Highlight data classes

[aremazeilles]

Fix #6

I restructured the file to highlight the data types, as commented in #6. I added a 4th class, for raw data file.

Some additional points to mention:

• the bipedal system section is unclear to me. I gathered all subsections related to it, and it does not seem to be completely consistent. Would need revision. I think we have several issues related to that aspect already: #5, #20, #21
• I simplified the testbed naming aspects. For me, a protocol involves one setting file, whatever is the equipment involved. So there is no need reminding the protocol name in the testbed name.
• If a testbed can change along time (like changing the slope), I suggest adding a number to identify it. We should also refer to that number in other recorded file. Other issue?

[aremazeilles]

Ready to merge @alfonsotecnalia ?

[alfonsotecnalia]

Done!