examples(vector_databases): migrate PolarDB, AnalyticDB, Hologres, redisjson to openai>=1.0 API

[kiwigitops]

Summary

Four more vector-database notebooks in examples/vector_databases/ still call the pre-v1 OpenAI Python SDK (openai.Embedding.create(...), openai.api_key = ..., dict-style response["data"][0]["embedding"]). The pre-v1 module-level API was removed in openai>=1.0.0 (November 2023), so these notebooks fail immediately on the current SDK with:

AttributeError: module 'openai' has no attribute 'Embedding'

This PR migrates all four to the v1+ OpenAI() client API. No model or parameter changes — pure SDK-surface fixes.

Notebooks touched

• examples/vector_databases/PolarDB/Getting_started_with_PolarDB_and_OpenAI.ipynb
• examples/vector_databases/analyticdb/Getting_started_with_AnalyticDB_and_OpenAI.ipynb
• examples/vector_databases/hologres/Getting_started_with_Hologres_and_OpenAI.ipynb
• examples/vector_databases/redis/redisjson/redisjson.ipynb

Per-notebook changes

For PolarDB / AnalyticDB / Hologres (same shape):

-import openai
+from openai import OpenAI
+client = OpenAI()

 def query_xxx(...):
-    embedded_query = openai.Embedding.create(
-        input=query,
-        model="text-embedding-3-small",
-    )["data"][0]["embedding"]
+    embedded_query = client.embeddings.create(
+        input=query,
+        model="text-embedding-3-small",
+    ).data[0].embedding

For redis/redisjson — client is already used for the Redis connection, so the OpenAI client is named openai_client to avoid the collision:

-import openai
-...
-openai.api_key = os.getenv("OPENAI_API_KEY")
+from openai import OpenAI
+...
+openai_client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))

 def get_vector(text, model="text-embedding-3-small"):
     text = text.replace("\n", " ")
-    return openai.Embedding.create(input = [text], model = model)['data'][0]['embedding']
+    return openai_client.embeddings.create(input=[text], model=model).data[0].embedding

Test plan

• All four notebooks still validate as JSON
• No remaining openai.Embedding.create, openai.api_key, ["data"][0]["embedding"], or other pre-v1 patterns in any code cell
• Same models, same parameters, same overall flow — purely a SDK-surface migration

Related

Follow-up to the same migration pattern from:

• examples(SingleStoreDB): migrate notebook to openai>=1.0 client API #2718 (SingleStoreDB)
• examples(elasticsearch): migrate notebook to openai>=1.0 client API #2719 (Elasticsearch)
• examples(pinecone): migrate GPT-4 retrieval notebook to openai>=1.0 client API #2720 (Pinecone)