reduce-historical/psl-1983/glisp/gluser.mss

@Make(Manual)
@Define(PE,FaceCode U)
@Begin(TitlePage)
@Begin(TitleBox)
@MajorHeading[GLISP User's Manual]
@BlankSpace(2)
@Center(
Gordon S. Novak Jr.
Computer Science Department
Stanford University
Stanford, California  94305)
@BlankSpace(3)
@BlankSpace(2)
@Center[@B<Revised:> @Value(Date)]
@End(TitleBox)
@Begin(ResearchCredit)
This research was supported in part by NSF grant SED-7912803 in the Joint
National Science Foundation - National Institute of Education Program
of Research on Cognitive Processes and the Structure of Knowledge in
Science and Mathematics, and in part by the Defense Advanced Research
Projects Agency under contract MDA-903-80-c-007.
@End(ResearchCredit)
@End(TitlePage)
@Chapter(Introduction)
@Section(Overview of GLISP)

     GLISP is a LISP-based language which provides high-level
language features not found in ordinary LISP.  The GLISP language
is implemented by means of a compiler which accepts GLISP as input and
produces ordinary LISP as output; this output can be further compiled
to machine code by the LISP compiler.  GLISP is available for several
LISP dialects, including Interlisp, Maclisp, UCI Lisp, ELISP, Franz
Lisp, and Portable Standard Lisp.

     The goal of GLISP is to allow structured objects to be referenced
in a convenient, succinct language, and to allow the structures of objects
to be changed without changing the code which references the objects.
GLISP provides both PASCAL-like and English-like syntaxes; much of the power
and brevity of GLISP derive from the compiler features necessary to
support the relatively informal, English-like language constructs.
The following example function illustrates how GLISP permits definite
reference to structured objects.
@Begin(ProgramExample)

(HourlySalaries (GLAMBDA ( (a DEPARTMENT) )
   (for each EMPLOYEE who is HOURLY
      (PRIN1 NAME) (SPACES 3) (PRINT SALARY) )  ))

@End(ProgramExample)
The features provided by GLISP include the following:
@Begin(Enumerate)

GLISP maintains knowledge of the "context" of the computation as the
program is executed.  Features of objects which are in context may be
referenced directly; the compiler will determine how to reference the
objects given the current context, and will add the newly referenced
objects to the context.  In the above example, the function's
argument, an object whose class is
DEPARTMENT, establishes an initial context relative to
which EMPLOYEEs can be found.  In the context of an EMPLOYEE, NAME
and SALARY can be found.

GLISP supports flexible object definition and reference with a
powerful abstract datatype facility.
Object classes are easily declared to the system.  An object
declaration includes a definition of the storage structure of the
object and declarations of properties of the object; these may be
declared in such a way that they compile open, resulting in efficient
object code.  GLISP supports object-centered programming, in which
processes are invoked by means of "messages" sent to objects.
Object structures may be LISP structures (for which code is
automatically compiled) or Units in the user's favorite representation
language (for which the user can supply compilation functions).

Loop constructs, such as
@ (FOR EACH <item> WITH <property> DO ...)@ ,
are compiled into loops of the appropriate form.

Compilation of infix expressions is provided for the arithmetic
operators and for additional operators which facilitate list manipulation.
Operators are interpreted appropriately for Lisp datatypes as well as
for numbers; operator overloading for user-defined objects is provided
using the message facility.

The GLISP compiler infers the types of objects when possible, and uses
this knowledge to generate efficient object code.  By performing
@I[ compilation relative to a knowledge base ], GLISP is able to perform
certain computations (e.g., inheritance of an attached procedure
from a parent class of an object
in a knowledge base) at compile time rather than at runtime, resulting
in much faster execution.

By separating object definitions from the code which references objects,
GLISP permits radical changes to object structures with no changes to
code.
@End(Enumerate)
@Section(Implementation)

     GLISP is implemented by means of a compiler, which produces a
normal Lisp EXPR from the GLISP code; the GLISP code is saved on the
function's property list, and the compiled definition replaces the
GLISP definition.  Use of GLISP entails the cost of a single
compilation, but otherwise is about as efficient as normal LISP.
The LISP code produced by GLISP can be further compiled to machine
code by the LISP compiler.

GLISP functions
are indicated by the use of GLAMBDA instead of LAMBDA in the function
definition.  When the Lisp interpreter sees the GLAMBDA, it
calls the GLISP compiler
to incrementally compile the GLISP function.
The compiled version replaces the GLISP version (which is saved on the
function name's property list), and is used thereafter.
This automatic compilation feature is currently implemented in Interlisp
and in Franz Lisp.  In other dialects, it is necessary for the user to
explicitly invoke compilation of GLISP functions by calling the compiler
function @PE[GLCC] for each one.

     To use GLISP, it is first necessary to load the compiler file into
Lisp.  Users' files containing structure descriptions and GLISP code
are then loaded.  Compilation of a GLISP function is requested by:
@Tabset(1.7 inch)
@Begin(Format)

@PE[(GLCC 'FN)]@\Compile @PE[FN].

@PE[(GLCP 'FN)]@\Compile @PE[FN] and prettyprint the result.

@PE[(GLP 'FN)]@\Print the compiled version of @PE[FN].
@End(Format)
In Interlisp, all the GLISP functions (beginning with GLAMBDA) in a file
can be compiled by invoking @PE[(GLCOMPCOMS@ <file>COMS)], where
@PE[<file>COMS] is the list of file package commands for the file.

