Implement tiledb_soma_healthcheck component

CHANGELOG.md

@@ -65,6 +65,8 @@ Warning: These experimental features are subject to change in future releases.
 
 * Added `annotate/singler`: Cell type annotation using SingleR (PR #1051).
 
+* Added `tiledb_soma_healthcheck` component (PR #1055). 
+
 * Added `tiledb/move_mudata_obsm_to_tiledb` (PR #1065).
 
 ## MAJOR CHANGES

src/query/tiledb_soma_healthcheck/config.vsh.yaml

@@ -0,0 +1,44 @@
+name: tiledb_healthcheck
+namespace: query
+scope: "public"
+description: |
+  Checks if a provided location can be queried as a TileDB-SOMA database.
+authors:
+  - __merge__: /src/authors/dries_schaumont.yaml
+    roles: [ author, maintainer ]
+argument_groups:
+  - name: Input database
+    description: "Open a TileDB SOMA database by URI."
+    arguments:
+      - name: "--input_uri"
+        type: string
+        description: "If specified, a URI containing the TileDB-SOMA objects."
+        required: true
+        example: "s3://bucket/path"
+      - name: "--s3_region"
+        description: |
+          Region where the TileDB-SOMA database is hosted.
+        type: string
+        required: true
+resources:
+  - type: python_script
+    path: script.py
+  - path: /src/utils/setup_logger.py
+test_resources:
+  - type: python_script
+    path: test.py
+engines:
+  - type: docker
+    image: python:3.12
+    setup:
+      - type: python
+        packages:
+          - tiledbsoma
+    test_setup:
+      - type: python
+        __merge__: [ /src/base/requirements/viashpy.yaml, .]
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [highmem, midcpu]

src/query/tiledb_soma_healthcheck/script.py

@@ -0,0 +1,35 @@
+import sys
+import tiledbsoma
+
+## VIASH START
+par = {
+    "input_uri": "s3://cellxgene-census-public-us-west-2/cell-census/2025-01-30/soma/",
+    "s3_region": "us-west-2",
+}
+meta = {"resources_dir": "src/utils"}
+## VIASH END
+
+sys.path.append(meta["resources_dir"])
+from setup_logger import setup_logger
+
+logger = setup_logger()
+
+
+def main(par):
+    logger.info("Component started")
+
+    tiledb_config = {
+        "vfs.s3.no_sign_request": "false",
+        "vfs.s3.region": par["s3_region"],
+    }
+    context = tiledbsoma.SOMATileDBContext(tiledb_config=tiledb_config)
+    logger.info(
+        "Trying to access '%s' in region '%s'", par["input_uri"], par["s3_region"]
+    )
+    with tiledbsoma.open(par["input_uri"], "w", context=context) as _:
+        logger.info("Connection successful!")
+    logger.info("Component finished")
+
+
+if __name__ == "__main__":
+    main(par)

src/query/tiledb_soma_healthcheck/test.py

@@ -0,0 +1,25 @@
+import sys
+import pytest
+
+## VIASH START
+meta = {
+    "resources_dir": "./resources_test/",
+    "executable": "./target/executable/query/tiledb_hea",
+    "config": "/home/di/code/openpipeline/src/query/cellxgene_census/config.vsh.yaml",
+}
+## VIASH END
+
+
+def test_healthcheck(run_component):
+    run_component(
+        [
+            "--input_uri",
+            "s3://cellxgene-census-public-us-west-2/cell-census/2025-01-30/soma/",
+            "--s3_region",
+            "us-west-2",
+        ]
+    )
+
+
+if __name__ == "__main__":
+    sys.exit(pytest.main([__file__]))