Parse <img> tags for crawling, if the option ParseImageTags is set.

[Kleissner]

Implement the code for #69 Parsing image tags to include them in crawling.
By default the new setting ParseImageTags is set to false, so the current default behavior does not change.

[mna]

Hello Peter,

Thanks for the PR, makes sense and looks good to me, could you just add a unit test that exercises that new feature (w.opts.ParseImageTags set to true, and checking that image source URLs have actually been added to the crawler)?

Thanks,
Martin