doc id update

transforms/universal/doc_id/ray/Dockerfile

@@ -10,8 +10,11 @@ RUN cd data-processing-lib-python && pip install --no-cache-dir -e .
 COPY --chown=ray:users data-processing-lib-ray/ data-processing-lib-ray/ 
 RUN cd data-processing-lib-ray    && pip install --no-cache-dir -e .
 
-COPY requirements.txt requirements.txt
-RUN pip install --no-cache-dir -r  requirements.txt
+# Install ray project source
+COPY --chown=ray:users src/ src/
+COPY --chown=ray:users pyproject.toml pyproject.toml
+COPY --chown=ray:users Readme.md Readme.md
+RUN pip install --no-cache-dir -e .
 
 # copy source data
 COPY ./src/doc_id_transform_ray.py .

transforms/universal/doc_id/ray/Makefile

@@ -30,20 +30,15 @@ publish:: publish-dist publish-image
 
 publish-image:: .transforms.publish-image-ray
 
-set-versions:
-	$(MAKE) TRANSFORM_PYTHON_VERSION=not-used  .transforms.set-versions
-
-build-dist:: set-versions .defaults.build-dist 
+setup:: .transforms.setup
 
-publish-dist:: .defaults.publish-dist
+# set the version of python transform that this depends on.
+set-versions:
+	$(MAKE) TRANSFORM_PYTHON_VERSION=${CODE2PARQUET_PYTHON_VERSION} TOML_VERSION=$(CODE2PARQUET_PYTHON_VERSION) .transforms.set-versions
 
-run-cli-sample: 
-	$(MAKE) RUN_FILE=$(TRANSFORM_NAME)_transform_ray.py \
-                RUN_ARGS="--run_locally True --data_local_config \"{ 'input_folder' : '../test-data/input', 'output_folder' : '../output'}\"  \
-                --doc_id_int True "	\
-                .transforms.run-src-file
+build-dist:: set-versions .defaults.build-dist
 
-run-local-python-only-sample: .transforms.run-local-sample
+publish-dist:: .defaults.publish-dist
 
 run-local-sample: .transforms.run-local-ray-sample
 
@@ -52,3 +47,10 @@ run-s3-sample: .transforms.run-s3-ray-sample
 minio-start:	.minio-start
 
 load-image:: .transforms.load-image
+
+run-cli-sample: 
+	$(MAKE) RUN_FILE=$(TRANSFORM_NAME)_transform_ray.py \
+                RUN_ARGS="--run_locally True --data_local_config \"{ 'input_folder' : '../test-data/input', 'output_folder' : '../output'}\"  \
+                --doc_id_int True "	\
+                .transforms.run-src-file
+

transforms/universal/doc_id/ray/pyproject.toml

@@ -0,0 +1,45 @@
+[project]
+name = "dpk_docid_transform_ray"
+version = "0.4.0.dev6"
+requires-python = ">=3.10"
+description = "docid Ray Transform"
+license = {text = "Apache-2.0"}
+readme = {file = "README.md", content-type = "text/markdown"}
+authors = [
+    { name = "David Wood", email = "dawood@us.ibm.com" },
+    { name = "Boris Lublinsky", email = "blublinsky@ibm.com" },
+]
+dependencies = [
+    "data-prep-toolkit-ray==0.2.0.dev6",
+]
+
+[build-system]
+requires = ["setuptools>=68.0.0", "wheel", "setuptools_scm[toml]>=7.1.0"]
+build-backend = "setuptools.build_meta"
+
+[project.optional-dependencies]
+dev = [
+    "twine",
+    "pytest>=7.3.2",
+    "pytest-dotenv>=0.5.2",
+    "pytest-env>=1.0.0",
+    "pre-commit>=3.3.2",
+    "pytest-cov>=4.1.0",
+    "pytest-mock>=3.10.0",
+    "moto==5.0.5",
+    "markupsafe==2.0.1",
+]
+
+[options]
+package_dir = ["src","test"]
+
+[options.packages.find]
+where = ["src/"]
+
+[tool.pytest.ini_options]
+# Currently we use low coverage since we have to run tests separately (see makefile)
+#addopts = "--cov --cov-report term-missing --cov-fail-under 25"
+markers = ["unit: unit tests", "integration: integration tests"]
+
+[tool.coverage.run]
+include = ["src/*"]

transforms/universal/ededup/ray/pyproject.toml

@@ -2,7 +2,7 @@
 name = "dpk_ededup_transform_ray"
 version = "0.4.0.dev6"
 requires-python = ">=3.10"
-description = "code2parquet Ray Transform"
+description = "ededup Ray Transform"
 license = {text = "Apache-2.0"}
 readme = {file = "README.md", content-type = "text/markdown"}
 authors = [

transforms/universal/fdedup/ray/pyproject.toml

@@ -2,7 +2,7 @@
 name = "dpk_fdedup_transform_ray"
 version = "0.4.0.dev6"
 requires-python = ">=3.10"
-description = "code2parquet Ray Transform"
+description = "fdedup Ray Transform"
 license = {text = "Apache-2.0"}
 readme = {file = "README.md", content-type = "text/markdown"}
 authors = [