Version: 0.9.80.dev.260506
后端： 1. LLM 独立服务与统一计费出口落地：新增 `cmd/llm`、`client/llm` 与 `services/llm/rpc`，补齐 BillingContext、CreditBalanceGuard、价格规则解析、stream usage 归集与 `credit.charge.requested` outbox 发布，active-scheduler / agent / course / memory / gateway fallback 全部改走 llm zrpc，不再各自本地初始化模型。 2. TokenStore 收口为 Credit 权威账本：新增 credit account / ledger / product / order / price-rule / reward-rule 能力与 Redis 快照缓存，扩展 tokenstore rpc/client 支撑余额快照、消耗看板、商品、订单、流水、价格规则和奖励规则，并接入 LLM charge 事件消费完成 Credit 扣费落账。 3. 计费旧链路下线与网关切口切换：`/token-store` 语义整体切到 `/credit-store`，agent chat 移除旧 TokenQuotaGuard，userauth 的 CheckTokenQuota / AdjustTokenUsage 改为废弃，聊天历史落库不再同步旧 token 额度账本，course 图片解析请求补 user_id 进入新计费口径。 前端： 4. 计划广场从 mock 数据切到真实接口：新增 forum api/types，首页支持真实列表、标签、搜索、防抖、点赞、导入和发布计划，详情页补齐帖子详情、评论树、回复和删除评论链路，同时补上“至少一个标签”的前后端约束与默认标签兜底。 5. 商店页切到 Credit 体系并重做展示：顶部改为余额 + Credit/Token 消耗看板，支持 24h/7d/30d/all 周期切换；套餐区展示原价与当前价；历史区改为当前用户 Credit 流水并支持查看更多，整体视觉和交互同步收口。 仓库： 6. 配置与本地启动体系补齐 llm / outbox 编排：`config.example.yaml` 增加 llm rpc 和统一 outbox service 配置，`dev-common.ps1` 把 llm 纳入多服务依赖并自动建 Kafka topic，`docker-compose.yml` 同步初始化 agent/task/memory/active-scheduler/notification/taskclass-forum/llm/token-store 全量 outbox topic。
This commit is contained in:
Losita
@@ -1,13 +1,10 @@
|
||||
package sv
|
||||
|
||||
import (
|
||||
"context"
|
||||
"errors"
|
||||
"strconv"
|
||||
"strings"
|
||||
|
||||
tokenmodel "github.com/LoveLosita/smartflow/backend/services/tokenstore/model"
|
||||
tokencontracts "github.com/LoveLosita/smartflow/backend/shared/contracts/tokenstore"
|
||||
storemodel "github.com/LoveLosita/smartflow/backend/services/tokenstore/model"
|
||||
"github.com/spf13/viper"
|
||||
)
|
||||
|
||||
@@ -32,85 +29,12 @@ type forumRewardDecision struct {
|
||||
Description string
|
||||
}
|
||||
|
||||
// RecordForumRewardGrant 负责把论坛点赞/导入奖励写入 token_grants。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只处理 forum_like / forum_import 两类奖励账本写入，不修改 users，也不调用 user/auth；
|
||||
// 2. 以 event_id 作为最终幂等边界，重复请求校验一致后返回既有 grant；
|
||||
// 3. 奖励金额优先读取 token_reward_rules，配置和代码默认值只作为兜底。
|
||||
func (s *Service) RecordForumRewardGrant(ctx context.Context, req tokencontracts.RecordForumRewardGrantRequest) (*tokencontracts.TokenGrantView, error) {
|
||||
if err := s.Ready(); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
normalized, err := normalizeForumRewardGrantRequest(req)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
// 1. 先按 event_id 回查，命中时直接视为成功，避免 outbox 重试重复写账本。
|
||||
// 2. 命中后必须校验用户、来源和来源业务 ID，避免错误复用 event_id 时静默吞掉错账。
|
||||
// 3. 校验通过才返回既有 grant，兼容“首次已成功、调用方超时后重试”的常见场景。
|
||||
existing, err := s.tokenDAO.FindGrantByEventID(ctx, normalized.EventID)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
if existing != nil {
|
||||
if err := validateExistingForumRewardGrant(*existing, normalized); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
view := grantViewFromModel(*existing)
|
||||
return &view, nil
|
||||
}
|
||||
|
||||
sourceRefID := normalized.SourceRefID
|
||||
decision, err := s.forumRewardDecision(ctx, normalized.Source)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
grant := tokenmodel.TokenGrant{
