Improve support for gen2 and gen3 tags

[gelotus]

Ok, my ocd wins..
There is a rework including @AdamLaurie commits.
Tested before and after on genuine and gen2.
Now you can rewrite gen2 with W, please see commit descs

[AdamLaurie]

Nice work @gelotus - thank you! But please modify unlock_card() so it does not write on READ commands. If this is the only way to detect Magic cards then we should print a warning that we are about to try to write to the card and ask if the user wants to go ahead or skip that test.

[gelotus]

Ok, fixed. About your second question, about one month ago I've bought a lot of 500 gen2 tags,for what I see, they have a random uid from factory. I haven't checked rigorously, but from my memory I remember so. Maybe when I'll buy another batch I'll check better. If we use is_directwrite() we break gen3 so the tag burn during check.
About the warning and the user intervention, this will break a possible scripted using of the program, for example I'm using it for quickly format that 500 tags what I have talked before. We can add an override switch but I think would be better to rework the args parser first. Anyway I don't see troubles whit this behaviour, if we try to W to a genuine or gen3 tag, writing simply fails, for gen1 it continues with its normal behaviour. If you think about a possible overwrite of the content, the same happens on a gen1.

[gelotus]

Sorry, I hadn't noticed that typo..

[AdamLaurie]

Note that I have merged this as the non-writing of full sectors is a bigger issue but there is still a problem here with 4k DirectWrite - on a read operation standard 4K cards are being recognised as DirectWrite - @gelotus want to take another look?

[gelotus]

Yes I'll could take a look , but unfortunately I haven't a 4k tag for testing. We'll see in the next days if I'll find one