@codilore/llm 1.15.13

package/src/protocols/openai-chat.ts

@@ -0,0 +1,431 @@
+import { Effect, Schema } from "effect"
+import { Route } from "../route/client"
+import { Auth } from "../route/auth"
+import { Endpoint } from "../route/endpoint"
+import { HttpTransport } from "../route/transport"
+import { Protocol } from "../route/protocol"
+import {
+  LLMEvent,
+  Usage,
+  type FinishReason,
+  type LLMRequest,
+  type ReasoningPart,
+  type TextPart,
+  type ToolCallPart,
+  type ToolDefinition,
+} from "../schema"
+import { isRecord, JsonObject, optionalArray, optionalNull, ProviderShared } from "./shared"
+import { OpenAIOptions } from "./utils/openai-options"
+import { Lifecycle } from "./utils/lifecycle"
+import { ToolStream } from "./utils/tool-stream"
+
+const ADAPTER = "openai-chat"
+export const DEFAULT_BASE_URL = "https://api.openai.com/v1"
+export const PATH = "/chat/completions"
+
+// =============================================================================
+// Request Body Schema
+// =============================================================================
+// The body schema is the provider-native JSON body. `fromRequest` below builds
+// this shape from the common `LLMRequest`, then `Route.make` validates and
+// JSON-encodes it before transport.
+const OpenAIChatFunction = Schema.Struct({
+  name: Schema.String,
+  description: Schema.String,
+  parameters: JsonObject,
+})
+
+const OpenAIChatTool = Schema.Struct({
+  type: Schema.tag("function"),
+  function: OpenAIChatFunction,
+})
+type OpenAIChatTool = Schema.Schema.Type<typeof OpenAIChatTool>
+
+const OpenAIChatAssistantToolCall = Schema.Struct({
+  id: Schema.String,
+  type: Schema.tag("function"),
+  function: Schema.Struct({
+    name: Schema.String,
+    arguments: Schema.String,
+  }),
+})
+type OpenAIChatAssistantToolCall = Schema.Schema.Type<typeof OpenAIChatAssistantToolCall>
+
+const OpenAIChatMessage = Schema.Union([
+  Schema.Struct({ role: Schema.Literal("system"), content: Schema.String }),
+  Schema.Struct({ role: Schema.Literal("user"), content: Schema.String }),
+  Schema.Struct({
+    role: Schema.Literal("assistant"),
+    content: Schema.NullOr(Schema.String),
+    tool_calls: optionalArray(OpenAIChatAssistantToolCall),
+    reasoning_content: Schema.optional(Schema.String),
+  }),
+  Schema.Struct({ role: Schema.Literal("tool"), tool_call_id: Schema.String, content: Schema.String }),
+]).pipe(Schema.toTaggedUnion("role"))
+type OpenAIChatMessage = Schema.Schema.Type<typeof OpenAIChatMessage>
+
+const OpenAIChatToolChoice = Schema.Union([
+  Schema.Literals(["auto", "none", "required"]),
+  Schema.Struct({
+    type: Schema.tag("function"),
+    function: Schema.Struct({ name: Schema.String }),
+  }),
+])
+
+export const bodyFields = {
+  model: Schema.String,
+  messages: Schema.Array(OpenAIChatMessage),
+  tools: optionalArray(OpenAIChatTool),
+  tool_choice: Schema.optional(OpenAIChatToolChoice),
+  stream: Schema.Literal(true),
+  stream_options: Schema.optional(Schema.Struct({ include_usage: Schema.Boolean })),
+  store: Schema.optional(Schema.Boolean),
+  reasoning_effort: Schema.optional(OpenAIOptions.OpenAIReasoningEffort),
+  max_tokens: Schema.optional(Schema.Number),
+  temperature: Schema.optional(Schema.Number),
+  top_p: Schema.optional(Schema.Number),
+  frequency_penalty: Schema.optional(Schema.Number),
+  presence_penalty: Schema.optional(Schema.Number),
+  seed: Schema.optional(Schema.Number),
+  stop: optionalArray(Schema.String),
+}
+const OpenAIChatBody = Schema.Struct(bodyFields)
+export type OpenAIChatBody = Schema.Schema.Type<typeof OpenAIChatBody>
+
+// =============================================================================
+// Streaming Event Schema
+// =============================================================================
+// The event schema is one decoded SSE `data:` payload. `Framing.sse` splits the
+// byte stream into strings, then `Protocol.jsonEvent` decodes each string into
+// this provider-native event shape.
