9df34abd6b
- ModelChannel 新增 proxy_url 字段，v48 迁移自动加列 - outbound 新增 OutboundHTTPClientWithProxy，socks5/socks5h/http/https 均支持； 目标主机名不做本地解析，直接交给代理服务器远端解析，本地开启 TUN/VPN 时仍可用 - 系统代理、Provider 出站、管理员模型测试路径均使用渠道级代理 - 渠道管理表单新增"代理地址"输入框，留空回退全局代理
246 lines
11 KiB
Go
246 lines
11 KiB
Go
// 供应商调用的错误类型与用户可读文案。
|
||
//
|
||
// 原始上游错误（状态码、响应体片段）只进诊断与调用日志；返回给用户的文案按类别归一，
|
||
// 不泄露密钥、内部地址或冗长响应体。
|
||
|
||
package app
|
||
|
||
import (
|
||
"context"
|
||
"encoding/json"
|
||
"errors"
|
||
"fmt"
|
||
"net/http"
|
||
"strconv"
|
||
"strings"
|
||
"time"
|
||
|
||
"yingce/backend/internal/model"
|
||
)
|
||
|
||
type providerError struct {
|
||
Message string `json:"message"`
|
||
}
|
||
|
||
// providerPayloadError 在进程内保留上游原始原因，供协议兼容分支做机器判断；
|
||
// 对调用方只暴露经过过滤的错误原因。Provider 正文可能包含密钥或内部诊断，
|
||
// 禁止原样进入用户错误和日志。
|
||
type providerPayloadError struct {
|
||
raw string
|
||
message string
|
||
}
|
||
|
||
func (e providerPayloadError) Error() string { return e.message }
|
||
|
||
type providerHTTPError struct {
|
||
StatusCode int
|
||
Status string
|
||
Body string
|
||
RetryAfter time.Duration
|
||
}
|
||
|
||
type providerResponseDecodeError struct {
|
||
Err error
|
||
}
|
||
|
||
func (e providerResponseDecodeError) Error() string { return e.Err.Error() }
|
||
|
||
func (e providerResponseDecodeError) Unwrap() error { return e.Err }
|
||
|
||
type providerCircuitOpenError struct{}
|
||
|
||
func (providerCircuitOpenError) Error() string {
|
||
return "当前渠道连续失败，已暂时熔断，请稍后重试"
|
||
}
|
||
|
||
type providerStatePendingError struct {
|
||
TaskID string
|
||
Cause error
|
||
}
|
||
|
||
func (e providerStatePendingError) Error() string {
|
||
return fmt.Sprintf("上游任务状态尚未同步，将继续查询原任务（任务 %s）", e.TaskID)
|
||
}
|
||
|
||
func (e providerStatePendingError) Unwrap() error { return e.Cause }
|
||
|
||
type providerAnalyticsKey struct{}
|
||
|
||
type providerAnalyticsContext struct {
|
||
Service *Service
|
||
Billing taskBillingLifecycle
|
||
UserID string
|
||
TaskID string
|
||
TraceID string
|
||
RequestID string
|
||
BillingOrderID string
|
||
BillingMode string
|
||
Capability string
|
||
Operation string
|
||
ChannelID string
|
||
ProxyURL string
|
||
Model string
|
||
VideoSeconds int
|
||
RequestKind string
|
||
ProviderRequestID string
|
||
ConcurrencyLimit int
|
||
}
|
||
|
||
func withProviderAnalytics(ctx context.Context, service *Service, task model.Task) context.Context {
|
||
metadata := providerAnalyticsContext{Service: service, UserID: task.UserID, TaskID: task.ID, TraceID: task.TraceID, RequestID: task.RequestID, BillingOrderID: task.BillingOrderID, Capability: capabilityFromTaskType(task.Type), Operation: task.Operation, Model: task.Model, ProviderRequestID: task.ProviderRequestID}
|
||
if service != nil {
|
||
metadata.Billing = service.taskBilling()
|
||
}
|
||
// 账单模式随请求上下文传递，流式协议据此只为 Token 计费开启 usage 终态块。
|
||
if service != nil && task.BillingOrderID != "" {
|
||
if order, err := service.repo.BillingOrder(task.BillingOrderID); err == nil {
|
||
metadata.BillingMode = order.BillingMode
|
||
}
|
||
}
|
||
var input struct {
|
||
Mode string `json:"mode"`
|
||
Config providerConfig `json:"config"`
|
||
}
|
||
if json.Unmarshal([]byte(task.InputJSON), &input) == nil {
