revert(qoder): drop the Responses usage plumbing from shared code
The merged Qoder work also rewrote shared translator/handler code so that /v1/responses clients got token usage on response.completed. That changed behaviour for every provider, not just Qoder: proxies saw input tokens rise by the 2000-token context buffer, and the plain token mapping was replaced by one that always adds input_tokens_details. A probe confirms the Qoder benefit does not depend on those edits: the executor's coalescer already emits one include_usage-style finish chunk, so a Claude client receives input_tokens and cache_read_input_tokens with every shared file at its original state. Only the Responses path relies on the shared translator, and that path has no Qoder-owned seam to put it in. Reverts the shared files to their pre-PR state and drops the Responses usage test. The Cline envelope unwrap in nonStreamingHandler.js, which landed after the PR in the same file, is kept.
This commit is contained in:
LLL
@@ -5,7 +5,7 @@
|
||||
import { register } from "../index.js";
|
||||
import { FORMATS } from "../formats.js";
|
||||
import { buildChunk } from "../concerns/chunk.js";
|
||||
import { buildUsage, toResponsesUsage } from "../concerns/usage.js";
|
||||
import { buildUsage } from "../concerns/usage.js";
|
||||
import { fallbackToolCallId } from "../concerns/toolCall.js";
|
||||
import { reasoningDelta, extractReasoningText } from "../concerns/reasoning.js";
|
||||
import { ROLE, OPENAI_BLOCK, RESPONSES_ITEM, OPENAI_FINISH, MODEL_FALLBACK } from "../schema/index.js";
|
||||
@@ -18,13 +18,7 @@ export function openaiToOpenAIResponsesResponse(chunk, state) {
|
||||
if (!chunk) {
|
||||
return flushEvents(state);
|
||||
}
|
||||
|
||||
// Usage riding on the finish chunk (include_usage style, e.g. coalesced Qoder frames):
|
||||
// remember it so response.completed can report tokens even outside stream.js.
|
||||
if (chunk.usage && typeof chunk.usage === "object" && !state.usage) {
|
||||
state.usage = chunk.usage;
|
||||
}
|
||||
|
||||
|
||||
if (!chunk.choices?.length) return [];
|
||||
|
||||
const events = [];
|
||||
@@ -374,19 +368,17 @@ function closeToolCall(state, emit, idx) {
|
||||
function sendCompleted(state, emit) {
|
||||
if (!state.completedSent) {
|
||||
state.completedSent = true;
|
||||
const response = {
|
||||
id: state.responseId,
|
||||
object: "response",
|
||||
created_at: state.created,
|
||||
status: "completed",
|
||||
background: false,
|
||||
error: null
|
||||
};
|
||||
// Carry provider usage (recorded by stream.js or from the finish chunk itself) in the
|
||||
// Responses shape; proxies such as sub2api/Codex read tokens only from here.
|
||||
const usage = toResponsesUsage(state.usage);
|
||||
if (usage) response.usage = usage;
|
||||
emit("response.completed", { type: "response.completed", response });
|
||||
emit("response.completed", {
|
||||
type: "response.completed",
|
||||
response: {
|
||||
id: state.responseId,
|
||||
object: "response",
|
||||
created_at: state.created,
|
||||
status: "completed",
|
||||
background: false,
|
||||
error: null
|
||||
}
|
||||
});
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@@ -67,50 +67,48 @@ function stopTextBlock(state, results) {
|
||||
state.textBlockStarted = false;
|
||||
}
|
||||
|
||||
function recordOpenAIUsage(chunk, state) {
|
||||
if (!chunk?.usage || typeof chunk.usage !== "object") return;
|
||||
|
||||
const promptTokens = typeof chunk.usage.prompt_tokens === "number" ? chunk.usage.prompt_tokens : 0;
|
||||
const outputTokens = typeof chunk.usage.completion_tokens === "number" ? chunk.usage.completion_tokens : 0;
|
||||
|
||||
// Extract cache tokens from prompt_tokens_details
|
||||
const cachedTokens = chunk.usage.prompt_tokens_details?.cached_tokens;
|
||||
const cacheCreationTokens = chunk.usage.prompt_tokens_details?.cache_creation_tokens;
|
||||
const cacheReadTokens = typeof cachedTokens === "number" ? cachedTokens : 0;
|
||||
const cacheCreateTokens = typeof cacheCreationTokens === "number" ? cacheCreationTokens : 0;
