feat(sdk): enable dependency-free runtime install of kfp

sdk/RELEASE.md

@@ -1,18 +1,8 @@
 # Current Version (in development)
 
-## Features
-
-## Breaking changes
-
-## Deprecations
-
-## Bug fixes and other changes
-
-## Documentation updates
-# 2.1.2
 
 ## Features
-* Create "dependency-free" runtime package (only `typing_extensions` required) for Lightweight Python Components to reduce runtime dependency resolution errors [\#9710](https://github.com/kubeflow/pipelines/pull/9710), [\#9738](https://github.com/kubeflow/pipelines/pull/9738)
+* Create "dependency-free" runtime package (only `typing_extensions` required) for Lightweight Python Components to reduce runtime dependency resolution errors [\#9710](https://github.com/kubeflow/pipelines/pull/9710), [\#9886](https://github.com/kubeflow/pipelines/pull/9886)
 
 ## Breaking changes
 

sdk/python/kfp/__init__.py

@@ -20,6 +20,12 @@
 
 TYPE_CHECK = True
 
-from kfp import components
-from kfp import dsl
-from kfp.client import Client
+import os
+
+# compile-time only dependencies
+if os.environ.get('_KFP_RUNTIME', 'false') != 'true':
+    # make `from kfp import components` and `from kfp import dsl` valid;
+    # related to namespace packaging issue
+    from kfp import components  # noqa: keep unused import
+    from kfp import dsl  # noqa: keep unused import
+    from kfp.client import Client  # noqa: keep unused import

sdk/python/kfp/cli/component.py

@@ -39,8 +39,6 @@
 
 _DOCKERFILE = 'Dockerfile'
 
-# TODO: merge kfp_package_path into runtime-requirements.txt, once we have
-# kfp_runtime package that is dependency-free.
 _DOCKERFILE_TEMPLATE = '''
 FROM {base_image}
 

sdk/python/kfp/dsl/__init__.py

@@ -14,31 +14,13 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+# runtime dependencies
 __all__ = [
-    'component',
-    'container_component',
-    'pipeline',
-    'importer',
-    'ContainerSpec',
-    'Condition',
-    'ExitHandler',
-    'ParallelFor',
-    'Collected',
     'Input',
     'Output',
     'InputPath',
     'OutputPath',
-    'IfPresentPlaceholder',
-    'ConcatPlaceholder',
     'PipelineTaskFinalStatus',
-    'PIPELINE_JOB_NAME_PLACEHOLDER',
-    'PIPELINE_JOB_RESOURCE_NAME_PLACEHOLDER',
-    'PIPELINE_JOB_ID_PLACEHOLDER',
-    'PIPELINE_TASK_NAME_PLACEHOLDER',
-    'PIPELINE_TASK_ID_PLACEHOLDER',
-    'PIPELINE_ROOT_PLACEHOLDER',
-    'PIPELINE_JOB_CREATE_TIME_UTC_PLACEHOLDER',
-    'PIPELINE_JOB_SCHEDULE_TIME_UTC_PLACEHOLDER',
     'Artifact',
     'ClassificationMetrics',
     'Dataset',
@@ -47,29 +29,18 @@
     'Metrics',
     'Model',
     'SlicedClassificationMetrics',
-    'PipelineTask',
+    'PIPELINE_JOB_NAME_PLACEHOLDER',
+    'PIPELINE_JOB_RESOURCE_NAME_PLACEHOLDER',
+    'PIPELINE_JOB_ID_PLACEHOLDER',
+    'PIPELINE_TASK_NAME_PLACEHOLDER',
+    'PIPELINE_TASK_ID_PLACEHOLDER',
+    'PIPELINE_ROOT_PLACEHOLDER',
+    'PIPELINE_JOB_CREATE_TIME_UTC_PLACEHOLDER',
+    'PIPELINE_JOB_SCHEDULE_TIME_UTC_PLACEHOLDER',
 ]
+import os
 
