Adding script for DOI updates

.github/workflows/run_tests.yaml

@@ -1,116 +1,69 @@
 name: Lint and Test
 
 on:
-    workflow_call:
+  workflow_call:
 
-jobs:
-    run_unit_tests:
-        runs-on: ubuntu-latest
-        if: 1 == 0
-        strategy:
-            fail-fast: false
-
-        steps:
-            - name: Repo checkout
-              uses: actions/checkout@main
-
-            - name: Set up Python 3.
-              uses: actions/setup-python@main
-              with:
-                python-version: 3.12
-
-            - name: Install uv
-              uses: astral-sh/setup-uv@v6
-
-            - name: Install dependencies
-              id: install_deps
-              run: uv sync
-
-            - name: run pytest -rP tests/
-              id: run_pytest
-              run: |
-                  uv run pytest --cov=credit_engine --cov-report=xml -rP tests/
-              continue-on-error: true
-
-            - name: Send to Codecov
-              id: send_to_codecov
-              uses: codecov/codecov-action@main
-              continue-on-error: true
-              with:
-                  files: ./coverage.xml
-                  fail_ci_if_error: true
-                  token: ${{ secrets.CODECOV_TOKEN }}
-
-            - name: outcome failure
-              if: steps.run_pytest.outcome != 'success' || steps.send_to_codecov.outcome != 'success'
-              run: |
-                  echo "Python tests: ${{ steps.run_pytest.outcome }}"
-                  echo "upload coverage: ${{ steps.send_to_codecov.outcome }}"
-                  exit 1
-
-            - name: outcome success
-              run: |
-                  echo "Python tests: ${{ steps.run_pytest.outcome }}"
-                  echo "upload coverage: ${{ steps.send_to_codecov.outcome }}"
-                  exit 0
-
-    run_linkml_tasks:
-        runs-on: ubuntu-latest
-        strategy:
-            fail-fast: false
+permissions:
+  contents: read
 
-        steps:
-            - name: Repo checkout
-              uses: actions/checkout@main
-
-            - name: Set up Python 3.
-              uses: actions/setup-python@main
-              with:
-                python-version: 3.12
-
-            - name: Install uv
-              uses: astral-sh/setup-uv@v6
-
-            - name: Install dependencies
-              id: install_deps
-              run: uv sync
-
-            - name: Lint linkml file
-              id: lint_linkml
-              run: |
-                  make lint-no-warn
-              continue-on-error: true
-
-            - name: Validate sample data against the schema
-              id: test_sample_data
-              run: |
-                  make test-sample-data
-              continue-on-error: true
-
-            - name: Validate sample data against JSONschema
-              id: test_sample_data_jsonschema
-              run: |
-                  make test-sample-data-jsonschema
-              continue-on-error: true
-
-            - name: Test documentation generation
-              id: test_docgen
-              run: |
-                  uv sync --group docs
-                  make gendoc
-              continue-on-error: true
-
-            - name: outcome failure
-              if: steps.lint_linkml.outcome != 'success' || steps.test_sample_data.outcome != 'success' || steps.test_sample_data_jsonschema.outcome != 'success' || steps.test_docgen.outcome != 'success'
-              run: |
-                  echo "linkml linting: ${{ steps.lint_linkml.outcome }}"
-                  echo "sample data validation: ${{ steps.test_sample_data.outcome }}"
-                  echo "sample data JSON Schema validation: ${{ steps.test_sample_data_jsonschema.outcome }}"
-                  echo "documentation generator: ${{ steps.test_docgen.outcome }}"
-                  exit 1
-
-            - name: outcome success
-              if: steps.lint_linkml.outcome == 'success' && steps.test_sample_data.outcome == 'success' && steps.test_sample_data_jsonschema.outcome == 'success' && steps.test_docgen.outcome == 'success'
-              run: |
-                  echo All tests passed successfully!
-                  exit 0
+jobs:
+  run_linkml_tasks:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+
+    steps:
+      - name: Repo checkout
+        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1
+
+      - name: Setup python
+        uses: actions/setup-python@6d4fd22b20465ccc0f7354476588e8df7c9ca0e0
+        with:
+          python-version: "3.14"
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e
+        with:
+          python-version: 3.14
+
+      - name: Install project
+        run: uv sync --frozen --group docs
+
+      - name: Lint linkml file
+        id: lint_linkml
+        run: |
+          make lint-no-warn
+        continue-on-error: true
+
+      - name: Validate sample data against the schema
+        id: test_sample_data
+        run: |
+          make test-sample-data
+        continue-on-error: true
+
+      - name: Validate sample data against JSONschema
+        id: test_sample_data_jsonschema
+        run: |
+          make test-sample-data-jsonschema
+        continue-on-error: true
+
+      - name: Test documentation generation
+        id: test_docgen
+        run: |
+          make gendoc
+        continue-on-error: true
+
+      - name: outcome failure
+        if: steps.lint_linkml.outcome != 'success' || steps.test_sample_data.outcome != 'success' || steps.test_sample_data_jsonschema.outcome != 'success' || steps.test_docgen.outcome != 'success'
+        run: |
+          echo "linkml linting: ${{ steps.lint_linkml.outcome }}"
+          echo "sample data validation: ${{ steps.test_sample_data.outcome }}"
+          echo "sample data JSON Schema validation: ${{ steps.test_sample_data_jsonschema.outcome }}"
+          echo "documentation generator: ${{ steps.test_docgen.outcome }}"
+          exit 1
+
+      - name: outcome success
+        if: steps.lint_linkml.outcome == 'success' && steps.test_sample_data.outcome == 'success' && steps.test_sample_data_jsonschema.outcome == 'success' && steps.test_docgen.outcome == 'success'
+        run: |
+          echo All tests passed successfully!
+          exit 0