|
||||
EventID: normalized.EventID,
|
||||
UserID: normalized.ReceiverUserID,
|
||||
Source: normalized.Source,
|
||||
SourceLabel: grantSourceLabel(normalized.Source, ""),
|
||||
SourceRefID: &sourceRefID,
|
||||
Amount: decision.Amount,
|
||||
Status: decision.Status,
|
||||
QuotaApplied: false,
|
||||
Description: decision.Description,
|
||||
}
|
||||
|
||||
// 1. 账本写入只依赖 token_grants.event_id 唯一约束兜底并发幂等。
|
||||
// 2. 若并发下插入触发唯一键冲突，立刻回查 event_id，把已有 grant 当作成功结果返回。
|
||||
// 3. 只有“冲突后仍查不到旧记录”这种异常态才上抛内部错误，避免吞掉真实一致性问题。
|
||||
if err := s.tokenDAO.CreateGrant(ctx, &grant); err != nil {
|
||||
if !isDuplicateKeyError(err) {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
existing, err := s.tokenDAO.FindGrantByEventID(ctx, normalized.EventID)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
if existing == nil {
|
||||
return nil, errors.New("forum reward grant duplicated but not found by event_id")
|
||||
}
|
||||
if err := validateExistingForumRewardGrant(*existing, normalized); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
view := grantViewFromModel(*existing)
|
||||
return &view, nil
|
||||
}
|
||||
|
||||
view := grantViewFromModel(grant)
|
||||
return &view, nil
|
||||
}
|
||||
|
||||
func normalizeForumRewardGrantRequest(req tokencontracts.RecordForumRewardGrantRequest) (forumRewardGrantRequest, error) {
|
||||
func normalizeForumRewardGrantRequest(req forumRewardGrantRequest) (forumRewardGrantRequest, error) {
|
||||
normalized := forumRewardGrantRequest{
|
||||
EventID: strings.TrimSpace(req.EventID),
|
||||
ReceiverUserID: req.ReceiverUserID,
|
||||
Source: strings.ToLower(strings.TrimSpace(req.Source)),
|
||||
SourceRefID: req.SourceRefID,
|
||||
}
|
||||
|
||||
switch {
|
||||
@@ -118,16 +42,12 @@ func normalizeForumRewardGrantRequest(req tokencontracts.RecordForumRewardGrantR
|
||||
return forumRewardGrantRequest{}, tokenStoreBadRequest("event_id 不能为空")
|
||||
case normalized.ReceiverUserID == 0:
|
||||
return forumRewardGrantRequest{}, tokenStoreBadRequest("receiver_user_id 不能为空")
|
||||
case normalized.SourceRefID == 0:
|
||||
return forumRewardGrantRequest{}, tokenStoreBadRequest("source_ref_id 不能为空")
|
||||
}
|
||||
|
||||
sourceRefID, err := parseForumRewardSourceRefID(req.SourceRefID)
|
||||
if err != nil {
|
||||
return forumRewardGrantRequest{}, err
|
||||
}
|
||||
normalized.SourceRefID = sourceRefID
|
||||
|
||||
switch normalized.Source {
|
||||
case tokenmodel.TokenGrantSourceForumLike, tokenmodel.TokenGrantSourceForumImport:
|
||||
case storemodel.CreditLedgerSourceForumLike, storemodel.CreditLedgerSourceForumImport:
|
||||
return normalized, nil
|
||||
default:
|
||||
return forumRewardGrantRequest{}, tokenStoreBadRequest("source 仅支持 forum_like 或 forum_import")
|
||||
@@ -147,69 +67,10 @@ func parseForumRewardSourceRefID(raw string) (uint64, error) {
|
||||
return parsed, nil
|
||||
}
|
||||
|
||||
// validateExistingForumRewardGrant 校验重复 event_id 是否真的是同一条论坛奖励。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只比较幂等所需的最小字段：接收人、来源和来源业务 ID；
|
||||
// 2. 不比较金额和状态，避免规则调整后重放旧事件被误判；
|
||||
// 3. 不一致时返回业务校验错误，让上游暴露这类错账风险。
|
||||
