[Chore] Put date into test index and collection names (#399)

## Problem

Sometimes indexes and collections aren't cleaned up properly after tests
run. It's difficult to clean up orphaned resources without potentially
impacting resources associated with tests that are currently running.

## Solution

Modify generated names to include a timestamp. This will pave the way
for automatically deleting orphaned resources after a certain period of
time.

## Type of Change

- [ ] Bug fix (non-breaking change which fixes an issue)
- [ ] New feature (non-breaking change which adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to not work as expected)
- [ ] This change requires a documentation update
- [x] Infrastructure change (CI configs, etc)
- [ ] Non-code change (docs, etc)
- [ ] None of the above: (explain here)

## Test Plan

Describe specific steps for validating this change.

Author: jhamon

.github/workflows/lint.yaml

@@ -1,5 +1,5 @@
 name: "Lint"
-on: [push, pull_request]
+on: [pull_request]
 
 jobs:
   lint:

scripts/create.py

@@ -1,6 +1,8 @@
 import os
+import re
 import random
 import string
+from datetime import datetime
 from pinecone import Pinecone
 
 
@@ -20,9 +22,44 @@ def write_gh_output(name, value):
         print(f"{name}={value}", file=fh)
 
 
+def generate_index_name(test_name: str) -> str:
+    github_actor = os.getenv("GITHUB_ACTOR", None)
+    user = os.getenv("USER", None)
+    index_owner = github_actor or user
+
+    formatted_date = datetime.now().strftime("%Y%m%d-%H%M%S%f")[:-3]
+
+    github_job = os.getenv("GITHUB_JOB", None)
+
+    if test_name.startswith("test_"):
+        test_name = test_name[5:]
+
+    # Remove trailing underscore, if any
+    if test_name.endswith("_"):
+        test_name = test_name[:-1]
+
+    name_parts = [index_owner, formatted_date, github_job, test_name]
+    index_name = "-".join([x for x in name_parts if x is not None])
+
+    # Remove invalid characters
+    replace_with_hyphen = re.compile(r"[\[\(_,\s]")
+    index_name = re.sub(replace_with_hyphen, "-", index_name)
+    replace_with_empty = re.compile(r"[\]\)\.]")
+    index_name = re.sub(replace_with_empty, "", index_name)
+
+    max_length = 45
+    index_name = index_name[:max_length]
+
+    # Trim final character if it is not alphanumeric
+    if index_name.endswith("_") or index_name.endswith("-"):
+        index_name = index_name[:-1]
+
+    return index_name.lower()
+
+
 def main():
     pc = Pinecone(api_key=read_env_var("PINECONE_API_KEY"))
-    index_name = read_env_var("NAME_PREFIX") + random_string(20)
+    index_name = generate_index_name(read_env_var("NAME_PREFIX") + random_string(20))
     pc.create_index(
         name=index_name,
         metric=read_env_var("METRIC"),

tests/integration/control/pod/conftest.py

@@ -1,9 +1,8 @@
 import pytest
 import random
-import string
 import time
 from pinecone import Pinecone, PodSpec
-from ...helpers import generate_index_name, get_environment_var
+from ...helpers import generate_index_name, generate_collection_name, get_environment_var
 
 
 @pytest.fixture()
@@ -69,17 +68,13 @@ def index_exists(index_name, client):
     return index_name in client.list_indexes().names()
 
 
-def random_string():
-    return "".join(random.choice(string.ascii_lowercase) for i in range(10))
-
-
 @pytest.fixture(scope="session")
 def reusable_collection():
     pc = Pinecone(
         api_key=get_environment_var("PINECONE_API_KEY"),
         additional_headers={"sdk-test-suite": "pinecone-python-client"},
     )
-    index_name = "temp-index-" + random_string()
+    index_name = generate_index_name("temp-index")
     dimension = int(get_environment_var("DIMENSION"))
     print(f"Creating index {index_name} to prepare a collection...")
     pc.create_index(
@@ -99,7 +94,7 @@ def reusable_collection():
     index = pc.Index(index_name)
     index.upsert(vectors=vectors)
 
-    collection_name = "reused-coll-" + random_string()
+    collection_name = generate_collection_name("reused-coll")
     pc.create_collection(name=collection_name, source=index_name)
 
     time_waited = 0

tests/integration/control/pod/test_collections.py

@@ -1,12 +1,8 @@
-import string
 import random
 import pytest
 import time
 from pinecone import PodSpec
-
-
-def random_string():
-    return "".join(random.choice(string.ascii_lowercase) for i in range(10))
+from ...helpers import generate_index_name, generate_collection_name
 
 
 class TestCollectionsHappyPath:
@@ -18,7 +14,7 @@ def test_index_to_collection_to_index_happy_path(
         vectors = [(str(i), random_vector()) for i in range(num_vectors)]
         index.upsert(vectors=vectors)
 
-        collection_name = "coll1-" + random_string()
+        collection_name = generate_collection_name("coll1")
         client.create_collection(name=collection_name, source=ready_index)
         desc = client.describe_collection(collection_name)
         assert desc["name"] == collection_name
@@ -51,7 +47,7 @@ def test_index_to_collection_to_index_happy_path(
         assert desc["size"] > 0
 
