Merge branch 'main' into 50395-2

* main:
  ci: add Docker build workflow for PR checks (airbytehq#385)
  fix: Update DEFAULT_ERROR_MAPPING for InvalidURL to RETRY (airbytehq#384)
  chore: use python base image 4.0.0 in SDM (airbytehq#382)
  ci: add check for 'do not merge' in PR titles (airbytehq#383)
  chore: remove embedded module (airbytehq#380)
  build: update pyarrow to ^19.0.0 (do not merge) (airbytehq#379)

Author: rpopov

.github/workflows/docker-build-check.yml

@@ -0,0 +1,44 @@
+name: Docker Build Check
+
+on:
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  docker-build-check:
+    name: SDM Docker Image Build  # Renamed job to be more descriptive
+    runs-on: ubuntu-24.04
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      # Build the Python package to create the wheel files needed for the Docker build
+      - name: Build Python Package
+        uses: hynek/build-and-inspect-python-package@v2
+        env:
+          POETRY_DYNAMIC_VERSIONING_BYPASS: "0.0.0dev0"
+
+      # Copy the wheel files to the dist directory
+      - name: Copy wheel files to dist directory
+        run: |
+          mkdir -p dist
+          cp /tmp/baipp/dist/*.whl dist/
+
+      - name: Set up QEMU for multi-platform builds
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Build Docker image for multiple platforms
+        id: docker-build
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          platforms: linux/amd64,linux/arm64
+          push: false
+          tags: airbyte/source-declarative-manifest:pr-${{ github.event.pull_request.number }}
+          outputs: type=image,name=target,annotation-index.org.opencontainers.image.description=SDM Docker image for PR ${{ github.event.pull_request.number }}

.github/workflows/semantic_pr_check.yml

@@ -39,3 +39,13 @@ jobs:
             Build
             tests
             Tests
+            
+      - name: Check for "do not merge" in PR title
+        if: ${{ github.event.pull_request.draft == false }}
+        uses: actions/github-script@v6
+        with:
+          script: |
+            const title = context.payload.pull_request.title.toLowerCase();
+            if (title.includes('do not merge') || title.includes('do-not-merge')) {
+              core.setFailed('PR title contains "do not merge" or "do-not-merge". Please remove this before merging.');
+            }

Dockerfile

@@ -5,7 +5,7 @@
 # A new version of source-declarative-manifest is built for every new Airbyte CDK release, and their versions are kept in sync.
 #
 
-FROM docker.io/airbyte/python-connector-base:3.0.0@sha256:1a0845ff2b30eafa793c6eee4e8f4283c2e52e1bbd44eed6cb9e9abd5d34d844
+FROM docker.io/airbyte/python-connector-base:4.0.0@sha256:d9894b6895923b379f3006fa251147806919c62b7d9021b5cd125bb67d7bbe22
 
 WORKDIR /airbyte/integration_code
 
@@ -24,8 +24,8 @@ RUN pip install dist/*.whl
 RUN mkdir -p source_declarative_manifest \
     && echo 'from source_declarative_manifest.run import run\n\nif __name__ == "__main__":\n    run()' > main.py \
     && touch source_declarative_manifest/__init__.py \
-    && cp /usr/local/lib/python3.10/site-packages/airbyte_cdk/cli/source_declarative_manifest/_run.py source_declarative_manifest/run.py \
-    && cp /usr/local/lib/python3.10/site-packages/airbyte_cdk/cli/source_declarative_manifest/spec.json source_declarative_manifest/
+    && cp /usr/local/lib/python3.11/site-packages/airbyte_cdk/cli/source_declarative_manifest/_run.py source_declarative_manifest/run.py \
+    && cp /usr/local/lib/python3.11/site-packages/airbyte_cdk/cli/source_declarative_manifest/spec.json source_declarative_manifest/
 
 # Remove unnecessary build files
 RUN rm -rf dist/ pyproject.toml poetry.lock README.md

airbyte_cdk/sources/embedded/__init__.py

@@ -19,9 +19,9 @@
         error_message="Invalid Protocol Schema: The endpoint that data is being requested from is using an invalid or insecure. Exception: requests.exceptions.InvalidSchema",
     ),
     InvalidURL: ErrorResolution(
-        response_action=ResponseAction.FAIL,
-        failure_type=FailureType.config_error,
-        error_message="Invalid URL specified: The endpoint that data is being requested from is not a valid URL. Exception: requests.exceptions.InvalidURL",
+        response_action=ResponseAction.RETRY,
+        failure_type=FailureType.transient_error,
+        error_message="Invalid URL specified or DNS error occurred: The endpoint that data is being requested from is not a valid URL. Exception: requests.exceptions.InvalidURL",
     ),
     RequestException: ErrorResolution(
         response_action=ResponseAction.RETRY,