-Change naming from UTC to iso

-Make precision to ms and add Z
-Ading pytest -n auto plugin
-Better devshell name
-Fix Tests to run on new python.

Author: huzaifams-svg

Dockerfile

@@ -7,29 +7,33 @@ ENV	LC_ALL C.UTF-8
 
 # TODO: keep requirements in one place
 RUN pip install \
+    'boto==2.49.0' \
+    'boto3==1.35.41' \
+    'botocore==1.35.64' \
+    'datalake<2' \
+    'flake8>=2.5.0,<4.1' \
+    'freezegun<1' \
+    'moto<3' \
+    'pytest<8' \
+    'pytest-xdist>=3.0.0' \
+    'responses<0.22.0' \
+    'tox>4,<5' \
+    # test requirements
     blinker>=1.4 \
-    boto3>=1.1.3 \
     click>=5.1 \
-    Flask>=0.10.1 \
     flask-swagger>=0.2.14 \
+    Flask>=0.10.1 \
     memoized_property>=1.0.1 \
+    pyinotify>=0.9.4, \
     python-dateutil>=2.4.2 \
     python-dotenv>=0.1.3 \
     pytz>=2015.4 \
-    sentry-sdk[flask]>=0.19.5 \
+    raven>=5.0.0 \
     requests>=2.5 \
+    sentry-sdk[flask]>=0.19.5 \
     simplejson>=3.3.1 \
-    six>=1.10.0 \
-    # test requirements
-    'flake8>=2.5.0,<4.1' \
-    'freezegun<1' \
-    'moto<3' \
-    'pytest<8' \
-    'responses<0.22.0' \
-    pyinotify>=0.9.4, \
-    raven>=5.0.0 \
-    'tox>4,<5' \
-    'datalake<2'
+    six>=1.10.0
+
 
 RUN mkdir -p /opt/
 COPY . /opt/

Makefile

@@ -7,8 +7,8 @@ IMAGE="$(REPO)/$(REPO_PATH):$(VERSION)"
 docker: version
 	docker build --build-arg VERSION=$(VERSION) -t $(IMAGE) .
 
-.PHONY: devshell  # Open a developer shell in the docker env
-devshell: docker
+.PHONY: dev  # Open a developer  in the docker env
+dev: docker
 	docker run --rm -it -v $$PWD:/opt --entrypoint /bin/bash $(IMAGE)
 
 test-client: docker

api/datalake_api/v0.py

@@ -18,6 +18,8 @@
 from flask import current_app as app
 import os
 import simplejson as json
+from datetime import datetime, timezone
+import decimal
 from .querier import ArchiveQuerier, Cursor, InvalidCursor, \
     DEFAULT_LOOKBACK_DAYS
 from .fetcher import ArchiveFileFetcher
@@ -29,6 +31,38 @@
 
 _archive_querier = None
 
+
+def add_utc_metadata(metadata):
+    """Add UTC timestamp fields to metadata
+
+    This function takes a metadata dict and adds start_iso and end_iso fields
+    containing ISO-formatted UTC timestamps corresponding to the millisecond
+    timestamps in the start and end fields.
+    Can be expanded to add any needed metadata to apis
+    """
+    if not metadata:
+        return metadata
+
+    start_iso = metadata['start']
+    end_iso = metadata['end']
+    if start_iso:
+        if isinstance(start_iso, decimal.Decimal):
+            start_iso = float(start_iso)
+        start_iso = datetime.fromtimestamp(
+            start_iso / 1000.0, tz=timezone.utc
+        ).isoformat(timespec='milliseconds').replace('+00:00', 'Z')
+    if end_iso:
+        if isinstance(end_iso, decimal.Decimal):
+            end_iso = float(end_iso)
+        end_iso = datetime.fromtimestamp(
+            end_iso / 1000.0, tz=timezone.utc
+        ).isoformat(timespec='milliseconds').replace('+00:00', 'Z')
+
+    metadata['start_iso'] = start_iso
+    metadata['end_iso'] = end_iso
+    return metadata
+
+
 def _get_aws_kwargs():
     kwargs = dict(
         region_name=app.config.get('AWS_REGION'),
@@ -305,6 +339,14 @@ def files_get():
                                   type: string
                                   description: 16-byte blake2 hash of the file
                                                content
+                                start_iso:
+                                  type: string
+                                  description: the start time of the file in ISO
+                                               format UTC iso timezone
+                                end_iso:
+                                  type: string
+                                  description: the end time of the file in ISO
+                                               format UTC iso timezone
 
               next:
                   type: string
@@ -349,7 +391,11 @@ def files_get():
                                    where=params.get('where'),
                                    cursor=params.get('cursor'))
 
-    [r.update(http_url=_get_canonical_http_url(r)) for r in results]
+    for r in results:
+        r.update(http_url=_get_canonical_http_url(r))
+        # Add UTC timestamps to each record's metadata
+        r['metadata'] = add_utc_metadata(r['metadata'])
+
     response = {
         'records': results,
         'next': _get_next_url(flask.request, results),
@@ -476,6 +522,7 @@ def file_get_metadata(file_id):
           id: DatalakeAPIError
     '''
     f = _get_file(file_id)
+    f.metadata = add_utc_metadata(f.metadata)
     return Response(json.dumps(f.metadata), content_type='application/json')
 
 
@@ -542,6 +589,8 @@ def latest_get(what, where):
     params = _validate_latest_params(params)
     f = _get_latest(what, where, params.get('lookback', DEFAULT_LOOKBACK_DAYS))
     f.update(http_url=_get_canonical_http_url(f))
+    # Add UTC timestamps to metadata
+    f['metadata'] = add_utc_metadata(f['metadata'])
     return Response(json.dumps(f), content_type='application/json')
 
 

api/setup.py

@@ -30,7 +30,7 @@ def get_version_from_pyver():
         if 'sdist' in sys.argv or 'bdist_wheel' in sys.argv:
             raise ImportError('You must install pyver to create a package')
         else:
-            return 'noversion'
+            return '0.0.0'
     version, version_info = pyver.get_version(pkg="datalake_api",
                                               public=True)
     return version

api/tests/test_metadata.py

@@ -11,6 +11,7 @@
 # WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
 # License for the specific language governing permissions and limitations under
 # the License.
+from datetime import datetime, timezone
 import pytest
 import simplejson as json
 
@@ -32,7 +33,16 @@ def test_get_metadata(metadata_getter, s3_file_maker, random_metadata):
     res = metadata_getter('12345')
     assert res.status_code == 200
     assert res.content_type == 'application/json'
-    assert json.loads(res.data) == random_metadata
+    res_data = json.loads(res.data)
+    for k, v in random_metadata.items():
+        if k == 'start_UTC' or k == 'end_UTC':
+            if v: # Null case
+                expected_v_utc = datetime.fromtimestamp(
+                    v / 1000.0, tz=timezone.utc
+                ).isoformat()
+                assert expected_v_utc == res_data[k]
+        else:
+            assert v == res_data[k]
 
 
 def test_no_such_metadata(s3_bucket_maker, metadata_getter):

ingester/setup.py

@@ -30,7 +30,7 @@ def get_version_from_pyver():
         if 'sdist' in sys.argv or 'bdist_wheel' in sys.argv:
             raise ImportError('You must install pyver to create a package')
         else:
-            return 'noversion'
+            return '0.0.0'
     version, version_info = pyver.get_version(pkg="datalake_ingester",
                                               public=True)
     return version