phantomjs is a server side headless web browser. Use it in conjunction with node js to take screenshots of web pages. I specifically need to crop an exact rectangle on a web page for which I will provide the coordinates and dimensions. Ie I wanna crop parts of web pages.
Go to the phantom js site and u will find examples. Here it is without the cropping: https://github.com/ariya/phantomjs/blob/master/examples/rasterize.js . Similar to the viewport property, there is a clipReq property where u can set top and left positions to start at.
I need to upload the result jpg to amazon s3. This needs to be a simple web service where I will pass a URL to take a screenshot and its dimensions and top/left starting coordinate. I will supply the s3 bucket as a parameter too.
Put it in a specified bucket in a folder named after the domain name of the site in a file named after the URL path and file name and dimensions and coordinates. I should be able to reproduce the image's URL based upon the location path on s3 if I needed. That's the idea.