YES

The TRS could be proven terminating. The proof took 165 ms.

The following DP Processors were used


Problem 1 was processed with processor DependencyGraph (74ms).
 | – Problem 2 was processed with processor SubtermCriterion (1ms).
 | – Problem 3 was processed with processor SubtermCriterion (1ms).

Problem 1: DependencyGraph



Dependency Pair Problem

Dependency Pairs

mark#(inf(X)) → mark#(X)mark#(take(X1, X2)) → mark#(X1)
a__eq#(s(X), s(Y)) → a__eq#(X, Y)mark#(eq(X1, X2)) → a__eq#(X1, X2)
mark#(take(X1, X2)) → a__take#(mark(X1), mark(X2))mark#(take(X1, X2)) → mark#(X2)
mark#(length(X)) → mark#(X)mark#(inf(X)) → a__inf#(mark(X))
mark#(length(X)) → a__length#(mark(X))

Rewrite Rules

a__eq(0, 0) → truea__eq(s(X), s(Y)) → a__eq(X, Y)
a__eq(X, Y) → falsea__inf(X) → cons(X, inf(s(X)))
a__take(0, X) → nila__take(s(X), cons(Y, L)) → cons(Y, take(X, L))
a__length(nil) → 0a__length(cons(X, L)) → s(length(L))
mark(eq(X1, X2)) → a__eq(X1, X2)mark(inf(X)) → a__inf(mark(X))
mark(take(X1, X2)) → a__take(mark(X1), mark(X2))mark(length(X)) → a__length(mark(X))
mark(0) → 0mark(true) → true
mark(s(X)) → s(X)mark(false) → false
mark(cons(X1, X2)) → cons(X1, X2)mark(nil) → nil
a__eq(X1, X2) → eq(X1, X2)a__inf(X) → inf(X)
a__take(X1, X2) → take(X1, X2)a__length(X) → length(X)

Original Signature

Termination of terms over the following signature is verified: a__take, a__length, inf, true, mark, a__eq, 0, s, a__inf, take, length, false, nil, cons, eq

Strategy


The following SCCs where found

mark#(inf(X)) → mark#(X)mark#(take(X1, X2)) → mark#(X1)
mark#(take(X1, X2)) → mark#(X2)mark#(length(X)) → mark#(X)

a__eq#(s(X), s(Y)) → a__eq#(X, Y)

Problem 2: SubtermCriterion



Dependency Pair Problem

Dependency Pairs

a__eq#(s(X), s(Y)) → a__eq#(X, Y)

Rewrite Rules

a__eq(0, 0) → truea__eq(s(X), s(Y)) → a__eq(X, Y)
a__eq(X, Y) → falsea__inf(X) → cons(X, inf(s(X)))
a__take(0, X) → nila__take(s(X), cons(Y, L)) → cons(Y, take(X, L))
a__length(nil) → 0a__length(cons(X, L)) → s(length(L))
mark(eq(X1, X2)) → a__eq(X1, X2)mark(inf(X)) → a__inf(mark(X))
mark(take(X1, X2)) → a__take(mark(X1), mark(X2))mark(length(X)) → a__length(mark(X))
mark(0) → 0mark(true) → true
mark(s(X)) → s(X)mark(false) → false
mark(cons(X1, X2)) → cons(X1, X2)mark(nil) → nil
a__eq(X1, X2) → eq(X1, X2)a__inf(X) → inf(X)
a__take(X1, X2) → take(X1, X2)a__length(X) → length(X)

Original Signature

Termination of terms over the following signature is verified: a__take, a__length, inf, true, mark, a__eq, 0, s, a__inf, take, length, false, nil, cons, eq

Strategy


Projection

The following projection was used:

Thus, the following dependency pairs are removed:

a__eq#(s(X), s(Y)) → a__eq#(X, Y)

Problem 3: SubtermCriterion



Dependency Pair Problem

Dependency Pairs

mark#(inf(X)) → mark#(X)mark#(take(X1, X2)) → mark#(X1)
mark#(take(X1, X2)) → mark#(X2)mark#(length(X)) → mark#(X)

Rewrite Rules

a__eq(0, 0) → truea__eq(s(X), s(Y)) → a__eq(X, Y)
a__eq(X, Y) → falsea__inf(X) → cons(X, inf(s(X)))
a__take(0, X) → nila__take(s(X), cons(Y, L)) → cons(Y, take(X, L))
a__length(nil) → 0a__length(cons(X, L)) → s(length(L))
mark(eq(X1, X2)) → a__eq(X1, X2)mark(inf(X)) → a__inf(mark(X))
mark(take(X1, X2)) → a__take(mark(X1), mark(X2))mark(length(X)) → a__length(mark(X))
mark(0) → 0mark(true) → true
mark(s(X)) → s(X)mark(false) → false
mark(cons(X1, X2)) → cons(X1, X2)mark(nil) → nil
a__eq(X1, X2) → eq(X1, X2)a__inf(X) → inf(X)
a__take(X1, X2) → take(X1, X2)a__length(X) → length(X)

Original Signature

Termination of terms over the following signature is verified: a__take, a__length, inf, true, mark, a__eq, 0, s, a__inf, take, length, false, nil, cons, eq

Strategy


Projection

The following projection was used:

Thus, the following dependency pairs are removed:

mark#(inf(X)) → mark#(X)mark#(take(X1, X2)) → mark#(X1)
mark#(length(X)) → mark#(X)mark#(take(X1, X2)) → mark#(X2)