.github/workflows/update_kbase_dois.yaml

@@ -0,0 +1,51 @@
+name: Update KBase DOI List
+
+on:
+    schedule:
+        # Runs weekly on Mondays at 08:00 UTC
+        - cron: "0 8 * * 1"
+    workflow_dispatch:
+
+permissions:
+    contents: write
+    pull-requests: write
+
+jobs:
+    update-kbase-dois:
+        runs-on: ubuntu-latest
+        name: Generate and update KBase DOI list
+        steps:
+            - name: Checkout repository
+              uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1
+              with:
+                  ref: main
+
+            - name: Setup python
+              uses: actions/setup-python@6d4fd22b20465ccc0f7354476588e8df7c9ca0e0
+              with:
+                  python-version: "3.14"
+
+            - name: Install uv
+              uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e
+              with:
+                  python-version: 3.14
+
+            - name: Install project
+              run: uv sync --frozen
+
+            - name: Generate KBase DOI list
+              run: uv run --frozen scripts/gen_kbase_doi_list.py
+
+            - name: Create Pull Request with updated KBase DOI list
+              uses: peter-evans/create-pull-request@5f6978faf089d4d20b00c7766989d076bb2fc7f1
+              with:
+                  token: ${{ secrets.GHCR_TOKEN || github.token }}
+                  add-paths: |
+                      src/docs/pages/kbase_dois.md
+                  base: main
+                  branch: update-kbase-dois
+                  delete-branch: true
+                  commit-message: "Update KBase FAIR Narratives DOI list"
+                  title: "Update KBase FAIR Narratives DOI list"
+                  body: |
+                      Automated weekly update of `src/docs/pages/kbase_dois.md` generated by `scripts/gen_kbase_doi_list.py`.