func validateExistingForumRewardGrant(existing tokenmodel.TokenGrant, req forumRewardGrantRequest) error {
|
||||
sourceRefID := uint64(0)
|
||||
if existing.SourceRefID != nil {
|
||||
sourceRefID = *existing.SourceRefID
|
||||
}
|
||||
if existing.UserID != req.ReceiverUserID || existing.Source != req.Source || sourceRefID != req.SourceRefID {
|
||||
return tokenStoreBadRequest("event_id 幂等冲突：已有奖励记录与本次论坛奖励请求不一致")
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
// forumRewardDecision 解析论坛奖励发放决策。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 优先读取 token_reward_rules，保持“从表里读”的 P0 口径；
|
||||
// 2. 规则停用或金额非正时写 skipped 账本，消费 outbox 但不增加 Token；
|
||||
// 3. 表规则缺失时再读取配置和代码默认值，兼容旧环境尚未 seed 的情况。
|
||||
func (s *Service) forumRewardDecision(ctx context.Context, source string) (forumRewardDecision, error) {
|
||||
rule, err := s.tokenDAO.FindRewardRuleBySource(ctx, source)
|
||||
if err != nil {
|
||||
return forumRewardDecision{}, err
|
||||
}
|
||||
if rule != nil {
|
||||
if strings.TrimSpace(rule.Status) != tokenmodel.TokenRewardRuleStatusActive {
|
||||
return skippedForumRewardDecision(source, "奖励规则已停用，未发放 Token"), nil
|
||||
}
|
||||
if rule.Amount <= 0 {
|
||||
return skippedForumRewardDecision(source, "奖励规则金额非正，未发放 Token"), nil
|
||||
}
|
||||
return recordedForumRewardDecision(source, rule.Amount), nil
|
||||
}
|
||||
|
||||
switch strings.TrimSpace(source) {
|
||||
case tokenmodel.TokenGrantSourceForumLike:
|
||||
return recordedForumRewardDecision(source, positiveConfigAmountOrDefault(forumLikeRewardConfigKey, defaultForumLikeRewardAmount)), nil
|
||||
case tokenmodel.TokenGrantSourceForumImport:
|
||||
return recordedForumRewardDecision(source, positiveConfigAmountOrDefault(forumImportRewardConfigKey, defaultForumImportRewardAmount)), nil
|
||||
default:
|
||||
return skippedForumRewardDecision(source, "未知论坛奖励来源，未发放 Token"), nil
|
||||
}
|
||||
}
|
||||
|
||||
func recordedForumRewardDecision(source string, amount int64) forumRewardDecision {
|
||||
if amount <= 0 {
|
||||
return skippedForumRewardDecision(source, "奖励金额非正，未发放 Token")
|
||||
}
|
||||
return forumRewardDecision{
|
||||
Amount: amount,
|
||||
Status: tokenmodel.TokenGrantStatusRecorded,
|
||||
Description: forumRewardDescription(source),
|
||||
}
|
||||
}
|
||||
|
||||
func skippedForumRewardDecision(source string, description string) forumRewardDecision {
|
||||
return forumRewardDecision{
|
||||
Amount: 0,
|
||||
Status: tokenmodel.TokenGrantStatusSkipped,
|
||||
Status: storemodel.CreditLedgerStatusSkipped,
|
||||
Description: strings.TrimSpace(description),
|
||||
}
|
||||
}
|
||||
@@ -224,9 +85,9 @@ func positiveConfigAmountOrDefault(configKey string, fallback int64) int64 {
|
||||
|
||||
func forumRewardDescription(source string) string {
|
||||
switch strings.TrimSpace(source) {
|
||||
case tokenmodel.TokenGrantSourceForumLike:
|
||||
case storemodel.CreditLedgerSourceForumLike:
|
||||
return "计划被点赞奖励"
|
||||
case tokenmodel.TokenGrantSourceForumImport:
|
||||
case storemodel.CreditLedgerSourceForumImport:
|
||||
return "计划被导入奖励"
|
||||
default:
|
||||
return "论坛奖励入账"
|
||||
|
||||
Reference in New Issue
Block a user