Please see SRFI 114 for the successor to this proposal.
SRFI 114 has been deprecated by its author. Here is the first draft of a simpler and better replacement, which will eventually become a new SRFI.
This SRFI provides comparators, procedure bundles that represent one or more of a total order, an equality predicate, and a hash function. By packaging these procedures together, along with a type test predicate, they can be treated as a single item for use in the implementation of data structures.
None at present.
The four procedures above have complex dependencies on one another, and it is inconvenient to have to pass them individually to other procedures that might or might not make use of all of them. For example, a set implementation by nature requires only an equality predicate, but if it is implemented using a hash table, an appropriate hash function is also required if the implementation does not provide one; alternatively, if it is implemented using a tree, a comparison procedure is required. By passing a comparator rather than a bare equality predicate, the set implementation can make use of whatever procedures are available and useful to it.
This SRFI is a simplified and enhanced rewrite of SRFI 114, and shares some of its design rationale and all of its acknowledgements. Special thanks to Taylan Ulrich Bayırlı/Kammer, whose insistence that SRFI 114 was over-complicated but also inadequate inspired this redesign.
The procedures in this SRFI are in the (srfi ???) library (or (srfi :???) on R6RS), but the sample implementation currently places them in the (comparators) library. This means it can't be used alongside SRFI 114, but there's no reason for anyone to do that.
A comparator is an object of a disjoint type. It is a bundle of procedures that are useful for comparing two objects either for equality or for ordering. There are four procedures in the bundle:
The type test predicate returns #t if its argument has the correct type to be passed as an argument to the other three procedures, and #f otherwise.
The equality predicate returns #t if the two objects are the same in the sense of the comparator, and #f otherwise. It is the programmer's responsibility to ensure that it is reflexive, symmetric, transitive, and can handle any arguments that satisfy the type test predicate.
The comparison procedure returns -1, 0, or 1 if the first object precedes the second, is equal to the second, or follows the second, respectively, in a total order defined by the comparator. It is the programmer's responsibility to ensure that it is reflexive, weakly antisymmetric, transitive, can handle any arguments that satisfy the type test predicate, and returns 0 iff the equality predicate returns #t. Comparison procedures are compatible with the compare procedures of SRFI 67; see SRFI 67 for the rationale for adopting this return convention.
The hash function takes one or more arguments and returns an exact non-negative integer; the first argument is the object to be hashed. It is the programmer's responsibility to ensure that it can handle any argument that satisfies the type test predicate, and that it returns the same value on two objects if the equality predicate says they are the same (but not necessarily the converse).
If a second argument is provided, it is a non-negative exact integer, hint from the caller of the hash function that the returned value should be bounded above by the second argument. There is no requirement that a hash
function support this hint, and the caller cannot count on its being applied. Supplying #f
as the second argument is equivalent to providing no argument.
If a third argument is provided, it is also a non-negative exact integer that serves as a salt value to be mixed in along with the first argument when performing the hash calculation. If the caller is careful to provide a sufficiently random salt, attacks on the hash function become more difficult. In addition, callers that need more than one hash function can easily obtain it by varying the salt. Like the upper bound, the salt is a hint: no hash function is required to make use of it.
It is also the programmer's responsibility to ensure that all four procedures provide the same result whenever they are applied to the same object(s) (in the sense of eqv?), unless the object(s) have been mutated since the last invocation. In particular, they must not depend in any way on memory addresses in implementations where the garbage collector can move objects in memory.
The comparator objects defined in this SRFI are not applicable to circular structure or to NaNs or objects containing them. Attempts to pass any such objects to any procedure defined here, or to any procedure that is part of a comparator defined here, is an error except as otherwise noted.
Predicates: comparator?, comparator-comparison-procedure?, comparator-hash-function?
Standard comparator constructors: make-comparator, make-char-ci-comparator, make-string-ci-comparator, make-real-number-comparator, make-complex-number-comparator, make-list-comparator, make-vector-comparator
Default comparators: default-comparator, comparator-register-default!
