<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% --> <!-- %% --> <!-- %A create.tex GAP manual Thomas Breuer --> <!-- %A Martin Schönert --> <!-- %% --> <!-- %% -->
<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->
<Chapter Label="Creating New Objects">
<Heading>Creating New Objects</Heading>
This chapter is divided into three parts.
<P/>
In the first part, it is explained how to create
objects with given type (see <Ref Sect="Creating Objects"/>).
<P/>
In the second part, first a few small examples are given,
for dealing with the usual cases of
component objects (see <Ref Sect="Component Objects"/>)
and positional objects (see <Ref Sect="Positional Objects"/>),
and for the implementation of new kinds of lists
(see <Ref Sect="Implementing New List Objects"/>
and <Ref Sect="Arithmetic Issues in the Implementation of New Kinds of Lists"/>).
Finally, the external representation of objects is introduced
(see <Ref Sect="External Representation"/>),
as a tool for representation independent access to an object.
<P/>
The third part deals with some rules concerning the organization
of the &GAP; library;
namely, some commands for creating global variables are explained
(see <Ref Sect="Global Variables in the Library"/>)
that correspond to the ones discussed in the first part of the chapter,
and the idea of distinguishing declaration and implementation part
of &GAP; packages is outlined (see <Ref Sect="Declaration and Implementation Part"/>).
<P/>
See also Chapter <Ref Chap="An Example -- Residue Class Rings"/> for examples
how the functions from the first part are used,
and why it is useful to have a declaration part and an implementation part.
<Description>
New objects are created by <Ref Func="Objectify"/>.
<A>data</A> must be a plain list or a plain record, and <A>type</A>
is the type that the desired object shall have.
<Ref Func="Objectify"/> turns <A>data</A> into an object with type
<A>type</A>.
That is, <A>data</A> is changed, and afterwards it will not be a list or a
record unless <A>type</A> is of type list resp. record.
<P/>
If <A>data</A> is a list then <Ref Func="Objectify"/> turns it into a
positional object, if <A>data</A> is a record then
<Ref Func="Objectify"/> turns it into a component object
(for examples, see <Ref Sect="Component Objects"/>
and <Ref Sect="Positional Objects"/>).
<P/>
<Ref Func="Objectify"/> does also return the object that it made out of
<A>data</A>.
<P/>
For examples where <Ref Func="Objectify"/> is used,
see <Ref Sect="Component Objects"/>,
<Ref Sect="Positional Objects"/>, and the example in
Chapter <Ref Chap="An Example -- Residue Class Rings"/>.
</Description>
</ManSection>
<Index Key='!.'><C>!.</C></Index>
A <E>component object</E> is an object in the representation
<Ref Filt="IsComponentObjectRep"/> or a subrepresentation of it.
Such an object <A>cobj</A> is built from subobjects that can be accessed via
<C><A>cobj</A>!.<A>name</A></C>, similar to components of a record.
Also analogously to records, values can be assigned to components of
<A>cobj</A> via <C><A>cobj</A>!.<A>name</A>:= <A>val</A></C>.
For the creation of component objects, see <Ref Sect="Creating Objects"/>.
One must be <E>very careful</E> when using the <C>!.</C> operator,
in order to interpret the component in the right way,
and even more careful when using the assignment to components using <C>!.</C>,
in order to keep the information stored in <A>cobj</A> consistent.
<P/>
First of all, in the access or assignment to a component as shown above,
<A>name</A> must be among the admissible component names
for the representation of <A>cobj</A>, see <Ref Func="NewRepresentation"/>.
Second, preferably only few low level functions should use <C>!.</C>,
whereas this operator should not occur in <Q>user interactions</Q>.
<P/>
Note that even if <A>cobj</A> claims that it is immutable, i.e.,
if <A>cobj</A> is not in the category <Ref Filt="IsMutable"/>,
access and assignment via <C>!.</C> and <C>!.:=</C> work.
This is necessary for being able to store newly discovered information
in immutable objects.
<P/>
The following example shows the implementation of an iterator
(see <Ref Sect="Iterators"/>) for the domain of integers,
which is represented as component object.
See <Ref Sect="Positional Objects"/> for an implementation using positional objects.
(In practice, such an iterator can be implemented more elegantly using
<Ref Func="IteratorByFunctions"/>,
see <Ref Sect="Example -- Constructing Iterators"/>.)
<P/>
The used succession of integers is <M>0, 1, -1, 2, -2, 3, -3, \ldots</M>,
that is, <M>a_n = n/2</M> if <M>n</M> is even,
and <M>a_n = (1-n)/2</M> otherwise.
<P/>
<Log><![CDATA[
DeclareRepresentation( "IsIntegersIteratorCompRep",
IsComponentObjectRep, [ "counter" ] );
]]></Log>
<P/>
The above command creates a new representation (see <Ref Func="NewRepresentation"/>)
<C>IsIntegersIteratorCompRep</C>,
as a subrepresentation of <Ref Filt="IsComponentObjectRep"/>,
and with one admissible component <C>counter</C>.
So no other components than <C>counter</C> will be needed.
