first commit

apps/miniprogram/pages/terminal/terminalStdoutRenderPolicy.test.ts

@@ -0,0 +1,292 @@
+import { describe, expect, it } from "vitest";
+
+const {
+  resolveTerminalStdoutOverlayDecision,
+  resolveTerminalStdoutRenderDecision,
+  shouldDeferTerminalStdoutRender
+} = require("./terminalStdoutRenderPolicy.js");
+
+describe("terminalStdoutRenderPolicy", () => {
+  it("小 backlog 不会延后视图提交", () => {
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 4096,
+        pendingReplayBytes: 1024,
+        nextSlicesSinceLastRender: 1,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toBe(false);
+  });
+
+  it("中等 backlog 会先累计到阈值再提交", () => {
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 24 * 1024,
+        pendingReplayBytes: 3 * 1024,
+        nextSlicesSinceLastRender: 3,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toBe(true);
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 24 * 1024,
+        pendingReplayBytes: 8 * 1024,
+        nextSlicesSinceLastRender: 3,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toBe(false);
+  });
+
+  it("大 backlog 会使用更高阈值，避免频繁整包 setData", () => {
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 512 * 1024,
+        pendingReplayBytes: 16 * 1024,
+        nextSlicesSinceLastRender: 12,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toBe(true);
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 512 * 1024,
+        pendingReplayBytes: 16 * 1024,
+        nextSlicesSinceLastRender: 32,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toBe(false);
+  });
+
+  it("用户输入或终端响应存在时必须立即提交", () => {
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 512 * 1024,
+        pendingReplayBytes: 1024,
+        nextSlicesSinceLastRender: 1,
+        pendingResponseCount: 1,
+        yieldedToUserInput: false
+      })
+    ).toBe(false);
+    expect(
+      shouldDeferTerminalStdoutRender({
+        remainingBytes: 512 * 1024,
+        pendingReplayBytes: 1024,
+        nextSlicesSinceLastRender: 1,
+        pendingResponseCount: 0,
+        yieldedToUserInput: true
+      })
+    ).toBe(false);
+  });
+
+  it("会给出本轮提交或延后的决策原因，便于诊断 render 频率", () => {
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 512 * 1024,
+        pendingReplayBytes: 1024,
+        nextSlicesSinceLastRender: 1,
+        pendingResponseCount: 0,
+        yieldedToUserInput: true
+      })
+    ).toMatchObject({
+      defer: false,
+      reason: "user_input",
+      policy: "interactive"
+    });
+
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 24 * 1024,
+        pendingReplayBytes: 3 * 1024,
+        nextSlicesSinceLastRender: 3,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toMatchObject({
+      defer: true,
+      reason: "defer_backlog",
+      policy: "medium_backlog"
+    });
+
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 24 * 1024,
+        pendingReplayBytes: 8 * 1024,
+        nextSlicesSinceLastRender: 3,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false
+      })
+    ).toMatchObject({
+      defer: false,
+      reason: "pending_bytes_threshold",
+      policy: "medium_backlog"
+    });
+  });
+
+  it("render 冷却期间即使进入尾段，也会继续延后视图提交", () => {
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 4096,
+        pendingReplayBytes: 2048,
+        nextSlicesSinceLastRender: 2,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false,
+        timeSinceLastRenderMs: 80,
+        taskDone: false
+      })
+    ).toMatchObject({
+      defer: true,
+      reason: "render_cooldown",
+      policy: "medium_backlog"
+    });
+
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 4096,
+        pendingReplayBytes: 2048,
+        nextSlicesSinceLastRender: 2,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false,
+        timeSinceLastRenderMs: 280,
+        taskDone: false
+      })
+    ).toMatchObject({
+      defer: false,
+      reason: "remaining_below_threshold",
+      policy: "medium_backlog"
+    });
+
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 0,
+        pendingReplayBytes: 1024,
+        nextSlicesSinceLastRender: 2,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false,
+        timeSinceLastRenderMs: 20,
+        taskDone: true
+      })
+    ).toMatchObject({
+      defer: false,
+      reason: "task_complete"
+    });
+  });
+
+  it("高 backlog 时会进入更激进的降级模式，优先丢弃中间帧", () => {
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 4096,
+        pendingReplayBytes: 4096,
+        nextSlicesSinceLastRender: 4,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false,
+        timeSinceLastRenderMs: 600,
+        taskDone: false,
+        totalRawBytes: 80 * 1024,
+        pendingStdoutBytes: 48 * 1024,
+        pendingStdoutSamples: 160,
+        schedulerWaitMs: 2400,
+        activeStdoutAgeMs: 1800
+      })
+    ).toMatchObject({
+      defer: true,
+      reason: "defer_critical_backlog",
+      policy: "critical_backlog"
+    });
+
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 4096,
+        pendingReplayBytes: 28 * 1024,
+        nextSlicesSinceLastRender: 4,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false,
+        timeSinceLastRenderMs: 600,
+        taskDone: false,
+        totalRawBytes: 80 * 1024,
+        pendingStdoutBytes: 48 * 1024,
+        pendingStdoutSamples: 160,
+        schedulerWaitMs: 2400,
+        activeStdoutAgeMs: 1800
+      })
+    ).toMatchObject({
+      defer: false,
+      reason: "pending_bytes_threshold",
+      policy: "critical_backlog"
+    });
+
+    expect(
+      resolveTerminalStdoutRenderDecision({
+        remainingBytes: 4096,
+        pendingReplayBytes: 4096,
+        nextSlicesSinceLastRender: 24,
+        pendingResponseCount: 0,
+        yieldedToUserInput: false,
+        timeSinceLastRenderMs: 600,
+        taskDone: false,
+        totalRawBytes: 80 * 1024,
+        pendingStdoutBytes: 48 * 1024,
+        pendingStdoutSamples: 160,
+        schedulerWaitMs: 2400,
+        activeStdoutAgeMs: 1800
+      })
+    ).toMatchObject({
+      defer: false,
+      reason: "slice_threshold",
+      policy: "critical_backlog"
+    });
+  });
+
+  it("stdout 持续输出时会对 overlay 做节流，但最终一帧仍会同步", () => {
+    expect(
+      resolveTerminalStdoutOverlayDecision({
+        isFinalRender: false,
+        yieldedToUserInput: false,
+        overlayPassCount: 0,
+        timeSinceLastOverlayMs: 0
+      })
+    ).toMatchObject({
+      sync: true,
+      reason: "first_render"
+    });
+
+    expect(
+      resolveTerminalStdoutOverlayDecision({
+        isFinalRender: false,
+        yieldedToUserInput: false,
+        overlayPassCount: 2,
+        timeSinceLastOverlayMs: 80
+      })
+    ).toMatchObject({
+      sync: false,
+      reason: "overlay_throttled"
+    });
+
+    expect(
+      resolveTerminalStdoutOverlayDecision({
+        isFinalRender: false,
+        yieldedToUserInput: false,
+        overlayPassCount: 2,
+        timeSinceLastOverlayMs: 260
+      })
+    ).toMatchObject({
+      sync: true,
+      reason: "overlay_cooldown_elapsed"
+    });
+
+    expect(
+      resolveTerminalStdoutOverlayDecision({
+        isFinalRender: true,
+        yieldedToUserInput: false,
+        overlayPassCount: 2,
+        timeSinceLastOverlayMs: 80
+      })
+    ).toMatchObject({
+      sync: true,
+      reason: "task_complete"
+    });
+  });
+});