fix: extract usage from ResponsesAPI streaming events

[varad-ahirwadkar]

What type of PR is this?
/kind bug

What this PR does / why we need it:
Responses API returns usage under response.usage, which was not previously handled by the streaming parser. This change adds support for that format while keeping compatibility with existing ChatCompletions and vLLM streaming responses.

Which issue(s) this PR fixes:
Fixes #2482

Does this PR introduce a user-facing change?:

NONE

[k8s-ci-robot]

Hi @varad-ahirwadkar. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work.

Tip

We noticed you've done this a few times! Consider joining the org to skip this step and gain /lgtm and other bot rights. We recommend asking approvers on your previous PRs to sponsor you.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[netlify]

✅ Deploy Preview for gateway-api-inference-extension ready!

Name	Link
🔨 Latest commit	107d3e4
🔍 Latest deploy log	https://app.netlify.com/projects/gateway-api-inference-extension/deploys/69bbb2fc8e230700081834f9
😎 Deploy Preview	https://deploy-preview-2573--gateway-api-inference-extension.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[varad-ahirwadkar]

/ok-to-test

[k8s-ci-robot]

@varad-ahirwadkar: Cannot trigger testing until a trusted user reviews the PR and leaves an /ok-to-test message.

Details

In response to this:

/ok-to-test

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[ahg-g]

/ok-to-test

[ahg-g]

Thanks, what kind of testing was done to validate the fix?

[ahg-g]

/assign @zetxqx

[zetxqx]

Looks good, can you add a unit test of this, with some real streaming responses data?

[varad-ahirwadkar]

what kind of testing was done to validate the fix?

I validated the fix by running the TestOpenAIParser_ParseResponse_Streaming tests:

go test -v ./pkg/epp/framework/plugins/requesthandling/parsers/openai/... -run TestOpenAIParser_ParseResponse_Streaming

=== RUN   TestOpenAIParser_ParseResponse_Streaming
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Single_data_chunk_with_usage
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Usage_and_DONE_in_the_same_multi-line_response
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Chunk_without_usage_returns_ParsedResponse_with_nil_usage
=== RUN   TestOpenAIParser_ParseResponse_Streaming/DONE_message_returns_error
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Malformed_JSON_in_stream_(skipped)
--- PASS: TestOpenAIParser_ParseResponse_Streaming (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Single_data_chunk_with_usage (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Usage_and_DONE_in_the_same_multi-line_response (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Chunk_without_usage_returns_ParsedResponse_with_nil_usage (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/DONE_message_returns_error (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Malformed_JSON_in_stream_(skipped) (0.00s)
PASS
ok  	sigs.k8s.io/gateway-api-inference-extension/pkg/epp/framework/plugins/requesthandling/parsers/openai	0.027s

I also ran the full test suite for the OpenAI parser package to ensure there were no regressions:

go test -v ./pkg/epp/framework/plugins/requesthandling/parsers/openai/...

However, when I sent a similar curl request as mentioned in this issue: #2482, but I still got the same results.

[varad-ahirwadkar]

Looks good, can you add a unit test of this, with some real streaming responses data?

Sure, will add a unit test. Thanks

[varad-ahirwadkar]

Hi @zetxqx,
I have added tests, PTAL

# go test -count=1  -v ./pkg/epp/framework/plugins/requesthandling/parsers/openai -run TestOpenAIParser_ParseResponse_Streaming 
=== RUN   TestOpenAIParser_ParseResponse_Streaming
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Single_data_chunk_with_usage
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Usage_and_DONE_in_the_same_multi-line_response
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Chunk_without_usage_returns_ParsedResponse_with_nil_usage
=== RUN   TestOpenAIParser_ParseResponse_Streaming/DONE_message_returns_error
=== RUN   TestOpenAIParser_ParseResponse_Streaming/Malformed_JSON_in_stream_(skipped)
=== RUN   TestOpenAIParser_ParseResponse_Streaming/ResponsesAPI_streaming_with_full_response
=== RUN   TestOpenAIParser_ParseResponse_Streaming/ResponsesAPI_without_response.completed_type_returns_nil
=== RUN   TestOpenAIParser_ParseResponse_Streaming/ResponsesAPI_with_multiple_events_extracts_from_completed
--- PASS: TestOpenAIParser_ParseResponse_Streaming (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Single_data_chunk_with_usage (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Usage_and_DONE_in_the_same_multi-line_response (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Chunk_without_usage_returns_ParsedResponse_with_nil_usage (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/DONE_message_returns_error (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/Malformed_JSON_in_stream_(skipped) (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/ResponsesAPI_streaming_with_full_response (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/ResponsesAPI_without_response.completed_type_returns_nil (0.00s)
    --- PASS: TestOpenAIParser_ParseResponse_Streaming/ResponsesAPI_with_multiple_events_extracts_from_completed (0.00s)
PASS
ok      sigs.k8s.io/gateway-api-inference-extension/pkg/epp/framework/plugins/requesthandling/parsers/openai        0.024s

[ahg-g]

/approve

I will leave the lgtm to @zetxqx

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ahg-g, varad-ahirwadkar

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [ahg-g]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment