def load_orthologs(fo: IO, metadata: dict):
"""Load orthologs into ArangoDB
Args:
fo: file obj - orthologs file
metadata: dict containing the metadata for orthologs
"""
version = metadata["metadata"]["version"]
# LOAD ORTHOLOGS INTO ArangoDB
with timy.Timer("Load Orthologs") as timer:
arango_client = arangodb.get_client()
if not arango_client:
print("Cannot load orthologs without ArangoDB access")
quit()
belns_db = arangodb.get_belns_handle(arango_client)
arangodb.batch_load_docs(belns_db,
orthologs_iterator(fo, version),
on_duplicate="update")
log.info("Load orthologs",
elapsed=timer.elapsed,
source=metadata["metadata"]["source"])
# Clean up old entries
remove_old_ortholog_edges = f"""
FOR edge in ortholog_edges
FILTER edge.source == "{metadata["metadata"]["source"]}"
FILTER edge.version != "{version}"
REMOVE edge IN ortholog_edges
"""
remove_old_ortholog_nodes = f"""
FOR node in ortholog_nodes
FILTER node.source == "{metadata["metadata"]["source"]}"
FILTER node.version != "{version}"
REMOVE node IN ortholog_nodes
"""
arangodb.aql_query(belns_db, remove_old_ortholog_edges)
arangodb.aql_query(belns_db, remove_old_ortholog_nodes)
# Add metadata to resource metadata collection
metadata["_key"] = f"Orthologs_{metadata['metadata']['source']}"
try:
belns_db.collection(arangodb.belns_metadata_name).insert(metadata)
except ArangoError as ae:
belns_db.collection(arangodb.belns_metadata_name).replace(metadata)
# Third Party Imports
import structlog
# Local Imports
import bel.db.arangodb as arangodb
import bel.db.elasticsearch
import bel.lang.belobj
import bel.utils
from bel.Config import config
log = structlog.getLogger(__name__)
es = bel.db.elasticsearch.get_client()
client = arangodb.get_client()
belapi_db = arangodb.get_belapi_handle(client)
bel_validations_name = arangodb.bel_validations_name
bel_validation_coll = belapi_db.collection(bel_validations_name)
def convert_msg_to_html(msg):
"""Convert \n into a <BR> for an HTML formatted message"""
msg = re.sub("\n", "<br />", msg, flags=re.MULTILINE)
return msg
def get_validation_for_hashes(hashes):
"""Get cached validations from validation cache database in arangodb"""
def load_terms(fo: IO, metadata: dict, forceupdate: bool):
"""Load terms into Elasticsearch and ArangoDB
Forceupdate will create a new index in Elasticsearch regardless of whether
an index with the resource version already exists.
Args:
fo: file obj - terminology file
metadata: dict containing the metadata for terminology
forceupdate: force full update - e.g. don't leave Elasticsearch indexes
alone if their version ID matches
"""
version = metadata["metadata"]["version"]
# LOAD TERMS INTO Elasticsearch
with timy.Timer("Load Terms") as timer:
es = bel.db.elasticsearch.get_client()
es_version = version.replace("T", "").replace("-", "").replace(":", "")
index_prefix = f"terms_{metadata['metadata']['namespace'].lower()}"
index_name = f"{index_prefix}_{es_version}"
# Create index with mapping
if not elasticsearch.index_exists(es, index_name):
elasticsearch.create_terms_index(es, index_name)
elif forceupdate: # force an update to the index
index_name += "_alt"
elasticsearch.create_terms_index(es, index_name)
else:
return # Skip loading if not forced and not a new namespace
terms_iterator = terms_iterator_for_elasticsearch(fo, index_name)
elasticsearch.bulk_load_docs(es, terms_iterator)
# Remove old namespace index
index_names = elasticsearch.get_all_index_names(es)
for name in index_names:
if name != index_name and index_prefix in name:
elasticsearch.delete_index(es, name)
# Add terms_alias to this index
elasticsearch.add_index_alias(es, index_name, terms_alias)
log.info(
"Load namespace terms",
elapsed=timer.elapsed,
namespace=metadata["metadata"]["namespace"],
)
# LOAD EQUIVALENCES INTO ArangoDB
with timy.Timer("Load Term Equivalences") as timer:
arango_client = arangodb.get_client()
if not arango_client:
print("Cannot load terms without ArangoDB access")
quit()
belns_db = arangodb.get_belns_handle(arango_client)
arangodb.batch_load_docs(belns_db,
terms_iterator_for_arangodb(fo, version),
on_duplicate="update")
log.info(
"Loaded namespace equivalences",
elapsed=timer.elapsed,
namespace=metadata["metadata"]["namespace"],
)
# Clean up old entries
remove_old_equivalence_edges = f"""
FOR edge in equivalence_edges
FILTER edge.source == "{metadata["metadata"]["namespace"]}"
FILTER edge.version != "{version}"
REMOVE edge IN equivalence_edges
"""
remove_old_equivalence_nodes = f"""
FOR node in equivalence_nodes
FILTER node.source == "{metadata["metadata"]["namespace"]}"
FILTER node.version != "{version}"
REMOVE node IN equivalence_nodes
"""
arangodb.aql_query(belns_db, remove_old_equivalence_edges)
arangodb.aql_query(belns_db, remove_old_equivalence_nodes)
# Add metadata to resource metadata collection
metadata["_key"] = f"Namespace_{metadata['metadata']['namespace']}"
try:
belns_db.collection(arangodb.belns_metadata_name).insert(metadata)
except ArangoError as ae:
belns_db.collection(arangodb.belns_metadata_name).replace(metadata)
