frivas/image-processing-api-project

Udacity Image Processing API Project

Project Specification

Comments

• The backend is inside the api folder.

• The scripts available are:

  • npm run test it will transpile and then execute jasmine.
  • npm run perfect it will execute prettier and lint.
  • npm run compile it will transpile the code using tsc.
  • npm run build it will clean the dist/ folder, execute the perfect, compile and compile scripts.
  • npm run start it will start nodemon on src/index.ts
  • You can also execute node dist/ to run the app.

• Endpoints:

  • http://localhost:3003/ = Will show a message stating the server is up and running.

  • http://localhost:3003/api/images = Accepts the following query parameters:

    • : Name of the file without extension that has to be placed in the full directory or uploaded through the upload endpoint. (Required)

    • : It is a number that specifies the width of the thumbnail file to be generated. (Optional. Default 200.)

    • : It is a number that specifies the height of the thumbnail file to be generated. (Optional. Default 200.)

    • This endpoint will return a message stating that a field is missing and it is required.

    • Sample requests:

      http://localhost:3003/api/images?filename=pollock&width=500&height=300

      http://localhost:3003/api/images?filename=fjord&width=500

      http://localhost:3003/api/images?filename=mountfuji&height=300

      http://localhost:3003/api/images?filename=santamonica

  • http://localhost:3003/api/images/upload = Will allow you to upload one or more files. It will show messages when:

    • There are no files to upload.
    • Files types are not supported. Only jpg, png and jpeg are supported.
    • Uploading is currently limited to 10 files at the time.

  • http://localhost:3003/api/images/list = Will return a JSON object that contains information about the files in the thumb directory.

• It is possible to create thumbs of the same images on different sizes.

• I had to do a lot of research to write the tests for the uploading endpoint, it is different when uploading 1 file and more than 1.

• There is a logger that shows a message on the console every time an endpoint is visited.

• I have included a copyAssets.ts script that copies all the images and needed files for the tests.

Note to reviewer:

• There is one test I would have loved to include. Testing multiple files upload, however I didn't manage to write the test for that one, I managed to do it for a single valid file, for an invalid file image and for an invalid file. Would you please guide me on this one?

  I wrote what I thought would be a correct test however it fails and I am not entirely sure why it is not working

  it('Upload a list of valid image files', async () => {
      const requestPOST = request.post('/api/images/upload');
      for (const file of validListFilePath) {
        requestPOST.attach('files', file);
      }
      const response = await requestPOST;
      expect(response.status).toBe(200);
      expect(response.text).toEqual(
        JSON.stringify({
          message: ['pollock.jpg', 'wassily-kandinsky.jpg']
        })
      );
    });

  In fact if I test that endpoint it returns what I expect. Please, notice the order of the items in the array. I also tried with .toContain() checking there is a message in the response however the result of the test is Error: Aborted.

  

Image Processing API

Setup and Architecture

Set up a project structure that promotes scalability

• Source code is kept separate from compiled code.
• All tests should be contained in their own folder.
• Separate modules are created for any processing.

Set up an npm project

• Package.json should contain both devDependencies, and dependencies.
• Scripts should be created for testing, linting/prettier, starting the server, and compiling TS.
• Build script should run without error.

Functionality

Add and use Express to a node.js project

• Start script should run without error.
• Provided endpoint should open in the browser with status 200.

Follow middleware documentation to use middleware to create an API

• Accessing the provided URL with image information should successfully resize an image and save it to disk on first access, then pull from disk on subsequent access attempts.
• An error message should be provided to the user when an image has failed to process or does not exist.

Code Quality

Write relevant unit tests with Jasmine and SuperTest to improve code quality and refactoring

• Test script runs and all tests created pass.
• There is at least 1 test per endpoint and at least one test for image processing.

Utilize TypeScript to avoid errors and improve maintainability

• All code in the SRC folder should use the .ts filetype.
• Functions should include typed parameters and return types and not use the any type.
• Import and Export used for modules.
• Build script should successfully compile TS to JS.

Write well-formatted linted code

• Prettier and Lint scripts should run without producing any error messages.
• An error message should be provided to the user when an image has failed to process or does not exist.

Suggestions to Make Your Project Stand Out!

• Add additional processing to accept and output other image formats than JPG.
• Modify the thumbnail filename to include the image size to allow for multiple sizes of the same image.
• Further explore the options in the Sharp module and add additional processing options.
• Add in logging to record when images are processed or accessed.
• Create a front-end for uploading more images to the full-size directory.
• Create a front-end that displays a thumbnail directory.
• Create a front-end that allows for the selection of how to process a selected image.