|
||
metadata.ChannelID = firstNonEmpty(input.Config.ChannelID, systemChannelIDFromBaseURL(input.Config.BaseURL))
|
||
metadata.Model = firstNonEmpty(input.Config.ChannelModelKey, input.Config.Model, metadata.Model)
|
||
metadata.VideoSeconds, _ = strconv.Atoi(input.Config.VideoSeconds)
|
||
if normalized := normalizeCapability(input.Mode); normalized != "" {
|
||
metadata.Capability = normalized
|
||
}
|
||
if metadata.ChannelID != "" && service != nil {
|
||
if ch, err := service.repo.SystemChannel(metadata.ChannelID); err == nil {
|
||
metadata.ProxyURL = ch.ProxyURL
|
||
}
|
||
}
|
||
}
|
||
return context.WithValue(ctx, providerAnalyticsKey{}, metadata)
|
||
}
|
||
|
||
func resumedProviderRequestID(ctx context.Context) string {
|
||
metadata, _ := ctx.Value(providerAnalyticsKey{}).(providerAnalyticsContext)
|
||
return strings.TrimSpace(metadata.ProviderRequestID)
|
||
}
|
||
|
||
func withProviderRequestKind(ctx context.Context, requestKind string) context.Context {
|
||
metadata, ok := ctx.Value(providerAnalyticsKey{}).(providerAnalyticsContext)
|
||
if !ok {
|
||
return ctx
|
||
}
|
||
metadata.RequestKind = requestKind
|
||
return context.WithValue(ctx, providerAnalyticsKey{}, metadata)
|
||
}
|
||
|
||
func (e providerHTTPError) Error() string {
|
||
if e.StatusCode == http.StatusBadRequest || e.StatusCode == http.StatusUnprocessableEntity {
|
||
return providerErrorWithDetail(e.summary(), e.Body)
|
||
}
|
||
if message := speechResourceDeniedUserMessage(e.Body); (e.StatusCode == http.StatusUnauthorized || e.StatusCode == http.StatusForbidden) && message != "" {
|
||
return message
|
||
}
|
||
return appendProviderErrorDetail(e.summary(), e.Body)
|
||
}
|
||
|
||
const volcengineSpeechResourceDeniedMessage = "语音合成服务未开通，或音色与模型版本不匹配。请在火山引擎控制台开通语音合成 1.0，并确认当前音色属于这一版本"
|
||
|
||
func speechResourceDeniedUserMessage(raw string) string {
|
||
if strings.Contains(strings.ToLower(raw), "requested resource not granted") {
|
||
return volcengineSpeechResourceDeniedMessage
|
||
}
|
||
return ""
|
||
}
|
||
|
||
func (e providerHTTPError) summary() string {
|
||
switch e.StatusCode {
|
||
case 524:
|
||
return "上游网关超时（524）：模型请求可能仍在服务端执行并产生费用，请勿立即重试，请先到供应商后台核对任务或账单"
|
||
case http.StatusBadRequest, http.StatusUnprocessableEntity:
|
||
return "模型服务拒绝了请求，请检查模型和参数"
|
||
case http.StatusUnauthorized, http.StatusForbidden:
|
||
return "模型服务鉴权失败，请检查 API Key 和模型权限"
|
||
case http.StatusNotFound:
|
||
return "模型或模型接口不存在，请检查渠道配置"
|
||
case http.StatusRequestTimeout, http.StatusGatewayTimeout:
|
||
return "模型服务响应超时，请稍后重试"
|
||
case http.StatusTooManyRequests:
|
||
return "模型服务请求过于频繁或额度不足，请稍后重试"
|
||
}
|
||
if e.StatusCode >= http.StatusInternalServerError {
|
||
return fmt.Sprintf("模型服务暂时不可用（HTTP %d）", e.StatusCode)
|
||
}
|
||
return fmt.Sprintf("模型服务请求失败（HTTP %d）", e.StatusCode)
|
||
}
|
||
|
||
func providerUserFacingErrorMessage(err error) string {
|
||
if err == nil {
|
||
return "模型服务请求失败"
|
||
}
|
||
if errors.Is(err, context.Canceled) {
|
||
return "模型请求已取消"
|
||
}
|
||
if errors.Is(err, context.DeadlineExceeded) {
|
||
return "模型服务响应超时，请稍后重试"
|
||
}
|
||
var appErr *AppError
|
||
if errors.As(err, &appErr) && strings.TrimSpace(appErr.Message) != "" {
