Support joins on prefixes of arbitrary length #48
Conversation
They belong on the methods, a lá `HashMap`.
| @@ -69,41 +69,41 @@ fn main() { | |||
| let requires = iteration2.variable::<(Region, Borrow, Point)>("requires"); | |||
| requires.insert(Vec::new().into()); | |||
|
|
|||
| let requires_rp = iteration2.variable::<((Region, Point), Borrow)>("requires_rp"); | |||
| let requires_bp = iteration2.variable::<((Borrow, Point), Region)>("requires_bp"); | |||
| let requires_rp = iteration2.variable::<(Region, Point, Borrow)>("requires_rp"); | |||
There was a problem hiding this comment.
Some of these are redundant now, although this is partly because it was written without leapjoin.
There was a problem hiding this comment.
I think Frank initially had these more like documentation of how to use the API than really optimal, runnable examples. There are docs and doctests for leapjoins IIRC so it's not terrible that this example doesn't use them. Don't feel obligated to update them.
We could have that as a "good first issue" "help wanted" issue if some contributor wanted to try and clean that up. It would teach them a bit about the API, contrast the example with the polonius analyses, and so on.
There was a problem hiding this comment.
A better example might switch requires to have the same layout as requires_rp or requires_bp and reorder it at the end. Without that and without leapjoin (which I agree doesn't really belong here), I think the current version is optimal.
There was a problem hiding this comment.
I guess maybe having a single iteration instead of two, or actually feeding input facts into the computation, and so on, could turn it into a slightly more realistic example if someone wanted to do so.
As for the other example, graspan1, I think there were some input facts in one of Franks's blog posts a few years back.
Currently,
datafrogrequires that all joins occur on tuples of the form(K, V). This becomes unpleasant for relations with more than two elements, as well as for joins on several elements. I found this limitation annoying when first trying to writedatafrogprograms.Because
datafrogis implemented on top of sorted data structures, we should be able to join on arbitrarily many tuple elements, so long as they are in order at the front of the tuple. This PR implements joins on arbitrary prefixes. As a result, you can specify all your variables and relations as a flat tuple, and wait to select what prefix you want for each individual join. For example,(A, B, C)could be joined with both(A, B, X)using(A, B)as the shared prefix and with(A, Y)usingAas the shared prefix without cloning it.Unfortunately, this is a breaking change, since we must copy prefixes and suffixes now instead of passing references to
KandV.