Support Driverstation log version 4.

Nobody has claimed this yet.

dsevents looks the same between version 3 and 4. By taking a small file, opening it in driverstation, and stepping through header, LV timestamp, LV string.
It seems to match up.

The python script did not work for the dslog. It read things in…but a few things were off:

* Looking in Driverstation, the voltage seemed to range between 12.5 to 12.7 volts from what I could see. * The Robot was never placed into autonomous mode. Driver station also showed 24 PDP’s and all of them were noted as zero.
* The python script though created a csv where auto was true, and voltage was all over the place.
* I opened the log using od -ha, and looked for repeating patterns.
* It seemed like after the 10 byte information of each line, there were 68 bytes before similar 10 bytes started repeating again.
* My example dslog was 489470 bytes long. With a 20 byte version/LV timestamp header, that left 48950 bytes that should be evenly divisible, and 48950/78=6275
* If I change the python to assign 68 bytes to the PDP, it’s still not correct, as some of the original 16 PDP’s are showing values.
* However the values for the first 10 bytes match with what would be reasonable for when the driver station was logging.

If I get closer to the format, I will open a pull request.

If I can get closer to what the format should be, I will open a pull request for the python script.

Start by locating the Python script that parses the dslog and compare its assumptions with the Driver Station version 4 observations in this issue. Use a small version 4 log and od -ha to verify the header, timestamp, string, PDP, voltage, and autonomous fields; done means version 4 values parse correctly without breaking existing logs.

Written by the indexing model from the issue text.