-try:
-    from typing import Annotated
-except ImportError:
-    from typing_extensions import Annotated
-
-from typing import TypeVar
-
-from kfp.dsl.component_decorator import component
-from kfp.dsl.container_component_decorator import container_component
-from kfp.dsl.for_loop import Collected
-from kfp.dsl.importer_node import importer
-from kfp.dsl.pipeline_context import pipeline
-from kfp.dsl.pipeline_task import PipelineTask
-from kfp.dsl.placeholders import ConcatPlaceholder
-from kfp.dsl.placeholders import IfPresentPlaceholder
-from kfp.dsl.structures import ContainerSpec
 from kfp.dsl.task_final_status import PipelineTaskFinalStatus
-from kfp.dsl.tasks_group import Condition
-from kfp.dsl.tasks_group import ExitHandler
-from kfp.dsl.tasks_group import ParallelFor
 from kfp.dsl.types.artifact_types import Artifact
 from kfp.dsl.types.artifact_types import ClassificationMetrics
 from kfp.dsl.types.artifact_types import Dataset
@@ -83,8 +54,14 @@
 from kfp.dsl.types.type_annotations import OutputAnnotation
 from kfp.dsl.types.type_annotations import OutputPath
 
-# hack: constants and custom type generics have to be defined here to be captured by autodoc and autodocsumm used in ./docs/conf.py
+try:
+    from typing import Annotated
+except ImportError:
+    from typing_extensions import Annotated
 
+from typing import TypeVar
+
+# hack: constants and custom type generics have to be defined here to be captured by autodoc and autodocsumm used in ./docs/conf.py
 PIPELINE_JOB_NAME_PLACEHOLDER = '{{$.pipeline_job_name}}'
 """A placeholder used to obtain a pipeline job name within a task at pipeline runtime.
 
@@ -247,3 +224,32 @@ def my_pipeline():
         producer_task = artifact_producer()
         artifact_consumer(model=producer_task.output)
 """
+
+# compile-time only dependencies
+if os.environ.get('_KFP_RUNTIME', 'false') != 'true':
+    from kfp.dsl.component_decorator import component
+    from kfp.dsl.container_component_decorator import container_component
+    from kfp.dsl.for_loop import Collected
+    from kfp.dsl.importer_node import importer
+    from kfp.dsl.pipeline_context import pipeline
+    from kfp.dsl.pipeline_task import PipelineTask
+    from kfp.dsl.placeholders import ConcatPlaceholder
+    from kfp.dsl.placeholders import IfPresentPlaceholder
+    from kfp.dsl.structures import ContainerSpec
+    from kfp.dsl.tasks_group import Condition
+    from kfp.dsl.tasks_group import ExitHandler
+    from kfp.dsl.tasks_group import ParallelFor
+    __all__.extend([
+        'component',
+        'container_component',
+        'pipeline',
+        'importer',
+        'ContainerSpec',
+        'Condition',
+        'ExitHandler',
+        'ParallelFor',
+        'Collected',
+        'IfPresentPlaceholder',
+        'ConcatPlaceholder',
+        'PipelineTask',
+    ])

sdk/python/kfp/dsl/component_factory.py

@@ -21,6 +21,7 @@
 import warnings
 
 import docstring_parser
+import kfp
 from kfp.dsl import container_component_artifact_channel
 from kfp.dsl import container_component_class
 from kfp.dsl import graph_component
@@ -109,24 +110,43 @@ def make_index_url_options(pip_index_urls: Optional[List[str]]) -> str:
 
 
 def _get_packages_to_install_command(
-        package_list: Optional[List[str]] = None,
-        pip_index_urls: Optional[List[str]] = None) -> List[str]:
+    kfp_package_path: Optional[str] = None,
+    pip_index_urls: Optional[List[str]] = None,
+    packages_to_install: Optional[List[str]] = None,
+    install_kfp_package: bool = True,
+    target_image: Optional[str] = None,
+) -> List[str]:
+    packages_to_install = packages_to_install or []
+    kfp_in_user_pkgs = any(pkg.startswith('kfp') for pkg in packages_to_install)
+    # if the user doesn't say "don't install", they aren't building a
+    # container component, and they haven't already specified a KFP dep
+    # themselves, we install KFP for them
+    inject_kfp_install = install_kfp_package and target_image is None and not kfp_in_user_pkgs
+    if inject_kfp_install:
+        if kfp_package_path:
+            packages_to_install.append(kfp_package_path)
+        else:
+            packages_to_install.extend(_get_injected_kfp_imports())
+
+    if packages_to_install:
+        concat_package_list = ' '.join(
+            [repr(str(package)) for package in packages_to_install])
+        index_url_options = make_index_url_options(pip_index_urls)
 
