Hash support about fsql

@kshvmdn This doesn't look too bad to get in, I haven't looked to much at the parsing logic but should be able to follow the logic for the other key words. Anything other than SHA-1 We'd like to see?

[kashav]

@jonesbrianc26 Sounds good!

Let's do this by introducing a new attribute, hash. I think the best way to specify which algorithm to use is with an attribute modifier, so perhaps

SELECT SHA1(hash) ...
... WHERE SHA256(hash) = ...

or

SELECT FORMAT(hash, SHA1) ...
... WHERE FORMAT(hash, SHA256) = ...

For the time being, lets stick with SHA-1 and see how that goes (crypto/sha1 will probably be useful for this). We should be able to easily abstract this to support multiple different algorithms in the future.

Turns out this isn't too bad, Though will probably need to rethink the output function a little bit to be smarter with spacing. I'll play around with it a little bit @kshvmdn what do you think?

[kashav]

Awesome!

The original output format was meant to match that of ls -l, I think we can stick with that?

[mode] [size] [time] [hash] [name]

Also, I think by default, we should show a truncated version of the hash (maybe 7-10 characters, like Git), and we can include a FULL function to show the full hash (perhaps even a TRUNCATE(, n) function to specify how many characters to show?). Feel free to leave it at the full hash for now though, we can add these functions later (or go ahead and implement it, up to you! 😄).

perhaps just {hashtype}(hash, n) though, TRUNCATE(,n) could be useful in general and would be a extremely reusable function so it would make sense to use that under the hood.

[kashav]

You're right, that works better!

If n isn't provided, let's default it to 7 for the time being, just so the output of all doesn't overflow.

Also, as far as formatting goes, what would you like to display in place of the hash for directories?

Currently I'm displaying 1 - for every character in the hash to preserve equal spacing, maybe once I get the PR up we can discuss that a bit more.

[kashav]

Sounds good to me!

[kashav]

Done in #31 – currently only supports SHA-1, feel free to suggest other functions.