test: improve type checks and declare typing support

.github/workflows/ci.yml

@@ -26,7 +26,7 @@ jobs:
         uses: astral-sh/setup-uv@v5
         with:
           # Install a specific version of uv and enable caching.
-          version: "0.6.2"
+          version: "0.7.2"
           enable-cache: true
 
       - name: Set up Python
@@ -39,7 +39,7 @@ jobs:
         run: uv run ruff format --check
 
       - name: Run mypy
-        run: uv run mypy src
+        run: uv run pyrefly check
 
       - name: Run pytest
         run: uv run pytest

pyproject.toml

@@ -35,7 +35,7 @@ Changelog = "https://github.com/BfArM-MVH/grz-pydantic-models/commits/main"
 [dependency-groups]
 dev = [
   "ruff >=0.11.0,<0.12",
-  "mypy >=1.13.0",
+  "pyrefly >=0.13.0,<0.14.0",
   "pytest == 8.*"
 ]
 
@@ -52,3 +52,6 @@ fixable = [
 select = [
   "I"
 ]
+
+[tool.pyrefly]
+search_path = ["src"]

src/grz_pydantic_models/pruefbericht/v0.py

@@ -22,18 +22,18 @@
 class DataCategory(enum.StrEnum):
     """Type of submission."""
 
-    clinical = enum.auto()
-    genomic = enum.auto()
+    clinical = "clinical"
+    genomic = "genomic"
 
 
 class LibraryType(enum.StrEnum):
     """Sequencing method, if applicable."""
 
-    panel = enum.auto()
-    wes = enum.auto()
-    wgs = enum.auto()
-    wgs_lr = enum.auto()
-    none = enum.auto()
+    panel = "panel"
+    wes = "wes"
+    wgs = "wgs"
+    wgs_lr = "wgs_lr"
+    none = "none"
 
 
 class SubmittedCase(StrictBaseModel):

src/grz_pydantic_models/submission/metadata/v1.py

@@ -727,7 +727,7 @@ class Donor(StrictBaseModel):
     @model_validator(mode="after")
     def warn_empty_sequence_data(self):
         for lab_datum in self.lab_data:
-            if not lab_datum.has_sequence_data():
+            if lab_datum.sequence_data is None:
                 log.warning(
                     f"No sequence data found for lab datum '{lab_datum.lab_data_name}' in donor '{self.donor_pseudonym}'. "
                     "Is this a submission without sequence data?"
@@ -750,7 +750,7 @@ def validate_target_bed_files_exist(self):
 
         for lab_datum in self.lab_data:
             if (
-                lab_datum.has_sequence_data()
+                lab_datum.sequence_data is not None
                 and lab_datum.library_type in lib_types
                 and not lab_datum.sequence_data.contains_files(FileType.bed)
             ):
@@ -766,7 +766,7 @@ def validate_vcf_file_exists(self):
         Check if there is a VCF file
         """
         for lab_datum in self.lab_data:
-            if lab_datum.has_sequence_data() and not lab_datum.sequence_data.contains_files(FileType.vcf):
+            if lab_datum.sequence_data is not None and not lab_datum.sequence_data.contains_files(FileType.vcf):
                 raise ValueError(
                     f"VCF file missing for lab datum '{lab_datum.lab_data_name}' in donor '{self.donor_pseudonym}'."
                 )
@@ -779,7 +779,7 @@ def validate_sequencing_file_exists(self):  # noqa: C901
         Check if there is a FASTQ or BAM file (depending on library type)
         """
         for lab_datum in self.lab_data:
-            if not lab_datum.has_sequence_data():
+            if lab_datum.sequence_data is None:
                 # Skip if no sequence data is present
                 continue
             fastq_files = lab_datum.sequence_data.list_files(FileType.fastq)
@@ -957,6 +957,7 @@ def validate_reference_genome_compatibility(self):
             (donor.donor_pseudonym, lab_datum.lab_data_name): lab_datum.sequence_data.reference_genome
             for donor in self.donors
             for lab_datum in donor.lab_data
+            if lab_datum.sequence_data is not None
         }
         unique_reference_genomes = set(reference_genomes.values())
         if len(unique_reference_genomes) > 1:
@@ -970,13 +971,12 @@ def validate_reference_genome_compatibility(self):
 
 
 def _check_thresholds(donor: Donor, lab_datum: LabDatum, thresholds: dict[str, Any]):
-    if not lab_datum.has_sequence_data():
+    if lab_datum.sequence_data is None:
         # Skip if no sequence data is present; warning issues in the validator `warn_empty_sequence_data` of `Donor`.
         return
     pseudonym = donor.donor_pseudonym
     lab_data_name = lab_datum.lab_data_name
-    # mypy cannot reason about the `has_sequence_data` check
-    sequence_data = typing.cast(SequenceData, lab_datum.sequence_data)
+    sequence_data = lab_datum.sequence_data
 
     mean_depth_of_coverage_t = thresholds.get("meanDepthOfCoverage")
     mean_depth_of_coverage_v = sequence_data.mean_depth_of_coverage