-    if not package_list:
-        return []
+        install_python_packages_script = _install_python_packages_script_template.format(
+            index_url_options=index_url_options,
+            concat_package_list=concat_package_list)
+        return ['sh', '-c', install_python_packages_script]
 
-    concat_package_list = ' '.join(
-        [repr(str(package)) for package in package_list])
-    index_url_options = make_index_url_options(pip_index_urls)
-    install_python_packages_script = _install_python_packages_script_template.format(
-        index_url_options=index_url_options,
-        concat_package_list=concat_package_list)
-    return ['sh', '-c', install_python_packages_script]
+    return []
 
 
-def _get_default_kfp_package_path() -> str:
-    import kfp
-    return f'kfp=={kfp.__version__}'
+def _get_injected_kfp_imports() -> List[str]:
+    return [
+        f'kfp=={kfp.__version__}',
+        '--no-deps',
+        'typing-extensions>=3.7.4,<5; python_version<"3.9"',
+    ]
 
 
 def _get_function_source_definition(func: Callable) -> str:
@@ -420,8 +440,9 @@ def _get_command_and_args_for_lightweight_component(
         '-ec',
         textwrap.dedent('''\
                     program_path=$(mktemp -d)
+
                     printf "%s" "$0" > "$program_path/ephemeral_component.py"
-                    python3 -m kfp.dsl.executor_main \
+                    _KFP_RUNTIME=true python3 -m kfp.dsl.executor_main \
                         --component_module_path \
                         "$program_path/ephemeral_component.py" \
                         "$@"
@@ -471,15 +492,14 @@ def create_component_from_func(
     The decorator is defined under component_decorator.py. See the
     decorator for the canonical documentation for this function.
     """
-    packages_to_install = packages_to_install or []
-
-    if install_kfp_package and target_image is None:
-        if kfp_package_path is None:
-            kfp_package_path = _get_default_kfp_package_path()
-        packages_to_install.append(kfp_package_path)
 
     packages_to_install_command = _get_packages_to_install_command(
-        package_list=packages_to_install, pip_index_urls=pip_index_urls)
+        install_kfp_package=install_kfp_package,
+        target_image=target_image,
+        kfp_package_path=kfp_package_path,
+        packages_to_install=packages_to_install,
+        pip_index_urls=pip_index_urls,
+    )
 
     command = []
     args = []

sdk/python/kfp/dsl/component_factory_test.py

@@ -28,31 +28,112 @@
 
 class TestGetPackagesToInstallCommand(unittest.TestCase):
 
-    def test_with_no_packages_to_install(self):
+    def test_with_no_user_packages_to_install(self):
         packages_to_install = []
 
         command = component_factory._get_packages_to_install_command(
-            packages_to_install)
+            packages_to_install=packages_to_install)
+
+        self.assertEqual(command, [
+            'sh', '-c',
+            '\nif ! [ -x "$(command -v pip)" ]; then\n    python3 -m ensurepip || python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1 python3 -m pip install --quiet     --no-warn-script-location \'kfp==2.1.2\' \'--no-deps\' \'typing-extensions>=3.7.4,<5; python_version<"3.9"\' && "$0" "$@"\n'
+        ])
+
+    def test_with_no_user_packages_to_install_and_install_kfp_false(self):
+        packages_to_install = []
+
+        command = component_factory._get_packages_to_install_command(
+            packages_to_install=packages_to_install,
+            install_kfp_package=False,
+        )
+        self.assertEqual(command, [])
+
+    def test_with_no_user_packages_to_install_and_kfp_package_path(self):
+        packages_to_install = []
+
+        command = component_factory._get_packages_to_install_command(
+            packages_to_install=packages_to_install,
+            kfp_package_path='git+https://github.com/kubeflow/pipelines.git@master#subdirectory=sdk/python'
+        )
+
+        self.assertEqual(command, [
+            'sh', '-c',
+            '\nif ! [ -x "$(command -v pip)" ]; then\n    python3 -m ensurepip || python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1 python3 -m pip install --quiet     --no-warn-script-location \'git+https://github.com/kubeflow/pipelines.git@master#subdirectory=sdk/python\' && "$0" "$@"\n'
+        ])
+
+    def test_with_no_user_packages_to_install_and_kfp_package_path_and_install_kfp_false(
+            self):
+        packages_to_install = []
+
+        command = component_factory._get_packages_to_install_command(
+            packages_to_install=packages_to_install,
+            kfp_package_path='git+https://github.com/kubeflow/pipelines.git@master#subdirectory=sdk/python',
+            install_kfp_package=False,
+        )
         self.assertEqual(command, [])
 
