Merge branch 'main' into templates-workflows

.pre-commit-config.yaml

@@ -8,7 +8,7 @@ repos:
   - id: trailing-whitespace
   - id: end-of-file-fixer
 - repo: https://github.com/astral-sh/ruff-pre-commit
-  rev: v0.2.0
+  rev: v0.5.0  # ruff version
   hooks:
     - id: ruff-format
     - id: ruff

pyproject.toml

@@ -30,7 +30,7 @@ dependencies = [
 
 [project.optional-dependencies]
 test = ["pytest>=6.0", "pytest-cov", "requests-mock"]
-dev = ["pre-commit", "ruff==0.2.0"]
+dev = ["pre-commit", "ruff==0.5.0"]
 docs = [
     "sphinx==6.1.3",
     "sphinx-autodoc-typehints==1.22.0",
@@ -84,16 +84,22 @@ select = [
     "DTZ",  # https://docs.astral.sh/ruff/rules/#flake8-datetimez-dtz
     "T10",  # https://docs.astral.sh/ruff/rules/#flake8-datetimez-dtz
     "EM",  # https://docs.astral.sh/ruff/rules/#flake8-errmsg-em
+    "LOG",  # https://docs.astral.sh/ruff/rules/#flake8-logging-log
     "G",  # https://docs.astral.sh/ruff/rules/#flake8-logging-format-g
+    "INP",  # https://docs.astral.sh/ruff/rules/#flake8-no-pep420-inp
     "PIE",  # https://docs.astral.sh/ruff/rules/#flake8-pie-pie
     "T20",  # https://docs.astral.sh/ruff/rules/#flake8-print-t20
     "PT",  # https://docs.astral.sh/ruff/rules/#flake8-pytest-style-pt
     "Q",  # https://docs.astral.sh/ruff/rules/#flake8-quotes-q
     "RSE",  # https://docs.astral.sh/ruff/rules/#flake8-raise-rse
     "RET",  # https://docs.astral.sh/ruff/rules/#flake8-return-ret
+    "SLF",  # https://docs.astral.sh/ruff/rules/#flake8-self-slf
     "SIM",  # https://docs.astral.sh/ruff/rules/#flake8-simplify-sim
+    "ARG",  # https://docs.astral.sh/ruff/rules/#flake8-unused-arguments-arg
     "PTH",  # https://docs.astral.sh/ruff/rules/#flake8-use-pathlib-pth
     "PGH",  # https://docs.astral.sh/ruff/rules/#pygrep-hooks-pgh
+    "PERF",  # https://docs.astral.sh/ruff/rules/#perflint-perf
+    "FURB",  # https://docs.astral.sh/ruff/rules/#refurb-furb
     "RUF",  # https://docs.astral.sh/ruff/rules/#ruff-specific-rules-ruf
 ]
 fixable = [
@@ -104,16 +110,20 @@ fixable = [
     "ANN",
     "B",
     "C4",
+    "LOG",
     "G",
     "PIE",
     "PT",
     "RSE",
     "SIM",
+    "PERF",
+    "FURB",
     "RUF"
 ]
 
-# ANN101 - missing-type-self
 # ANN003 - missing-type-kwargs
+# ANN101 - missing-type-self
+# ANN102 - missing-type-cls
 # D203 - one-blank-line-before-class
 # D205 - blank-line-after-summary
 # D206 - indent-with-spaces*
@@ -129,7 +139,7 @@ fixable = [
 # S321 - suspicious-ftp-lib-usage
 # *ignored for compatibility with formatter
 ignore = [
-    "ANN101", "ANN003",
+    "ANN003", "ANN101", "ANN102",
     "D203", "D205", "D206", "D213", "D300", "D400", "D415",
     "E111", "E114", "E117", "E501",
     "W191",
@@ -139,15 +149,20 @@ ignore = [
 [tool.ruff.lint.per-file-ignores]
 # ANN001 - missing-type-function-argument
 # ANN2 - missing-return-type
-# ANN201 - Missing type annotation
-# ANN102 - missing-type-cls
-# N805 - invalid-first-argument-name-for-method
-# PT018 - pytest-composite-assertion
 # S101 - assert
-"tests/*" = ["ANN001", "ANN102", "ANN2", "S101"]
+# B011 - assert-false
+# INP001 - implicit-namespace-package
+# PT018 - pytest-composite-assertion
+# ARG001 - unused-function-argument
+# ARG005 - unused-lambda-argument
+# T201 - print
+"tests/*" = ["ANN001", "ANN2", "S101", "B011", "INP001"]
 "tests/test_hemonc.py" = ["PT018"]
 "tests/test_gtop.py" = ["PT018"]
 "tests/test_do.py" = ["PT018"]
-"*__init__.py" = ["F401"]
-"src/wags_tails/base_source.py" = ["ANN102"]
+"tests/test_custom.py" = ["ARG001", "ARG005"]
 "src/wags_tails/utils/downloads.py" = ["T201"]
+
+
+[tool.ruff.format]
+docstring-code-format = true