+const OpenAIChatUsage = Schema.Struct({
+  prompt_tokens: Schema.optional(Schema.Number),
+  completion_tokens: Schema.optional(Schema.Number),
+  total_tokens: Schema.optional(Schema.Number),
+  prompt_tokens_details: optionalNull(
+    Schema.Struct({
+      cached_tokens: Schema.optional(Schema.Number),
+    }),
+  ),
+  completion_tokens_details: optionalNull(
+    Schema.Struct({
+      reasoning_tokens: Schema.optional(Schema.Number),
+    }),
+  ),
+})
+
+const OpenAIChatToolCallDeltaFunction = Schema.Struct({
+  name: optionalNull(Schema.String),
+  arguments: optionalNull(Schema.String),
+})
+
+const OpenAIChatToolCallDelta = Schema.Struct({
+  index: Schema.Number,
+  id: optionalNull(Schema.String),
+  function: optionalNull(OpenAIChatToolCallDeltaFunction),
+})
+type OpenAIChatToolCallDelta = Schema.Schema.Type<typeof OpenAIChatToolCallDelta>
+
+const OpenAIChatDelta = Schema.Struct({
+  content: optionalNull(Schema.String),
+  reasoning_content: optionalNull(Schema.String),
+  tool_calls: optionalNull(Schema.Array(OpenAIChatToolCallDelta)),
+})
+
+const OpenAIChatChoice = Schema.Struct({
+  delta: optionalNull(OpenAIChatDelta),
+  finish_reason: optionalNull(Schema.String),
+})
+
+const OpenAIChatEvent = Schema.Struct({
+  choices: Schema.Array(OpenAIChatChoice),
+  usage: optionalNull(OpenAIChatUsage),
+})
+type OpenAIChatEvent = Schema.Schema.Type<typeof OpenAIChatEvent>
+type OpenAIChatRequestMessage = LLMRequest["messages"][number]
+
+interface ParserState {
+  readonly tools: ToolStream.State<number>
+  readonly toolCallEvents: ReadonlyArray<LLMEvent>
+  readonly usage?: Usage
+  readonly finishReason?: FinishReason
+  readonly lifecycle: Lifecycle.State
+}
+
+const invalid = ProviderShared.invalidRequest
+
+// =============================================================================
+// Request Lowering
+// =============================================================================
+// Lowering is the only place that knows how common LLM messages map onto the
+// OpenAI Chat wire format. Keep provider quirks here instead of leaking native
+// fields into `LLMRequest`.
