test(retry): 钉住声明式重试判据与流式握手重试语义

- config：ShouldRetry/DelayDuration 表驱动（nil/次数=1/状态不匹配/body 不匹配/
  空子串/多信号/非法间隔）
- provider：命中策略就重试直到 200、body 不匹配**不许**重试（400 也常用于真错误）、
  流式只重试握手（首帧之后不重试，避免重复输出）

internal/config/retry_policy_test.go

@@ -0,0 +1,59 @@
+package config
+
+import (
+	"testing"
+	"time"
+)
+
+// ShouldRetry 是纯数据判据：状态码列表 + 响应体子串。它决定"这次失败值不值得
+// 在网关内部等一会儿重试"。厂商知识绝不能进内核，所以判据必须能用一个表格测清楚。
+func TestRetryPolicyShouldRetry(t *testing.T) {
+	cases := []struct {
+		name   string
+		policy *RetryPolicy
+		status int
+		body   string
+		want   bool
+	}{
+		{"nil policy", nil, 500, "anything", false},
+		{"max_attempts=1 disables", &RetryPolicy{OnStatus: []int{400}, MaxAttempts: 1}, 400, "x", false},
+		{"status matches, no body filter",
+			&RetryPolicy{OnStatus: []int{429}, MaxAttempts: 3}, 429, "", true},
+		{"status mismatch",
+			&RetryPolicy{OnStatus: []int{429}, MaxAttempts: 3}, 400, "429", false},
+		{"status + body match",
+			&RetryPolicy{OnStatus: []int{400}, OnBodyContains: []string{"TM.00001041"}, MaxAttempts: 3},
+			400, `{"error_code":"TM.00001041"}`, true},
+		{"status matches but body does not",
+			&RetryPolicy{OnStatus: []int{400}, OnBodyContains: []string{"TM.00001041"}, MaxAttempts: 3},
+			400, `{"error_code":"TM.00001005"}`, false},
+		{"empty substring never matches",
+			&RetryPolicy{OnStatus: []int{400}, OnBodyContains: []string{""}, MaxAttempts: 3}, 400, "x", false},
+		{"any of several body signals",
+			&RetryPolicy{OnStatus: []int{400, 429}, OnBodyContains: []string{"TM.00001041", "TPM limit"}, MaxAttempts: 3},
+			429, `{"message":"TPM limit exceeded"}`, true},
+	}
+	for _, tc := range cases {
+		if got := tc.policy.ShouldRetry(tc.status, tc.body); got != tc.want {
+			t.Errorf("%s: ShouldRetry(%d, %q) = %v, want %v", tc.name, tc.status, tc.body, got, tc.want)
+		}
+	}
+}
+
+func TestRetryPolicyDelay(t *testing.T) {
+	cases := []struct {
+		name   string
+		policy *RetryPolicy
+		want   time.Duration
+	}{
+		{"nil", nil, 10 * time.Second},
+		{"explicit", &RetryPolicy{Delay: "250ms"}, 250 * time.Millisecond},
+		{"garbage falls back", &RetryPolicy{Delay: "not-a-duration"}, 10 * time.Second},
+		{"negative falls back", &RetryPolicy{Delay: "-5s"}, 10 * time.Second},
+	}
+	for _, tc := range cases {
+		if got := tc.policy.DelayDuration(); got != tc.want {
+			t.Errorf("%s: DelayDuration() = %v, want %v", tc.name, got, tc.want)
+		}
+	}
+}

