Skip to content

Latest commit

 

History

History
147 lines (110 loc) · 4.69 KB

README.md

File metadata and controls

147 lines (110 loc) · 4.69 KB
Raw

HTML PDF Converter

Docker Repository on Quay Build Status Build Status

Uses Chrome Headless to convert HTML to a PDF

Send a HTML or Mustache template and recieve a PDF stream as the response.

Install and start

Node App - Running a local html-pdf-instance in a docker container

Navigate to quay.io/ukhomeofficedigital/html-pdf-converter to find latest the tagged version. Docker will pull whichever version you specify.

For example, if the latest tagged version v2.4.3 then this command will need to be run:

docker pull quay.io/ukhomeofficedigital/html-pdf-converter:v2.4.3

Once completed you can check the image is available locally by running:

docker image list

All HOF forms run locally on port 8080 and in some cases port 8081 may also be in use; so the html-pdf-converter should be run on another port. Currently port 8082 is recommended.

docker run -t -i -p 8082:8080 quay.io/ukhomeofficedigital/html-pdf-converter:**<tag>**

Observe following in terminal:

2023-09-13T11:18:07.061Z - info: Listening on localhost:8080

Note: The terminal will say that the application is listening on port 8080, however you can verify which port the html-pdf-converter container is using by running:

docker ps -a

The Node app that is using the html to pdf converter will need to be run locally too. The service will need the following env variable:

  • PDF_CONVERTER_URL: If you are running a local PDF converter this is the url and port it is running on. This URL should be in the format PDF_CONVERTER_URL=http://localhost:<PORT>/convert.

  • In this example the PDF_CONVERTER_URL would be PDF_CONVERTER_URL=http://localhost:8082/convert

Upon a successful html to pdf conversion the response should look something like this:

2023-09-12T15:31:30.249Z - info:  status=201, method=POST, url=/convert, response_time=392, content_length=39644

Development

git clone [email protected]:UKHomeOffice/html-pdf-converter.git
cd html-pdf-converter
npm install
npm start

Example usage

Mustache and Data

curl -H "Content-Type:application/json" \
     -d '{
            "template": "'"\
              <html>\
                <head>\
                  <title>{{title}}</title>\
                </head>\
                <body>\
                  <h1>{{header}}</h1>\
                  <p>{{para}}</p>\
                </body>\
              </html>\
            "'",
            "data": {
              "title": "My title",
              "header": "My header",
              "para": "My content"
            }
         }' \
     -i localhost:8080/convert

HTML

curl -H "Content-Type:application/json" \
     -d '{
            "template": "'"\
              <html>\
                <head>\
                  <title>My title</title>\
                </head>\
                <body>\
                  <p>Hello world</p>\
                </body>\
              </html>\
            "'"
         }' \
     -i localhost:8080/convert

Response (example)

%PDF-1.4\n1 0 obj\n<<\n/Title ...

PDF Options

Chrome can accept a number of options to its PDF render function. These are documented here: https://github.com/GoogleChrome/puppeteer/blob/master/docs/api.md#pagepdfoptions

These can be set on a per-request basis by passing a pdfOptions object as part of your request body.

{
  "template":"<h1>Hello World!</h1>",
  "pdfOptions": {
    "printBackground": true
  }
}

External Resources

This service cannot resolve external resources such as linked CSS, JavaScript or images. If your template includes links to any of these resources, we suggest you use https://github.com/remy/inliner. The source for Inliner can be a URL, a file location or an HTML string.

Environment Variables

APP_PORT:    Defaults to 8080
APP_HOST:    Defaults to 'localhost'

Troubleshooting

If you get the following error locally, html-pdf-converter: Handling error message=Could not find browser revision 756035. Run "npm install" or "yarn install" to download a browser binary.

Then you may need to manually install puppeteer npm i puppeteer