Properties of compiled functions are stored on the property list of
the function name:
@Begin(Format)
@PE[GLORIGINALEXPR]@\Original (GLISP) version of the function.@FOOT[The
original definition is saved as EXPR in Interlisp.]
@PE[GLCOMPILED]@\GLISP-compiled version of the function.
@PE[GLRESULTTYPE]@\Type of the result of the function.
@PE[GLARGUMENTTYPES]@\Types of the arguments of the function.
@End(format)
Properties of GLISP functions can be examined with the function
@PE[(GLED '<name>)], which calls the Lisp editor on the property
list of @PE[<name>].  @PE[(GLEDF '<name>)] calls the Lisp editor on the
original (GLISP) definition of @PE[<name>].

@Section(Error Messages)
GLISP provides detailed error messages when compilation errors are
detected; many careless errors such as misspellings will be caught
by the compiler.  When the source program contains errors, the
compiled code generates runtime errors upon execution of the
erroneous expressions.

@Section(Interactive Features of GLISP)
Several features of GLISP are available interactively, as well as in
compiled functions:
@Enumerate{
The @PE[A] function, which creates structured objects from a readable
property/value list, is available as an interactive function.

Messages to objects can be executed interactively.

A display editor/inspector, GEV, is available for use with bitmap
graphics terminals.@Foot[GEV is currently implemented only for Xerox
Lisp machines.]  GEV interprets objects according to their GLISP
structure descriptions; it allows the user to inspect objects, edit
them, interactively construct programs which operate on them, display
computed properties, send messages to objects, and "push down" to
inspect data values.}


@Chapter(Object Descriptions)
@Section(Declaration of Object Descriptions)
An @I(Object Description) in GLISP is a description of the structure
of an object in terms of named substructures, together with definitions
of ways of referencing the object.  The latter may include
@I( properties )
(i.e., data whose values are not stored, but are computed
from the values of stored data), adjectival predicates, and
@I(messages) which the object can receive; the messages can be used to
implement operator overloading and other compilation features.

Object Descriptions are obtained by GLISP in several ways:
@Begin(Enumerate)
The descriptions of basic datatypes (e.g., INTEGER) are automatically
known to the compiler.

Structure descriptions (but not full object descriptions) may be used
directly as @I(types) in function definitions.

The user may declare object descriptions to the system using the
function GLISPOBJECTS; the names of the object types may then be
used as @I[ types ] in function definitions and definitions of other
structures.

Object descriptions may be included as part of a knowledge
representation language, and are then furnished to GLISP by the
interface package written for that representation language.
@End(Enumerate)

LISP data structures are declared using the function GLISPOBJECTS@Foot{
Once declared, object descriptions may be included in INTERLISP program
files by including in the <file>COMS a statement of the form:
@PE[(GLISPOBJECTS@ <object-name@-(1)>@ ...@ <object-name@-(n)>)]},
which takes one or more object
descriptions as arguments (assuming the descriptions to be quoted).
Since GLISP compilation is performed relative to the knowledge base
of object descriptions, the object descriptions must be declared
prior to GLISP compilation of functions using those descriptions.
The format of each description is as follows:
@Begin(ProgramExample)

(<object name>   <structure description>
          PROP   <property descriptions>
          ADJ    <adjective descriptions>
          ISA    <predicate descriptions>
          MSG    <message descriptions>
          SUPERS <list of superclasses>
          VALUES <list of values>              )

@End(ProgramExample)
The <object name> and <structure description> are required; the other
property/value pairs are optional, and may appear in any order.
The following example illustrates some of the
declarations which might be made to describe the object type
@PE(VECTOR).
@Begin(ProgramExample)

(GLISPOBJECTS

   (VECTOR   (CONS (X NUMBER) (Y NUMBER))

      PROP   ( (MAGNITUDE  ((SQRT X*X + Y*Y))) )

      ADJ    ( (ZERO       (X IS ZERO AND Y IS ZERO))
               (NORMALIZED (MAGNITUDE = 1.0)) )

      MSG    ( (+          VECTORPLUS OPEN T)
               (-          VECTORDIFFERENCE) )

     ))

@End(ProgramExample)

@Subsection(Property Descriptions)
Each @PE[<description>] specified with PROP, ADJ, ISA, or MSG
has the following format:
@Begin(ProgramExample)

(<name>  <response>  <prop@-[1]> <value@-[1]> ... <prop@-[n]> <value@-[n]>)

@END(ProgramExample)
where @PE[<name>] is the (atomic) name of the property, @PE[<response>]
is a function name or a list of GLISP code to be compiled in place
of the property, and the @PE[<prop>@ <value>] pairs are optional
properties which affect compilation.  All four kinds of
properties are compiled in a similar fashion, as
described in the section "Compilation of Messages".

@Subsection(Supers Description)
The SUPERS list specifies a list of @I[ superclasses ], i.e., the names
of other object descriptions from which the object may inherit PROP,
ADJ, ISA, and MSG properties.  Inheritance from superclasses can be
recursive, as described under "Compilation of Messages".

@Subsection(Values Description)
The VALUES list is a list of pairs, @PE[ (<name> <value>) ], which is
used to associate symbolic names with constant values for an object
type.  If VALUES are defined for the type of the @I[ selector ] of a
CASE statement, the corresponding symbolic names may be used as the
selection values for the clauses of the CASE statement.

