Creating a parser of news from websites

We need a web programmer. You need to create a parser of news from news sites. We need a parser of news from news sites, at the moment we have 72 sites in the list from which you can take news, but it is important that the news is not only parsed, but also compared, that is, that the script automatically checks how much unique news is among the existing sites in the database and if the news is only on 5% of sites, then it is necessary to automatically add it to our site under the control of DataLife Engine.
News should be taken with all the news and videos and should cut out various inserts belonging to a particular news site, such as "our portal *** previously reported about" and so on.
Watermarks in the pictures will need to be cropped and at the same time bring the images to a certain format.
I am interested in the cost and timing of this parser.