🖨 losen-pdf

PDF generator based on headless Chrome and puppeteer for making good looking PDFs for wizards made with losen.

How it works

Expects a POST request to /pdf with a JSON body with the following properties in the body:

• url - The url to generate a PDF of. Type: string. Required
• localStorageKey - The local storage key to set data for in the browser before printing. Must be set in combination with. localStorageData. Type: object from local storage.
• localStorageData – The data to set in local storage before printing. Must be set in combination with localStorageKey. Type: string.

Example

Payload

{
  "url": "https://dibk-prodok.firebaseapp.com",
  "localStorageKey": "reduxPersist:@WIZARD_STATE",
  "localStorageData: {
    "page": "page1",
    "competence": {
      "experience": 5,
      "education": 4
    },
    "companytype": "norwegian",
    "orgnr": {
      "orgid": "916898908",
      "status": "",
      "validApprovalAreas": "",
      "dataSG": false,
      "name": "NETLIFE DESIGN AS",
      "postcode": "0184",
      "postplace": "OSLO",
      "address": "Stenersgata 8",
      "dataOrg": true,
      "invalidOrg": false
    },
    "sgdata": false,
    "contactperson": {
      "name": "Kristoffer Brabrand",
      "email": "[email protected]",
      "phone": "98765432"
    }
  }
}

Request

POST /pdf HTTP/1.1
Host: dibk-pdf-lt0crv82h.now.sh
Content-Type: application/json
Accept: */*
Content-Length: 105
{
  "url": "https://dibk.no/apekatt",
  "localStorageKey": "hei",
  "localStorageData": {
    "blah": 123
  }
}

Response

The response is a binary blob with content type application/pdf.

🚀 Deployment

The PDF generator is built for hosting as a serverless function on now. It's super easy to get up and running, free for reasonable amounts of traffic and built on top of AWS lambda.

🚦 Before deploying

You want to fork this repo and edit the now.json file to give the app a name and set the environment variable ALLOWED_DOMAINS to a comma separated list of domains that you want to allow generating PDFs from.

If you want to, you can use .* as a wildcard in the urls, like this: .*.ngrok.io.

The process

1. Install now by running npm i -g now (only first time)
2. Issue the command now when standing in the project folder. Now will start building the service, give you an url and deploy it. If you haven't deployed with now before, or if it's a long time since last time you'll be prompted to sign in.
3. You should get a response similar to this when you deploy:

➜ now
> Deploying ~/development/netlife/dibk/pdf/losen-pdf under kbrabrand
> Using project dibk-pdf
> Synced 1 file (1.17KB) [1s]
> https://dibk-pdf-lt0crv82h.now.sh [v2] [in clipboard] [9s]
┌ pdf.js        Ready               [37s]
└── λ pdf.js (35.15MB) [bru1]
> Success! Deployment ready [47s]

Custom domain

Now supports pointing a custom domain to the deployed application through the use of the now alias function. If you want to point my.custom.domain.com to the instance, you issue this command

now alias https://dibk-pdf-lt0crv82h.now.sh my.custom.domain.com

now will check if the domain you want to use has been pointed to them already (which it need to be in order for the traffic to get there) and will guide you through the process if not.

Integrating with losen

In order to have a wizard use the pdf service you need to add the print service url to the meta portion of your schema, like this:

{
  "meta": {
    ...
    "pdfGeneratorUrl": "https://dibk-pdf-lt0crv82h.now.sh",
    "localStorageKey": "reduxPersist:@WIZARD_STATE"
  },
  "schema": [
    ...
  ]
}

With this in place, losen will use the print service to generate PDFs instead of triggering a regular print command in the browser.