mirror of
https://github.com/RooVetGit/Roo-Code.git
synced 2026-09-05 08:10:14 +00:00
fix: remove i18n dependencies from embedders and vector store for worker compatibility
- Remove i18n imports from all embedder implementations (openai, openai-compatible, ollama) - Remove i18n import from qdrant-client vector store - Replace all translation calls with plain English strings - This allows these modules to be used in worker threads where vscode APIs are not available
This commit is contained in:
hannesrudolph
parent
ce0d891777
commit
cb4bb8476f
4 changed files with 23 additions and 69 deletions
|
|
@ -2,7 +2,6 @@ import { ApiHandlerOptions } from "../../../shared/api"
|
|||
import { EmbedderInfo, EmbeddingResponse, IEmbedder } from "../interfaces"
|
||||
import { getModelQueryPrefix } from "../../../shared/embeddingModels"
|
||||
import { MAX_ITEM_TOKENS } from "../constants"
|
||||
import { t } from "../../../i18n"
|
||||
|
||||
/**
|
||||
* Implements the IEmbedder interface using a local Ollama instance.
|
||||
|
|
@ -39,11 +38,7 @@ export class CodeIndexOllamaEmbedder implements IEmbedder {
|
|||
const estimatedTokens = Math.ceil(prefixedText.length / 4)
|
||||
if (estimatedTokens > MAX_ITEM_TOKENS) {
|
||||
console.warn(
|
||||
t("embeddings:textWithPrefixExceedsTokenLimit", {
|
||||
index,
|
||||
estimatedTokens,
|
||||
maxTokens: MAX_ITEM_TOKENS,
|
||||
}),
|
||||
`Text at index ${index} with prefix exceeds token limit (${estimatedTokens} > ${MAX_ITEM_TOKENS})`,
|
||||
)
|
||||
// Return original text if adding prefix would exceed limit
|
||||
return text
|
||||
|
|
@ -67,18 +62,14 @@ export class CodeIndexOllamaEmbedder implements IEmbedder {
|
|||
})
|
||||
|
||||
if (!response.ok) {
|
||||
let errorBody = t("embeddings:ollama.couldNotReadErrorBody")
|
||||
let errorBody = "Could not read error body"
|
||||
try {
|
||||
errorBody = await response.text()
|
||||
} catch (e) {
|
||||
// Ignore error reading body
|
||||
}
|
||||
throw new Error(
|
||||
t("embeddings:ollama.requestFailed", {
|
||||
status: response.status,
|
||||
statusText: response.statusText,
|
||||
errorBody,
|
||||
}),
|
||||
`Ollama request failed with status ${response.status} ${response.statusText}: ${errorBody}`,
|
||||
)
|
||||
}
|
||||
|
||||
|
|
@ -87,7 +78,7 @@ export class CodeIndexOllamaEmbedder implements IEmbedder {
|
|||
// Extract embeddings using 'embeddings' key as requested
|
||||
const embeddings = data.embeddings
|
||||
if (!embeddings || !Array.isArray(embeddings)) {
|
||||
throw new Error(t("embeddings:ollama.invalidResponseStructure"))
|
||||
throw new Error("Invalid response structure: expected 'embeddings' array")
|
||||
}
|
||||
|
||||
return {
|
||||
|
|
@ -97,7 +88,7 @@ export class CodeIndexOllamaEmbedder implements IEmbedder {
|
|||
// Log the original error for debugging purposes
|
||||
console.error("Ollama embedding failed:", error)
|
||||
// Re-throw a more specific error for the caller
|
||||
throw new Error(t("embeddings:ollama.embeddingFailed", { message: error.message }))
|
||||
throw new Error(`Ollama embedding failed: ${error.message}`)
|
||||
}
|
||||
}
|
||||
|
||||
|
|
|
|||
|
|
@ -7,7 +7,6 @@ import {
|
|||
INITIAL_RETRY_DELAY_MS as INITIAL_DELAY_MS,
|
||||
} from "../constants"
|
||||
import { getDefaultModelId, getModelQueryPrefix } from "../../../shared/embeddingModels"
|
||||
import { t } from "../../../i18n"
|
||||
|
||||
interface EmbeddingItem {
|
||||
embedding: string | number[]
|
||||
|
|
@ -89,11 +88,7 @@ export class OpenAICompatibleEmbedder implements IEmbedder {
|
|||
const estimatedTokens = Math.ceil(prefixedText.length / 4)
|
||||
if (estimatedTokens > MAX_ITEM_TOKENS) {
|
||||
