1 files changed, 26 insertions, 0 deletions
diff --git a/bh20sequploader/qc_metadata.py b/bh20sequploader/qc_metadata.py
new file mode 100644
index 0000000..e477f21
--- /dev/null
+++ b/bh20sequploader/qc_metadata.py
@@ -0,0 +1,26 @@
+import schema_salad.schema
+import schema_salad.ref_resolver
+import logging
+import pkg_resources
+import logging
+import traceback
+
+def qc_metadata(metadatafile):
+    schema_resource = pkg_resources.resource_stream(__name__, "bh20seq-schema.yml")
+    cache = {"https://raw.githubusercontent.com/arvados/bh20-seq-resource/master/bh20sequploader/bh20seq-schema.yml": schema_resource.read().decode("utf-8")}
+    (document_loader,
+     avsc_names,
+     schema_metadata,
+     metaschema_loader) = schema_salad.schema.load_schema("https://raw.githubusercontent.com/arvados/bh20-seq-resource/master/bh20sequploader/bh20seq-schema.yml", cache=cache)
+
+    if not isinstance(avsc_names, schema_salad.avro.schema.Names):
+        print(avsc_names)
+        return False
+
+    try:
+        doc, metadata = schema_salad.schema.load_and_validate(document_loader, avsc_names, metadatafile, True)
+        return True
+    except Exception as e:
+        traceback.print_exc()
+        logging.warn(e)
+    return False