|
||
return appErr.Message
|
||
}
|
||
var httpErr providerHTTPError
|
||
if errors.As(err, &httpErr) {
|
||
return httpErr.Error()
|
||
}
|
||
return "连接模型服务失败，请检查渠道地址和网络"
|
||
}
|
||
|
||
// providerPayloadErrorCategory 把上游失败正文归类为固定的用户可见原因。
|
||
// 第二个返回值为 false 表示正文无法归类，调用方应退回到更通用的提示，
|
||
// 不要因为归类失败就把正文本身当作错误信息。
|
||
// 正文可能包含密钥或内部诊断信息，只能参与归类，不得回传用户或写入日志。
|
||
func providerPayloadErrorCategory(raw string) (string, bool) {
|
||
normalized := strings.ToLower(strings.TrimSpace(raw))
|
||
if normalized == "" {
|
||
return "", false
|
||
}
|
||
switch {
|
||
// 真人肖像类目只匹配供应商错误码里的稳定标识，不扫描自然语言。
|
||
// 正文常常回显用户提示词，"likeness"、"肖像"这类词单独出现并不能证明
|
||
// 上游是因为真人形象拒绝，按词判断会把普通参数错误误报成肖像问题。
|
||
// 该类目排在安全审核之前：错误码已经足够具体，比通用审核提示更可行动。
|
||
case strings.Contains(normalized, "privacyinformation"), strings.Contains(normalized, "sensitivecontentdetected"):
|
||
return "输入素材疑似包含真人形象，该模型拒绝生成，请更换为非真人素材或改用其他模型", true
|
||
case strings.Contains(normalized, "safety"), strings.Contains(normalized, "moderation"), strings.Contains(normalized, "content policy"), strings.Contains(normalized, "blocked"):
|
||
return "请求内容未通过模型服务安全审核，请调整后重试", true
|
||
// 工具调用与工具结果不配对：上游要求 assistant 消息声明的每一个 tool_call_id 都在
|
||
// 紧随其后的 tool 消息里被回应。这是**我们组装请求**的问题——用户改提示词或查额度
|
||
// 都没用——所以文案指向反馈而不是"调整输入"。
|
||
//
|
||
// 必须排在额度类目之前：DeepSeek 的原文含 "insufficient"，
|
||
// "An assistant message with 'tool_calls' must be followed by tool messages
|
||
// responding to each 'tool_call_id'. (insufficient tool messages following
|
||
// tool_calls message)" 落到额度类目就会把协议错误报成"渠道余额不足"，
|
||
// 掩盖真正的原因（历史里的工具结果不连续）。
|
||
case strings.Contains(normalized, "must be followed by tool messages"),
|
||
strings.Contains(normalized, "insufficient tool messages following"):
|
||
return "会话里的工具调用与结果不匹配，本轮已停止；这不是额度或提示词问题，如反复出现请反馈", true
|
||
case strings.Contains(normalized, "quota"), strings.Contains(normalized, "insufficient"), strings.Contains(normalized, "balance"), strings.Contains(normalized, "billing"):
|
||
return "模型服务额度不足，请检查渠道余额或配额", true
|
||
case strings.Contains(normalized, "model") && (strings.Contains(normalized, "not found") || strings.Contains(normalized, "permission") || strings.Contains(normalized, "access")):
|
||
return "模型不存在或当前渠道未获得模型权限", true
|
||
// 推理/思考模式模型通常禁止强制指定工具调用：DeepSeek 思考模式返回
|
||
// "Thinking mode does not support this tool_choice"，其他 OpenAI 兼容
|
||
// 供应商措辞类似。归为固定可行动原因；显式思考模式会在出站前省略
|
||
// tool_choice，未声明但由上游隐式开启思考时再按兼容序列重试。排在
|
||
// 通用参数类目之前，避免稳定标识落回笼统的"请检查模型和参数"。
|
||
case (strings.Contains(normalized, "thinking") || strings.Contains(normalized, "reasoning")) && strings.Contains(normalized, "tool_choice"),
|
||
strings.Contains(normalized, "tool_choice") && (strings.Contains(normalized, "not support") || strings.Contains(normalized, "unsupported")):
|
||
return "当前模型为思考/推理模式，不支持强制工具调用（tool_choice=required），请改用自动工具选择或更换非思考模式模型", true
|
||
case strings.Contains(normalized, "invalid"), strings.Contains(normalized, "parameter"), strings.Contains(normalized, "argument"):
|
||
return "模型服务拒绝了请求，请检查模型和参数", true
|
||
}
|
||
return "", false
|
||
}
|
||
|
||
func providerPayloadErrorMessage(raw string) string {
|
||
return providerErrorWithDetail("模型服务返回失败，请检查请求内容或渠道配置", raw)
|
||
}
|