src/wags_tails/__init__.py

@@ -1,4 +1,5 @@
 """Data acquisition tools for Wagnerds."""
+
 from .base_source import DataSource, RemoteDataError
 from .chembl import ChemblData
 from .chemidplus import ChemIDplusData

src/wags_tails/base_source.py

@@ -3,11 +3,12 @@
 All source classes should inherit - directly or indirectly - from ``DataSource``. Each
 class defined here is an ``abstract base class`` and cannot be instantiated directly.
 """
+
 import abc
 import datetime
 import logging
+from collections.abc import Generator
 from pathlib import Path
-from typing import Generator, Optional, Tuple
 
 import requests
 
@@ -33,7 +34,7 @@ class DataSource(abc.ABC):
     _filetype: str
     _versioned: bool = True
 
-    def __init__(self, data_dir: Optional[Path] = None, silent: bool = True) -> None:
+    def __init__(self, data_dir: Path | None = None, silent: bool = True) -> None:
         """Set common class parameters.
 
         :param data_dir: direct location to store data files in, if specified. See
@@ -71,7 +72,7 @@ def _download_data(self, version: str, outfile: Path) -> None:
 
     def get_latest(
         self, from_local: bool = False, force_refresh: bool = False
-    ) -> Tuple[Path, str]:
+    ) -> tuple[Path, str]:
         """Get path to latest version of data.
 
         Provides logic for both versioned and unversioned data here, rather than in the

src/wags_tails/chembl.py

@@ -1,4 +1,5 @@
 """Provide source fetching for ChEMBL."""
+
 import fnmatch
 import re
 import tarfile

src/wags_tails/chemidplus.py

@@ -1,4 +1,5 @@
 """Provide source fetching for ChemIDplus."""
+
 import datetime
 import re
 from pathlib import Path
@@ -38,7 +39,7 @@ def _get_latest_version() -> str:
             .strftime(DATE_VERSION_PATTERN)
         )
 
-    def _download_data(self, version: str, outfile: Path) -> None:
+    def _download_data(self, version: str, outfile: Path) -> None:  # noqa: ARG002
         """Download data file to specified location. ChemIDplus data is no longer
         updated, so versioning is irrelevant.
 

src/wags_tails/custom.py

@@ -9,8 +9,9 @@
 The :ref:`documentation <custom_data_source>` provides more explanation and an in-depth
 example.
 """
+
+from collections.abc import Callable
 from pathlib import Path
-from typing import Callable, Optional
 
 from .base_source import DataSource
 
@@ -24,8 +25,8 @@ def __init__(
         filetype: str,
         latest_version_cb: Callable[[], str],
         download_cb: Callable[[str, Path], None],
-        data_dir: Optional[Path] = None,
-        file_name: Optional[str] = None,
+        data_dir: Path | None = None,
+        file_name: str | None = None,
         versioned: bool = True,
         silent: bool = False,
     ) -> None:

src/wags_tails/do.py

@@ -1,4 +1,5 @@
 """Provide source fetching for Human Disease Ontology."""
+
 import datetime
 import tarfile
 from pathlib import Path

src/wags_tails/drugbank.py

@@ -1,7 +1,7 @@
 """Provide source fetching for DrugBank."""
+
 import logging
 from pathlib import Path
-from typing import Tuple
 
 import requests
 
@@ -19,7 +19,7 @@ class DrugBankData(DataSource):
     _filetype = "csv"
 