<P/>
<Log><![CDATA[
InstallMethod( Iterator, "method for `Integers'",
[ IsIntegers ],
function( Integers )
return Objectify( NewType( IteratorsFamily,
IsIterator
and IsIntegersIteratorCompRep ),
rec( counter := 0 ) );
end );
]]></Log>
<P/>
After the above method installation, one can already ask for
<C>Iterator( Integers )</C>.
Note that exactly the domain of integers is described by
the filter <Ref Filt="IsIntegers"/>.
<P/>
By the call to <Ref Func="NewType"/>, the returned object lies in the family
containing all iterators, which is <C>IteratorsFamily</C>,
it lies in the category <Ref Filt="IsIterator"/>
and in the representation <C>IsIntegersIteratorCompRep</C>;
furthermore, it has the component <C>counter</C> with value <C>0</C>.
<P/>
What is missing now are methods for the two basic operations
of iterators, namely <Ref Oper="IsDoneIterator"/> and
<Ref Oper="NextIterator"/>.
The former must always return <K>false</K>, since there are infinitely
many integers.
The latter must return the next integer in the iteration,
and update the information stored in the iterator,
that is, increase the value of the component <C>counter</C>.
<P/>
<Log><![CDATA[
InstallMethod( IsDoneIterator, "method for iterator of `Integers'",
[ IsIterator and IsIntegersIteratorCompRep ],
ReturnFalse );
InstallMethod( NextIterator, "method for iterator of `Integers'",
[ IsIntegersIteratorCompRep ],
function( iter )
iter!.counter:= iter!.counter + 1;
if iter!.counter mod 2 = 0 then
return iter!.counter / 2;
else
return ( 1 - iter!.counter ) / 2;
fi;
end );
]]></Log>
<Index Key='![]'><C>![]</C></Index>
A <E>positional object</E> is an object in the representation
<Ref Filt="IsPositionalObjectRep"/> or a subrepresentation of it.
Such an object <A>pobj</A> is built from subobjects that can be accessed via
<C><A>pobj</A>![<A>pos</A>]</C>, similar to positions in a list.
Also analogously to lists, values can be assigned to positions of
<A>pobj</A> via <C><A>pobj</A>![<A>pos</A>]:= <A>val</A></C>.
For the creation of positional objects, see <Ref Sect="Creating Objects"/>.
<P/>
One must be <E>very careful</E> when using the <C>![]</C> operator,
in order to interpret the position in the right way,
and even more careful when using the assignment to positions using <C>![]</C>,
in order to keep the information stored in <A>pobj</A> consistent.
<P/>
First of all, in the access or assignment to a position as shown above,
<A>pos</A> must be among the admissible positions
for the representation of <A>pobj</A>, see <Ref Func="NewRepresentation"/>.
Second, preferably only few low level functions should use <C>![]</C>,
whereas this operator should not occur in <Q>user interactions</Q>.
<P/>
Note that even if <A>pobj</A> claims that it is immutable, i.e.,
if <A>pobj</A> is not in the category <Ref Filt="IsMutable"/>,
access and assignment via <C>![]</C> work.
This is necessary for being able to store newly discovered information
in immutable objects.
<P/>
The following example shows the implementation of an iterator
(see <Ref Sect="Iterators"/>) for the domain of integers,
which is represented as positional object.
See <Ref Sect="Component Objects"/> for an implementation using component objects,
and more details.
<P/>
<Log><![CDATA[
DeclareRepresentation( "IsIntegersIteratorPosRep",
IsPositionalObjectRep, [ 1 ] );
]]></Log>
<P/>
The above command creates a new representation (see <Ref Func="NewRepresentation"/>)
<C>IsIntegersIteratorPosRep</C>,
as a subrepresentation of <Ref Filt="IsPositionalObjectRep"/>,
and with only the first position being admissible for storing data.
<P/>
<Log><![CDATA[
InstallMethod( Iterator, "method for `Integers'",
[ IsIntegers ],
function( Integers )
return Objectify( NewType( IteratorsFamily,
IsIterator
and IsIntegersIteratorPosRep ),
[ 0 ] );
end );
]]></Log>
<P/>
After the above method installation, one can already ask for
<C>Iterator( Integers )</C>.
Note that exactly the domain of integers is described by
the filter <Ref Filt="IsIntegers"/>.
<P/>
By the call to <Ref Func="NewType"/>, the returned object lies in the family
containing all iterators, which is <C>IteratorsFamily</C>,
it lies in the category <Ref Filt="IsIterator"/>
and in the representation <C>IsIntegersIteratorPosRep</C>;
furthermore, the first position has value <C>0</C>.
<P/>
What is missing now are methods for the two basic operations
of iterators, namely <Ref Oper="IsDoneIterator"/>
and <Ref Oper="NextIterator"/>.
The former must always return <K>false</K>, since there are infinitely
many integers.
The latter must return the next integer in the iteration,
and update the information stored in the iterator,
that is, increase the value stored in the first position.