         # Create index from collection
-        index_name = "index-from-collection-" + collection_name
+        index_name = generate_index_name("index-from-collection-" + collection_name)
         print(f"Creating index {index_name} from collection {collection_name}...")
         client.create_index(
             name=index_name,
@@ -91,7 +87,7 @@ def test_create_index_with_different_metric_from_orig_index(
         metrics = ["cosine", "euclidean", "dotproduct"]
         target_metric = random.choice([x for x in metrics if x != metric])
 
-        index_name = "from-coll-" + random_string()
+        index_name = generate_index_name("from-" + reusable_collection)
         client.create_index(
             name=index_name,
             dimension=dimension,

tests/integration/control/pod/test_collections_errors.py

@@ -1,19 +1,15 @@
-import string
 import random
 import pytest
 from pinecone import PodSpec
-
-
-def random_string():
-    return "".join(random.choice(string.ascii_lowercase) for i in range(10))
+from ...helpers import generate_collection_name, generate_index_name, random_string
 
 
 class TestCollectionErrorCases:
     def test_create_index_with_nonexistent_source_collection(
         self, client, dimension, metric, environment
     ):
         with pytest.raises(Exception) as e:
-            index_name = "from-nonexistent-coll-" + random_string()
+            index_name = generate_index_name("from-nonexistent-coll-" + random_string(10))
             client.create_index(
                 name=index_name,
                 dimension=dimension,
@@ -43,7 +39,7 @@ def test_create_index_in_mismatched_environment(
         target_env = random.choice([x for x in envs if x != environment])
 
         with pytest.raises(Exception) as e:
-            index_name = "from-coll-" + random_string()
+            index_name = generate_index_name("from-coll-" + random_string(10))
             client.create_index(
                 name=index_name,
                 dimension=dimension,
@@ -59,7 +55,7 @@ def test_create_index_with_mismatched_dimension(
     ):
         with pytest.raises(Exception) as e:
             client.create_index(
-                name="from-coll-" + random_string(),
+                name=generate_index_name("from-coll-" + random_string(10)),
                 dimension=dimension + 1,
                 metric=metric,
                 spec=PodSpec(environment=environment, source_collection=reusable_collection),
@@ -91,13 +87,13 @@ def test_create_index_with_mismatched_dimension(
     #     assert 'Source collection is not ready' in str(e.value)
 
     def test_create_collection_from_not_ready_index(self, client, notready_index):
-        name = "coll3-" + random_string()
+        name = generate_collection_name("coll3")
         with pytest.raises(Exception) as e:
             client.create_collection(name, notready_index)
         assert "Source index is not ready" in str(e.value)
 
     def test_create_collection_with_invalid_index(self, client):
-        name = "coll4-" + random_string()
+        name = generate_collection_name("coll4")
         with pytest.raises(Exception) as e:
             client.create_collection(name, "invalid_index")
         assert "Resource invalid_index not found" in str(e.value)

tests/integration/data/conftest.py

@@ -2,7 +2,7 @@
 import os
 import time
 import json
-from ..helpers import get_environment_var, random_string
+from ..helpers import get_environment_var, random_string, generate_index_name
 from .seed import setup_data, setup_list_data, setup_weird_ids_data
 
 # Test matrix needs to consider the following dimensions:
@@ -57,7 +57,7 @@ def spec():
 
 @pytest.fixture(scope="session")
 def index_name():
-    return "dataplane-" + random_string(20)
+    return generate_index_name("dataplane")
 
 
 @pytest.fixture(scope="session")

tests/integration/helpers/__init__.py

@@ -3,6 +3,7 @@
     get_environment_var,
     random_string,
     generate_index_name,
+    generate_collection_name,
     poll_stats_for_namespace,
     poll_fetch_for_ids_in_namespace,
 )

tests/integration/helpers/helpers.py

@@ -4,26 +4,34 @@
 import random
 import string
 from typing import Any
+from datetime import datetime
 
 
 def random_string(length):
     return "".join(random.choice(string.ascii_lowercase) for i in range(length))
 
 
-def generate_index_name(test_name: str) -> str:
-    buildNumber = os.getenv("GITHUB_BUILD_NUMBER", None)
+def generate_collection_name(label):
+    return generate_index_name(label)
 
-    if test_name.startswith("test_"):
-        test_name = test_name[5:]
 
-    # Trim name length to save space for other info in name
-    test_name = test_name[:20]
+def generate_index_name(label: str) -> str:
+    github_actor = os.getenv("GITHUB_ACTOR", None)
+    user = os.getenv("USER", None)
+    index_owner = github_actor or user
+
+    formatted_date = datetime.now().strftime("%Y%m%d-%H%M%S%f")[:-3]
+
+    github_job = os.getenv("GITHUB_JOB", None)
+
+    if label.startswith("test_"):
+        label = label[5:]
 
     # Remove trailing underscore, if any
-    if test_name.endswith("_"):
-        test_name = test_name[:-1]
+    if label.endswith("_"):
+        label = label[:-1]
 
-    name_parts = [buildNumber, test_name, random_string(45)]
+    name_parts = [index_owner, formatted_date, github_job, label]
     index_name = "-".join([x for x in name_parts if x is not None])
 
     # Remove invalid characters
@@ -36,8 +44,8 @@ def generate_index_name(test_name: str) -> str:
     index_name = index_name[:max_length]
 
     # Trim final character if it is not alphanumeric
-    if test_name.endswith("_") or test_name.endswith("-"):
-        test_name = test_name[:-1]
+    if index_name.endswith("_") or index_name.endswith("-"):
+        index_name = index_name[:-1]
 
     return index_name.lower()