refactor(#637): Refactor BodyPix and CharRNN tests to jest

.eslintrc.js

@@ -13,7 +13,9 @@ module.exports = {
   env: {
     browser: true,
     jasmine: true,
+    "jest/globals": true,
   },
+  plugins: ["jest"],
   overrides: [
     {
       files: ["examples/**"],

package.json

@@ -68,6 +68,7 @@
     "eslint-loader": "2.0.0",
     "eslint-nibble": "^5.1.0",
     "eslint-plugin-import": "2.9.0",
+    "eslint-plugin-jest": "^26.0.0",
     "extract-text-webpack-plugin": "4.0.0-beta.0",
     "ghooks": "2.0.2",
     "html-webpack-plugin": "^3.0.7",
@@ -118,4 +119,4 @@
     "face-api.js": "~0.22.2",
     "onchange": "^6.1.0"
   }
-}
+}

src/BodyPix/__tests__/BodyPix.test.js

@@ -1,10 +1,10 @@
+/* eslint-disable no-use-before-define */
 // Copyright (c) 2018 ml5
 //
 // This software is released under the MIT License.
 // https://opensource.org/licenses/MIT
-import bodyPix from "../index";
-import * as tf from "@tensorflow/tfjs";
 import * as tfBodyPix from "@tensorflow-models/body-pix";
+import bodyPix from "../index";
 
 jest.mock("@tensorflow-models/body-pix");
 
@@ -14,6 +14,70 @@ const BODYPIX_DEFAULTS = {
   segmentationThreshold: 0.5,
 };
 
+const mockBodyPix = {
+  mobileNet: {
+    predict: jest.fn(),
+    convToOutput: jest.fn(),
+    dispose: jest.fn(),
+  },
+  predictForSegmentation: jest.fn(),
+  predictForPartMap: jest.fn(),
+  estimatePersonSegmentationActivation: jest.fn(),
+  estimatePersonSegmentation: jest.fn(),
+  estimatePartSegmentationActivation: jest.fn(),
+  estimatePartSegmentation: jest.fn(),
+  dispose: jest.fn(),
+};
+
+describe("bodyPix", () => {
+  let bp;
+
+  beforeAll(async () => {
+    tfBodyPix.load = jest.fn().mockResolvedValue(mockBodyPix);
+    bp = await bodyPix();
+  });
+
+  afterAll(async () => {
+    jest.clearAllMocks();
+  });
+
+  it("Should create bodyPix with all the defaults", async () => {
+    expect(bp.config.multiplier).toBe(BODYPIX_DEFAULTS.multiplier);
+    expect(bp.config.outputStride).toBe(BODYPIX_DEFAULTS.outputStride);
+    expect(bp.config.segmentationThreshold).toBe(BODYPIX_DEFAULTS.segmentationThreshold);
+  });
+
+  it("segment calls segmentInternal and returns", async () => {
+    bp.segmentInternal = jest.fn().mockResolvedValue({
+      segmentation: {
+        data: [],
+        width: 200,
+        height: 50,
+      },
+    });
+    const img = await getImageData();
+    const results = await bp.segment(img);
+    expect(bp.segmentInternal).toHaveBeenCalledTimes(1);
+    expect(results.segmentation.width).toBe(200);
+    expect(results.segmentation.height).toBe(50);
+  });
+
+  it("segmentWithParts calls segmentWithPartsInternal and returns", async () => {
+    bp.segmentWithPartsInternal = jest.fn().mockResolvedValue({
+      segmentation: {
+        data: [],
+        width: 200,
+        height: 50,
+      },
+    });
+    const img = await getImageData();
+    const results = await bp.segmentWithParts(img);
+    expect(bp.segmentWithPartsInternal).toHaveBeenCalledTimes(1);
+    expect(results.segmentation.width).toBe(200);
+    expect(results.segmentation.height).toBe(50);
+  });
+});
+
 async function getImage() {
   const img = new Image();
   img.crossOrigin = true;
@@ -49,43 +113,3 @@ async function getImageData() {
   const img = new ImageData(arr, 200);
   return img;
 }
-
-describe("bodyPix", () => {
-  let bp;
-
-  beforeAll(async () => {
-    tfBodyPix.load = jest.fn().mockResolvedValue({});
-    bp = await bodyPix();
-  });
-
-  it("Should create bodyPix with all the defaults", async () => {
-    console.log(bp);
-    expect(bp.config.multiplier).toBe(BODYPIX_DEFAULTS.multiplier);
-    expect(bp.config.outputStride).toBe(BODYPIX_DEFAULTS.outputStride);
-    expect(bp.config.segmentationThreshold).toBe(BODYPIX_DEFAULTS.segmentationThreshold);
-  });
-
-  // it("segment takes ImageData", async () => {
-  //   const img = await getImageData();
-  //   const results = await bp.segment(img);
-  //   // 200 * 50 == 10,000 * 4 == 40,000 the size of the array
-  //   expect(results.segmentation.width).toBe(200);
-  //   expect(results.segmentation.height).toBe(50);
-  // });
-
-  // describe("segmentation", () => {
-  //   it("Should segment an image of a Harriet Tubman with a width and height of 128", async () => {
-  //     const img = await getImage();
-  //     await bp.segment(img).then(results => {
-  //       expect(results.segmentation.width).toBe(128);
-  //       expect(results.segmentation.height).toBe(128);
-
-  //       expect(results.segmentation.width).toBe(128);
-  //       expect(results.segmentation.height).toBe(128);
-
-  //       expect(results.segmentation.width).toBe(128);
-  //       expect(results.segmentation.height).toBe(128);
-  //     });
-  //   });
-  // });
-});