Merge pull request #116 from scaleapi/dwong-cleanup-slices-script

add Slice cleanup script

Author: ardila

scripts/cleanup_slices.py

@@ -0,0 +1,43 @@
+"""
+Cleans up messy test/demo artifacts. 
+- Delete empty Slices with <1k Items
+- Delete Slices with name like "test" (often repeated demo flows)
+"""
+import argparse
+import logging
+
+import nucleus
+
+log = logging.getLogger()
+log.setLevel(logging.INFO)
+
+
+def cleanup_slices(api_key, dataset_id):
+    client = nucleus.NucleusClient(api_key)
+
+    dataset = client.get_dataset(dataset_id)
+
+    log.info("======\nChecking for Slices to delete\n======")
+    for slc_id in dataset.slices:
+        slc = client.get_slice(slc_id)
+        slc_info = slc.info()
+        slc_name = slc_info["name"]
+
+        if any(kw in slc_name.lower() for kw in ["test", "night", "police", "truck"]):
+            log.info("Deleting Slice '%s' because it contains night/police/truck keyword. Slice id: %s",
+                     slc_name, slc_id)
+            client.delete_slice(slc_id)
+
+        if len(slc_info['dataset_items']) < 1000:
+            log.info("Deleting Slice '%s' because it contains fewer than 1000 items. Slice id: %s",
+                     slc_name, slc_id)
+            client.delete_slice(slc_id)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument('api_key', type=str)
+    parser.add_argument('dataset_id', type=str)
+    args = parser.parse_args()
+
+    cleanup_slices(args.api_key, args.dataset_id)