<P/>
<Log><![CDATA[
InstallMethod( IsDoneIterator, "method for iterator of `Integers'",
[ IsIterator and IsIntegersIteratorPosRep ],
ReturnFalse );
InstallMethod( NextIterator, "method for iterator of `Integers'",
[ IsIntegersIteratorPosRep ],
function( iter )
iter![1]:= iter![1] + 1;
if iter![1] mod 2 = 0 then
return iter![1] / 2;
else
return ( 1 - iter![1] ) / 2;
fi;
end );
]]></Log>
<P/>
It should be noted that one can of course install both the methods shown
in Section <Ref Sect="Component Objects"/> and <Ref Sect="Positional Objects"/>.
The call <C>Iterator( Integers )</C> will cause one of the methods to be
selected, and for the returned iterator, which will have one of the
representations we constructed,
the right <Ref Oper="NextIterator"/> method will be chosen.
</Section>
<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->
<Section Label="Implementing New List Objects">
<Heading>Implementing New List Objects</Heading>
This section gives some hints for the quite usual situation that one wants
to implement new objects that are lists.
More precisely, one either wants to deal with lists that have additional
features, or one wants that some objects also behave as lists.
An example can be found in <Ref Sect="Example -- Constructing Enumerators"/>.
<P/>
A <E>list</E> in &GAP; is an object in the category
<Ref Filt="IsList"/>.
Basic operations for lists are <Ref Attr="Length"/>,
<Ref Oper="\[\]"/>,
and <Ref Oper="IsBound\[\]"/>
(see <Ref Sect="Basic Operations for Lists"/>).
<P/>
Note that the access to the position <A>pos</A> in the list <A>list</A>
via <C><A>list</A>[<A>pos</A>]</C> is handled by the call
<C>\[\]( <A>list</A>, <A>pos</A> )</C>
to the operation <Ref Oper="\[\]"/>.
To explain the somewhat strange name <C>\[\]</C> of this operation,
note that non-alphanumeric characters like <C>[</C> and <C>]</C> may occur in
&GAP; variable names only if they are escaped by a <C>\</C> character.
<P/>
Analogously, the check <C>IsBound( <A>list</A>[<A>pos</A>] )</C> whether the position
<A>pos</A> of the list <A>list</A> is bound is handled by the call
<C>IsBound\[\]( <A>list</A>, <A>pos</A> )</C> to the operation
<Ref Oper="IsBound\[\]"/>.
<P/>
For mutable lists, also assignment to positions and unbinding of
positions via the operations <Ref Oper="\[\]\:\="/>
and <Ref Oper="Unbind\[\]"/>
are basic operations.
The assignment <C><A>list</A>[<A>pos</A>]:= <A>val</A></C>
is handled by the call
<C>\[\]\:\=( <A>list</A>, <A>pos</A>, <A>val</A> )</C>,
and <C>Unbind( <A>list</A>[<A>pos</A>] )</C> is handled by the call
<C>Unbind\[\]( <A>list</A>, <A>pos</A> )</C>.
<P/>
All other operations for lists, e.g., <Ref Oper="Add"/>,
<Ref Oper="Append"/>, <Ref Func="Sum"/>,
are based on these operations.
This means that it is sufficient to install methods for the new list
objects only for the basic operations.
<P/>
So if one wants to implement new list objects then one creates them
as objects in the category <Ref Filt="IsList"/>,
and installs methods for <Ref Attr="Length"/>,
<Ref Oper="\[\]"/>,
and <Ref Oper="IsBound\[\]"/>.
If the new lists shall be mutable, one needs to install also methods
for <Ref Oper="\[\]\:\="/> and
<Ref Oper="Unbind\[\]"/>.
<P/>
One application for this is the implementation of <E>enumerators</E>
for domains.
An enumerator for the domain <M>D</M> is a dense list whose entries are
in bijection with the elements of <M>D</M>.
If <M>D</M> is large then it is not useful to write down all elements.
Instead one can implement such a bijection implicitly.
This works also for infinite domains.
<P/>
In this situation, one implements a new representation of the
lists that are already available in &GAP;,
in particular the family of such a list is the same as the family of
the domain <M>D</M>.
<P/>
But it is also possible to implement new kinds of lists that lie in
new families, and thus are not equal to lists that were available
in &GAP; before.
An example for this is the implementation of matrices
whose multiplication via <Q><C>*</C></Q> is the Lie product of matrices.
<P/>
In this situation, it makes no sense to put the new matrices into the
same family as the original matrices.
Note that the product of two Lie matrices shall be defined but not the
product of an ordinary matrix and a Lie matrix.
So it is possible to have two lists that have the same entries but that
are not equal w.r.t. <Q><C>=</C></Q> because they lie in different families.
When dealing with countable sets, a usual task is to define enumerations,
i.e., bijections to the positive integers.
In &GAP;, this can be implemented via <E>enumerators</E>
(see <Ref Sect="Enumerators"/>).
These are lists containing the elements in a specified ordering,
and the operations <Ref Oper="Position"/>
and list access via <Ref Oper="\[\]"/> define the
desired bijection.
For implementing such an enumerator, one mainly needs to install the
appropriate functions for these operations.
<P/>
A general setup for creating such lists is given by
<Ref Func="EnumeratorByFunctions" Label="for a domain and a record"/>.
