Use new bio2zarr zarr_format option

.github/workflows/ci.yml

@@ -152,7 +152,7 @@ jobs:
         run: |
           uv run pytest
         env:
-          BIO2ZARR_ZARR_FORMAT: ${{ matrix.zarr-format }}
+          BIO2ZARR_DEFAULT_ZARR_FORMAT: ${{ matrix.zarr-format }}
 
   test-azure-emulator:
     name: Test Azure emulator
@@ -183,4 +183,4 @@ jobs:
         run: |
           uv run pytest -q tests/test_azure.py
         env:
-          BIO2ZARR_ZARR_FORMAT: "3"
+          BIO2ZARR_DEFAULT_ZARR_FORMAT: "3"

README.md

@@ -85,8 +85,8 @@ NA00003
 # Create some VCZ data
 % rm -rf data
 % mkdir data
-% BIO2ZARR_ZARR_FORMAT=3 uv run vcf2zarr convert --no-progress --samples-chunk-size=4 tests/data/vcf/sample-part1.vcf.gz data/sample-part1.vcf.vcz
-% BIO2ZARR_ZARR_FORMAT=3 uv run vcf2zarr convert --no-progress tests/data/vcf/sample-part2.vcf.gz data/sample-part2.vcf.vcz
+% uv run vcf2zarr convert --no-progress --zarr-format=3 --samples-chunk-size=4 tests/data/vcf/sample-part1.vcf.gz data/sample-part1.vcf.vcz
+% uv run vcf2zarr convert --no-progress --zarr-format=3 tests/data/vcf/sample-part2.vcf.gz data/sample-part2.vcf.vcz
 
 # Copy first vcz to an icechunk store
 % uv run vczstore copy-store-to-icechunk data/sample-part1.vcf.vcz data/store.vcz

tests/test_append.py

@@ -131,7 +131,7 @@ def test_append_icechunk(tmp_path):
 
     # note that vcz1 is in icechunk, but the dataset being appended, vcz2, needn't be
     vcz1 = convert_vcf_to_vcz_icechunk("sample-part1.vcf.gz", tmp_path)
-    vcz2 = convert_vcf_to_vcz("sample-part2.vcf.gz", tmp_path)
+    vcz2 = convert_vcf_to_vcz("sample-part2.vcf.gz", tmp_path, zarr_format=3)
 
     print(vcz1)
     print(vcz2)

tests/test_azure.py

@@ -56,12 +56,12 @@ def _convert_vcf_to_vcz_zarr3(vcf_name, tmp_path):
     vcf_path = pathlib.Path("tests/data/vcf") / vcf_name
     output = (tmp_path / vcf_path.name).with_suffix(".vcz")
     env = dict(os.environ)
-    env["BIO2ZARR_ZARR_FORMAT"] = "3"
     completed = subprocess.run(
         [
             "vcf2zarr",
             "convert",
             "--no-progress",
+            "--zarr-format=3",
             "--worker-processes=0",
             str(vcf_path),
             str(output),

tests/test_normalise.py

@@ -350,7 +350,7 @@ def test_normalise_and_append_icechunk(tmp_path):
 
     # note that vcz0 is in icechunk, but the others needn't be
     vcz0 = convert_vcf_to_vcz_icechunk("sample-variants.vcf.gz", tmp_path, ploidy=2)
-    vcz1 = convert_vcf_to_vcz("sample-part1.vcf.gz", tmp_path)
+    vcz1 = convert_vcf_to_vcz("sample-part1.vcf.gz", tmp_path, zarr_format=3)
     vcz1_norm = zarr.storage.MemoryStore()
 
     with icechunk_transaction(vcz0, "main", message="append") as store:

tests/utils.py

@@ -227,7 +227,12 @@ def vcz_path_cache(vcf_path):
 
 
 def convert_vcf_to_vcz(
-    vcf_name, tmpdir, variants_chunk_size=None, samples_chunk_size=None, ploidy=None
+    vcf_name,
+    tmpdir,
+    variants_chunk_size=None,
+    samples_chunk_size=None,
+    ploidy=None,
+    zarr_format=None,
 ):
     vcf_path = pathlib.Path("tests/data/vcf") / vcf_name
     output = (pathlib.Path(tmpdir) / vcf_path.name).with_suffix(".vcz")
@@ -239,6 +244,7 @@ def convert_vcf_to_vcz(
         worker_processes=0,
         local_alleles=False,
         ploidy=ploidy,
+        zarr_format=zarr_format,
     )
     return output
 
@@ -300,6 +306,7 @@ def convert_vcf_to_vcz_icechunk(
         variants_chunk_size=variants_chunk_size,
         samples_chunk_size=samples_chunk_size,
         ploidy=ploidy,
+        zarr_format=3,
     )
 
     ic_tmp_path = tmp_path / "icechunk"