From 12b9e16bb4f2054fe074945e3e68d8d2fd78f5e5 Mon Sep 17 00:00:00 2001
From: Marcin Wojtyczka <marcin.wojtyczka@databricks.com>
Date: Tue, 4 Feb 2025 15:14:05 +0100
Subject: [PATCH] disable integration tests on forks Require pyspark
 dependencies only for tests and cli

---
 .github/workflows/acceptance.yml | 8 ++++++--
 labs.yml                         | 1 +
 pyproject.toml                   | 5 ++++-
 3 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/acceptance.yml b/.github/workflows/acceptance.yml
index f40d6d57..66f529bc 100644
--- a/.github/workflows/acceptance.yml
+++ b/.github/workflows/acceptance.yml
@@ -20,7 +20,9 @@ concurrency:
 
 jobs:
   integration:
-    if: github.event_name == 'pull_request' && github.event.pull_request.draft == false
+    # Only run this job for PRs from branches on the main repository and not from forks.
+    # Workflows triggered by PRs from forks don't have access to the tool environment.
+    if: github.event_name == 'pull_request' && !github.event.pull_request.draft && !github.event.pull_request.head.repo.fork
     environment: tool
     runs-on: larger
     steps:
@@ -55,7 +57,9 @@ jobs:
           ARM_TENANT_ID: ${{ secrets.ARM_TENANT_ID }}
 
   serverless_integration:
-    if: github.event_name == 'pull_request' && github.event.pull_request.draft == false
+    # Only run this job for PRs from branches on the main repository and not from forks.
+    # Workflows triggered by PRs from forks don't have access to the tool environment.
+    if: github.event_name == 'pull_request' && !github.event.pull_request.draft && !github.event.pull_request.head.repo.fork
     environment: tool
     runs-on: larger
     env:
diff --git a/labs.yml b/labs.yml
index 24d6e2bb..a5419efd 100644
--- a/labs.yml
+++ b/labs.yml
@@ -3,6 +3,7 @@ name: dqx
 description: Data Quality Framework for PySpark Workloads
 install:
   script: src/databricks/labs/dqx/installer/install.py
+  extras: cli
 uninstall:
   script: src/databricks/labs/dqx/installer/uninstall.py
 entrypoint: src/databricks/labs/dqx/cli.py
diff --git a/pyproject.toml b/pyproject.toml
index 2df80299..cc04efce 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -30,9 +30,11 @@ classifiers = [
 dependencies = ["databricks-labs-blueprint>=0.9.1,<0.10",
                 "databricks-sdk~=0.30",
                 "databricks-labs-lsql>=0.5,<0.15",
-                "pyspark~=3.5.0" # required by the CLI to work correctly
 ]
 
+[project.optional-dependencies]
+cli = ["pyspark~=3.5.0"]
+
 [project.entry-points.databricks]
 runtime = "databricks.labs.dqx.runtime:main"
 
@@ -70,6 +72,7 @@ dependencies = [
     "types-PyYAML~=6.0.12",
     "types-requests~=2.31.0",
     "databricks-connect~=15.4",
+    "pyspark~=3.5.0",
 ]
 
 python="3.10"