@Section(Structure Descriptions)
     Much of the power of GLISP is derived from its use of Structure
Descriptions.  A Structure Description (abbreviated "<sd>") is a means
of describing a LISP data structure and giving names to parts of the
structure; it is similar in concept to a Record declaration in PASCAL.
Structure descriptions are used by the GLISP compiler to generate code
to retrieve and store parts of structures.
@Subsection(Syntax of Structure Descriptions)

     The syntax of structure
descriptions is recursively defined in terms of basic types and
composite types which are built up from basic types.  The syntax of
structure descriptions is as follows:
@Foot[The names of the basic types and the structuring operators must
be all upper-case or lower-case, depending on the case which is usual for
the underlying Lisp system.  In general, other GLISP keywords and
user program names may be in upper-case, lower-case, or mixed-case,
if mixed cases are permitted by the Lisp system.]
@Begin(Enumerate)

The following basic types are known to the compiler:
@Begin(Format)
@Tabdivide(3)
@B(ATOM)
@B(INTEGER)
@B(REAL)
@B(NUMBER)@\(either INTEGER or REAL)
@B(STRING)
@B(BOOLEAN)@\(either T or NIL)
@B(ANYTHING)@\(an arbitrary structure)
@End(Format)

An object type which is known to the compiler, either from a GLISPOBJECTS
declaration or because it is a Class of units in the user's knowledge
representation language, is a valid type for use in a structure
description.  The <name>@  of such an object type may be specified
directly as <name> or, for readability, as @ @B[(A]@ <name>@B[)]@ 
or @ @B[(AN]@ <name>@B[)].
@Foot[Whenever the form @B<(A ...)> is allowed in GLISP, the form
@B<(AN ...)> is also allowed.]@ 


Any substructure can be named by enclosing it
in a list prefixed by the name: @ @B[(]<name>@ @ <sd>@B[)]@ .
This allows the same substructure to have multiple names.
"A", "AN", and the names used in forming composite types (given below)
are treated as reserved words, and may not be used as names.

Composite Structures:@  
Structured data types composed of other structures are described using
the following structuring operators:
@Begin(Enumerate)

(@B[CONS]@ @ <sd@-[1]>@ @ <sd@-[2]>)
@*
The CONS of two structures whose descriptions
are <sd@-[1]> and <sd@-[2]>.

(@B[LIST]@ @ <sd@-[1]>@ @ <sd@-[2]>@ @ ...@ @ <sd@-[n]>)
@*
A list of exactly the elements
whose descriptions are <sd@-[1]>@ <sd@-[2]>@ ...@ <sd@-[n]>.

(@B[LISTOF]@ @ <sd>)
@*
A list of zero or more elements, each of which has
the description <sd>.

(@B[ALIST]@ @ (<name@-[1]>@ <sd@-[1]>)@ ...@ (<name@-[n]>@ <sd@-[n]>))
@*
An association list
in which the atom <name@-[i]>, if present, is associated with a structure
whose description is <sd@-[i]>.

(@B[PROPLIST]@ @ (<name@-[1]>@ <sd@-[1]>)@ ...@ (<name@-[n]>@ <sd@-[n]>))
@*
An association list in "property-list format" (alternating names and
values)
in which the atom <name@-[i]>, if present, is associated with a structure
whose description is <sd@-[i]>.

(@B[ATOM]@ @ @ (@B[BINDING]@ @ <sd>)
@ @ @ @ (@B[PROPLIST]@ @ (<pname@-[1]>@ <sd@-[1]>)@ ...@ @~
(<pname@-[n]>@ <sd@-[n]>)@ ))
@*
This describes an atom with its binding and/or its property list;
either the BINDING or the PROPLIST group may be omitted.
Each property name <pname@-[i]> is treated as a property list indicator
as well as the name of the substructure.  When creation of such a
structure is specified, GLISP will compile code to create a GENSYM atom.

(@B[RECORD]@ @ <recordname>@ @ (<name@-[1]>@ <sd@-[1]>)@ @ ...@ @ (<name@-[n]>@ <sd@-[n]>))
@*
RECORD specifies the use of contiguous records for data storage.
<recordname> is the name of the record type; it is optional,
and is not used in some Lisp dialects.@Foot[RECORDs are
implemented using RECORDs in Interlisp, HUNKs in Maclisp and Franz Lisp,
VECTORs in Portable Standard Lisp, and lists in UCI Lisp and ELISP.
In Interlisp, appropriate RECORD declarations must be made to the system
by the user in addition to the GLISP declarations.]

(@B[TRANSPARENT]@ @ <type>)
@*
An object of type <type> is incorporated into the structure being
defined in @I[transparent mode], which means that all fields and
properties of the object of type <type> can be directly referenced
as if they were properties of the object being defined.  A substructure
which is a named @I[ type ] and which is not declared to be TRANSPARENT
is assumed to be opaque, i.e., its internal structure cannot be seen
unless an access path explicitly names the subrecord.@Foot{For example,
a PROFESSOR record might contain some fields which are unique to
professors, plus a pointer to an EMPLOYEE record.  If the declaration
in the PROFESSOR record were @PE[(EMPREC@ (TRANSPARENT@ EMPLOYEE))],
then a field of the employee record, say SALARY, could be referenced
directly from a variable P which points to a PROFESSOR record as
@PE[ P:SALARY ]; if the declaration were @PE[(EMPREC@ EMPLOYEE)],
it would be necessary to say @PE[P:EMPREC:SALARY].}
The object
of type <type> may also contain TRANSPARENT objects; the graph of
TRANSPARENT object references must of course be acyclic.

