Fix setParsableMimeTypes()

[superpenguin612]

Addresses issues #352 and #369, and PR #432.

The setParsableMimeTypes option does not actually prevent a crawl of a URL, it only prevents URLs that originate from that page from being crawled.

If executeJavascript is enabled in tandem, the setParsableMimeTypes option becomes completely broken; as it turns out, the method with which mime types are ignored conflicts with Browsershot's JavaScript execution, since it overwrites the $body variable:

if ($this->crawler->mayExecuteJavaScript()) {
    $body = $this->getBodyAfterExecutingJavaScript($crawlUrl->url);
    $response = $response->withBody(Utils::streamFor($body));
}

This PR prevents pages with non-parsable mime types from being crawled completely, edits the test for mime types to be more representative of what this feature should accomplish, and adds an additional test to validate that the issue is fixed when executing JS, as well.

[freekmurze]

Thanks!