From a0ed1dbab5beacb034efafb442a853c9e61f0ea6 Mon Sep 17 00:00:00 2001
From: Alex Bednarek <alex.bednarek@planet.com>
Date: Wed, 29 May 2024 15:56:30 -0400
Subject: [PATCH] WIP: Addressing LATEST_MAX_LOOKBACK with the new latest
 functionality.

---
 api/datalake_api/app.py           |  1 +
 api/datalake_api/querier.py       | 15 +++++++++------
 api/datalake_api/settings.py      |  6 ++++++
 api/datalake_api/v0.py            |  4 ++++
 api/tests/test_archive_querier.py | 17 ++++++++++++++++-
 5 files changed, 36 insertions(+), 7 deletions(-)

diff --git a/api/datalake_api/app.py b/api/datalake_api/app.py
index 0d61e2b..af94d74 100644
--- a/api/datalake_api/app.py
+++ b/api/datalake_api/app.py
@@ -23,6 +23,7 @@
 from datalake_api import settings
 
 
+logging.basicConfig(level=logging.INFO)
 LOGGER = logging.getLogger(__name__)
 
 
diff --git a/api/datalake_api/querier.py b/api/datalake_api/querier.py
index 141c735..18bdd94 100644
--- a/api/datalake_api/querier.py
+++ b/api/datalake_api/querier.py
@@ -21,9 +21,7 @@
 import os
 
 import logging
-logging.basicConfig(level=logging.INFO)
 log = logging.getLogger(__name__)
-log.setLevel(logging.INFO)
 
 
 '''the maximum number of results to return to the user
@@ -179,12 +177,17 @@ def _unpack(self, result):
 
 class ArchiveQuerier(object):
 
-    def __init__(self, table_name, latest_table_name=None, dynamodb=None):
+    def __init__(self, table_name,
+                 latest_table_name=None,
+                 use_latest_table=None,
+                 latest_max_lookback=30,
+                 dynamodb=None):
         self.table_name = table_name
         self.latest_table_name = latest_table_name
+        self.use_latest_table = use_latest_table
+        self.latest_max_lookback = latest_max_lookback
         self.dynamodb = dynamodb
-        self.use_latest_table = os.environ.get("DATALAKE_USE_LATEST_TABLE",
-                                               "false").lower() == "true"
+        
 
     def query_by_work_id(self, work_id, what, where=None, cursor=None):
         kwargs = self._prepare_work_id_kwargs(work_id, what)
@@ -353,7 +356,7 @@ def query_latest(self, what, where, lookback_days=DEFAULT_LOOKBACK_DAYS):
                 KeyConditionExpression=Key('what_where_key').eq(f'{what}:{where}')
             )
             items = response.get('Items', [])
-            if not items:
+            if not items and self.latest_max_lookback > 0:
                 return self._default_latest(what, where, lookback_days)
 
             latest_item = items[0]
diff --git a/api/datalake_api/settings.py b/api/datalake_api/settings.py
index e55c6d1..3d79650 100644
--- a/api/datalake_api/settings.py
+++ b/api/datalake_api/settings.py
@@ -11,10 +11,16 @@
 # WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
 # License for the specific language governing permissions and limitations under
 # the License.
+import os
 
 # default settings
+
 DYNAMODB_TABLE = 'test'
 DYNAMODB_LATEST_TABLE = 'test_latest'
+DATALAKE_USE_LATEST_TABLE = \
+    os.environ.get("DATALAKE_USE_LATEST_TABLE", "false").lower() == "true"
+LATEST_MAX_LOOKBACK = int(os.environ.get("LATEST_MAX_LOOKBACK", "30"))
+
 AWS_REGION = 'us-west-2'
 AWS_ACCESS_KEY_ID = None
 AWS_SECRET_ACCESS_KEY = None
diff --git a/api/datalake_api/v0.py b/api/datalake_api/v0.py
index 927997d..aba95f1 100644
--- a/api/datalake_api/v0.py
+++ b/api/datalake_api/v0.py
@@ -51,8 +51,12 @@ def get_archive_querier():
     if not hasattr(app, 'archive_querier'):
         table_name = app.config.get('DYNAMODB_TABLE')
         latest_table_name = app.config.get('DYNAMODB_LATEST_TABLE')
+        use_latest_table = app.config.get('DATALAKE_USE_LATEST_TABLE')
+        latest_max_lookback = app.config.get("LATEST_MAX_LOOKBACK")
         app.archive_querier = ArchiveQuerier(table_name,
                                              latest_table_name,
+                                             use_latest_table,
+                                             latest_max_lookback,
                                              dynamodb=get_dynamodb())
     return app.archive_querier
 
diff --git a/api/tests/test_archive_querier.py b/api/tests/test_archive_querier.py
index 2c6fe37..00daa7c 100644
--- a/api/tests/test_archive_querier.py
+++ b/api/tests/test_archive_querier.py
@@ -123,6 +123,10 @@ def query_latest(self, what, where):
         return HttpRecord(**record)
 
 
+
+"""
+Incorporate LATEST_MAX_LOOKBACK HERE
+"""
 @pytest.fixture(params=[
     ('archive', 'use_latest'),
     ('archive', 'use_default'),
@@ -565,4 +569,15 @@ def test_latest_table_query(table_maker, querier, record_maker):
     table_maker(records)
     querier.use_latest_table = True
     result = querier.query_latest('boo', 'hoo0')
-    _validate_latest_result(result, what='boo', where='hoo0')
\ No newline at end of file
+    _validate_latest_result(result, what='boo', where='hoo0')
+
+"""
+Write tests:
+With setup of latest table records, 
+with DYNAMODB_LATEST_TABLE set, with DATALAKE_USE_LATEST_TABLE=true, with LATEST_MAX_LOOKBACK=0, record is found
+
+With setup of latest table records, 
+with DYNAMODB_LATEST_TABLE set, with DATALAKE_USE_LATEST_TABLE=false, with LATEST_MAX_LOOKBACK=0, record is not found
+
+2-4
+"""
\ No newline at end of file