(@B[OBJECT]@ @ (<name@-[1]>@ <sd@-[1]>)@ ...@ (<name@-[n]>@ <sd@-[n]>))
@*(@B[ATOMOBJECT]@ @ (<name@-[1]>@ <sd@-[1]>)@ ...@ (<name@-[n]>@ <sd@-[n]>))
@*(@B[LISTOBJECT]@ @ (<name@-[1]>@ <sd@-[1]>)@ ...@ (<name@-[n]>@ <sd@-[n]>))
@*These declarations describe @I[ Objects ], data structures which can
receive messages at runtime.  The three types of objects are implemented
as records, atoms, or lists, respectively.  In each case, the system
adds to the object
a @PE[CLASS] datum which points to the name of the type of the
object.  An object declaration may only appear as the top-level
declaration of a named object type.
@End(Enumerate)
@End(Enumerate)
@Subsection(Examples of Structure Descriptions)
     The following examples illustrate the use of Structure Descriptions.
@Begin(ProgramExample)

(GLISPOBJECTS

    (CAT (LIST (NAME ATOM)
               (PROPERTIES (LIST (CONS (SEX ATOM)
                                       (WEIGHT INTEGER))
                                 (AGE INTEGER)
                                 (COLOR ATOM)))
               (LIKESCATNIP BOOLEAN)))

    (PERSON (ATOM
              (PROPLIST
                (CHILDREN (LISTOF (A PERSON)))
                (AGE INTEGER)
                (PETS (LIST (CATS (LISTOF CAT))
                            (DOGS (LISTOF (A DOG))) ))
             )))
   )

@End(ProgramExample)
     The first structure, CAT, is entirely composed of list structure.
An CAT structure might look like:
@Begin(ProgramExample)
(PUFF ((MALE . 10) 5 CALICO) T)
@End(ProgramExample)
Given a CAT object X, we could ask for its WEIGHT [equivalent to
(CDAADR X)] or for a subrecord such as PROPERTIES [equivalent
to (CADR X)].  Having set a variable Y to the PROPERTIES,
we could also ask for the WEIGHT from Y [equivalent to (CDAR Y)].
In general, whenever a subrecord is accessed, the structure description
of the subrecord is associated with it by the compiler,
enabling further accesses to parts of the
subrecord.  Thus, the meaning
of a subrecord name depends on the type of record from which the
subrecord is retrieved.  The subrecord AGE has two different
meanings when applied to PERSONs and CATs.
     The second structure, PERSON, illustrates a description of
an object which is a Lisp atom with properties stored on its property
list.  Whereas no structure names appear in an actual CAT structure,
the substructures of a PROPLIST operator must be named, and
the names appear in the actual structures.  For example, if X is a
PERSON structure, retrieval of the AGE of X is equivalent to
@PE[(GETPROP@ X@ 'AGE)].
A subrecord of a PROPLIST record can be referenced directly; e.g., one
can ask for the DOGS of a PERSON directly, without cognizance of
the fact that DOGS is part of the PETS property.

@Section(Editing of Object Descriptions)

An object description can be edited by calling @PE[ (GLEDS TYPE) ],
where @PE[ TYPE ] is the name of the object type.  This will cause the
Lisp editor to be called on the object description of @PE[ TYPE ].

@Section(Interactive Editing of Objects)

An interactive structure inspector/editor, GEV, is available for the
Xerox 1100-series lisp machines.  GEV allows the user to inspect and
edit any structures which are described by GLISP object descriptions,
to "zoom in" on substructures of interest, and to display the values
of computed properties automatically or on demand.  GEV is described
in a separate document.

@Section(Global Variables)

The types of free variables can be declared within the functions which
reference them.  Alternatively, the types of global variables can be
declared to the compiler using the
form:@Foot[@PE{(GLISPGLOBALS@ <name@-(1)>@ ...@ <name@-(n)>)}
is defined as a file package command for Interlisp.]
@Begin(ProgramExample)

(GLISPGLOBALS  (<name> <type>) ... )

@End(ProgramExample)
Following such a declaration, the compiler will assume a free variable
<name> is of the corresponding <type>.  A GLOBAL object does not have
to actually exist as a storage structure; for example, one could define
a global object "MOUSE" or "SYSTEM" whose properties are actually
implemented by calls to the operating system.

@Section(Compile-Time Constants and Conditional Compilation)
The values and types of compile-time constants can be declared to the
compiler using the
form:@Foot[@PE{(GLISPCONSTANTS@ <name@-(1)>@ ...@ <name@-(n)>)}
is defined as a file package command for Interlisp.]
@Programexample[

(GLISPCONSTANTS  (<name> <value-expression> <type>) ... )

]
The <name> and <type> fields are assumed to be quoted.  The
@PE[ <value-expression> ] field is a GLISP expression which is
parsed and evaluated; this allows constants to be defined by expressions
involving previously defined constants.

