Fix openai streaming wrong span bug

[hmstepanek]

Overview

To reproduce:

import newrelic.agent

@newrelic.agent.background_task()
def streamingOpenAI(prompt):
    import openai

    print(f'prompt with openai: {prompt}')

    response = openai.chat.completions.create(
        model='gpt-3.5-turbo',
        messages=[
            {'role': 'user', 'content': prompt}
        ],
        temperature=0,
        stream=True,
    )

    message = ""
    i = 1
    for chunk in response:
        if 'choices' in chunk:
            i+=1
            for choice in chunk['choices']:
                if 'delta' in choice and 'content' in choice['delta']:
                    message += choice['delta']['content']

    print(f'output ({i}): {message}\n\n')

if __name__ == "__main__":
    # Enable New Relic Python agent
    newrelic.agent.initialize('newrelic.ini')
    newrelic.agent.register_application(timeout=10)

    prompt = "What is 2 + 2?"

    streamingOpenAI(prompt)
    # Allow the New Relic agent to send final messages as part of shutdown
    # The agent by default can send data up to a minute later
    newrelic.agent.shutdown_agent(60)

    print("Agent run finished!")

When you click on the AI details view of the openai span openai create, previously the details would error and not show up. After this fix they do.

[github-actions[bot]]

🦙 MegaLinter status: ❌ ERROR

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
❌ PYTHON	bandit	19		1	10.93s
✅ PYTHON	black	71	6	0	8.71s
❌ PYTHON	flake8	71		9	4.4s
✅ PYTHON	isort	71	9	0	0.52s
❌ PYTHON	pylint	71		152	33.99s

See detailed report in MegaLinter reports _Set VALIDATE_ALL_CODEBASE: true in mega-linter.yml to validate all sources, not only the diff_

_MegaLinter is graciously provided by _