<P/>
If the set in question is a domain <A>D</A> for which a
<Ref Attr="Size"/> method is
available then all one has to do is to write down the functions for
computing the <M>n</M>-th element of the list and for computing the position
of a given &GAP; object in the list, to put them into the components
<C>ElementNumber</C> and <C>NumberElement</C> of a record, and to call
<Ref Func="EnumeratorByFunctions" Label="for a domain and a record"/>
with the domain <A>D</A> and this record as arguments.
For example, the following lines of code install an
<Ref Attr="Enumerator"/> method
for the case that <A>D</A> is the domain of rational integers.
(Note that <Ref Filt="IsIntegers"/> is a filter
that describes exactly the domain of rational integers.)
<P/>
<Log><![CDATA[
InstallMethod( Enumerator, "for integers",
[ IsIntegers ],
Integers -> EnumeratorByFunctions( Integers, rec(
ElementNumber := function( e, n ) ... end,
NumberElement := function( e, x ) ... end ) ) );
]]></Log>
<P/>
The bodies of the functions have been omitted above;
here is the code that is actually used in &GAP;.
(The ordering coincides with that for the iterators for the domain of
rational integers that have been discussed in <Ref Sect="Component Objects"/>
and <Ref Sect="Positional Objects"/>.)
<P/>
<Example><![CDATA[
gap> enum:= Enumerator( Integers );
<enumerator of Integers>
gap> Print( enum!.NumberElement, "\n" );
function ( e, x )
local pos;
if not IsInt( x ) then
return fail;
elif 0 < x then
pos := 2 * x;
else
pos := -2 * x + 1;
fi;
return pos;
end
gap> Print( enum!.ElementNumber, "\n" );
function ( e, n )
if n mod 2 = 0 then
return n / 2;
else
return (1 - n) / 2;
fi;
return;
end
]]></Example>
<P/>
The situation becomes slightly more complicated if the set <M>S</M> in question
is not a domain.
This is because one must provide also at least a method for computing the
length of the list, and because one has to determine the family in which
it lies (see <Ref Sect="Creating Objects"/>).
The latter should usually not be a problem since either <M>S</M> is nonempty and
all its elements lie in the same family –in this case one takes the
collections family of any element in <M>S</M>– or the family of the enumerator
must be <C>ListsFamily</C>.
<P/>
An example in the &GAP; library is an enumerator for the set of <M>k</M>-tuples
over a finite set; the function is called
<Ref Func="EnumeratorOfTuples"/>. <!-- % The functions <C>ExtendedVectors</C> and <C>OneDimSubspacesTransversal</C> are --> <!-- % also examples but are currently also undocumented ... -->
<P/>
<Example><![CDATA[
gap> Print( EnumeratorOfTuples, "\n" );
function ( set, k )
local enum;
if k = 0 then
return Immutable( [ [ ] ] );
elif IsEmpty( set ) then
return Immutable( [ ] );
fi;
enum
:= EnumeratorByFunctions( CollectionsFamily( FamilyObj( set ) ),
rec(
ElementNumber := function ( enum, n )
local nn, t, i;
nn := n - 1;
t := [ ];
for i in [ 1 .. enum!.k ] do
t[i] := RemInt( nn, Length( enum!.set ) ) + 1;
nn := QuoInt( nn, Length( enum!.set ) );
od;
if nn <> 0 then
Error( "[", n, "] must have an assigned value" );
fi;
nn := enum!.set{Reversed( t )};
MakeImmutable( nn );
return nn;
end,
NumberElement := function ( enum, elm )
local n, i;
if not IsList( elm ) then
return fail;
fi;
elm := List( elm, function ( x )
return Position( enum!.set, x );
end );
if fail in elm or Length( elm ) <> enum!.k then
return fail;
fi;
n := 0;
for i in [ 1 .. enum!.k ] do
n := Length( enum!.set ) * n + elm[i] - 1;
od;
return n + 1;
end,
Length := function ( enum )
return Length( enum!.set ) ^ enum!.k;
end,
PrintObj := function ( enum )
Print( "EnumeratorOfTuples( ", enum!.set, ", ",
enum!.k, " )" );
return;
end,
set := Set( set ),
k := k ) );
SetIsSSortedList( enum, true );
return enum;
end
]]></Example>
<P/>
We see that the enumerator is a homogeneous list that stores individual
functions <C>ElementNumber</C>, <C>NumberElement</C>,
<C>Length</C>, and <C>PrintObj</C>;
besides that, the data components <M>S</M> and <M>k</M> are contained.
Iterators are a kind of objects that is implemented for several collections
in the &GAP; library and which might be interesting also in other cases,
see <Ref Sect="Iterators"/>.
A general setup for implementing new iterators is provided by
<Ref Func="IteratorByFunctions"/>.
<P/>
All one has to do is to write down the functions for
<Ref Oper="NextIterator"/>,
<Ref Oper="IsDoneIterator"/>,
and <Ref Oper="ShallowCopy"/>, and to call
<Ref Func="IteratorByFunctions"/> with this record as argument.
For example, the following lines of code install an
<Ref Oper="Iterator"/> method
for the case that the argument is the domain of rational integers.
