Added first_key and last_key to dataset

[sguada]

This PR depends on #1238.

Just added two more methods to dataset to be able to know which is the first_key and last_key. This would be very useful when the keys are generated sequentially.

@kmatzen Could you take a look?

[shelhamer]

@sguada before follow-ups shouldn't we first merge #1238? You'll just need to rebase this once the merge commit is in.

[sguada]

@shelhamer agreed, I'm waiting the confirmation from @kmatzen to merge #1238

[shelhamer]

We already recognize the beginning and end records for restarting prefetching, could you explain the use case for needing the first and last key instead?

[sguada]

To be able to know which is the first_key and the last_key without needing
to iterate.
In cases in which the keys are generated sequentially, that's all you need
to know the ranges of the keys, and in that way you can ask for any key in
that range.

Sergio

2014-10-18 10:06 GMT-07:00 Evan Shelhamer notifications@github.com:

We already recognize the beginning and end records for restarting
prefetching, could you explain the use case for needing the first and last
key instead?

—
Reply to this email directly or view it on GitHub
#1288 (comment).