+const lowerTool = (tool: ToolDefinition): OpenAIChatTool => ({
+  type: "function",
+  function: {
+    name: tool.name,
+    description: tool.description,
+    parameters: tool.inputSchema,
+  },
+})
+
+const lowerToolChoice = (toolChoice: NonNullable<LLMRequest["toolChoice"]>) =>
+  ProviderShared.matchToolChoice("OpenAI Chat", toolChoice, {
+    auto: () => "auto" as const,
+    none: () => "none" as const,
+    required: () => "required" as const,
+    tool: (name) => ({ type: "function" as const, function: { name } }),
+  })
+
+const lowerToolCall = (part: ToolCallPart): OpenAIChatAssistantToolCall => ({
+  id: part.id,
+  type: "function",
+  function: {
+    name: part.name,
+    arguments: ProviderShared.encodeJson(part.input),
+  },
+})
+
+const openAICompatibleReasoningContent = (native: unknown) =>
+  isRecord(native) && typeof native.reasoning_content === "string" ? native.reasoning_content : undefined
+
+const lowerUserMessage = Effect.fn("OpenAIChat.lowerUserMessage")(function* (message: OpenAIChatRequestMessage) {
+  const content: TextPart[] = []
+  for (const part of message.content) {
+    if (!ProviderShared.supportsContent(part, ["text"]))
+      return yield* ProviderShared.unsupportedContent("OpenAI Chat", "user", ["text"])
+    content.push(part)
+  }
+  return { role: "user" as const, content: ProviderShared.joinText(content) }
+})
+
+const lowerAssistantMessage = Effect.fn("OpenAIChat.lowerAssistantMessage")(function* (
+  message: OpenAIChatRequestMessage,
+) {
+  const content: TextPart[] = []
+  const reasoning: ReasoningPart[] = []
+  const toolCalls: OpenAIChatAssistantToolCall[] = []
+  for (const part of message.content) {
+    if (!ProviderShared.supportsContent(part, ["text", "reasoning", "tool-call"]))
+      return yield* ProviderShared.unsupportedContent("OpenAI Chat", "assistant", ["text", "reasoning", "tool-call"])
+    if (part.type === "text") {
+      content.push(part)
+      continue
+    }
+    if (part.type === "reasoning") {
+      reasoning.push(part)
+      continue
+    }
+    if (part.type === "tool-call") {
+      toolCalls.push(lowerToolCall(part))
+      continue
+    }
+  }
+  return {
+    role: "assistant" as const,
+    content: content.length === 0 ? null : ProviderShared.joinText(content),
+    tool_calls: toolCalls.length === 0 ? undefined : toolCalls,
+    reasoning_content:
+      reasoning.length > 0
+        ? reasoning.map((part) => part.text).join("")
+        : openAICompatibleReasoningContent(message.native?.openaiCompatible),
+  }
+})
+
+const lowerToolMessages = Effect.fn("OpenAIChat.lowerToolMessages")(function* (message: OpenAIChatRequestMessage) {
+  const messages: OpenAIChatMessage[] = []
+  for (const part of message.content) {
+    if (!ProviderShared.supportsContent(part, ["tool-result"]))
+      return yield* ProviderShared.unsupportedContent("OpenAI Chat", "tool", ["tool-result"])
+    messages.push({ role: "tool", tool_call_id: part.id, content: ProviderShared.toolResultText(part) })
+  }
+  return messages
+})
+
+const lowerMessage = Effect.fn("OpenAIChat.lowerMessage")(function* (message: OpenAIChatRequestMessage) {
+  if (message.role === "user") return [yield* lowerUserMessage(message)]
+  if (message.role === "assistant") return [yield* lowerAssistantMessage(message)]
+  return yield* lowerToolMessages(message)
+})
+
+const lowerMessages = Effect.fn("OpenAIChat.lowerMessages")(function* (request: LLMRequest) {
+  const system: OpenAIChatMessage[] =
+    request.system.length === 0 ? [] : [{ role: "system", content: ProviderShared.joinText(request.system) }]
+  const messages = [...system]
+  for (const message of request.messages) {
+    if (message.role === "system") {
+      const part = yield* ProviderShared.wrappedSystemUpdate("OpenAI Chat", message)
+      const previous = messages.at(-1)
+      if (previous?.role === "user")
+        messages[messages.length - 1] = { role: "user", content: `${previous.content}\n${part.text}` }
+      else messages.push({ role: "user", content: part.text })
+      continue
+    }
+    messages.push(...(yield* lowerMessage(message)))
+  }
+  return messages
+})
+
+const lowerOptions = Effect.fn("OpenAIChat.lowerOptions")(function* (request: LLMRequest) {
+  const store = OpenAIOptions.store(request)
+  const reasoningEffort = OpenAIOptions.reasoningEffort(request)
+  if (reasoningEffort && !OpenAIOptions.isReasoningEffort(reasoningEffort))
+    return yield* invalid(`OpenAI Chat does not support reasoning effort ${reasoningEffort}`)
+  return {
+    ...(store !== undefined ? { store } : {}),
+    ...(reasoningEffort ? { reasoning_effort: reasoningEffort } : {}),
+  }
+})
+
+const fromRequest = Effect.fn("OpenAIChat.fromRequest")(function* (request: LLMRequest) {
+  // `fromRequest` returns the provider body only. Endpoint, auth, framing,
+  // validation, and HTTP execution are composed by `Route.make`.