<P/>
(Note that <Ref Filt="IsIntegers"/> is a filter
that describes exactly the domain of rational integers.)
<P/>
<Log><![CDATA[
InstallMethod( Iterator, "for integers",
[ IsIntegers ],
Integers -> IteratorByFunctions( rec(
NextIterator:= function( iter ) ... end,
IsDoneIterator := ReturnFalse,
ShallowCopy := function( iter ) ... end ) ) );
]]></Log>
<P/>
The bodies of two of the functions have been omitted above;
here is the code that is actually used in &GAP;.
(The ordering coincides with that for the iterators for the domain of
rational integers that have been discussed in <Ref Sect="Component Objects"/>
and <Ref Sect="Positional Objects"/>.)
<P/>
<Example><![CDATA[
gap> iter:= Iterator( Integers );
<iterator of Integers at 0>
gap> Print( iter!.NextIterator, "\n" );
function ( iter )
iter!.counter := iter!.counter + 1;
if iter!.counter mod 2 = 0 then
return iter!.counter / 2;
else
return (1 - iter!.counter) / 2;
fi;
return;
end
gap> Print( iter!.ShallowCopy, "\n" );
function ( iter )
return rec(
counter := iter!.counter );
end
]]></Example>
<P/>
Note that the <C>ShallowCopy</C> component of the record must be a function
that does not return an iterator but a record that can be used as the
argument of <Ref Func="IteratorByFunctions"/>
in order to create the desired shallow copy.
</Section>
<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->
<Section Label="Arithmetic Issues in the Implementation of New Kinds of Lists">
<Heading>Arithmetic Issues in the Implementation of New Kinds of Lists</Heading>
When designing a new kind of list objects in &GAP;,
defining the arithmetic behaviour of these objects is an issue.
<P/>
There are situations where arithmetic operations of list objects
are unimportant in the sense that adding two such lists need not be
represented in a special way.
In such cases it might be useful either to support no arithmetics at all
for the new lists, or to enable the default arithmetic methods.
The former can be achieved by not setting the filters
<Ref Filt="IsGeneralizedRowVector"/> and
<Ref Filt="IsMultiplicativeGeneralizedRowVector"/>
in the types of the lists,
the latter can be achieved by setting the filter
<Ref Filt="IsListDefault"/>.
(for details,
see <Ref Sect="Filters Controlling the Arithmetic Behaviour of Lists"/>).
An example for <Q>wrapped lists</Q> with default behaviour are vector space
bases;
they are lists with additional properties concerning the computation of
coefficients, but arithmetic properties are not important.
So it is no loss to enable the default methods for these lists.
<P/>
However, often the arithmetic behaviour of new list objects is important,
and one wants to keep these lists away from default methods for addition,
multiplication etc.
For example, the sum and the product of (compatible) block matrices shall
be represented as a block matrix, so the default methods for sum and
product of matrices shall not be applicable,
although the results will be equal to those of the default methods
in the sense that their entries at corresponding positions are equal.
<P/>
So one does not set the filter <Ref Filt="IsListDefault"/>
in such cases,
and thus one can implement one's own methods for arithmetic operations. <!-- % It should be stated explicitly what <Q>arithmetic operations</Q> means! -->
(Of course <Q>can</Q> means on the other hand that one <E>must</E> implement such
methods if one is interested in arithmetics of the new lists.)
<P/>
The specific binary arithmetic methods for the new lists will usually cover
the case that both arguments are of the new kind,
and perhaps also the interaction between a list of the new kind and certain
other kinds of lists may be handled if this appears to be useful.
<P/>
For the last situation, interaction between a new kind of lists and other
kinds of lists, &GAP; provides already a setup.
Namely, there are the categories
<Ref Filt="IsGeneralizedRowVector"/> and
<Ref Filt="IsMultiplicativeGeneralizedRowVector"/>,
which are concerned with the
additive and the multiplicative behaviour, respectively, of lists.
For lists in these filters, the structure of the results of arithmetic
operations is prescribed (see <Ref Sect="Additive Arithmetic for Lists"/> and
<Ref Sect="Multiplicative Arithmetic for Lists"/>).
<P/>
For example,
if one implements block matrices in
<Ref Filt="IsMultiplicativeGeneralizedRowVector"/>
then automatically the product of such a block matrix and a (plain) list
of such block matrices will be defined as the obvious list of matrix
products, and a default method for plain lists will handle this
multiplication.
(Note that this method will rely on a method for computing the product of
the block matrices, and of course no default method is available for that.)
Conversely, if the block matrices are not in
<Ref Filt="IsMultiplicativeGeneralizedRowVector"/>
then the product of a block matrix
and a (plain) list of block matrices is not defined.
(There is no default method for it, and one can define the result and
provide a method for computing it.)
<P/>
Thus if one decides to set the filters
<Ref Filt="IsGeneralizedRowVector"/> and
<Ref Filt="IsMultiplicativeGeneralizedRowVector"/>
for the new lists,
on the one hand one loses freedom in defining arithmetic behaviour,
but on the other hand one gains several default methods for a more
or less natural behaviour.