The GLISP compiler will perform many kinds of computations on
constants at compile time, reducing the size of the compiled code and
improving execution speed.@Foot[Ordinary Lisp functions are evaluated
on constant arguments if the property @PE(GLEVALWHENCONST) is set to T on
the property list of the function name.  This property is set by the
compiler for the basic arithmetic functions.]
In particular, arithmetic, comparison,
logical, conditional, and CASE function calls are optimized, with
elimination of dead code.  This permits conditional compilation in
a clean form.  Code can be written which tests the values of flags
in the usual way; if the flag values are then declared to be
compile-time constants using GLISPCONSTANTS,
the tests will be performed at compile time, and the unneeded code
will vanish.

@Chapter(Reference To Objects)
@Section(Accessing Objects)

The problem of reference is the problem of determining what object,
or feature of a structured object, is referred to by some part of
a statement in a language.  Most programming languages solve the
problem of reference by unique naming: each distinct object in a
program unit has a unique name, and is referenced by that name.
Reference to a part of a structured object is done by giving the name
of the variable denoting that object and a path specification which
tells how to get to the desired part from the whole.

GLISP permits reference by unique naming and path specification,
but in addition permits @I[definite reference relative to context.]
A @I[definite reference] is a reference to an object which has not
been explicitly named before, but which can be understood relative
to the current context of computation.  If, for example, an object
of type VECTOR (as defined earlier) is in context, the program
statement
@Begin(ProgramExample)
(IF X IS NEGATIVE ...
@End(ProgramExample)
contains a definite reference to "X", which may be interpreted as the
X substructure of the VECTOR which is in context.  The definition of
the computational context and the way in which definite references
are resolved are covered in a later section of this manual.

In the following section, which describes the syntaxes of reference
to objects in GLISP, the following notation is used.  "<var>" refers
to a variable name in the usual LISP sense, i.e., a LAMBDA variable,
PROG variable, or GLOBAL variable; the variable is assumed to point
to (be bound to) an object.  "<type>" refers to the type of object
pointed to by a variable.  "<property>" refers to a property or subrecord of
an object.

     Two syntaxes are available for reference to objects: an
English-like syntax, and a PASCAL-like syntax.
The two are equivalent, and may be intermixed freely within a GLISP
function.  The allowable forms of references in the two syntaxes are
shown in the table below.
@Begin(Format)
@TabDivide(3)
@U("PASCAL" Syntax)@\@U("English" Syntax)@\@U(Meaning)

<var>@\<var>@\The object denoted
@\@\by <var>
@B[:]<type>@\@B[The] <type>@\The object whose type
@\@\is <type>
@B[:]<property>@\@B[The] <property>@\The <property> of
@I[or] <property>@\@\some object
<var>@B[:]<property>@\@B[The] <property> @B[of] <var>@\The <property> of the
@\@\object denoted by <var>
@End(Format)
These forms can be extended to specify longer paths in the obvious way,
as in "The AGE of the SPOUSE of the HEAD of the DEPARTMENT" or
"DEPARTMENT:HEAD:SPOUSE:AGE".  Note that there is no distinction
between reference to substructures and reference to properties as
far as the syntax of the referencing code is concerned; this
facilitates hiding the internal structures of objects.

@Section(Creation of Objects)
GLISP allows the creation of structures to be specified by expressions
of the form:
@BlankSpace(1)
@B[(A] <type> @P[with] <property@-[1]> @P[=] <value@-[1]> @P[,] ... @P[,] @~
<property@-[n]> @P[=] <value@-[n]>@B[)]
@BlankSpace(1)
In this expression, the "@I[with]", "=", and "," are allowed for
readability, but may be omitted if desired@Foot[Some Lisp dialects,
e.g. Maclisp, will interpret commas as "backquote" commands and generate
error messages.  In such dialects, the commas must be omitted or be
"slashified".]; if present, they must all
be delimited on both sides by blanks.
In response to such an expression, GLISP will generate code to create
a new instance of
the specified structure.  The <property> names may be specified in any
order.  Unspecified properties are defaulted according to the
following rules:
@Begin(Enumerate)

Basic types are defaulted to 0 for INTEGER and NUMBER, 0.0 for REAL,
and NIL for other types.

Composite structures are created from the defaults of their
components, except that missing PROPLIST and ALIST items which
would default to NIL are omitted.
@End(Enumerate)
Except for missing PROPLIST and ALIST elements, as noted above, a
newly created LISP structure will contain all of the fields specified
in its structure description.

@Section(Interpretive Creation of Objects)

The "A" function is defined for interpretive use as well as for use
within GLISP functions.

@Section(Predicates on Objects)
Adjectives defined for structures using the @PE[ADJ] and @PE[ISA]
specifications may be used in predicate expressions on objects in
@B[If] and @B[For] statements.  The syntax of basic predicate
expressions is:
@Begin(ProgramExample)
<object> @b[is] <adjective>
<object> @B[is a] <isa-adjective>
@End(ProgramExample)
Basic predicate expressions may be combined using AND, OR, NOT or ~, and
grouping parentheses.

The compiler automatically recognizes the LISP adjectives
ATOMIC, NULL, NIL, INTEGER,
REAL, ZERO, NUMERIC, NEGATIVE, MINUS, and BOUND, and the ISA-adjectives
ATOM, LIST, NUMBER, INTEGER, SYMBOL, STRING, ARRAY, and
BIGNUM@Foot[where applicable.]; user definitions have precedence
over these pre-defined adjectives.