+  const generation = request.generation
+  return {
+    model: request.model.id,
+    messages: yield* lowerMessages(request),
+    tools: request.tools.length === 0 ? undefined : request.tools.map(lowerTool),
+    tool_choice: request.toolChoice ? yield* lowerToolChoice(request.toolChoice) : undefined,
+    stream: true as const,
+    stream_options: { include_usage: true },
+    max_tokens: generation?.maxTokens,
+    temperature: generation?.temperature,
+    top_p: generation?.topP,
+    frequency_penalty: generation?.frequencyPenalty,
+    presence_penalty: generation?.presencePenalty,
+    seed: generation?.seed,
+    stop: generation?.stop,
+    ...(yield* lowerOptions(request)),
+  }
+})
+
+// =============================================================================
+// Stream Parsing
+// =============================================================================
+// Streaming parsers are small state machines: every event returns a new state
+// plus the common `LLMEvent`s produced by that event. Tool calls are accumulated
+// because OpenAI streams JSON arguments across multiple deltas.
+const mapFinishReason = (reason: string | null | undefined): FinishReason => {
+  if (reason === "stop") return "stop"
+  if (reason === "length") return "length"
+  if (reason === "content_filter") return "content-filter"
+  if (reason === "function_call" || reason === "tool_calls") return "tool-calls"
+  return "unknown"
+}
+
+// OpenAI Chat reports `prompt_tokens` (inclusive total) with a
+// `cached_tokens` subset, and `completion_tokens` (inclusive total) with
+// a `reasoning_tokens` subset. We pass the inclusive totals through and
+// derive the non-cached breakdown so the `LLM.Usage` contract is
+// satisfied on both sides.
+const mapUsage = (usage: OpenAIChatEvent["usage"]): Usage | undefined => {
+  if (!usage) return undefined
+  const cached = usage.prompt_tokens_details?.cached_tokens
+  const reasoning = usage.completion_tokens_details?.reasoning_tokens
+  const nonCached = ProviderShared.subtractTokens(usage.prompt_tokens, cached)
+  return new Usage({
+    inputTokens: usage.prompt_tokens,
+    outputTokens: usage.completion_tokens,
+    nonCachedInputTokens: nonCached,
+    cacheReadInputTokens: cached,
+    reasoningTokens: reasoning,
+    totalTokens: ProviderShared.totalTokens(usage.prompt_tokens, usage.completion_tokens, usage.total_tokens),
+    providerMetadata: { openai: usage },
+  })
+}
+
+const step = (state: ParserState, event: OpenAIChatEvent) =>
+  Effect.gen(function* () {
+    const events: LLMEvent[] = []
+    const usage = mapUsage(event.usage) ?? state.usage
+    const choice = event.choices[0]
+    const finishReason = choice?.finish_reason ? mapFinishReason(choice.finish_reason) : state.finishReason
+    const delta = choice?.delta
+    const toolDeltas = delta?.tool_calls ?? []
+    let tools = state.tools
+
+    let lifecycle = state.lifecycle
+
+    if (delta?.reasoning_content)
+      lifecycle = Lifecycle.reasoningDelta(lifecycle, events, "reasoning-0", delta.reasoning_content)
+
+    if (delta?.content) lifecycle = Lifecycle.textDelta(lifecycle, events, "text-0", delta.content)
+
+    for (const tool of toolDeltas) {
+      const result = ToolStream.appendOrStart(
+        ADAPTER,
+        tools,
+        tool.index,
+        { id: tool.id ?? undefined, name: tool.function?.name ?? undefined, text: tool.function?.arguments ?? "" },
+        "OpenAI Chat tool call delta is missing id or name",
+      )
+      if (ToolStream.isError(result)) return yield* result
+      tools = result.tools
+      if (result.events.length) lifecycle = Lifecycle.stepStart(lifecycle, events)
+      events.push(...result.events)
+    }
+
+    // Finalize accumulated tool inputs eagerly when finish_reason arrives so
+    // JSON parse failures fail the stream at the boundary rather than at halt.
