Encoding issues with Shodan HTTP content

[3c7]

Shodan does some weird things with UTF-8 HTTP content which results in an error during string encoding (before hashing). Temporarily, the content gets replaced with an empty string, as I couldn't come up with a proper solution. The censys http response for the same host is fine:

shodan.json
censys.json

Thanks to @lo-chr for finding and reporting the bug.