     @staticmethod
-    def _get_latest_version() -> Tuple[str, str]:
+    def _get_latest_version() -> tuple[str, str]:
         """Retrieve latest version value
 
         :return: latest release value and base download URL
@@ -35,7 +35,7 @@ def _get_latest_version() -> Tuple[str, str]:
             msg = "Unable to parse latest DrugBank version number from releases API endpoint"
             raise RemoteDataError(msg) from e
 
-    def _get_latest_local_file(self, glob: str) -> Tuple[Path, str]:
+    def _get_latest_local_file(self, glob: str) -> tuple[Path, str]:
         """Get most recent locally-available file. DrugBank uses versioning that isn't
         easily sortable by default so we have to use some extra magic.
 
@@ -72,7 +72,7 @@ def _download_data(self, url: str, outfile: Path) -> None:
 
     def get_latest(
         self, from_local: bool = False, force_refresh: bool = False
-    ) -> Tuple[Path, str]:
+    ) -> tuple[Path, str]:
         """Get path to latest version of data, and its version value
 
         :param from_local: if True, use latest available local file

src/wags_tails/drugsatfda.py

@@ -1,4 +1,5 @@
 """Provide source fetching for Drugs@FDA."""
+
 import datetime
 from pathlib import Path
 
@@ -38,7 +39,7 @@ def _get_latest_version() -> str:
             .strftime(DATE_VERSION_PATTERN)
         )
 
-    def _download_data(self, version: str, outfile: Path) -> None:
+    def _download_data(self, version: str, outfile: Path) -> None:  # noqa: ARG002
         """Download latest data file to specified location.
 
         :param version: version to acquire

src/wags_tails/ensembl.py

@@ -1,4 +1,5 @@
 """Provide data management for Ensembl genomic data."""
+
 from pathlib import Path
 
 import requests

src/wags_tails/ensembl_transcript_mappings.py

@@ -14,7 +14,7 @@ class EnsemblTranscriptMappingData(UnversionedDataSource):
     _src_name = "ensembl_transcript_mappings"
     _filetype = "tsv"
 
-    def _download_data(self, version: str, outfile: Path) -> None:
+    def _download_data(self, version: str, outfile: Path) -> None:  # noqa: ARG002
         """Download data file to specified location.
 
         :param version: version to acquire

src/wags_tails/guide_to_pharmacology.py

@@ -1,8 +1,9 @@
 """Provide source fetching for Guide To Pharmacology."""
+
 import logging
 import re
 from pathlib import Path
-from typing import NamedTuple, Tuple
+from typing import NamedTuple
 
 import requests
 
@@ -67,7 +68,7 @@ def _download_data(self, file_paths: GtoPLigandPaths) -> None:
 
     def get_latest(
         self, from_local: bool = False, force_refresh: bool = False
-    ) -> Tuple[GtoPLigandPaths, str]:
+    ) -> tuple[GtoPLigandPaths, str]:
         """Get path to latest version of data, and its version value
 
         :param from_local: if True, use latest available local file

src/wags_tails/hemonc.py

@@ -1,9 +1,10 @@
 """Provide source fetching for HemOnc."""
+
 import logging
 import os
 import zipfile
 from pathlib import Path
-from typing import NamedTuple, Tuple
+from typing import NamedTuple
 
 import requests
 
@@ -69,7 +70,7 @@ def _download_handler(self, dl_path: Path, file_paths: HemOncPaths) -> None:
                         zip_ref.extract(file, path.parent)
         dl_path.unlink()
 
-    def _download_data(self, version: str, outfile_paths: HemOncPaths) -> None:
+    def _download_data(self, version: str, outfile_paths: HemOncPaths) -> None:  # noqa: ARG002
         """Download data file to specified location.
 