+    const finished =
+      finishReason !== undefined && state.finishReason === undefined && Object.keys(tools).length > 0
+        ? yield* ToolStream.finishAll(ADAPTER, tools)
+        : undefined
+
+    return [
+      {
+        tools: finished?.tools ?? tools,
+        toolCallEvents: finished?.events ?? state.toolCallEvents,
+        usage,
+        finishReason,
+        lifecycle,
+      },
+      events,
+    ] as const
+  })
+
+const finishEvents = (state: ParserState): ReadonlyArray<LLMEvent> => {
+  const events: LLMEvent[] = []
+  const hasToolCalls = state.toolCallEvents.length > 0
+  const reason = state.finishReason === "stop" && hasToolCalls ? "tool-calls" : state.finishReason
+  const lifecycle = state.toolCallEvents.length ? Lifecycle.stepStart(state.lifecycle, events) : state.lifecycle
+  events.push(...state.toolCallEvents)
+  if (reason) Lifecycle.finish(lifecycle, events, { reason, usage: state.usage })
+  return events
+}
+
+// =============================================================================
+// Protocol And OpenAI Route
+// =============================================================================
+/**
+ * The OpenAI Chat protocol — request body construction, body schema, and the
+ * streaming-event state machine. Reused by every route that speaks OpenAI Chat
+ * over HTTP+SSE: native OpenAI, DeepSeek, TogetherAI, Cerebras, Baseten,
+ * Fireworks, DeepInfra, and (once added) Azure OpenAI Chat.
+ */
+export const protocol = Protocol.make({
+  id: ADAPTER,
+  body: {
+    schema: OpenAIChatBody,
+    from: fromRequest,
+  },
+  stream: {
+    event: Protocol.jsonEvent(OpenAIChatEvent),
+    initial: () => ({ tools: ToolStream.empty<number>(), toolCallEvents: [], lifecycle: Lifecycle.initial() }),
+    step,
+    onHalt: finishEvents,
+  },
+})
+
+export const httpTransport = HttpTransport.sseJson.with<OpenAIChatBody>()
+
+export const route = Route.make({
+  id: ADAPTER,
+  provider: "openai",
+  protocol,
+  endpoint: Endpoint.path(PATH, { baseURL: DEFAULT_BASE_URL }),
+  auth: Auth.none,
+  transport: httpTransport,
+})
+
+export * as OpenAIChat from "./openai-chat"

package/src/protocols/openai-compatible-chat.ts

@@ -0,0 +1,24 @@
+import { Route, type RouteRoutedModelInput } from "../route/client"
+import { Endpoint } from "../route/endpoint"
+import { Framing } from "../route/framing"
+import * as OpenAIChat from "./openai-chat"
+
+const ADAPTER = "openai-compatible-chat"
+
+export type OpenAICompatibleChatModelInput = RouteRoutedModelInput
+
+/**
+ * Route for non-OpenAI providers that expose an OpenAI Chat-compatible
+ * `/chat/completions` endpoint. Reuses `OpenAIChat.protocol` end-to-end and
+ * overrides only the route id so providers can be resolved per-family without
+ * colliding with native OpenAI. Provider helpers configure the route endpoint
+ * before model selection.
+ */
+export const route = Route.make({
+  id: ADAPTER,
+  protocol: OpenAIChat.protocol,
+  endpoint: Endpoint.path("/chat/completions"),
+  framing: Framing.sse,
+})
+
+export * as OpenAICompatibleChat from "./openai-compatible-chat"