|
||||
|
||||
// input_tokens = prompt_tokens - cached_tokens - cache_creation_tokens
|
||||
// Because OpenAI's prompt_tokens includes all prompt-side tokens
|
||||
const inputTokens = promptTokens - cacheReadTokens - cacheCreateTokens;
|
||||
|
||||
state.usage = {
|
||||
input_tokens: inputTokens,
|
||||
output_tokens: outputTokens
|
||||
};
|
||||
|
||||
if (cacheReadTokens > 0) {
|
||||
state.usage.cache_read_input_tokens = cacheReadTokens;
|
||||
}
|
||||
if (cacheCreateTokens > 0) {
|
||||
state.usage.cache_creation_input_tokens = cacheCreateTokens;
|
||||
}
|
||||
}
|
||||
|
||||
// Convert OpenAI stream chunk to Claude format
|
||||
export function openaiToClaudeResponse(chunk, state) {
|
||||
if (!chunk) return null;
|
||||
|
||||
// Track usage from OpenAI chunk if available
|
||||
if (chunk.usage && typeof chunk.usage === "object") {
|
||||
recordOpenAIUsage(chunk, state);
|
||||
}
|
||||
|
||||
if (!chunk.choices?.[0]) return null;
|
||||
if (!chunk || !chunk.choices?.[0]) return null;
|
||||
|
||||
const results = [];
|
||||
const choice = chunk.choices[0];
|
||||
const delta = choice.delta;
|
||||
|
||||
// Track usage from OpenAI chunk if available
|
||||
if (chunk.usage && typeof chunk.usage === "object") {
|
||||
const promptTokens = typeof chunk.usage.prompt_tokens === "number" ? chunk.usage.prompt_tokens : 0;
|
||||
const outputTokens = typeof chunk.usage.completion_tokens === "number" ? chunk.usage.completion_tokens : 0;
|
||||
|
||||
// Extract cache tokens from prompt_tokens_details
|
||||
const cachedTokens = chunk.usage.prompt_tokens_details?.cached_tokens;
|
||||
const cacheCreationTokens = chunk.usage.prompt_tokens_details?.cache_creation_tokens;
|
||||
const cacheReadTokens = typeof cachedTokens === "number" ? cachedTokens : 0;
|
||||
const cacheCreateTokens = typeof cacheCreationTokens === "number" ? cacheCreationTokens : 0;
|
||||
|
||||
// input_tokens = prompt_tokens - cached_tokens - cache_creation_tokens
|
||||
// Because OpenAI's prompt_tokens includes all prompt-side tokens
|
||||
const inputTokens = promptTokens - cacheReadTokens - cacheCreateTokens;
|
||||
|
||||
state.usage = {
|
||||
input_tokens: inputTokens,
|
||||
output_tokens: outputTokens
|
||||
};
|
||||
|
||||
// Add cache_read_input_tokens if present
|
||||
if (cacheReadTokens > 0) {
|
||||
state.usage.cache_read_input_tokens = cacheReadTokens;
|
||||
}
|
||||
|
||||
// Add cache_creation_input_tokens if present
|
||||
if (cacheCreateTokens > 0) {
|
||||
state.usage.cache_creation_input_tokens = cacheCreateTokens;
|
||||
}
|
||||
|
||||
// Note: completion_tokens_details.reasoning_tokens is already included in output_tokens
|
||||
// No need to add separately as Claude expects total output_tokens
|
||||
}
|
||||
|
||||
// First chunk - ALWAYS send message_start first
|
||||
if (!state.messageStartSent) {
|
||||
state.messageStartSent = true;
|
||||
@@ -223,9 +221,8 @@ export function openaiToClaudeResponse(chunk, state) {
|
||||
}
|
||||
}
|
||||
|
||||
// Finish (OpenAI puts this on the choice; Qoder often puts it on delta)
|
||||
const finishReason = choice.finish_reason || delta?.finish_reason;
|
||||
if (finishReason) {
|
||||
// Finish
|
||||
if (choice.finish_reason) {
|
||||
stopThinkingBlock(state, results);
|
||||
stopTextBlock(state, results);
|
||||
|
||||
@@ -247,13 +244,13 @@ export function openaiToClaudeResponse(chunk, state) {
|
||||
}
|
||||
|
||||
// Mark finish for later usage injection in stream.js
|
||||
state.finishReason = finishReason;
|
||||
state.finishReason = choice.finish_reason;
|
||||
|
||||
// Use tracked usage (will be estimated in stream.js if not valid)
|
||||
const finalUsage = state.usage || { input_tokens: 0, output_tokens: 0 };
|
||||
results.push({
|
||||
type: "message_delta",
|
||||
delta: { stop_reason: convertFinishReason(finishReason) },
|
||||
delta: { stop_reason: convertFinishReason(choice.finish_reason) },
|
||||
usage: finalUsage
|
||||
});
|
||||
results.push({ type: "message_stop" });
|
||||
|
||||
Reference in New Issue
Block a user