<P/>
If a list in the filter <Ref Filt="IsGeneralizedRowVector"/>
(<Ref Filt="IsMultiplicativeGeneralizedRowVector"/>)
lies in <Ref Filt="IsAttributeStoringRep"/>,
the values of additive (multiplicative) nesting depth is stored in
the list and need not be calculated for each arithmetic operation.
One can then store the value(s) already upon creation of the lists,
with the effect that the default arithmetic operations will access
elements of these lists only if this is unavoidable.
For example, the sum of two plain lists of <Q>wrapped matrices</Q> with
stored nesting depths are computed via the method for adding two such
wrapped lists, and without accessing any of their rows
(which might be expensive).
In this sense, the wrapped lists are treated as black boxes.
An operation is defined for elements rather than for objects in the sense
that if the arguments are replaced by objects that are equal to the old
arguments w.r.t. the equivalence relation <Q><C>=</C></Q> then the result must be
equal to the old result w.r.t. <Q><C>=</C></Q>.
<P/>
But the implementation of many methods is representation dependent in the
sense that certain representation dependent subobjects are accessed.
<P/>
For example, a method that implements the addition of univariate
polynomials may access coefficients lists of its arguments
only if they are really stored,
while in the case of sparsely represented polynomials a different approach
is needed.
<P/>
In spite of this, for many operations one does not want to write an own
method for each possible representations of each argument,
for example because none of the methods could in fact take advantage
of the actually given representations of the objects.
Another reason could be that one wants to install first a representation
independent method, and then add specific methods as they are needed to
gain more efficiency, by really exploiting the fact that the arguments
have certain representations.
<P/>
For the purpose of admitting representation independent code,
one can define an <E>external representation</E> of objects in a given family,
install methods to compute this external representation for each
representation of the objects,
and then use this external representation of the objects whenever they
occur.
<P/>
We cannot provide conversion functions that allow us to first convert
any object in question to one particular <Q>standard representation</Q>,
and then access the data in the way defined for this representation,
simply because it may be impossible to choose such a <Q>standard
representation</Q> uniformly for all objects in the given family.
<P/>
So the aim of an external representation of an object <A>obj</A> is a
different one, namely to describe the data from which <A>obj</A> is composed.
In particular, the external representation of <A>obj</A> is <E>not</E> one possible
(<Q>standard</Q>) representation of <A>obj</A>,
in fact the external representation of <A>obj</A> is in general different
from <A>obj</A> w.r.t. <Q><C>=</C></Q>,
first of all because the external representation of <A>obj</A> does in general
not lie in the same family as <A>obj</A>.
<P/>
For example the external representation of a rational function is a list
of length two or three, the first entry being the zero coefficient,
the second being a list describing the coefficients and monomials of the
numerator, and the third, if bound, being a list describing the coefficients
and monomials of the denominator.
In particular, the external representation of a polynomial is a list
and not a polynomial.
<P/>
The other way round, the external representation of <A>obj</A> encodes <A>obj</A>
in such a way that from this data and the family of <A>obj</A>,
one can create an object that is equal to <A>obj</A>.
Usually the external representation of an object is a list or a record.
<P/>
Although the external representation of <A>obj</A> is by definition independent
of the actually available representations for <A>obj</A>,
it is usual that a representation of <A>obj</A> exists for which the
computation of the external representation is obtained by just
<Q>unpacking</Q> <A>obj</A>,
in the sense that the desired data is stored in a component or a position
of <A>obj</A>, if <A>obj</A> is a component object (see <Ref Sect="Component Objects"/>)
or a positional object (see <Ref Sect="Positional Objects"/>).
<P/>
To implement an external representation means to install methods for the
following two operations.
<Description>
<Ref Oper="ExtRepOfObj"/> returns the external representation of its argument,
and <Ref Oper="ObjByExtRep"/> returns an object in the family <A>fam</A>
that has external representation <A>data</A>.
<P/>
Of course,
<C>ObjByExtRep( FamilyObj( <A>obj</A> ), ExtRepOfObj( <A>obj</A> ) )</C>
must be equal to <A>obj</A> w.r.t. the operation
<Ref Oper="\="/>.
But it is <E>not</E> required that equal objects have equal external
representations.
<P/>
Note that if one defines a new representation of objects for which an
external representation does already exist
then one <E>must</E> install a method to compute this external representation
for the objects in the new representation.
</Description>
</ManSection>
</Section>
<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->
<Section Label="Mutability and Copying">
<Heading>Mutability and Copying</Heading>
Any &GAP; object is either mutable or immutable. This can be tested
with the function <Ref Filt="IsMutable"/>.
The intended meaning of (im)mutability is a mathematical one:
an immutable object should never change in
such a way that it represents a different Element. Objects <E>may</E>
change in other ways, for instance to store more information, or
represent an element in a different way.
<P/>
Immutability is enforced in different ways for built-in objects (like
records, or lists) and for external objects (made using
<Ref Func="Objectify"/>).
<P/>
For built-in objects which are immutable, the kernel will prevent
you from changing them. Thus
<P/>
<Example><![CDATA[
gap> l := [1,2,4];
[ 1, 2, 4 ]
gap> MakeImmutable(l);
[ 1, 2, 4 ]
gap> l[3] := 5;
Error, List Assignment: <list> must be a mutable list
]]></Example>
<P/>
For external objects, the situation is different. An external object which
claims to be immutable (i.e. its type does not contain
<Ref Filt="IsMutable"/>)
should not admit any methods which change the element it represents.