Wrapped equality predicates: make-eq-comparator, make-eqv-comparator, make-equal-comparator
Accessors: comparator-type-test-procedure, comparator-equality-predicate, comparator-comparison-procedure, comparator-hash-function, comparator-check-type
Comparison predicates: =?, <?, >?, <=?, >=?
(comparator? obj)
Returns #t if obj is a comparator, and #f otherwise.
(comparator-comparison-procedure? comparator)
Returns #t if comparator has a supplied comparison procedure, and #f otherwise.
(comparator-hash-function? comparator)
Returns #t if comparator has a supplied hash function, and #f otherwise.
The following comparator constructors all supply appropriate type test functions, equality predicates and comparison procedures, and hash functions based on the supplied arguments. They are allowed to cache their results: they need not return a newly allocated object, since comparators are pure and functional.
(make-comparator type-test equality compare hash)
Returns a comparator which bundles the type-test, equality, compare, and hash procedures provided. As a convenience, the following additional values are accepted:
#t, a type-test procedure that accepts any arguments is provided.
#t, an equality predicate is provided that returns #t iff compare returns 0.
#f, a procedure is provided that signals an error on application. The predicates comparator-comparison-procedure? and/or comparator-hash-function?, respectively, will return #f in these cases.
(make-char-ci-comparator)
Returns a comparator that compares characters using the total order implied by char-ci<? On R6RS and R7RS systems, this is Unicode codepoint order after the characters have been folded to lower case.
(make-string-ci-comparator)
Returns a comparator that compares strings using the total order implied by string-ci<?. Note that this order is implementation-dependent on R5RS and R7RS systems, but on R6RS systems it is the lexicographical extension of Unicode codepoint order after the characters have been folded to lower case.
(make-real-number-comparator [ type-test ])
Returns a comparator that compares real numbers using < and =. The
type-test serves as the type test predicate for the comparator returned. Its default value is real?. Different comparators returned by this procedure must be compatible with the R5RS numerical tower in the following sense: If S is a subtype of the numerical type T and the two objects are members of S, then the equality predicate and comparison procedures (but not necessarily the hash function) of S-comparator and T-comparator compute the same results on those objects.
(make-list-comparator element-comparator type-test empty? head tail)
Returns a comparator which compares two sequences that satisfy type-test lexicographically, as follows:
null? by default) compares equal to itself.car by default). If the heads are not equal when compared using element-comparator (a default comparator by default), then the result is the result of that comparison. Otherwise, the results of calling the tail procedure (cdr by default) are compared recursively.
In particular, (make-list-comparator) returns a comparator that compares lists using a default comparator for the elements, and (make-list-comparator element-comparator) returns a comparator that compares lists using element-comparator for the elements.
(make-vector-comparator element-comparator type-test length ref)
Returns a comparator which compares two sequences that satisfy type-test as vectors, using the length procedure (vector-length by default) to determine the lengths of the sequences, and the ref procedure (vector-ref by default) to access a particular elements. If one sequence is shorter than the other, it compares less than the other. If the sequences are the same length, they are compared element-wise using element-comparator (a default comparator by default) until they are exhausted.
In particular, (make-vector-comparator returns a comparator that compares vectors using a default comparator for the elements, and (make-vector-comparator element-comparator) returns a comparator that compares vectors using element-comparator. Similarly, (make-vector-comparator (make-default-comparator) bytevector? bytevector-length bytevector-u8-ref) returns a comparator that compares bytevectors.
(make-default-comparator)
Returns a comparator known as a default comparator that accepts any two Scheme values (with the exceptions listed in the Limitations section) and orders them in some implementation-defined way, subject to the following conditions:
The following ordering between types must hold: the empty list precedes other lists, which precede booleans, which precede characters, which precede strings, which precede symbols, which precede numbers, which precede vectors, which precede bytevectors, which precede all other objects.
When applied to lists, it must behave the same as a comparator returned by make-list-comparator when applied to no arguments.
