HTMLShots

yet another phantom.js screenshots app

HTMLShots is a simple phantom.js app that creates screenshots from HTML files. Unlike other similar screenshot solutions, HTMLShots makes use of the WebServer module so there is no need to spawn one process for each screenshots request.

requirements

PhantomJS >= 1.6.0
ttf-mscorefonts-installer (optional, only if you want Windows fonts)

install

package

We maintain packages for Ubuntu 12.04 LTS (precise) in our PPA:

$ sudo apt-get install -y python-software-properties
$ sudo add-apt-repository ppa:templateria/htmlshots
$ sudo apt-get update
$ sudo apt-get install htmlshots

Note: Ubuntu 12.04 LTS (precise) ships PhantomJS 1.4.0, so our PPA repository provides an updated version (1.6.0), backported from Ubuntu 12.10 (quantal).

By default the HTMLShots service is disabled. To enable it, open /etc/default/htmlshots and set ENABLED=1, then:

$ sudo start htmlshots

Note: HTMLShots supports running multiple instances (processes) with the INSTANCES parameter in /etc/default/htmlshots. This is particularly useful if you are building a service on top of HTMLShots and wants to take advantage of running multiple processes.

If you want to build the package by yourself:

$ git clone https://github.com/templateria/htmlshots
$ cd htmlshots
$ dpkg-buildpackage

git

When installing directly from git, first make sure you have installed PhantomJS 1.6.0, then:

$ git clone https://github.com/templateria/htmlshots

To run HTMLShots on port 3000 (default):

$ cd htmlshots
$ phantomjs --config=conf/htmlshots.json htmlshots.js

If you want HTMLShots to listen in a different port, pass an additional parameter:

$ phantomjs --config=conf/htmlshots.json htmlshots.js 8080

how to use

Note: All screenshots are PNG images returned as base64-encoded strings.

HTMLShots is not supposed to be a REST API, just a simple way to retrieve screenshots from HTML files.

Currently, the following methods are supported:

PUT

Send a PUT request with the HTML contents in the request body. This example uses cURL's -T (upload file) parameter to simplify things:

$ curl -T index.html http://localhost:3000/ | base64 -d > screenshot.png

POST

Send a POST request with the HTML contents URL-encoded in the html parameter.

$ curl -X POST -d html=contents http://localhost:3000/ | base64 -d > screenshot.png

TODO: Send a POST request with the URL for the desired page in the url parameter.

$ curl http://localhost:3000/?url=http://www.google.com/ | base64 -d > screenshot.png

ppadron/htmlshots