Merge pull request #737 from Ewan-Keith/add-databricks-support

 Add databricks support

Author: k1LoW

.gitignore

@@ -1,5 +1,6 @@
 tbls
 tbls.exe
+tbls.yml
 dist/
 dbdoc/
 coverage.out

Makefile

@@ -37,7 +37,7 @@ db_sqlite:
 	sqlite3 $(PWD)/testdata/testdb.sqlite3 < testdata/ddl/sqlite.sql
 
 test:
-	go test ./... -tags 'bq clickhouse dynamo mariadb mongodb mssql mysql postgres redshift snowflake spanner sqlite' -coverprofile=coverage.out -covermode=count
+	go test ./... -tags 'bq clickhouse databricks dynamo mariadb mongodb mssql mysql postgres redshift snowflake spanner sqlite' -coverprofile=coverage.out -covermode=count
 
 test-no-db:
 	go test ./... -coverprofile=coverage.out -covermode=count

README.md

@@ -607,6 +607,32 @@ dsn: clickhouse://dbuser:dbpass@hostname:9000/dbname
 
 See also: https://pkg.go.dev/github.com/ClickHouse/clickhouse-go
 
+**Databricks (Experimental):**
+
+Personal Access Token (PAT) Authentication:
+```yaml
+# .tbls.yml
+dsn: databricks://your_databricks_workspace_id.cloud.databricks.com:443/sql/1.0/warehouses/your_warehouse_id?catalog=your_catalog&schema=your_schema&token=your_token
+```
+
+OAuth Client Credentials Authentication:
+```yaml
+# .tbls.yml
+dsn: databricks://your_databricks_workspace_id.cloud.databricks.com:443/sql/1.0/warehouses/your_warehouse_id?catalog=your_catalog&schema=your_schema&client_id=your_client_id&client_secret=your_client_secret
+```
+
+Required parameters:
+- `your_databricks_workspace_id`: databricks workspace ID
+- `your_warehouse_id`: databricks sql warehouse ID
+- `your_catalog`: Unity Catalog name
+- `your_schema`: Schema/database name within the catalog
+
+Authentication (choose one):
+- PAT Authentication: `token` - Personal access token for authentication
+- OAuth Authentication: `client_id` and `client_secret` - OAuth M2M client credentials
+
+See also: https://pkg.go.dev/github.com/databricks/databricks-sql-go
+
 **JSON:**
 
 The JSON file output by the `tbls out -t json` command can be read as a datasource (JSON Schema is [here](spec/tbls.schema.json_schema.json)).

datasource/databricks.go

@@ -0,0 +1,83 @@
+package datasource
+
+import (
+	"database/sql"
+	"errors"
+	"fmt"
+	"net/url"
+
+	"github.com/k1LoW/tbls/drivers/databricks"
+	"github.com/k1LoW/tbls/schema"
+)
+
+// AnalyzeDatabricks analyze `databricks://`
+func AnalyzeDatabricks(urlstr string) (_ *schema.Schema, err error) {
+	s := &schema.Schema{}
+
+	u, err := url.Parse(urlstr)
+	if err != nil {
+		return nil, err
+	}
+
+	catalog := u.Query().Get("catalog")
+	if catalog == "" {
+		return nil, errors.New("no catalog name in the connection string")
+	}
+	schemaName := u.Query().Get("schema")
+	if schemaName == "" {
+		return nil, errors.New("no schema name in the connection string")
+	}
+
+	// Extract authentication parameters
+	token := u.Query().Get("token")
+	clientID := u.Query().Get("client_id")
+	clientSecret := u.Query().Get("client_secret")
+
+	// Validate authentication parameter combinations
+	hasToken := token != ""
+	hasOAuth := clientID != "" && clientSecret != ""
+
+	if !hasToken && !hasOAuth {
+		return nil, errors.New("authentication required: provide either 'token' for PAT authentication or both 'client_id' and 'client_secret' for OAuth authentication")
+	}
+
+	if hasToken && hasOAuth {
+		return nil, errors.New("conflicting authentication methods: provide either 'token' for PAT authentication OR 'client_id'/'client_secret' for OAuth authentication, not both")
+	}
+
+	if clientID != "" && clientSecret == "" {
+		return nil, errors.New("incomplete OAuth credentials: 'client_secret' is required when 'client_id' is provided")
+	}
+
+	if clientSecret != "" && clientID == "" {
+		return nil, errors.New("incomplete OAuth credentials: 'client_id' is required when 'client_secret' is provided")
+	}
+
+	s.Name = fmt.Sprintf("%s.%s", catalog, schemaName)
+
+	// Build databricks driver DSN based on authentication method
+	var databricksDSN string
+	if hasToken {
+		// PAT token authentication: token:TOKEN@host:port/path?catalog=CATALOG&schema=SCHEMA
+		databricksDSN = fmt.Sprintf("token:%s@%s%s?catalog=%s&schema=%s", token, u.Host, u.Path, catalog, schemaName)
+	} else {
+		// OAuth client credentials authentication: host:port/path?authType=OauthM2M&clientID=ID&clientSecret=SECRET&catalog=CATALOG&schema=SCHEMA
+		databricksDSN = fmt.Sprintf("%s%s?authType=OauthM2M&clientID=%s&clientSecret=%s&catalog=%s&schema=%s", 
+			u.Host, u.Path, clientID, clientSecret, catalog, schemaName)
+	}
+
+	db, err := sql.Open("databricks", databricksDSN)
+	if err != nil {
+		return nil, err
+	}
+	defer func() {
+		_ = db.Close()
+	}()
+
+	driver := databricks.New(db)
+	if err := driver.Analyze(s); err != nil {
+		return nil, err
+	}
+
+	return s, nil
+}

datasource/datasource.go

@@ -68,6 +68,9 @@ func Analyze(dsn config.DSN) (_ *schema.Schema, err error) {
 	if strings.HasPrefix(urlstr, "mongodb://") || strings.HasPrefix(urlstr, "mongo://") {
 		return AnalyzeMongodb(urlstr)
 	}
+	if strings.HasPrefix(urlstr, "databricks://") {
+		return AnalyzeDatabricks(urlstr)
+	}
 	s := &schema.Schema{}
 	u, err := dburl.Parse(urlstr)
 	if err != nil || !slices.Contains(supportDriversWithDburl, u.Driver) {