Merge pull request #19 from 4dn-dcic/0.3.3

0.3.3

dcicutils/_version.py

@@ -1,4 +1,4 @@
 """Version information."""
 
 # The following line *must* be the last in the module, exactly as formatted:
-__version__ = "0.3.2"
+__version__ = "0.3.3"

dcicutils/ff_utils.py

@@ -6,7 +6,11 @@
 import copy
 import boto3
 from uuid import UUID
-from dcicutils import s3_utils, submit_utils
+from dcicutils import (
+    s3_utils,
+    submit_utils,
+    es_utils
+)
 import requests
 # urlparse import differs between py2 and 3
 if sys.version_info[0] < 3:
@@ -330,6 +334,29 @@ def delete_field(obj_id, del_field, key=None, ff_env=None):
     return get_response_json(response)
 
 
+def get_es_metadata(uuid, schema_name, es_client=None, key=None, ff_env=None):
+    """
+    Given string item uuid and schema name (e.g. "file_fastq"), will return a
+    dictionary response of the full ES ecord for that item (or an empty
+    dictionary if the item doesn't exist/ is not indexed)
+    You can pass in an Elasticsearch client (initialized by create_es_client)
+    through the es_client param to save init time.
+    Same auth mechanism as the other metadata functions
+    """
+    from elasticsearch.exceptions import TransportError
+    if es_client is None:
+        # need to know ES server location and item type
+        auth = get_authentication_with_server(key, ff_env)
+        health_res = authorized_request(auth['server'] + '/health', auth=auth, verb='GET')
+        es_url = get_response_json(health_res)['elasticsearch']
+        es_client = es_utils.create_es_client(es_url, use_aws_auth=True)
+    try:
+        es_res = es_client.get(index=schema_name, doc_type=schema_name, id=uuid)
+    except TransportError:
+        return {}
+    return es_res.get('_source', {})
+
+
 #####################
 # Utility functions #
 #####################

test/test_ff_utils.py

@@ -490,6 +490,32 @@ def get_search_generator(integrated_ff):
     assert len(all_gen3_uuids) == len(all_gen3)
 
 
+@pytest.mark.integrated
+def test_get_es_metadata(integrated_ff):
+    from dcicutils import es_utils
+    # use this test biosource
+    test_item = '331111bc-8535-4448-903e-854af460b254'
+    res = ff_utils.get_es_metadata(test_item, 'biosource', key=integrated_ff['ff_key'])
+    assert res['uuid'] == test_item
+    assert res['item_type'] == 'biosource'
+    assert isinstance(res['embedded'], dict)
+    assert isinstance(res['links'], dict)
+
+    # you can also pass in your own elasticsearch client
+    # ugly here because we need to get it from health page
+    health_res = ff_utils.authorized_request(integrated_ff['ff_key']['server'] + '/health',
+                                             auth=integrated_ff['ff_key'])
+    es_url = ff_utils.get_response_json(health_res)['elasticsearch']
+    es_client = es_utils.create_es_client(es_url, use_aws_auth=True)
+    res2 = ff_utils.get_es_metadata(test_item, 'biosource', es_client=es_client,
+                                    key=integrated_ff['ff_key'])
+    assert res2['uuid'] == res['uuid']
+
+    # bad item returns empty dict
+    res = ff_utils.get_es_metadata('blahblah', 'biosource', key=integrated_ff['ff_key'])
+    assert res == {}
+
+
 @pytest.mark.integrated
 def test_fdn_connection(integrated_ff):
     connection = ff_utils.fdn_connection(key={'default': integrated_ff['ff_key']})