@Subsection(Self-Recognition Adjectives)
If the ISA-adjective @PE[ self ] is defined for an object type, the
type name may be used as an ISA-adjective to test whether a given
object is a member of that type.  Given a predicate phrase of the
form "@PE[@ X@ is@ a@ Y@ ]", the compiler first looks at the definition
of the object type of @PE[ X ] to see if @PE[ Y ] is defined as an
ISA-adjective for such objects.
If no such ISA-adjective is found, and @PE[ Y ]
is a type name, the compiler looks to see if @PE[ self ]
is defined as an ISA-adjective for @PE[ Y ], and if so, compiles it.

If a @PE[ self ] ISA-adjective predicate is compiled as the test of an
@B[If], @B[While], or @B[For] statement, and the tested object is a
simple variable, the variable will be known to be of that type within
the scope of the test.  For example, in the statement
@Begin(ProgramExample)

   (If X is a FOO then (_ X Print) ...

@End(ProgramExample)
the compiler will know that X is a FOO if the test succeeds, and will
compile the Print message appropriate for a FOO, even if the type of
X was declared as something other than FOO earlier.  This feature is
useful in implementing disjunctive types, as discussed in a later
section.

@Subsection(Testing Object Classes)
For those data types which are defined using one of the OBJECT
structuring operators, the Class name is automatically defined as an
ISA-adjective.  The ISA test is implemented by runtime examination of
the CLASS datum of the object.

@Chapter(GLISP Program Syntax)
@Section(Function Syntax)
     GLISP function syntax is essentially the same as that of LISP
with the addition of type information and RESULT and GLOBAL declarations.
The basic function syntax is:
@Foot[The PROG is not required.  In Lisp dialects other than Interlisp,
LAMBDA may be used instead of GLAMBDA.]
@Begin(ProgramExample)

(<function-name> (@B[GLAMBDA] (<arguments>)
                         @P[(RESULT] <result-description>@P[)]
                         @P[(GLOBAL] <global-variable-descriptions>@P[)]
      (PROG (<prog-variables>)
            <code>   )))

@End(ProgramExample)
     The RESULT declaration is optional; in many cases, the compiler
will infer the result type automatically.  The main use of the RESULT
declaration is to allow the compiler to determine the result type
without compiling the function, which may be useful when compiling
another function which calls it.  The <result-description> is a
standard structure description or <type>.

     The GLOBAL declaration is used to inform the compiler of the
types of free variables.  The function GLISPGLOBALS can be used to
declare the types of global variables, making GLOBAL declarations
within individual functions unnecessary.

     The major difference between a GLISP function definition and a
standard LISP definition is the presence of type declarations for
variables, which are in PASCAL-like syntax of the following forms:
@Begin(ProgramExample)

<variable>@B[:]<type>
<variable>@B[:(A] <type>@B[)]
<variable>@B[,]<variable>@B[,]...@B[:]<type>
<variable>@B[,]<variable>@B[,]...@B[:(A] <type>@B[)]
          @B[:]<type>
           @B[(A] <type>@B[)]

@End(ProgramExample)
In addition to declared <type>s, a Structure Description may be
used directly as a <type> in a variable declaration.

Type declarations are required only for variables whose subrecords or
properties will be referenced.  In general, if the value of a variable is
computed in such a way that the type of the value can be inferred, the
variable will receive the appropriate type automatically; in such
cases, no type declaration is necessary.  Since GLISP maintains a
@I[context] of the computation, it is often unnecessary to name a
variable which is an argument of a function;
in such cases, it is only necessary to specify the <type> of
the argument, as shown in the latter two syntax forms above.
PROG and GLOBAL declarations must always specify variable
names (with optional types); the ability to directly reference features
of objects reduces the number of PROG variables needed in many cases.

Initial values for PROG variables may be specified, as in Interlisp,
by enclosing the variable and its initial value in a list@Foot[This
feature is available in all Lisp dialects.]:
@ProgramExample{

(PROG (X (N 0) Y) ...)
}
However, the syntax of variable declarations does not permit the type
of a variable and its initial value to both be specified.

@Section(Expressions)
GLISP provides translation of infix expressions of the kind usually
found in programming languages.  In addition, it provides additional
operators which facilitate list manipulation and other operations.
Overloading of operators for user-defined types is provided by means
of the @I[message] facility.

Expressions may be written directly in-line within function references,
as in
@PE[ (SQRT X*X + Y*Y) ],
or they may be written within parentheses; parentheses may be used for
grouping in the usual way.  Operators may be written with or without
delimiting spaces, @I[except for the "-" operator, which @P(must) be delimited
by spaces].
@Foot[The "-" operator is required to be delimited by spaces since "-" is
often used as a hyphen within variable names.  The "-" operator will be
recognized within "atom" names if the flag GLSEPMINUS is set to T.]
Expression parsing is done by an operator precedence parser, using the
same precedence ordering as in FORTRAN.
@Foot[The precedence of compound operators is higher than assignment
but lower than that of all other operators.  The operators
@PE[^ _ _+ +_ _- -_] are right-associative; all others are left-associative.]
The operators which are recognized are as follows:@Foot<In Maclisp, the
operator @PE[/] must be written @PE[//].>
@Begin(Format)
@TabDivide(3)
Assignment@\@PE(_) @I[ or ] @PE[:=]
Arithmetic@\@PE[+  -  *  /  ^]
Comparison@\@PE[=  @R<~>= <> <  <=  >  >=]
Logical@\@PE[AND  OR  NOT  @R<~>]
Compound@\@PE(_+  _-  +_  -_)
@End(Format)

@Subsection(Interpretation of Operators)
In addition to the usual interpretation of operators when used with
numeric arguments, some of the operators are interpreted appropriately
for other Lisp types.

