fix: support int | str group ids

.github/workflows/ci.yml

@@ -29,7 +29,7 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
       - name: Test
         run: docker build --target=test .
   release:
@@ -42,7 +42,7 @@ jobs:
       - name: Setup Node
         uses: actions/setup-node@v3
         with:
-          node-version: 16
+          node-version: 18
       - name: Install semantic-release
         run: npm i [email protected] [email protected]
       - name: Release

.github/workflows/publish.yml

@@ -22,7 +22,7 @@ jobs:
       - name: Update pyproject.toml version
         run: sed -i 's/0\.0\.0/${{ github.event.release.tag_name }}/' pyproject.toml
       - name: Login to Registry
-        uses: docker/login-action@v2
+        uses: docker/login-action@v3
         with:
           registry: ${{ env.REGISTRY }}
           username: ${{ secrets.GH_USERNAME }}
@@ -36,7 +36,8 @@ jobs:
         uses: docker/build-push-action@v4
         with:
           context: .
+          labels: ${{ steps.meta.outputs.labels }}
           push: true
           tags: ${{ steps.meta.outputs.tags }}
           target: "base"
-          labels: ${{ steps.meta.outputs.labels }}
+

fixtures.py

@@ -1,13 +1,12 @@
 from pathlib import Path
 from types import SimpleNamespace
-from typing import List
 
 from pyfixtures import fixture
 from virtool_workflow.data_model.indexes import WFIndex
 
 
 @fixture
-def index(indexes: List[WFIndex]):
+def index(indexes: list[WFIndex]):
     return indexes[0]
 
 

pathoscope.py

@@ -4,7 +4,7 @@
 import rust_utils
 from functools import cached_property
 from pathlib import Path
-from typing import Any, Dict, Generator, List
+from typing import Any, Generator
 
 
 class SamLine:
@@ -36,7 +36,7 @@ def read_length(self) -> int:
         return len(self.fields[9])
 
     @cached_property
-    def fields(self) -> List[Any]:
+    def fields(self) -> list[Any]:
         """
         The SAM fields
         """
@@ -135,7 +135,7 @@ def rescale_samscore(u, nu, max_score, min_score):
     return u, nu
 
 
-def find_sam_align_score(fields: List[Any]) -> float:
+def find_sam_align_score(fields: list[Any]) -> float:
     """
     Find the Bowtie2 alignment score for the given split line (``fields``).
 
@@ -494,7 +494,7 @@ def write_report(
     return results
 
 
-def calculate_coverage(sam_path: Path, ref_lengths: Dict[str, int]):
+def calculate_coverage(sam_path: Path, ref_lengths: list[str, int]):
     coverage_dict = {}
     pos_length_list = []
 

pyproject.toml

@@ -7,20 +7,20 @@ license = "MIT"
 
 [tool.poetry.dependencies]
 python = "~3.10"
-virtool-workflow = "^5.3.1"
+virtool-workflow = "^5.4.2"
 rust = "^0.1.1"
 
 [tool.poetry.group.dev.dependencies]
+black = "^23.1.0"
+pre-commit = "^3.0.4"
+pydantic-factories = "^1.15.0"
 pytest = "^6.2.0"
 pytest-aiohttp = "^0.3.0"
 pytest-asyncio = "^0.15.1"
 pytest-datafiles = "^2.0"
-syrupy = "^1.7.4"
 pytest-mock = "^3.7.0"
 pytest-regressions = "^2.2.0"
-pydantic-factories = "^1.15.0"
-pre-commit = "^3.0.4"
-black = "^23.1.0"
+syrupy = "^1.7.4"
 
 [tool.pytest.ini_options]
 markers = ["datafiles: load datafiles", "asyncio_mode: auto"]

workflow.py

@@ -2,14 +2,14 @@
 import shlex
 import shutil
 from collections import defaultdict
-from logging import getLogger
 from pathlib import Path
 from types import SimpleNamespace
-from typing import Any, Dict, List, TextIO, Set
+from typing import Any, TextIO
 
-import rust_utils
 import aiofiles
 import aiofiles.os
+import rust_utils
+from structlog import get_logger
 from virtool_workflow import hooks, step
 from virtool_workflow.data_model.analysis import WFAnalysis
 from virtool_workflow.data_model.indexes import WFIndex
@@ -20,7 +20,7 @@
 from pathoscope import run as run_pathoscope
 from pathoscope import write_report
 
-logger = getLogger("pathoscope")
+logger = get_logger("pathoscope")
 
 BAD_FIRST_SAM_CHARACTERS = {"\n", "@", "#"}
 
@@ -41,7 +41,7 @@ def read_fastq_grouped_lines(fastq_file: TextIO):
 
 
 def subtract_fastq(
-    current_fastq_path: Path, new_fastq_path: Path, subtracted_reads: Set[str]
+    current_fastq_path: Path, new_fastq_path: Path, subtracted_reads: set[str]
 ):
     with open(current_fastq_path, "r") as current_fastq_file, open(
         new_fastq_path, "w"
@@ -116,7 +116,7 @@ async def stdout_handler(line: bytes):
         stdout_handler=stdout_handler,
     )
 
-    logger.info("Found %i potential OTUs.", len(intermediate.to_otus))
+    logger.info("Found potential OTUs", count=len(intermediate.to_otus))
 
 
 @step
@@ -221,9 +221,9 @@ async def eliminate_subtraction(
     isolate_fastq_path: Path,
     isolate_sam_path: Path,
     proc: int,
-    results: Dict[str, Any],
+    results: dict[str, Any],
     run_subprocess: RunSubprocess,
-    subtractions: List[WFSubtraction],
+    subtractions: list[WFSubtraction],
     subtracted_sam_path: Path,
     work_path: Path,
 ):
@@ -322,10 +322,10 @@ async def eliminate_subtraction(
         await asyncio.to_thread(shutil.copyfile, new_fastq_path, current_fastq_path)
 
         logger.info(
-            "Subtracted reads that mapped better to a subtraction subtraction_id=%s subtraction_name=%s count=%i",
-            subtraction.id,
-            subtraction.name,
-            subtracted_count,
+            "Some reads mapped better to a subtraction and were removed",
+            id=subtraction.id,
+            name=subtraction.name,
+            count=subtracted_count,
         )
 
     results["subtracted_count"] = subtracted_count
@@ -350,9 +350,9 @@ async def reassignment(
     reassigned_path = work_path / "reassigned.sam"
 
     logger.info(
-        "Running Pathoscope subtracted_sam_path=%s reassigned_path=%s",
-        subtracted_sam_path,
-        reassigned_path,
+        "Running Pathoscope",
+        subtracted_path=subtracted_sam_path,
+        reassigned_path=reassigned_path,
     )
 
     (
@@ -376,7 +376,7 @@ async def reassignment(
 
     report_path = work_path / "report.tsv"
 
-    logger.info("Writing report report_path=%s", report_path)
+    logger.info("Writing report", path=report_path)
 
     report = await asyncio.to_thread(
         write_report,