console.warn(
|
||||
t("embeddings:textWithPrefixExceedsTokenLimit", {
|
||||
index,
|
||||
estimatedTokens,
|
||||
maxTokens: MAX_ITEM_TOKENS,
|
||||
}),
|
||||
`Text at index ${index} with prefix exceeds token limit (${estimatedTokens} > ${MAX_ITEM_TOKENS})`,
|
||||
)
|
||||
// Return original text if adding prefix would exceed limit
|
||||
return text
|
||||
|
|
@ -115,14 +110,8 @@ export class OpenAICompatibleEmbedder implements IEmbedder {
|
|||
const text = remainingTexts[i]
|
||||
const itemTokens = Math.ceil(text.length / 4)
|
||||
|
||||
if (itemTokens > this.maxItemTokens) {
|
||||
console.warn(
|
||||
t("embeddings:textExceedsTokenLimit", {
|
||||
index: i,
|
||||
itemTokens,
|
||||
maxTokens: this.maxItemTokens,
|
||||
}),
|
||||
)
|
||||
if (itemTokens > MAX_ITEM_TOKENS) {
|
||||
console.warn(`Text at index ${i} exceeds token limit (${itemTokens} > ${MAX_ITEM_TOKENS})`)
|
||||
processedIndices.push(i)
|
||||
continue
|
||||
}
|
||||
|
|
@ -278,13 +267,7 @@ export class OpenAICompatibleEmbedder implements IEmbedder {
|
|||
|
||||
if (isRateLimitError && hasMoreAttempts) {
|
||||
const delayMs = INITIAL_DELAY_MS * Math.pow(2, attempts)
|
||||
console.warn(
|
||||
t("embeddings:rateLimitRetry", {
|
||||
delayMs,
|
||||
attempt: attempts + 1,
|
||||
maxRetries: MAX_RETRIES,
|
||||
}),
|
||||
)
|
||||
console.warn(`Rate limit hit, retrying in ${delayMs}ms (attempt ${attempts + 1}/${MAX_RETRIES})`)
|
||||
await new Promise((resolve) => setTimeout(resolve, delayMs))
|
||||
continue
|
||||
}
|
||||
|
|
@ -293,7 +276,7 @@ export class OpenAICompatibleEmbedder implements IEmbedder {
|
|||
console.error(`OpenAI Compatible embedder error (attempt ${attempts + 1}/${MAX_RETRIES}):`, error)
|
||||
|
||||
// Provide more context in the error message using robust error extraction
|
||||
let errorMessage = t("embeddings:unknownError")
|
||||
let errorMessage = "Unknown error"
|
||||
if (httpError?.message) {
|
||||
errorMessage = httpError.message
|
||||
} else if (typeof error === "string") {
|
||||
|
|
@ -302,25 +285,25 @@ export class OpenAICompatibleEmbedder implements IEmbedder {
|
|||
try {
|
||||
errorMessage = String(error)
|
||||
} catch {
|
||||
errorMessage = t("embeddings:unknownError")
|
||||
errorMessage = "Unknown error"
|
||||
}
|
||||
}
|
||||
|
||||
const statusCode = httpError?.status || httpError?.response?.status
|
||||
|
||||
if (statusCode === 401) {
|
||||
throw new Error(t("embeddings:authenticationFailed"))
|
||||
throw new Error("Authentication failed. Please check your API key.")
|
||||
} else if (statusCode) {
|
||||
throw new Error(
|
||||
t("embeddings:failedWithStatus", { attempts: MAX_RETRIES, statusCode, errorMessage }),
|
||||
`Failed to create embeddings after ${MAX_RETRIES} attempts. Status: ${statusCode}. Error: ${errorMessage}`,
|
||||
)
|
||||
} else {
|
||||
throw new Error(t("embeddings:failedWithError", { attempts: MAX_RETRIES, errorMessage }))
|
||||
throw new Error(`Failed to create embeddings after ${MAX_RETRIES} attempts. Error: ${errorMessage}`)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
throw new Error(t("embeddings:failedMaxAttempts", { attempts: MAX_RETRIES }))
|
||||
throw new Error(`Failed to create embeddings after ${MAX_RETRIES} attempts`)
|
||||
}
|
||||
|
||||
/**
|
||||
|
|
|
|||
|
|
@ -9,7 +9,6 @@ import {
|
|||
INITIAL_RETRY_DELAY_MS as INITIAL_DELAY_MS,
|
||||
} from "../constants"
|
||||
import { getModelQueryPrefix } from "../../../shared/embeddingModels"
|
||||
import { t } from "../../../i18n"
|
||||
|
||||
/**
|
||||
* OpenAI implementation of the embedder interface with batching and rate limiting
|
||||
|
|