@Paragraph(Operations on Strings)
For operands of type STRING, the operator @PE[ + ] performs
concatenation.  All of the comparison operators are defined for STRINGs.

@Paragraph(Operations on Lists)
Several operators are defined in such a way that they perform set
operations on lists of the form @PE[ (LISTOF@ <type>) ], where
@PE[ <type> ] is considered to be the element type.  The following
table shows the interpretations of the operators:
@Begin(Format)
@Tabdivide(3)
@PE[<list> + <list>]@\Set Union
@PE[<list> - <list>]@\Set Difference
@PE[<list> * <list>]@\Set Intersection

@PE[<list>     +   <element>]@\CONS
@PE[<element>  +   <list>]@\CONS
@PE[<list>     -   <element>]@\REMOVE
@PE[<element>  <=  <list>]@\MEMBER or MEMB
@PE[<list>     >=  <element>]@\MEMBER or MEMB
@End(Format)

@Paragraph(Compound Operators)
Each compound operator performs an operation involving the arguments
of the operator and assigns a value to the left-hand argument;
compound operators are therefore thought of as "destructive change"
operators.
The meaning of a compound operator depends on the type of its
left-hand argument, as shown in the following table:
@Begin(Group)
@Begin(Format)
@TabDivide(5)
@U(Operator)@\@U(Mnemonic)@\@U(NUMBER)@\@U(LISTOF)@\@U(BOOLEAN)
@B[@PE(_+)]@\@I(Accumulate)@\PLUS@\NCONC1@\OR
@B[@PE(_-)]@\@I(Remove)@\DIFFERENCE@\REMOVE@\AND NOT
@B[@PE(+_)]@\@I(Push)@\PLUS@\PUSH@\OR
@B[@PE(-_)]@\@I(Pop)@\@\POP@Foot[For the Pop operator, the arguments are in
the reverse of the usual order, i.e., (TOP@ @PE(-_)@ STACK) will pop the
top element off STACK and assign the element removed to TOP.]
@End(Format)
@End(Group)
As an aid in remembering the list operators, the arrow may be
thought of as representing the list, with the head of the arrow being
the front of the list and the operation (+ or -) appearing where the
operation occurs on the list.  Thus, for example, @PE(_+) adds an element
at the end of the list, while @PE(+_) adds an element at the front of the
list.

Each of the compound operators performs an assignment to its left-hand
side; the above table shows an abbreviation of the operation which is
performed prior to the assignment.
The following examples show the effects of the operator "@PE(_+)" on
local variables of different types:
@Begin(Format)
@TabDivide(3)
@U(Type)@\@U(Source Code)@\@U(Compiled Code)

INTEGER@\@PE(I _+ 5)@\@PE[(SETQ I (IPLUS I 5))]
BOOLEAN@\@PE(P _+ Q)@\@PE[(SETQ P (OR P Q))]
LISTOF@\@PE(L _+ ITEM)@\@PE[(SETQ L (NCONC1 L ITEM))]
@END(Format)

When the compound operators are not specifically defined for a type,
they are interpreted as specifying the operation (@PE[+] or @PE[-])
on the two operands, followed by assignment of the result to the
left-hand operand.

@Paragraph(Assignment)
Assignment of a value to the left-hand argument of an assignment
operator is relatively flexible in GLISP.  The following kinds of
operands are allowed on the left-hand side of an assignment operator:
@Begin(Enumerate)
Variables.

Stored substructures of a structured type.

PROPerties of a structured type, whenever the interpretation of the PROPerty
would be a legal left-hand side.

Algebraic expressions involving numeric types, @I[ provided ] that
the expression ultimately involves only one occurrence of a variable
or stored value.@Foot{For example, @PE[(X^2 _ 2.0)] is acceptable,
but @PE[(X*X@ _@ 2.0)] is not because the variable @PE[X] occurs twice.}
@End(Enumerate)

For example, consider the following Object Description for a CIRCLE:
@ProgramExample{

(CIRCLE (LIST (START VECTOR) (RADIUS REAL))
  PROP  ((PI            (3.1415926))
         (DIAMETER      (RADIUS*2))
         (CIRCUMFERENCE (PI*DIAMETER))
         (AREA          (PI*RADIUS^2))) )
}
Given this description, and a CIRCLE @PE[ C ],
the following are legal assignments:
@Programexample{

(C:RADIUS _ 5.0)
(C:AREA _ 100.0)
(C:AREA _ C:AREA*2)
(C:AREA _+ 100.0)
}

@Paragraph(Self-Assignment Operators
@Foot[This section may be skipped by the casual user of GLISP.])

There are some cases where it would be desirable to let an object
perform an assignment of its own value.  For example, the user might
want to define @I[PropertyList] as an abstract datatype, with messages
such as GETPROP and PUTPROP, and use PropertyLists as substructures
of other datatypes.  However, a message such as PUTPROP may cause the
PropertyList object to modify its own structure, perhaps even changing
its structure from NIL to a non-NIL value.  If the function which
implements PUTPROP performs a normal assignment to its "self" variable,
the assignment will affect only the local variable, and will not modify
the PropertyList component of the containing structure.  The purpose
of the Self-Assignment Operators is to allow such modification of the
value within the containing structure.