internal/provider/retry_policy_test.go

@@ -0,0 +1,131 @@
+package provider
+
+import (
+	"context"
+	"fmt"
+	"net/http"
+	"net/http/httptest"
+	"sync/atomic"
+	"testing"
+
+	"llmsproxy/internal/config"
+	"llmsproxy/internal/types"
+)
+
+// 上游重试是**内核的通用机制**，判据纯数据（status + body 子串），
+// 目的是让"账号级限流、等一会儿就好"这类拒绝不冒到客户端。
+//
+// 真实事故（2026-10-20）：CodeArts 账户并发会话上限 3 个，连续第 3~4 个请求
+// 回 400 TM.00001041，最后一个请求结束后约 44~60 秒才释放；流式更糟 ——
+// 上游用 HTTP 200 + error 帧表达拒绝，内核只能报 `empty stream`。
+// 参考实现的处理是每 10s 重试一次 chat 直到放行，本测试钉住同一语义。
+func TestChatRetriesDeclaredUpstreamRejection(t *testing.T) {
+	var hits int32
+	up := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		n := atomic.AddInt32(&hits, 1)
+		if n <= 2 {
+			w.WriteHeader(http.StatusBadRequest)
+			fmt.Fprint(w, `{"error_code":"TM.00001041","error_msg":"并发会话数已达上限(3个)"}`)
+			return
+		}
+		fmt.Fprint(w, `{"choices":[{"message":{"content":"hi"},"finish_reason":"stop"}]}`)
+	}))
+	defer up.Close()
+
+	s := src("codearts-like", up.URL, "openai", "m1")
+	s.Retry = &config.RetryPolicy{
+		OnStatus:       []int{400},
+		OnBodyContains: []string{"TM.00001041"},
+		Delay:          "1ms",
+		MaxAttempts:    4,
+	}
+	p := newTestProvider(t, s)
+	resp, err := p.Chat(context.Background(), &types.ChatRequest{
+		Model:    "m1",
+		Messages: []types.ChatMessage{{Role: "user", Content: types.StringContent("hi")}},
+	})
+	if err != nil {
+		t.Fatalf("命中重试策略后应成功，实得错误: %v", err)
+	}
+	if resp.Content != "hi" {
+		t.Fatalf("content = %q", resp.Content)
+	}
+	if got := atomic.LoadInt32(&hits); got != 3 {
+		t.Fatalf("应恰好尝试 3 次（2 次失败 + 1 次成功），实得 %d", got)
+	}
+}
+
+// 同一状态码但 body 不匹配 ⇒ **不许重试**（400 也常用于真正的请求错误，
+// 无差别重试只会把错误延迟 3 分钟才暴露给用户）。
+func TestChatDoesNotRetryUnmatchedBody(t *testing.T) {
+	var hits int32
+	up := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt32(&hits, 1)
+		w.WriteHeader(http.StatusBadRequest)
+		fmt.Fprint(w, `{"error_code":"TM.00001005","error_msg":"无正常可用套餐"}`)
+	}))
+	defer up.Close()
+
+	s := src("codearts-like", up.URL, "openai", "m1")
+	s.Retry = &config.RetryPolicy{
+		OnStatus:       []int{400},
+		OnBodyContains: []string{"TM.00001041"},
+		Delay:          "1ms",
+		MaxAttempts:    5,
+	}
+	p := newTestProvider(t, s)
+	_, err := p.Chat(context.Background(), &types.ChatRequest{
+		Model:    "m1",
+		Messages: []types.ChatMessage{{Role: "user", Content: types.StringContent("hi")}},
+	})
+	if err == nil {
+		t.Fatal("不可重试的 body 必须失败")
+	}
+	if got := atomic.LoadInt32(&hits); got != 1 {
+		t.Fatalf("body 不匹配时只能尝试 1 次，实得 %d", got)
+	}
+}
+
+// 流式：只重试**握手**。命中策略 → 重试直到 200 才开始派发；一旦开始流式
+// 就不再重试（重复输出比报错更糟）。
+func TestChatStreamRetriesHandshakeOnly(t *testing.T) {
+	var hits int32
+	up := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		n := atomic.AddInt32(&hits, 1)
+		if n == 1 {
+			w.WriteHeader(http.StatusBadRequest)
+			fmt.Fprint(w, `{"error_code":"TM.00001041","error_msg":"并发会话数已达上限(3个)"}`)
+			return
+		}
+		w.Header().Set("Content-Type", "text/event-stream")
+		fmt.Fprint(w, "data: {\"choices\":[{\"delta\":{\"content\":\"hi\"},\"finish_reason\":\"stop\"}]}\n\n")
+		fmt.Fprint(w, "data: [DONE]\n\n")
+	}))
+	defer up.Close()
+
+	s := src("codearts-like", up.URL, "openai", "m1")
+	s.Retry = &config.RetryPolicy{
+		OnStatus:       []int{400},
+		OnBodyContains: []string{"TM.00001041"},
+		Delay:          "1ms",
+		MaxAttempts:    4,
+	}
+	p := newTestProvider(t, s)
+	ch, err := p.ChatStream(context.Background(), &types.ChatRequest{
+		Model:    "m1",
+		Messages: []types.ChatMessage{{Role: "user", Content: types.StringContent("hi")}},
+	})
+	if err != nil {
+		t.Fatalf("握手重试后应成功: %v", err)
+	}
+	var content string
+	for ck := range ch {
+		content += ck.Content
+	}
+	if content != "hi" {
+		t.Fatalf("content = %q", content)
+	}
+	if got := atomic.LoadInt32(&hits); got != 2 {
+		t.Fatalf("应握手重试 1 次（共 2 次请求），实得 %d", got)
+	}
+}