-    def test_with_packages_to_install_and_no_pip_index_url(self):
+    def test_with_user_packages_to_install_and_kfp_package_path_and_install_kfp_false(
+            self):
+        packages_to_install = ['sklearn']
+
+        command = component_factory._get_packages_to_install_command(
+            packages_to_install=packages_to_install,
+            kfp_package_path='git+https://github.com/kubeflow/pipelines.git@master#subdirectory=sdk/python',
+            install_kfp_package=False,
+        )
+
+        self.assertEqual(command, [
+            'sh', '-c',
+            '\nif ! [ -x "$(command -v pip)" ]; then\n    python3 -m ensurepip || python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1 python3 -m pip install --quiet     --no-warn-script-location \'sklearn\' && "$0" "$@"\n'
+        ])
+
+    def test_with_no_user_packages_to_install_and_kfp_package_path_and_target_image(
+            self):
+        packages_to_install = []
+
+        command = component_factory._get_packages_to_install_command(
+            packages_to_install=packages_to_install,
+            target_image='gcr.io/my-kfp-image',
+            kfp_package_path='./sdk/python')
+
+        self.assertEqual(command, [])
+
+    def test_with_no_user_packages_to_install_and_kfp_package_path_and_target_image_and_install_kfp_false(
+            self):
+        packages_to_install = []
+
+        command = component_factory._get_packages_to_install_command(
+            packages_to_install=packages_to_install,
+            target_image='gcr.io/my-kfp-image',
+            kfp_package_path='./sdk/python',
+            install_kfp_package=False)
+
+        self.assertEqual(command, [])
+
+    def test_with_user_packages_to_install_and_no_pip_index_url(self):
         packages_to_install = ['package1', 'package2']
 
         command = component_factory._get_packages_to_install_command(
-            packages_to_install)
-        concat_command = ' '.join(command)
-        for package in packages_to_install:
-            self.assertTrue(package in concat_command)
+            packages_to_install=packages_to_install)
+
+        self.assertEqual(command, [
+            'sh', '-c',
+            '\nif ! [ -x "$(command -v pip)" ]; then\n    python3 -m ensurepip || python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1 python3 -m pip install --quiet     --no-warn-script-location \'package1\' \'package2\' \'kfp==2.1.2\' \'--no-deps\' \'typing-extensions>=3.7.4,<5; python_version<"3.9"\' && "$0" "$@"\n'
+        ])
 
     def test_with_packages_to_install_with_pip_index_url(self):
         packages_to_install = ['package1', 'package2']
         pip_index_urls = ['https://myurl.org/simple']
 
         command = component_factory._get_packages_to_install_command(
-            packages_to_install, pip_index_urls)
-        concat_command = ' '.join(command)
-        for package in packages_to_install + pip_index_urls:
-            self.assertTrue(package in concat_command)
+            packages_to_install=packages_to_install,
+            pip_index_urls=pip_index_urls,
+        )
+
+        self.assertEqual(command, [
+            'sh', '-c',
+            '\nif ! [ -x "$(command -v pip)" ]; then\n    python3 -m ensurepip || python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1 python3 -m pip install --quiet     --no-warn-script-location --index-url https://myurl.org/simple --trusted-host https://myurl.org/simple \'package1\' \'package2\' \'kfp==2.1.2\' \'--no-deps\' \'typing-extensions>=3.7.4,<5; python_version<"3.9"\' && "$0" "$@"\n'
+        ])
 
 
 class TestInvalidParameterName(unittest.TestCase):