Run tests against fondant package with tox

.github/workflows/pipeline.yaml

@@ -22,26 +22,11 @@ jobs:
         run: |
           pip install --upgrade pip
           pip install poetry==1.4.0
-          poetry install --all-extras
-      - name: Test with pre-commit
-        run: poetry run pre-commit run --all-files --show-diff-on-failure
-      - name: Check licenses
-        run: |
-          poetry export -f requirements.txt --output /tmp/requirements.txt
-          poetry run liccheck -s license_strategy.ini -r /tmp/requirements.txt -l PARANOID
-      - name: Shadow-run pylint
-        # Check if pylint catches errors not caught by our pre-commit checks to validate switch
-        # to ruff
-        run: |
-          pip install pylint==2.16.4
-          pylint fondant --exit-zero
-      - name: Test with pytest
-        run: |
-          poetry run pytest tests --cov fondant --cov-report term-missing
+          poetry install --all-extras --with test
+      - name: Test with tox
+        run: poetry run tox
       - name: Coveralls
-        run: |
-          pip install "coveralls<4"
-          coveralls --service github
+        run: poetry run coveralls --service github
         env:
           COVERALLS_PARALLEL: true
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

components/caption_images/src/main.py

@@ -6,11 +6,10 @@
 import numpy as np
 import pandas as pd
 import torch
+from fondant.component import PandasTransformComponent
 from PIL import Image
 from transformers import BatchEncoding, BlipForConditionalGeneration, BlipProcessor
 
-from fondant.component import PandasTransformComponent
-
 logger = logging.getLogger(__name__)
 
 

components/download_images/src/main.py

@@ -11,9 +11,8 @@
 import urllib
 
 import dask.dataframe as dd
-from resizer import Resizer
-
 from fondant.component import DaskTransformComponent
+from resizer import Resizer
 
 logger = logging.getLogger(__name__)
 

components/embedding_based_laion_retrieval/src/main.py

@@ -7,7 +7,6 @@
 
 import pandas as pd
 from clip_client import ClipClient, Modality
-
 from fondant.component import PandasTransformComponent
 
 logger = logging.getLogger(__name__)

components/filter_comments/src/main.py

@@ -5,9 +5,8 @@
 import logging
 
 import dask.dataframe as dd
-from utils.text_extraction import get_comments_to_code_ratio
-
 from fondant.component import DaskTransformComponent
+from utils.text_extraction import get_comments_to_code_ratio
 
 logger = logging.getLogger(__name__)
 

components/filter_image_resolution/src/main.py

@@ -5,7 +5,6 @@
 
 import numpy as np
 import pandas as pd
-
 from fondant.component import PandasTransformComponent
 
 logger = logging.getLogger(__name__)

components/filter_line_length/src/main.py

@@ -2,7 +2,6 @@
 import logging
 
 import dask.dataframe as dd
-
 from fondant.component import DaskTransformComponent
 
 logger = logging.getLogger(__name__)

components/image_cropping/src/main.py

@@ -5,11 +5,10 @@
 
 import dask.dataframe as dd
 import numpy as np
+from fondant.component import DaskTransformComponent
 from image_crop import remove_borders
 from PIL import Image
 
-from fondant.component import DaskTransformComponent
-
 logger = logging.getLogger(__name__)
 
 

components/image_embedding/src/main.py

@@ -6,11 +6,10 @@
 import numpy as np
 import pandas as pd
 import torch
+from fondant.component import PandasTransformComponent
 from PIL import Image
 from transformers import CLIPProcessor, CLIPVisionModelWithProjection
 
-from fondant.component import PandasTransformComponent
-
 logger = logging.getLogger(__name__)
 
 

components/image_resolution_extraction/src/main.py

@@ -6,7 +6,6 @@
 import imagesize
 import numpy as np
 import pandas as pd
-
 from fondant.component import PandasTransformComponent
 
 logger = logging.getLogger(__name__)

components/language_filter/src/main.py

@@ -3,7 +3,6 @@
 
 import fasttext
 import pandas as pd
-
 from fondant.component import PandasTransformComponent
 
 logger = logging.getLogger(__name__)

components/language_filter/tests/language_filter_component_test.py

@@ -1,8 +1,8 @@
 """Unit test for language filter component."""
 import pandas as pd
+from fondant.component_spec import ComponentSpec
 
 from components.language_filter.src.main import LanguageFilterComponent
-from fondant.component_spec import ComponentSpec
 
 
 def test_run_component_test():

components/load_from_hf_hub/src/main.py

@@ -3,7 +3,6 @@
 import typing as t
 
 import dask.dataframe as dd
-
 from fondant.component import LoadComponent
 
 logger = logging.getLogger(__name__)

components/pii_redaction/src/main.py

@@ -4,11 +4,10 @@
 import logging
 
 import dask.dataframe as dd
+from fondant.component import DaskTransformComponent
 from pii_detection import scan_pii
 from pii_redaction import redact_pii
 
-from fondant.component import DaskTransformComponent
-
 logger = logging.getLogger(__name__)
 
 

components/prompt_based_laion_retrieval/src/main.py

@@ -6,7 +6,6 @@
 
 import pandas as pd
 from clip_client import ClipClient, Modality
-
 from fondant.component import PandasTransformComponent
 
 logger = logging.getLogger(__name__)

components/segment_images/src/main.py

@@ -6,12 +6,11 @@
 import numpy as np
 import pandas as pd
 import torch
+from fondant.component import PandasTransformComponent
 from palette import palette
 from PIL import Image
 from transformers import AutoModelForSemanticSegmentation, BatchFeature, SegformerImageProcessor
 