pyproject.toml

@@ -1,44 +1,57 @@
 [project]
 name = "credit-engine"
 version = "0.1.0"
-description = "Add your description here"
+description = "Standardised citation metadata for scientific datasets."
 readme = "README.md"
 requires-python = ">=3.12"
 dependencies = [
-    "check-jsonschema>=0.33.0",
-    "linkml>=1.9.1",
-]
-
-[dependency-groups]
-dev = [
-    "ruff>=0.11.0",
-]
-data_import = [
-    "asteval>=1.0.6",
-    "dlt[deltalake,duckdb,parquet]>=1.10.0",
-    "httpx>=0.28.1",
-    "ipykernel>=6.29.5",
-    "json-merge-patch>=0.3.0",
-    "jsonpatch>=1.33",
-    "pyyaml>=6.0.2",
-    "streamlit>=1.45.0",
-    "vcrpy>=7.0.0",
-    "xmlschema>=4.0.1",
-]
-erdantic = [
-    "erdantic>=1.1.0.post1",
-    "pygraphviz>=1.14",
-]
-docs = [
-    "mkdocs-material>=9.6.9",
-    "mkdocs-mermaid2-plugin>=1.2.1",
+    "check-jsonschema>=0.38.2",
+    "dlt>=1.31.0",
+    "jsonschema[format]>=4.26.0",
+    "linkml>=1.12.1",
+    "linkml-runtime>=1.12.1",
 ]
 
 [tool.ruff]
 line-length = 120
 
-# Assume Python 3.12.
-target-version = "py312"
+# Assume Python 3.14.
+target-version = "py314"
+
+# Allow autofix for all enabled rules (when `--fix`) is provided.
+fixable = ["ALL"]
+unfixable = []
+
+# Exclude a variety of commonly ignored directories.
+exclude = [
+    "__pypackages__",
+    "_build",
+    ".bzr",
+    ".direnv",
+    ".eggs",
+    ".git-rewrite",
+    ".git",
+    ".github",
+    ".hg",
+    ".mypy_cache",
+    ".nox",
+    ".pants.d",
+    ".pytype",
+    ".ruff_cache",
+    ".svn",
+    ".tox",
+    ".venv",
+    "*.pyc",
+    "buck-out",
+    "build",
+    "deps",
+    "dist",
+    "node_modules",
+    "other_schema",
+    "python-coverage",
+    "sample_data",
+    "venv",
+]
 
 [tool.ruff.lint]
 select = [
@@ -91,12 +104,12 @@ select = [
     "TD", # flake8-todos
     "FIX", # flake8-fixme
     "ERA", # eradicate
-    # "PD", # pandas-vet
+    "PD", # pandas-vet
     "PGH", # pygrep-hooks
     "PL", # Pylint
     "TRY", # tryceratops
     "FLY", # flynt
-    # "NPY", # NumPy-specific rules
+    "NPY", # NumPy-specific rules
     "AIR", # Airflow
     "PERF", # Perflint
     "FURB", # refurb
@@ -116,42 +129,23 @@ ignore = [
     "E203",
     "E501",
     "ISC001",
-    "S101",
 ]
 
-# Allow autofix for all enabled rules (when `--fix`) is provided.
-fixable = ["ALL"]
-unfixable = []
+[tool.ruff.lint.per-file-ignores]
+"tests/**/*.py" = ["S101"] # use of assert
 
+[tool.ruff.lint.mccabe]
+# Flag errors (`C901`) whenever the complexity level exceeds 15.
+max-complexity = 15
 
-# Exclude a variety of commonly ignored directories.
-exclude = [
-    "__pypackages__",
-    "_build",
-    ".bzr",
-    ".direnv",
-    ".eggs",
-    ".git-rewrite",
-    ".git",
-    ".github",
-    ".hg",
-    ".mypy_cache",
-    ".nox",
-    ".pants.d",
-    ".pytype",
-    ".ruff_cache",
-    ".svn",
-    ".tox",
-    ".venv",
-    "*.pyc",
-    "buck-out",
-    "build",
-    "deps",
-    "dist",
-    "node_modules",
-    "other_schema",
-    "python-coverage",
-    "sample_data",
-    "venv",
+[dependency-groups]
+dev = [
+    "ruff>=0.17.2",
+]
+docs = [
+    "mkdocs-material>=9.7.7",
+    "mkdocs-mermaid2-plugin>=1.2.3",
+]
+erd = [
+    "erdantic>=1.2.1",
 ]
-per-file-ignores = {}