refactor: openai and mistral audio refactored

[Pratham-Mishra04]

Summary

Refactors audio processing code to reduce duplication and improve maintainability across providers by extracting shared functionality into reusable handler functions.

Changes

• Consolidated duplicate code in Mistral's transcription functionality by merging createMistralTranscriptionStreamMultipartBody and createMistralTranscriptionMultipartBody into a single function
• Added a Stream parameter to the Mistral transcription request structure for better control
• Created reusable handler functions in OpenAI provider for speech, speech streaming, transcription, and transcription streaming operations
• Fixed raw request handling in OpenAI's speech endpoint to properly return request data when enabled
• Improved error handling and logging consistency across audio processing functions

Type of change

• Bug fix
• Feature
• Refactor
• Documentation
• Chore/CI

Affected areas

• Core (Go)
• Transports (HTTP)
• Providers/Integrations
• Plugins
• UI (Next.js)
• Docs

How to test

Test the audio processing functionality with both OpenAI and Mistral providers:

# Run all tests
go test ./core/providers/mistral/... ./core/providers/openai/...

# Test specific transcription functionality
go test ./core/providers/mistral -run TestCreateMistralTranscriptionMultipartBody

Breaking changes

• No
• Yes

Related issues

Improves code maintainability and reduces duplication in audio processing code.

Security considerations

No security implications as this is a refactoring of existing functionality.

Checklist

• I added/updated tests where appropriate
• I verified builds succeed (Go and UI)
• I verified the CI pipeline passes locally if applicable

[coderabbitai]

Caution

Review failed

The pull request is closed.

📝 Walkthrough

Summary by CodeRabbit

• New Features

  • Mistral transcription service added support for streaming transcription requests.

• Refactor

  • OpenAI provider implementation restructured with unified request handlers for speech synthesis, transcription, and embedding services. New centralized handler functions manage HTTP request construction, response parsing, error handling, post-response transformation converters, and structured logging across all service endpoints.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

Walkthrough

Mistral transcription gained an optional Stream flag and unified multipart form construction for streaming and non-streaming requests. OpenAI provider logic was refactored to extract centralized handlers for Speech, Transcription, and Embedding (including streaming variants) with shared parameters and optional converter hooks.

Changes

Cohort / File(s)	Summary
Mistral streaming & parsing consolidation core/providers/mistral/types.go, core/providers/mistral/mistral.go, core/providers/mistral/transcription.go, core/providers/mistral/transcription_test.go	Added Stream *bool to MistralTranscriptionRequest. TranscriptionStream now sets Stream = true and uses the unified createMistralTranscriptionMultipartBody. Consolidated parsing into parseTranscriptionFormDataBodyFromRequest which conditionally writes the stream field; updated tests to call the unified body builder.
OpenAI centralized handlers & refactor core/providers/openai/openai.go	Extracted centralized handlers: HandleOpenAISpeechRequest, HandleOpenAISpeechStreamRequest, HandleOpenAITranscriptionRequest, HandleOpenAITranscriptionStreamRequest, HandleOpenAIEmbeddingRequest. Refactored provider methods to delegate to these handlers, threading client, headers, auth, logger, and optional post-request/response converters. Removed or replaced several in-file helper/parsing duplications.
OpenAI transcription multipart helper core/providers/openai/transcription.go	Added internal parseTranscriptionFormDataBodyFromRequest to build multipart form bodies (file field with filename, model, optional language, prompt, response_format, and conditional stream), included multipart imports and provider error handling.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

• Review centralized OpenAI handler signatures and ensure all call sites pass correct clients, headers, auth, and converter hooks.
• Verify streaming behavior and conditional stream field handling in Mistral multipart construction and tests.
• Inspect error propagation and logging changes in streaming paths.

Possibly related PRs

• refactor: openai and mistral audio refactored #1112 — Appears to touch the same Mistral transcription streaming code and the OpenAI transcription/speech refactors; likely closely related.

Suggested reviewers

• akshaydeo
• danpiths

Poem

🐰 I hopped in code, ears all a-twitch,
Stream flags set, multipart stitch by stitch.
OpenAI handlers lined in a row,
Converters twirl, streams start to flow —
A tiny rabbit cheers: nice and slick! 🎉

Pre-merge checks and finishing touches

❌ Failed checks (1 inconclusive)

Check name	Status	Explanation	Resolution
Title check	❓ Inconclusive	The title 'refactor: openai and mistral audio refactored' is partially related to the changeset—it identifies the providers and general area (audio), but is somewhat vague and redundant ('refactored' appears twice).	Revise to be more specific about the main change, e.g., 'refactor: consolidate audio handler functions across openai and mistral' or 'refactor: extract shared audio processing logic into reusable handlers'.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description check	✅ Passed	The PR description covers all key template sections with sufficient detail: clear summary, specific changes listed, type marked, affected areas checked, testing instructions provided, no breaking changes noted, and checklist items completed.
Docstring Coverage	✅ Passed	Docstring coverage is 90.00% which is sufficient. The required threshold is 80.00%.

---

📜 Recent review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between f7fd3bf and e8b50ea.

📒 Files selected for processing (6)

• core/providers/mistral/mistral.go (1 hunks)
• core/providers/mistral/transcription.go (2 hunks)
• core/providers/mistral/transcription_test.go (2 hunks)
• core/providers/mistral/types.go (3 hunks)
• core/providers/openai/openai.go (20 hunks)
• core/providers/openai/transcription.go (2 hunks)

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[Pratham-Mishra04]

• feat: added audio support to chat completions #1114
• feat: add transcription and speech support for Azure #1113
• refactor: openai and mistral audio refactored #1112 👈 (View in Graphite)
• feat: add handling for HTML and empty responses from providers #1105
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[akshaydeo]

Merge activity

• Dec 17, 3:51 PM UTC: A user started a stack merge that includes this pull request via Graphite.
• Dec 17, 3:53 PM UTC: @akshaydeo merged this pull request with Graphite.