-from fondant.component import PandasTransformComponent
-
 logger = logging.getLogger(__name__)
 
 

components/text_normalization/src/main.py

@@ -5,7 +5,6 @@
 from typing import List
 
 import pandas as pd
-
 from fondant.component import PandasTransformComponent
 
 logger = logging.getLogger(__name__)

components/write_to_hf_hub/src/main.py

@@ -9,9 +9,8 @@
 # Define the schema for the struct using PyArrow
 import huggingface_hub
 from datasets.features.features import generate_from_arrow_type
-from PIL import Image
-
 from fondant.component import WriteComponent
+from PIL import Image
 
 logger = logging.getLogger(__name__)
 

pyproject.toml

@@ -55,10 +55,10 @@ pandas = { version = ">= 1.3.5", optional = true }
 pipelines = ["kfp", "kubernetes"]
 
 [tool.poetry.group.test.dependencies]
-liccheck = "^0.7.3"
 pre-commit = "^3.1.1"
-pytest = "^7.2.2"
-pytest-cov = "^4.1.0"
+tox = "^4.6.4"
+tox-gh-actions = "^3.1.3"
+coveralls = "^3.3.1"
 
 [tool.poetry.group.docs.dependencies]
 mkdocs-material = "^9.1.8"

scripts/pre-build.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# This script copies the components/ directory to fondant/components, replacing the symlink
+# This script copies the components/ directory to src/fondant/components, replacing the symlink
 # It should be run before building the fondant package'
 # This script makes changes to the local files, which should not be committed to git
 set -e
@@ -8,6 +8,6 @@ scripts_path=$( cd "$(dirname "${BASH_SOURCE[0]}")" ; pwd -P )
 root_path=$(dirname "$scripts_path")
 
 pushd "$root_path"
-rm fondant/components
-cp -r components/ fondant/
+rm -rf src/fondant/components
+cp -r components/ src/fondant/
 popd

tests/test_cli.py

@@ -3,7 +3,6 @@
 from unittest.mock import patch
 
 import pytest
-
 from fondant.cli import ImportFromStringError, compile, pipeline_from_string, run
 from fondant.pipeline import Pipeline
 

tests/test_compiler.py

@@ -3,7 +3,6 @@
 
 import pytest
 import yaml
-
 from fondant.compiler import DockerCompiler
 from fondant.pipeline import ComponentOp, Pipeline
 

tests/test_component.py

@@ -8,7 +8,6 @@
 import pandas as pd
 import pytest
 import yaml
-
 from fondant.component import (
     Component,
     DaskTransformComponent,

tests/test_component_specs.py

@@ -6,7 +6,6 @@
 
 import pytest
 import yaml
-
 from fondant.component_spec import ComponentSpec, ComponentSubset, KubeflowComponentSpec
 from fondant.exceptions import InvalidComponentSpec
 from fondant.schema import Type

tests/test_data_io.py

@@ -2,7 +2,6 @@
 
 import dask.dataframe as dd
 import pytest
-
 from fondant.component_spec import ComponentSpec
 from fondant.data_io import DaskDataLoader, DaskDataWriter
 from fondant.manifest import Manifest

tests/test_import_utils.py

@@ -3,7 +3,6 @@
 from unittest import mock
 
 import pytest
-
 from fondant.import_utils import (
     is_datasets_available,
     is_kfp_available,

tests/test_manifest.py

@@ -3,7 +3,6 @@
 from pathlib import Path
 
 import pytest
-
 from fondant.exceptions import InvalidManifest
 from fondant.manifest import Field, Index, Manifest, Subset, Type
 

tests/test_manifest_evolution.py

@@ -3,7 +3,6 @@
 
 import pytest
 import yaml
-
 from fondant.component_spec import ComponentSpec
 from fondant.manifest import Manifest
 

tests/test_pipeline.py

@@ -3,7 +3,6 @@
 
 import pytest
 import yaml
-
 from fondant.exceptions import InvalidPipelineDefinition
 from fondant.pipeline import ComponentOp, ComponentSpec, Pipeline
 

tests/test_schema.py

@@ -1,6 +1,5 @@
 import pyarrow as pa
 import pytest
-
 from fondant.exceptions import InvalidTypeSchema
 from fondant.schema import Type
 

tox.ini

@@ -0,0 +1,42 @@
+[tox]
+isolated_build = True
+envlist =
+    pre-commit
+    check-licenses
+    py{38,39,310,311}
+
+[gh-actions]
+python =
+    3.8: py38
+    3.9: py39
+    3.10: pre-commit,check-licenses, py310
+    3.11: py311
+
+[testenv:pre-commit]
+deps=pre-commit
+commands=pre-commit run --all-files --show-diff-on-failure
+
+[testenv:check-licenses]
+deps=
+    liccheck==0.7.3
+commands=
+    poetry lock
+    poetry install --all-extras
+    poetry show
+    poetry export -f requirements.txt --output /tmp/requirements.txt
+    poetry run liccheck -s license_strategy.ini -r /tmp/requirements.txt -l PARANOID
+
+[testenv]
+setenv=PYTHONPATH = {toxinidir}:{toxinidir}
+deps=
+    pytest==7.4.0
+    pytest-cov==4.1.0
+allowlist_externals=
+    poetry
+    /usr/bin/bash
+commands=
+    bash ./scripts/pre-build.sh
+    poetry lock
+    poetry install --all-extras
+    poetry show
+    poetry run python -m pytest tests --cov fondant --cov-report term-missing