aboutsummaryrefslogtreecommitdiff
path: root/bh20sequploader/qc_metadata.py
diff options
context:
space:
mode:
Diffstat (limited to 'bh20sequploader/qc_metadata.py')
-rw-r--r--bh20sequploader/qc_metadata.py15
1 files changed, 12 insertions, 3 deletions
diff --git a/bh20sequploader/qc_metadata.py b/bh20sequploader/qc_metadata.py
index 2b57991..27657b1 100644
--- a/bh20sequploader/qc_metadata.py
+++ b/bh20sequploader/qc_metadata.py
@@ -8,15 +8,20 @@ import traceback
from rdflib import Graph, Namespace
from pyshex.evaluate import evaluate
+metadata_schema = None
def qc_metadata(metadatafile):
+ global metadata_schema
log = logging.getLogger(__name__ )
- 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")}
+ if metadata_schema is None:
+ 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")}
+ metadata_schema = schema_salad.schema.load_schema("https://raw.githubusercontent.com/arvados/bh20-seq-resource/master/bh20sequploader/bh20seq-schema.yml", cache=cache)
+
(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)
+ metaschema_loader) = metadata_schema
shex = pkg_resources.resource_stream(__name__, "bh20seq-shex.rdf").read().decode("utf-8")
@@ -27,6 +32,10 @@ def qc_metadata(metadatafile):
g = schema_salad.jsonld_context.makerdf("workflow", doc, document_loader.ctx)
rslt, reason = evaluate(g, shex, doc["id"], "https://raw.githubusercontent.com/arvados/bh20-seq-resource/master/bh20sequploader/bh20seq-shex.rdf#submissionShape")
+ # As part of QC make sure serialization works too, this will raise
+ # an exception if there are invalid URIs.
+ g.serialize(format="ntriples")
+
if not rslt:
raise Exception(reason)