Ignore http GET data after the useful data.

What's the easiest way to strip the stuff before the useful data?

The easiest way is to not store the useless data. It appears that the "useful" data is preceded by two carriage return/line feed combination. So, if a character is not a CR, then you have not yet reached the useful data. If it is, then maybe you are getting close. If the next character is not a LF, you are not.

If the next character is not a CR, you have reached the start of a new record of "useless" data. Repeat reading and dumping data until you see CR, LF, CR, LF. When you see the second LF, with nothing before it but the CR, LF, CR sequence, start saving the data.