The kernel does <E>not</E> prevent the use of <C>!.</C> and <C>![</C>
to change the underlying data structure.
This is used for instance by the code that stores attribute values for reuse.
In general, these <C>!</C> operations should only be used in methods
which depend on the representation of the object.
Furthermore, we would <E>not</E>
recommend users to install methods which depend on the representations of
objects created by the library or by &GAP; packages, as there is certainly no
guarantee of the representations being the same in future versions of &GAP;.
<P/>
Here we see an immutable object (the group <M>S_4</M>), in which we improperly
install a new component.
<P/>
<Example><![CDATA[
gap> g := SymmetricGroup(IsPermGroup,4);
Sym( [ 1 .. 4 ] )
gap> IsMutable(g);
false
gap> NamesOfComponents(g);
[ "Size", "NrMovedPoints", "MovedPoints", "GeneratorsOfMagmaWithInverses" ]
gap> g!.silly := "rubbish"; "rubbish"
gap> NamesOfComponents(g);
[ "Size", "NrMovedPoints", "MovedPoints", "GeneratorsOfMagmaWithInverses", "silly" ]
gap> g!.silly; "rubbish"
]]></Example>
<P/>
On the other hand, if we form an immutable externally represented list, we
find that &GAP; will not let us change the object.
<P/>
<Example><![CDATA[
gap> e := Enumerator(g);
<enumerator of perm group>
gap> IsMutable(e);
false
gap> IsList(e);
true
gap> e[3];
(1,2,4)
gap> e[3] := false;
Error, The list you are trying to assign to is immutable
]]></Example>
<P/>
When we consider copying objects, another filter
<Ref Filt="IsCopyable"/>, enters the game and we find that
<Ref Oper="ShallowCopy"/> and
<Ref Func="StructuralCopy"/> behave quite
differently. Objects can be divided for this purpose into three:
mutable objects, immutable but copyable objects, and non-copyable
objects (called constants).
<P/>
A mutable or copyable object should have a method for the operation
<Ref Oper="ShallowCopy"/>,
which should make a new mutable object, sharing its top-level
subobjects with the original. The exact definition of top-level subobject may
be defined by the implementor for new kinds of object.
<P/>
<Ref Oper="ShallowCopy"/> applied to a constant
simply returns the constant.
<P/>
<Ref Func="StructuralCopy"/> is expected to be much less used
than <Ref Oper="ShallowCopy"/>.
Applied to a mutable object, it returns a new mutable
object which shares no mutable sub-objects with the input. Applied to
an immutable object (even a copyable one), it just returns the
object. It is not an operation (indeed, it's a rather special kernel
function).
<P/>
<Example><![CDATA[
gap> e1 := StructuralCopy(e);
<enumerator of perm group>
gap> IsMutable(e1);
false
gap> e2 := ShallowCopy(e);
[ (), (1,4), (1,2,4), (1,3,4), (2,4), (1,4,2), (1,2), (1,3,4,2),
(2,3,4), (1,4,2,3), (1,2,3), (1,3)(2,4), (3,4), (1,4,3), (1,2,4,3),
(1,3), (2,4,3), (1,4,3,2), (1,2)(3,4), (1,3,2), (2,3), (1,4)(2,3),
(1,2,3,4), (1,3,2,4) ]
]]></Example>
<P/>
There are two other related functions:
<Ref Func="Immutable"/>, which makes a new
immutable object which shares no mutable subobjects with its input and
<Ref Func="MakeImmutable"/> which changes an object and its
mutable subobjects <E>in place</E> to be immutable.
It should only be used on <Q>new</Q> objects that
you have just created, and which cannot share mutable subobjects with
anything else.
<P/>
Both <Ref Func="Immutable"/> and
<Ref Func="MakeImmutable"/> work on external objects by just
resetting the <Ref Filt="IsMutable"/> filter
in the object's type. This should make
ineligible any methods that might change the object. As a consequence,
you must allow for the possibility of immutable versions of any
objects you create.
<P/>
So, if you are implementing your own external objects. The rules amount to the
following:
<P/>
<Enum>
<Item>
You decide if your objects should be mutable or copyable or constants, by
fixing whether their type includes <Ref Filt="IsMutable"/>
or <Ref Filt="IsCopyable"/>.
</Item>
<Item>
You install methods for your objects respecting that decision:
<List>
<Mark>for constants:</Mark>
<Item>
no methods change the underlying elements;
</Item>
<Mark>for copyables:</Mark>
<Item>
you provide a method for <Ref Oper="ShallowCopy"/>;
</Item>
<Mark>for mutables:</Mark>
<Item>
you may have methods that change the underlying elements
and these should explicitly require
<Ref Filt="IsMutable"/>.
</Item>
</List>
</Item>
</Enum>
</Section>
<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->
<Section Label="Global Variables in the Library">
<Heading>Global Variables in the Library</Heading>
Global variables in the &GAP; library are usually read-only in order to
prevent them from being overwritten accidentally.
