Adds a requirements.txt file to the course directory with dependencies (currently tied to version),and pytest integration tests to test the code works.

Ideally, requirements.txt will be versionless to test for breaking changes or deprecation notices on current versions and notify of any changes.

For the remote repos (llm-chatbot-python, etc), these can be run via a scheduled github action to regularly test on the current version.

Additional Updates

• Update langchain-community imports to langchain-neo4j in contained code
• Added example tests for any python code used in the course
• Added standard naming convention for the OPENAI_API_KEY, loaded from .env file

Updates to imports:

OPENAI_API_KEY variable

This will establish a convention across all courses that is in line with other platforms (deeplearning, DataCamp)

Replace:

llm = OpenAI(openai_api_key="sk-...")

With:

import os 

OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
llm = OpenAI(openai_api_key=OPENAI_API_KEY)

TODO:

• All python scripts should have a test or be removed from the repo
• Add pytest integration tests to course repos:
  • llm-chatbot-python
  • llm-knowledge-graph-construction
  • llm-vectors-unstructured
  • genai-workshop
  • genai-workshop-graphrag
• Prefer langchain-neo4j changes from Updates courses to use the LangChain Neo4j package #284