When applied to booleans, it must compare them using the total order #f < #t.
When applied to characters, it must compare them using char<?. In R5RS, this is an implementation-dependent order that is typically the same as Unicode codepoint order; in R6RS and R7RS, it is Unicode codepoint order.
When applied to strings, it must compare them using string<?. In R5RS, this is lexicographic order on the implementation-dependent order defined by char<?; in R6RS it is lexicographic order order on Unicode codepoint order; in R7RS it is implementation-defined order.
When applied to symbols, it must compare them using an implementation-dependent total order. One possibility is to use the order obtained applying symbol->string to the symbols and comparing them using the total order implied by string<?. It is not a requirement that the hash function of symbol-comparator be consistent with the hash function of string-comparator, however.
When applied to numbers where either number is complex, since non-real numbers cannot be compared with <, the following least-surprising ordering is defined: If the real parts are < or >, so are the numbers; otherwise, the numbers are ordered by their imaginary parts. This can still produce somewhat surprising results if one real part is exact and the other is inexact.
When applied to real numbers, it must behave the same as a comparator returned by make-real-number-comparator when applied to no arguments.
When applied to lists, it must behave the same as a comparator returned by make-list-comparator when applied to no arguments.
When applied to bytevectors, its behavior must be the same as that of a comparator created by (make-vector-comparator (make-default-comparator) bytevector? bytevector-length bytevector-u8-ref).
When applied to types registered with comparator-register-default!, its behavior is the same as the comparator registered using that function.
Given disjoint types a and b, one of three conditions must hold:
(comparator-register-default! comparator)
Registers comparator for use by default comparators, such that if the objects being compared both satisfy the type test predicate of comparator, it will be employed by all default comparators to compare them.
(make-eq-comparator)
(make-eqv-comparator)
(make-equal-comparator)
The equality predicates of these comparators are eq?, eqv?, and equal? respectively. When their comparison procedures are applied to non-equal objects, their behavior is implementation-defined. The type test predicates always return #t.
These comparators accept circular structure (in the case of
equal-comparator, provided the implementation's equal? predicate
does so) and NaNs.
(comparator-type-test-procedure comparator)
Returns the type test predicate of comparator.
(comparator-equality-predicate comparator)
Returns the equality predicate of comparator.
(comparator-comparison-procedure comparator)
Returns the comparison procedure of comparator.
(comparator-hash-function comparator)
Returns the hash function of comparator.
(comparator-check-type comparator obj)
Invokes the type test predicate of comparator on obj and returns true if it returns true and signals an error otherwise.
(=? comparator object1 object2 object3 ...)
(<? comparator object1 object2 object3 ...)
(>? comparator object1 object2 object3 ...)
(<=? comparator object1 object2 object3 ...)
(>=? comparator object1 object2 object3 ...)
These procedures are analogous to the number, character, and string comparison predicates of Scheme. They allow the convenient use of comparators to handle arbitrary data types.
These procedures apply the comparison procedure of comparator to the objects as follows. If the specified relation returns #t for all objecti and objectj where n is the number of objects and 1 <= i < j <= n, then the procedures return #t, but otherwise #f.
The order in which the values are compared is unspecified. Because the relations are transitive, it suffices to compare each object with its successor.
The sample implementation contains the following files: FIXME
basics.scm - the syntax, record type definition, and simple constructors and proceduresdefault.scm - a simple implementation of the default constructor, which should be improved by implementers to handle records and implementation-specific typesconstructors.scm - most of the constructorsadvanced.scm - the more complex constructorsr7rs-shim.scm - procedures for R7RS compatibility, including a trivial implementation of bytevectors on top of SRFI 4 u8vectorscomplex-shim.scm - a trivial implementation of real-part and imag-part for Schemes that don't have complex numberscomparators.sld - an R7RS librarycomparators.scm - a Chicken libraryA future release will include a test program using the Chicken test egg, which is available on Chibi as the (chibi test) library.