Compatibility library for term manipulation predicates. Most predicates
in this library are provided as SWI-Prolog built-ins.
- - YAP, SICStus, Quintus. Not all versions of this library define
exactly the same set of predicates, but defined predicates are
- term_size(@Term, -Size) is det
- True if Size is the size in cells occupied by Term on the
global (term) stack. A cell is 4 bytes on 32-bit machines and
8 bytes on 64-bit machines. The calculation does take sharing
into account. For example:
?- A = a(1,2,3), term_size(A,S).
S = 4.
?- A = a(1,2,3), term_size(a(A,A),S).
S = 7.
?- term_size(a(a(1,2,3), a(1,2,3)), S).
S = 11.
Note that small objects such as atoms and small integers have a
size 0. Space is allocated for floats, large integers, strings
and compound terms.
- variant(@Term1, @Term2) is semidet
- Same as SWI-Prolog
Term1 =@= Term2.
- subsumes_chk(@Generic, @Specific)
- True if Generic can be made equivalent to Specific without
- - Replace by subsumes_term/2.
- subsumes(+Generic, @Specific)
- True if Generic is unified to Specific without changing
- - It turns out that calls to this predicate almost
always should have used subsumes_term/2. Also the name is
misleading. In case this is really needed, one is adviced to
follow subsumes_term/2 with an explicit unification.
- term_subsumer(+Special1, +Special2, -General) is det
- General is the most specific term that is a generalisation of
Special1 and Special2. The implementation can handle cyclic
- - Inspired by LOGIC.PRO by Stephen Muggleton
- - SICStus
- lgg_safe(+S1, +S2, -G, +Map0, -Map) is det[private]
- Cycle-safe version of the above. The difference is that we
insert compounds into the mapping table and check the mapping
table before going into a compound.
- term_factorized(+Term, -Skeleton, -Substiution)
- Is true when Skeleton is Term where all subterms that appear
multiple times are replaced by a variable and Substitution is a
list of Var=Value that provides the subterm at the location Var.
I.e., After unifying all substitutions in Substiutions, Term ==
Skeleton. Term may be cyclic. For example:
?- X = a(X), term_factorized(b(X,X), Y, S).
Y = b(_G255, _G255),
S = [_G255=a(_G255)].
- mapargs(:Goal, ?Term1, ?Term2)
- Term1 and Term2 have the same functor (name/arity) and for each
matching pair of arguments
call(Goal, A1, A2) is true.
- mapsubterms(:Goal, +Term1, -Term2) is det
- Recursively map sub terms of Term1 into subterms of Term2 for every
pair for which
call(Goal, ST1, ST2) succeeds. Procedurably, the
mapping for each (sub) term pair
T1/T2 is defined as:
- If T1 is a variable, Unify T2 with T1.
call(Goal, T1, T2) succeeds we are done. Note that the
mapping does not continue in T2. If this is desired, Goal
must call mapsubterms/3 explicitly as part of it conversion.
- If T1 is a dict, map all values, i.e., the tag and keys
are left untouched.
- If T1 is a list, map all elements, i.e., the list structure
is left untouched.
- If T1 is a compound, use same_functor/3 to instantiate T2
and recurse over the term arguments left to right.
- Otherwise T2 is unified with T1.
- same_functor(?Term1, ?Term2) is semidet
- same_functor(?Term1, ?Term2, -Arity) is semidet
- same_functor(?Term1, ?Term2, ?Name, ?Arity) is semidet
- True when Term1 and Term2 are terms that have the same functor
(Name/Arity). The arguments must be sufficiently instantiated, which
means either Term1 or Term2 must be bound or both Name and Arity
must be bound.
If Arity is 0, Term1 and Term2 are unified with Name for
- - SICStus
The following predicates are exported, but not or incorrectly documented.
- term_hash(Arg1, Arg2)
- term_hash(Arg1, Arg2, Arg3, Arg4)
- term_variables(Arg1, Arg2, Arg3)