fix(llmobs): fix content arg extraction for vertex ai integration [backport 2.18]

[github-actions]

Backport 67a5a9c from #12034 to 2.18.

In MLOS-42 a customer was experiencing the following error:

ddtrace.internal.utils.ArgumentError: contents (at position 0)

where the content argument was not being extracted properly from the list of keyword arguments inputted into the chat.send_message method.

This is because in the Vertex AI integration, we look for the contents keyword argument. However, the content field is titled content in the send_message method and contents in the generate_content method, so it is necessary to differentiate between these two cases.

This PR is a small fix that corrects this error by differentiating between chat and completion requests in order to extract either content or contents respectively.

Testing

Automatic Testing

I edited some of the currently existing tests to use the keyword argument extraction rather than the positional argument extraction to get the content in order to confirm that this fix resolves the error.

Manual Testing

Running the following code reproduced the error; furthermore, I confirmed that with this fix, the error is no longer present and the request completes successfully.

llm = GenerativeModel("gemini-1.5-flash")
chat = llm.start_chat()
resp = chat.send_message(content="hello")

I also verified that the following code which uses the generate_content method is not impacted (continues to work as before) as a result of this fix.

llm = GenerativeModel("gemini-1.5-flash")
resp = llm.generate_content(contents="hello")

Checklist

• PR author has checked that all the criteria below are met
• The PR description includes an overview of the change
• The PR description articulates the motivation for the change
• The change includes tests OR the PR description describes a testing strategy
• The PR description notes risks associated with the change, if any
• Newly-added code is easy to change
• The change follows the library release note guidelines
• The change includes or references documentation updates if necessary
• Backport labels are set (if applicable)

Reviewer Checklist

• Reviewer has checked that all the criteria below are met
• Title is accurate
• All changes are related to the pull request's stated goal
• Avoids breaking API changes
• Testing strategy adequately addresses listed risks
• Newly-added code is easy to change
• Release note makes sense to a user of the library
• If necessary, author has acknowledged and discussed the performance implications of this PR as reported in the benchmarks PR comment
• Backport labels are set in a manner that is consistent with the release branch maintenance policy

[datadog-dd-trace-py-rkomorn]

Datadog Report

Branch report: backport-12034-to-2.18
Commit report: a35c150
Test service: dd-trace-py

✅ 0 Failed, 1468 Passed, 0 Skipped, 25m 35.23s Total duration (12m 36.13s time saved)

[pr-commenter]

Benchmarks

Benchmark execution time: 2025-01-24 16:18:53

Comparing candidate commit a35c150 in PR branch backport-12034-to-2.18 with baseline commit 0e3612f in branch 2.18.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 394 metrics, 2 unstable metrics.

[github-actions]

CODEOWNERS have been resolved as:

releasenotes/notes/fix-vertexai-content-extraction-b216207bd8192e5f.yaml  @DataDog/apm-python
ddtrace/contrib/internal/vertexai/_utils.py                             @DataDog/ml-observability
ddtrace/contrib/internal/vertexai/patch.py                              @DataDog/ml-observability
ddtrace/llmobs/_integrations/vertexai.py                                @DataDog/ml-observability
tests/contrib/vertexai/test_vertexai.py                                 @DataDog/ml-observability
tests/contrib/vertexai/test_vertexai_llmobs.py                          @DataDog/ml-observability