create-dql-app 1.6.13 → 1.6.14

package/README.md

@@ -6,6 +6,9 @@ The fastest way to start a DQL project.
 npx create-dql-app@latest my-analytics
 cd my-analytics
 npm install
+npm install --prefix .dql/connectors duckdb       # DuckDB/local files only
+# npm install --prefix .dql/connectors snowflake-sdk  # Snowflake only
+# Databricks does not need an extra package.
 npm run notebook
 ```
 
@@ -13,16 +16,28 @@ Opens a running notebook at <http://127.0.0.1:3474> in under 5 minutes on a
 clean machine. No global install required; the template installs
 `@duckcodeailabs/dql-cli` locally and exposes it through npm scripts.
 
-The scaffolded project is a clean starter: `dql.config.json` (DuckDB
-in-memory by default — bring your own warehouse), a welcome notebook, and npm
-scripts for `notebook`, `compile`, `lineage`, `doctor`, and `validate`. If a
-sibling dbt project is detected, it is wired into `dql.config.json`
-automatically so `dql sync dbt` works out of the box.
+The scaffolded project is a clean starter: `dql.config.json`, a welcome
+notebook, and npm scripts for `notebook`, `compile`, `lineage`, `doctor`, and
+`validate`. If a sibling dbt project is detected, it is wired into
+`dql.config.json` automatically so `dql sync dbt` works out of the box.
+
+Database drivers are project-local so the base install stays fast:
+
+| Database | Extra install | Notes |
+| --- | --- | --- |
+| Databricks SQL | none | Built into DQL through HTTPS |
+| DuckDB or local files | `npm install --prefix .dql/connectors duckdb` | Needed for DuckDB, CSV, Parquet, and JSON |
+| Snowflake | `npm install --prefix .dql/connectors snowflake-sdk` | Needed for Snowflake connections |
+
+You can also install DuckDB or Snowflake from the notebook Connections page.
 
 Want a ready-made dbt project to try DQL on? Clone the example repo and add
 DQL to it — the same steps you'd use on your own dbt repo:
 [github.com/duckcode-ai/jaffle-shop-duckdb](https://github.com/duckcode-ai/jaffle-shop-duckdb).
 
+Already installed `@duckcodeailabs/dql-cli` globally? That only installs the
+command. From an existing dbt repo, run `dql init ./dql` before `cd dql`.
+
 ## Flags
 
 | Flag | Default | Meaning |

package/bin/create-dql-app.mjs

@@ -167,6 +167,12 @@ async function main() {
   const pm = detectPackageManager();
   const installCmd = pm === 'npm' ? 'npm install' : `${pm} install`;
   const runCmd = pm === 'npm' ? 'npm run notebook' : `${pm} notebook`;
+  const connectorTip = [
+    c.dim('# Optional database driver install before running queries:'),
+    c.dim('#   DuckDB/local files: npm install --prefix .dql/connectors duckdb'),
+    c.dim('#   Snowflake:       npm install --prefix .dql/connectors snowflake-sdk'),
+    c.dim('#   Databricks:      no extra package'),
+  ].join('\n  ');
   const dbtTip = dbtSibling
     ? `\n\n${c.dim('Tip:')} run ${c.bold('dbt parse')} inside ${c.dim(dbtSibling)}\n     first, then ${c.bold(pm === 'npm' ? 'npm run sync' : `${pm} sync`)} to import the dbt DAG.`
     : '';
@@ -176,6 +182,7 @@ ${c.green('✓ Ready.')} Next steps:
 
   ${c.bold(`cd ${args.dir}`)}
   ${c.bold(installCmd)}
+  ${connectorTip}
   ${c.bold(runCmd)}${dbtTip}
 
 Your notebook will open at ${c.cyan('http://127.0.0.1:3474')}.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "create-dql-app",
-  "version": "1.6.13",
+  "version": "1.6.14",
   "description": "Scaffold a new DQL project. Run with: npx create-dql-app <name>",
   "license": "MIT",
   "author": "DuckCode AI Labs",

package/templates/starter/README.md

@@ -17,17 +17,29 @@ connects technical lineage with business lineage.
 
 ## Connect your warehouse
 
-Edit `dql.config.json` — DQL ships 15 drivers out of the box:
-[docs/reference/connectors.md](https://github.com/duckcode-ai/dql/blob/main/docs/reference/connectors.md).
-Add your warehouse connection in the notebook or config. For local files, add a
-DuckDB/file connection and query CSV or Parquet data with functions such as
-`read_csv_auto('./file.csv')`.
+Edit `dql.config.json` or open the notebook connection panel. DQL uses a flat
+install by default: Databricks SQL works through the built-in HTTP connector,
+and DuckDB or Snowflake drivers can be installed project-locally when needed.
+Add your warehouse connection in the notebook or config.
 
 ```bash
 npm install
 npm run doctor
 ```
 
+Choose one database path before running queries:
+
+| Database | Extra install | Notes |
+| --- | --- | --- |
+| Databricks SQL | none | Built into DQL through HTTPS |
+| DuckDB or local CSV/Parquet/JSON files | `npm install --prefix .dql/connectors duckdb` | Needed for `duckdb` and `file` connections |
+| Snowflake | `npm install --prefix .dql/connectors snowflake-sdk` | Needed for Snowflake connections |
+
+The notebook Connections page can also install DuckDB or Snowflake into
+`.dql/connectors/`. For local files, add a DuckDB/file connection and query CSV
+or Parquet data with functions such as `read_csv_auto('./file.csv')`.
+See the [connector reference](https://github.com/duckcode-ai/dql/blob/main/docs/reference/connectors.md).
+
 ## Start the notebook
 
 ```bash

package/templates/starter/apps/.gitkeep

@@ -0,0 +1 @@
+

package/templates/starter/blocks/.gitkeep

@@ -0,0 +1 @@
+

package/templates/starter/business-views/.gitkeep

@@ -0,0 +1 @@
+

package/templates/starter/package.json

@@ -11,7 +11,7 @@
     "validate": "dql validate"
   },
   "devDependencies": {
-    "@duckcodeailabs/dql-cli": "^1.6.13"
+    "@duckcodeailabs/dql-cli": "^1.6.14"
   },
   "engines": {
     "node": ">=20 <23"

package/templates/starter/semantic-layer/.gitkeep

@@ -0,0 +1 @@
+

package/templates/starter/terms/.gitkeep

@@ -0,0 +1 @@
+