See also Section <Ref Sect="More About Global Variables"/>.
<Description>
For global variables, sometimes code needs to reference them before
a value can sensibly be assigned to them. For example, consider the
following definition of a recursive function:
<Log><![CDATA[
BindGlobal( "fun", function(n)
if n > 0 then
return 2*fun(n-1);
fi;
return 1;
end );
]]></Log>
The problem with that code is that it triggers a syntax warning about access
to an unbound global variable, as <C>fun</C> only gets assigned after the
complete statement has been parsed, yet that statement references <C>fun</C>.
<P/>
To resolve this, one can declare the variable with <Ref
Func="DeclareGlobalName"/> before assigning it via <Ref Func="BindGlobal"/>.
This informs &GAP; that a global variable with the specified name will
eventually be defined, and that thus no syntax warnings pertaining to its use
should be printed.
<P/>
We recommend using <Ref Func="DeclareGlobalName"/> instead of <Ref
Func="DeclareGlobalVariable"/> or <Ref Func="DeclareGlobalFunction"/>
whenever possible.
<P/>
<Ref Func="DeclareGlobalName"/> shall be used in the declaration part
of the respective package
(see <Ref Sect="Declaration and Implementation Part"/>),
values can then be assigned to the new variable as usual, preferably via
<Ref Func="BindGlobal"/>, in the implementation part
(again, see <Ref Sect="Declaration and Implementation Part"/>).
</Description>
</ManSection>
<!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->
<Section Label="Declaration and Implementation Part">
<Heading>Declaration and Implementation Part</Heading>
Each package of &GAP; code consists of two parts,
the <E>declaration part</E> that defines the new categories and operations for
the objects the package deals with,
and the <E>implementation part</E> where the corresponding methods are
installed.
The declaration part should be representation independent,
representation dependent information should be dealt with in the
implementation part.
<P/>
&GAP; functions that are not operations and that are intended to be
called by users should be notified to &GAP; in the declaration part via
<Ref Func="DeclareGlobalFunction"/>.
Values for these functions can be installed in the implementation part
via <Ref Func="InstallGlobalFunction"/>.
<P/>
Calls to the following functions belong to the declaration part.
<List>
<Item><Ref Func="DeclareAttribute"/>,</Item>
<Item><Ref Func="DeclareCategory"/>,</Item>
<Item><Ref Func="DeclareFilter"/>,</Item>
<Item><Ref Func="DeclareOperation"/>,</Item>
<Item><Ref Func="DeclareGlobalFunction"/>,</Item>
<Item><Ref Func="DeclareGlobalName"/>,</Item>
<Item><Ref Func="DeclareGlobalVariable"/>,</Item>
<Item><Ref Func="DeclareSynonym"/>,</Item>
<Item><Ref Func="DeclareSynonymAttr"/>,</Item>
<Item><Ref Func="DeclareProperty"/>,</Item>
<Item><Ref Func="InstallTrueMethod"/>.</Item>
</List>
Calls to the following functions belong to the implementation part.
<List>
<Item><Ref Func="DeclareRepresentation"/>,</Item>
<Item><Ref Func="InstallGlobalFunction"/>,</Item>
<Item><Ref Func="InstallValue"/>,</Item>
<Item><Ref Func="InstallMethod"/>,</Item>
<Item><Ref Func="InstallImmediateMethod"/>,</Item>
<Item><Ref Func="InstallOtherMethod"/>,</Item>
<Item><Ref Func="NewFamily"/>,</Item>
<Item><Ref Func="NewType"/>,</Item>
<Item><Ref Func="Objectify"/>.</Item>
</List>
<Index Key="DeclareRepresentation" Subkey="belongs to implementation part">
<C>DeclareRepresentation</C></Index>
Whenever both a <C>New</C><A>Something</A> and a
<C>Declare</C><A>Something</A> variant of a function exist
(see <Ref Sect="Global Variables in the Library"/>),
the use of <C>Declare</C><A>Something</A> is recommended
because this protects the variables in question from being overwritten.
Note that there are <E>no</E> functions <C>DeclareFamily</C> and
<C>DeclareType</C> since families and types are created dynamically,
hence usually no global variables are associated to them.
Further note that <Ref Func="DeclareRepresentation"/> is regarded as
belonging to the implementation part,
because usually representations of objects are accessed only in very
few places, and all code that involves a particular representation
is contained in one file;
additionally, representations of objects are often not interesting
for the user, so there is no need to provide a user interface
or documentation about representations.
<P/>
It should be emphasized that <Q>declaration</Q> means only an explicit
notification of mathematical or technical terms or of concepts to &GAP;.
For example, declaring a category or property with name <C>IsInteresting</C>
does of course not tell &GAP; what this shall mean,
and it is necessary to implement possibilities to create objects that
know already that they lie in <C>IsInteresting</C> in the case that it is a
category, or to install implications or methods in order to
compute for a given object whether <C>IsInteresting</C> is <K>true</K> or <K>false</K>
for it in the case that <C>IsInteresting</C> is a property.
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung ist noch experimentell.