         :param version: version to acquire
@@ -84,13 +85,13 @@ def _download_data(self, version: str, outfile_paths: HemOncPaths) -> None:
             self.data_dir,
             headers={"X-Dataverse-key": api_key},
             # provide save_path arg for API consistency, but don't use it
-            handler=lambda dl_path, save_path: self._download_handler(
+            handler=lambda dl_path, save_path: self._download_handler(  # noqa: ARG005
                 dl_path, outfile_paths
             ),
             tqdm_params=self._tqdm_params,
         )
 
-    def _get_local_files(self) -> Tuple[HemOncPaths, str]:
+    def _get_local_files(self) -> tuple[HemOncPaths, str]:
         """Acquire locally-available data files.
 
         :return: HemOnc file paths and their version
@@ -108,7 +109,7 @@ def _get_local_files(self) -> Tuple[HemOncPaths, str]:
 
     def get_latest(
         self, from_local: bool = False, force_refresh: bool = False
-    ) -> Tuple[HemOncPaths, str]:
+    ) -> tuple[HemOncPaths, str]:
         """Get path to latest version of data, and its version value
 
         :param from_local: if True, use latest available local file

src/wags_tails/hgnc.py

@@ -1,4 +1,5 @@
 """Provide data fetching for HGNC."""
+
 import ftplib
 from pathlib import Path
 
@@ -26,7 +27,7 @@ def _get_latest_version(self) -> str:
             timestamp = ftp.voidcmd(f"MDTM {self._directory_path}{self._host_filename}")
         return timestamp[4:12]
 
-    def _download_data(self, version: str, outfile: Path) -> None:
+    def _download_data(self, version: str, outfile: Path) -> None:  # noqa: ARG002
         """Download data file to specified location.
 
         :param version: version to acquire

src/wags_tails/mondo.py

@@ -1,8 +1,8 @@
 """Provide source fetching for Mondo Disease Ontology."""
+
 import datetime
 import logging
 from pathlib import Path
-from typing import Tuple
 
 import requests
 
@@ -22,7 +22,7 @@ class MondoData(GitHubDataSource):
     _repo = "monarch-initiative/mondo"
 
     @staticmethod
-    def _get_latest_version() -> Tuple[str, str]:
+    def _get_latest_version() -> tuple[str, str]:
         """Retrieve latest version value, and download URL, from GitHub release data.
 
         :param asset_name: name of file asset, if needed
@@ -71,7 +71,7 @@ def _download_data(self, version: str, outfile: Path) -> None:
 
     def get_latest(
         self, from_local: bool = False, force_refresh: bool = False
-    ) -> Tuple[Path, str]:
+    ) -> tuple[Path, str]:
         """Get path to latest version of data. Overwrite inherited method because
         final downloads depend on information gleaned from the version API call.
 

src/wags_tails/ncbi.py

@@ -1,9 +1,10 @@
 """Provide data fetching for NCBI gene data sources."""
+
 import ftplib
 import logging
 import re
 from pathlib import Path
-from typing import NamedTuple, Tuple
+from typing import NamedTuple
 
 from wags_tails.base_source import DataSource, RemoteDataError
 from wags_tails.utils.downloads import download_ftp, handle_gzip
@@ -142,7 +143,7 @@ def _download_data(self, file_paths: NcbiGenePaths) -> None:
 
     def get_latest(
         self, from_local: bool = False, force_refresh: bool = False
-    ) -> Tuple[NcbiGenePaths, str]:
+    ) -> tuple[NcbiGenePaths, str]:
         """Get path to latest version of data, and its version value
 
         :param from_local: if True, use latest available local file

src/wags_tails/ncbi_lrg_refseqgene.py

@@ -1,4 +1,5 @@
 """Fetches NCBI LRG_RefSeqGene data."""
+
 import re
 from pathlib import Path
 
@@ -36,7 +37,7 @@ def _get_latest_version(self) -> str:
             raise RemoteDataError(msg)
         return match[0].replace("-", "")
 
-    def _download_data(self, version: str, outfile: Path) -> None:
+    def _download_data(self, version: str, outfile: Path) -> None:  # noqa: ARG002
         """Download data file to specified location.
 
         :param version: version to acquire