert data into the table, or update if the row already exists.
• truncate: Just truncate the table.
• delete: Delete rows in the table that matches the dataset's primary keys.

_operations:
  user: clear-insert
  access_log_2025_*: truncate
  login_history: delete

user:
- { id: 10, name: Frank, email: frank@example.com }
- { id: 11, name: Grace, email: grace@example.com }
- { id: 12, name: Heidi, email: heidi@example.com }
- { id: 13, name: Ivan}

login_history
- { user_id: 10, time: 2024-12-14 }

There are two options for matching rules.

• exact(default): The table should have exact same rows with the data set. But if the fields not in the data set is ignored (like system fields).
• sub: If the table has extra rows that are not in the data set, it is still valid.

And there are special field values for matching:

• [null]: It assumes the value is NULL. it is as same as null.
• [notnull]: It assumes the value is not NULL.
• [any]: It matches any value.
• [currentdate, <duration>]: It matches when the actual value is within the specified duration from the current time. The duration is optional and defaults to 1m (Go duration format, e.g., 30s, 2m, 1h). The actual value can be time.Time or a string parseable by RFC3339 or common SQL timestamp formats.
• [regexp, <pattern>]: It matches when the stringified actual value matches the given regular expression pattern.

_match:
  user: exact
  access_log_2025_*: sub

user:
- { id: 10, name: Frank, email: null }

login_history
- { user_id: 10,. time: [notnull]}

audit_log:
- { id: 1, created_at: [currentdate, 2m], message: [regexp, "^User .+ logged in$"] }

Each row can have tags. You can filter the rows by tags when loading:

member:
- { id: 1, name: Frank, _tags: [admin, user] }
- { id: 2, name: Grace, _tags: [user] }
- { id: 3, name: Ivy, _tags: [ex_user] }

# you can use `-i`/`-e` as a shortcut 
$ dbtestify seed testdata/users.yaml --include-tag=user --exclude-tag=admin
$ dbtestify assert testdata/users.yaml --include-tag=user --exclude-tag=admin

You can use Form data or JSON to specify the tags. For curl command, use form data is easier than JSON. But if request fromfetch or request(Playwright), you should feel using JSON is easier.

$ curl -X POST -d "i=user" -d "" http://localhost:8000/api/seed/users.yaml
$ curl -X POST -H "Content-Type: application/json" -d '{"include_tags": ["user"], "exclude_tags": ["admin"]}' http://localhost:8000/api/seed/users.yaml
$ curl -d "i=user" -d "" http://localhost:8000/api/assert/users.yaml

$ dbtestify assert testdata/users.yaml --tags user

• AGPL-3.0