@ -50,11 +49,7 @@ export class OpenAiEmbedder extends OpenAiNativeHandler implements IEmbedder {
|
|||
const estimatedTokens = Math.ceil(prefixedText.length / 4)
|
||||
if (estimatedTokens > MAX_ITEM_TOKENS) {
|
||||
console.warn(
|
||||
t("embeddings:textWithPrefixExceedsTokenLimit", {
|
||||
index,
|
||||
estimatedTokens,
|
||||
maxTokens: MAX_ITEM_TOKENS,
|
||||
}),
|
||||
`Text at index ${index} with prefix exceeds token limit (${estimatedTokens} > ${MAX_ITEM_TOKENS})`,
|
||||
)
|
||||
// Return original text if adding prefix would exceed limit
|
||||
return text
|
||||
|
|
@ -77,13 +72,7 @@ export class OpenAiEmbedder extends OpenAiNativeHandler implements IEmbedder {
|
|||
const itemTokens = Math.ceil(text.length / 4)
|
||||
|
||||
if (itemTokens > MAX_ITEM_TOKENS) {
|
||||
console.warn(
|
||||
t("embeddings:textExceedsTokenLimit", {
|
||||
index: i,
|
||||
itemTokens,
|
||||
maxTokens: MAX_ITEM_TOKENS,
|
||||
}),
|
||||
)
|
||||
console.warn(`Text at index ${i} exceeds token limit (${itemTokens} > ${MAX_ITEM_TOKENS})`)
|
||||
processedIndices.push(i)
|
||||
continue
|
||||
}
|
||||
|
|
@ -143,13 +132,7 @@ export class OpenAiEmbedder extends OpenAiNativeHandler implements IEmbedder {
|
|||
|
||||
if (isRateLimitError && hasMoreAttempts) {
|
||||
const delayMs = INITIAL_DELAY_MS * Math.pow(2, attempts)
|
||||
console.warn(
|
||||
t("embeddings:rateLimitRetry", {
|
||||
delayMs,
|
||||
attempt: attempts + 1,
|
||||
maxRetries: MAX_RETRIES,
|
||||
}),
|
||||
)
|
||||
console.warn(`Rate limit hit, retrying in ${delayMs}ms (attempt ${attempts + 1}/${MAX_RETRIES})`)
|
||||
await new Promise((resolve) => setTimeout(resolve, delayMs))
|
||||
continue
|
||||
}
|
||||
|
|
@ -174,18 +157,18 @@ export class OpenAiEmbedder extends OpenAiNativeHandler implements IEmbedder {
|
|||
const statusCode = error?.status || error?.response?.status
|
||||
|
||||
if (statusCode === 401) {
|
||||
throw new Error(t("embeddings:authenticationFailed"))
|
||||
throw new Error("Authentication failed. Please check your API key.")
|
||||
} else if (statusCode) {
|
||||
throw new Error(
|
||||
t("embeddings:failedWithStatus", { attempts: MAX_RETRIES, statusCode, errorMessage }),
|
||||
`Failed to create embeddings after ${MAX_RETRIES} attempts. Status: ${statusCode}. Error: ${errorMessage}`,
|
||||
)
|
||||
} else {
|
||||
throw new Error(t("embeddings:failedWithError", { attempts: MAX_RETRIES, errorMessage }))
|
||||
throw new Error(`Failed to create embeddings after ${MAX_RETRIES} attempts. Error: ${errorMessage}`)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
throw new Error(t("embeddings:failedMaxAttempts", { attempts: MAX_RETRIES }))
|
||||
throw new Error(`Failed to create embeddings after ${MAX_RETRIES} attempts`)
|
||||
}
|
||||
|
||||
get embedderInfo(): EmbedderInfo {
|
||||
|
|
|
|||
|
|
@ -5,7 +5,6 @@ import { getWorkspacePath } from "../../../utils/path"
|
|||
import { IVectorStore } from "../interfaces/vector-store"
|
||||
import { Payload, VectorStoreSearchResult } from "../interfaces"
|
||||
import { MAX_SEARCH_RESULTS, SEARCH_MIN_SCORE } from "../constants"
|
||||
import { t } from "../../../i18n"
|
||||
|
||||
/**
|
||||
* Qdrant implementation of the vector store interface
|
||||
|
|
@ -205,9 +204,7 @@ export class QdrantVectorStore implements IVectorStore {
|
|||
)
|
||||
|
||||
// Provide a more user-friendly error message that includes the original error
|
||||
throw new Error(
|
||||
t("embeddings:vectorStore.qdrantConnectionFailed", { qdrantUrl: this.qdrantUrl, errorMessage }),
|
||||
)
|
||||
throw new Error(`Failed to connect to Qdrant at ${this.qdrantUrl}: ${errorMessage}`)
|
||||
}
|
||||
}
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue