Add trait for tuple splitting

Author: ecstatic-morse

src/lib.rs

@@ -15,6 +15,7 @@ mod iteration;
 mod join;
 mod map;
 mod merge;
+mod prefix;
 mod relation;
 mod test;
 mod treefrog;
@@ -23,6 +24,7 @@ mod variable;
 pub use crate::{
     iteration::Iteration,
     join::JoinInput,
+    prefix::Split,
     relation::Relation,
     treefrog::{
         extend_anti::ExtendAnti,

src/prefix.rs

@@ -0,0 +1,90 @@
+/// Trait for splitting a tuple into two at some index.
+///
+/// `P` is a tuple containing only the desired fields prior to the index. For some type `(A, B)`,
+/// `(A, B)`, `(A,)`, and `()` are all valid choices for `P`, but `(B,)` is not. `A` is also not a
+/// valid prefix, since `P` must be a tuple.
+pub trait Split<P>: Sized {
+    /// The remaining fields after the input has been split.
+    type Suffix: Sized;
+
+    /// Splits a tuple into a prefix and a suffix.
+    fn split(self) -> (P, Self::Suffix);
+
+    /// Returns the prefix of the tuple.
+    fn prefix(self) -> P {
+        self.split().0
+    }
+
+    /// Returns the suffix of the tuple.
+    fn suffix(self) -> Self::Suffix {
+        self.split().1
+    }
+}
+
+macro_rules! rev {
+    ($($T:ident)*) => {
+        rev!(@REV [$($T)*])
+    };
+    (@REV [$H:ident $($T:ident)*] $($R:ident)*) => {
+        rev!(@REV [$($T)*] $H $($R)*)
+    };
+    (@REV [$H:ident] $($R:ident)*) => {
+        rev!(@REV [] $H $($R)*)
+    };
+    (@REV [] $($R:ident)*) => { ($($R,)*) };
+}
+
+macro_rules! impls {
+    ($($t:ident)*) => {
+        impls!(@outer [$($t)*]);
+    };
+
+    (@outer [$h:ident $($t:ident)+]) => {
+        impls!(@inner [$h $($t)*] [$h $($t)*] []);
+        impls!(@outer [$($t)*]);
+    };
+    (@outer [$h:ident]) => {
+        impls!(@inner [$h] [$h] []);
+        impls!(@outer []);
+    };
+    (@outer []) => {};
+
+    (@inner [$($t:ident)+] [$p:ident $($pt:ident)*] [$($r:ident)*]) => {
+        impls!(@imp [$($t)*] [$p $($pt)*] [$($r)*]);
+        impls!(@inner [$($t)*] [$($pt)*] [$p $($r)*]);
+    };
+    (@inner [$($t:ident)+] [$p:ident] [$($r:ident)*]) => {
+        impls!(@imp [$($t)*] [$p] [$($r)*]);
+        impls!(@inner [$($t)*] [] [$p $($r)*]);
+    };
+    (@inner [$($t:ident)+] [] [$($r:ident)*]) => {
+        impls!(@imp [$($t)*] [] [$($r)*]);
+    };
+
+    (@imp [$($t:ident)+] [$($p:ident)*] [$($r:ident)*]) => {
+        impl<$($t),*> Split<rev!($($p)*)> for rev!($($t)*) {
+            type Suffix = ($($r,)*);
+
+            fn split(self) -> (rev!($($p)*), Self::Suffix) {
+                let rev!($($t)*) = self;
+                (rev!($($p)*), ($($r,)*))
+            }
+        }
+    };
+    (@imp [$($t:ident),+] [] []) => {
+        impl<$($t),*> Split<()> for rev!($($t),*) {
+            type Suffix = Self;
+
+            fn split(self) -> ((), Self) {
+                ((), self)
+            }
+        }
+    };
+}
+
+#[allow(non_snake_case)]
+mod impls {
+    use super::*;
+
+    impls!(F E D C B A);
+}