mirror of
https://github.com/fawney19/Aether.git
synced 2026-10-08 10:27:46 +08:00
refactor: 大规模模块拆分与代码精简,新增 ai-pipeline/data-contracts 独立 crate
- 新增 aether-ai-pipeline 和 aether-data-contracts crate,将 pipeline 逻辑与数据契约从 gateway 中解耦 - 重构 admin handlers:拆分单体模块为 auth/billing/endpoint/features/model/observability/provider/system 等独立子模块 - 合并 chat/cli 重复代码路径:精简 conversion、finalize、planner 中的 sync/chat/cli 分支 - 重构 scheduler/executor/data 层,引入 facade 模式降低模块间耦合 - 移除冗余的 intent 模块,将 plan_fallback/policy/stream_path/sync_path 迁移至 executor - 前端适配:调整 admin API 调用和 provider 模型测试对话框
This commit is contained in:
@@ -1,17 +1,17 @@
|
||||
use std::time::Duration;
|
||||
|
||||
use aether_contracts::{ExecutionPlan, ExecutionResult, ProxySnapshot};
|
||||
use aether_data::repository::candidate_selection::StoredMinimalCandidateSelectionRow;
|
||||
use aether_data::repository::candidates::{StoredRequestCandidate, UpsertRequestCandidateRecord};
|
||||
use aether_data::repository::global_models::{
|
||||
use aether_data_contracts::repository::candidates::{
|
||||
StoredRequestCandidate, UpsertRequestCandidateRecord,
|
||||
};
|
||||
use aether_data_contracts::repository::global_models::{
|
||||
AdminGlobalModelListQuery, AdminProviderModelListQuery, StoredAdminGlobalModelPage,
|
||||
StoredAdminProviderModel, UpsertAdminProviderModelRecord,
|
||||
};
|
||||
use aether_data::repository::provider_catalog::{
|
||||
use aether_data_contracts::repository::provider_catalog::{
|
||||
StoredProviderCatalogEndpoint, StoredProviderCatalogKey, StoredProviderCatalogProvider,
|
||||
};
|
||||
use aether_data::repository::quota::StoredProviderQuotaSnapshot;
|
||||
use aether_data::DataLayerError;
|
||||
use aether_data_contracts::repository::quota::StoredProviderQuotaSnapshot;
|
||||
use aether_model_fetch::{
|
||||
aggregate_models_for_cache, model_fetch_interval_minutes, ModelFetchAssociationStore,
|
||||
ModelFetchTransportRuntime,
|
||||
@@ -27,10 +27,10 @@ use crate::provider_transport::{
|
||||
resolve_transport_proxy_snapshot_with_tunnel_affinity, GatewayProviderTransportSnapshot,
|
||||
LocalResolvedOAuthRequestAuth,
|
||||
};
|
||||
use crate::scheduler::{
|
||||
GatewayMinimalCandidateSelectionCandidate, SchedulerCandidateSelectionRowSource,
|
||||
SchedulerRequestCandidateRuntimeState, SchedulerRuntimeState,
|
||||
use crate::request_candidate_runtime::{
|
||||
RequestCandidateRuntimeReader, RequestCandidateRuntimeWriter,
|
||||
};
|
||||
use crate::scheduler::state::SchedulerRuntimeState;
|
||||
use crate::{execution_runtime, provider_transport};
|
||||
|
||||
#[async_trait]
|
||||
@@ -237,17 +237,20 @@ impl ModelFetchAssociationStore for AppState {
|
||||
}
|
||||
|
||||
#[async_trait]
|
||||
impl SchedulerRequestCandidateRuntimeState for AppState {
|
||||
fn has_request_candidate_data_writer(&self) -> bool {
|
||||
AppState::has_request_candidate_data_writer(self)
|
||||
}
|
||||
|
||||
impl RequestCandidateRuntimeReader for AppState {
|
||||
async fn read_request_candidates_by_request_id(
|
||||
&self,
|
||||
request_id: &str,
|
||||
) -> Result<Vec<StoredRequestCandidate>, GatewayError> {
|
||||
AppState::read_request_candidates_by_request_id(self, request_id).await
|
||||
}
|
||||
}
|
||||
|
||||
#[async_trait]
|
||||
impl RequestCandidateRuntimeWriter for AppState {
|
||||
fn has_request_candidate_data_writer(&self) -> bool {
|
||||
AppState::has_request_candidate_data_writer(self)
|
||||
}
|
||||
|
||||
async fn upsert_request_candidate(
|
||||
&self,
|
||||
@@ -257,28 +260,6 @@ impl SchedulerRequestCandidateRuntimeState for AppState {
|
||||
}
|
||||
}
|
||||
|
||||
#[async_trait]
|
||||
impl SchedulerCandidateSelectionRowSource for AppState {
|
||||
async fn read_minimal_candidate_selection_rows_for_api_format_and_global_model(
|
||||
&self,
|
||||
api_format: &str,
|
||||
global_model_name: &str,
|
||||
) -> Result<Vec<StoredMinimalCandidateSelectionRow>, DataLayerError> {
|
||||
self.data
|
||||
.list_minimal_candidate_selection_rows(api_format, global_model_name)
|
||||
.await
|
||||
}
|
||||
|
||||
async fn read_minimal_candidate_selection_rows_for_api_format(
|
||||
&self,
|
||||
api_format: &str,
|
||||
) -> Result<Vec<StoredMinimalCandidateSelectionRow>, DataLayerError> {
|
||||
self.data
|
||||
.list_minimal_candidate_selection_rows_for_api_format(api_format)
|
||||
.await
|
||||
}
|
||||
}
|
||||
|
||||
#[async_trait]
|
||||
impl SchedulerRuntimeState for AppState {
|
||||
async fn read_provider_quota_snapshot(
|
||||
@@ -309,23 +290,6 @@ impl SchedulerRuntimeState for AppState {
|
||||
AppState::read_recent_request_candidates(self, limit).await
|
||||
}
|
||||
|
||||
async fn read_minimal_candidate_selection(
|
||||
&self,
|
||||
api_format: &str,
|
||||
global_model_name: &str,
|
||||
require_streaming: bool,
|
||||
auth_snapshot: Option<&crate::data::auth::GatewayAuthApiKeySnapshot>,
|
||||
) -> Result<Vec<GatewayMinimalCandidateSelectionCandidate>, GatewayError> {
|
||||
AppState::read_minimal_candidate_selection(
|
||||
self,
|
||||
api_format,
|
||||
global_model_name,
|
||||
require_streaming,
|
||||
auth_snapshot,
|
||||
)
|
||||
.await
|
||||
}
|
||||
|
||||
fn provider_key_rpm_reset_at(&self, key_id: &str, now_unix_secs: u64) -> Option<u64> {
|
||||
AppState::provider_key_rpm_reset_at(self, key_id, now_unix_secs)
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user