The Self-Assignment Operators are @PE[__], @PE[__+], @PE[_+_], and
@PE[__-], corresponding to the operators @PE[_], @PE[_+], @PE[+_],
and @PE[_-], respectively.  The meaning of these operators is that
the assignment is performed to the object on the left-hand side of
the operator, @I[as seen from the structure containing the object].

The use of these operators is highly restricted; any use of a
Self-Assignment Operator must meet all of the following conditions:
@Begin(Enumerate)
A Self-Assignment Operator can only be used within a Message function
which is compiled OPEN.

The left-hand side of the assignment must be a simple variable which
is an argument of the function.

The left-hand-side variable must be given a unique (unusual) name to
prevent accidental aliasing with a user variable name.
@End(Enumerate)

As an example, the PUTPROP message for a PropertyList datatype could
be implemented as follows:
@Begin(ProgramExample)

 (PropertyList.PUTPROP (GLAMBDA (PropertyListPUTPROPself prop val)
      (PropertyListPUTPROPself __
                (LISTPUT PropertyListPUTPROPself prop val)) ))

@End(ProgramExample)

@Section(Control Statements)
GLISP provides several PASCAL-like control statements.
@Subsection(IF Statement)
The syntax of the IF statement is as follows:
@Begin(ProgramExample)
(@B[IF]         <condition@-[1]> @P[THEN] <action@-[11]>@ ...@ <action@-[1i]>
    @P[ELSEIF] <condition@-[2]> @P[THEN] <action@-[21]>@ ...@ <action@-[2j]>
    ...
    @P[ELSE]   <action@-[m1]>@ ...@ <action@-[mk]>)
@End(ProgramExample)
Such a statement is translated to a COND of the obvious form.  The
"THEN" keyword is optional, as are the "ELSEIF" and "ELSE" clauses.

@Subsection(CASE Statement)
The CASE statement selects a set of actions based on an atomic selector
value; its syntax is:
@Begin(ProgramExample)
(@B[CASE]     <selector> @B[OF]
          (<case@-[1]> <action@-[11]>@ ...@ <action@-[1i]>)
          (<case@-[2]> <action@-[21]>@ ...@ <action@-[2j]>)
          ...
          @P[ELSE]   <action@-[m1]>@ ...@ <action@-[mk]>)
@End(ProgramExample)
The @PE[<selector>] is evaluated, and is compared with the given
@PE[<case>] specifications.  Each @PE[<case>] specification is either
a single, atomic specification, or a list of atomic specifications.
All @PE[<case>] specifications are assumed to be quoted.  The "ELSE"
clause is optional; the "ELSE" actions are executed if @PE[<selector>]
does not match any @PE[<case>].

If the @I[ type ] of the @PE[<selector>] has a VALUES specification,
@PE[<case>] specifications which match the VALUES for that type will
be translated into the corresponding values.

@Subsection(FOR Statement)
The FOR statement generates a loop through a set of elements (typically
a list).  Two syntaxes of the FOR statement are provided:
@Begin(ProgramExample)

(@B[FOR EACH] <set> @P[DO] <action@-[1]>@ ...@ <action@-[n]>)

(@B[FOR] <variable> @B[IN] <set> @P[DO] <action@-[1]>@ ...@ <action@-[n]>)
@End(ProgramExample)
The keyword "DO" is optional.  In the first form of the FOR statement,
the singular form of the <set> is specified; GLISP will convert the
given set name to the plural form.
@Foot[For names with irregular plurals, the plural form should be put
on the property list of the singular form under the property name
PLURAL, e.g., @PE<(PUTPROP 'MAN 'PLURAL 'MEN)>.]
The <set> may be qualified by an
adjective or predicate phrase in the first form; the allowable syntaxes
for such qualifying phrases are shown below:
@Begin(ProgramExample)
<set> @B[WITH] <predicate>
<set> @B[WHICH IS] <adjective>
<set> @B[WHO IS]   <adjective>
<set> @B[THAT IS]  <adjective>
@End(ProgramExample)
The <predicate> and <adjective> phrases may be combined with AND, OR, NOT,
and grouping parentheses.  These phrases may be followed by a qualifying
phrase of the form:
@Begin(ProgramExample)
@B[WHEN] <expression>
@End(ProgramExample)
The "WHEN" expression is ANDed with the other qualifying expressions to
determine when the loop body will be executed.

Within the FOR loop, the current member of
the <set> which is being examined is automatically put into @I[context]
at the highest level of priority.
For example, suppose that the current context contains a substructure
whose description is:
@Begin(ProgramExample)
(PLUMBERS (LISTOF EMPLOYEE))
@END(ProgramExample)
Assuming that EMPLOYEE contains the appropriate definitions, the
following FOR loop could be written:
@Begin(ProgramExample)
(FOR EACH PLUMBER WHO IS NOT A TRAINEE DO SALARY _+ 1.50)
@End(ProgramExample)

To simplify the collection of features of a group of objects, the
<action>s in the FOR loop may be replaced by the CLISP-like construct:
@Begin(ProgramExample)
      ... @B[COLLECT] <form>)
@End(ProgramExample)

@Subsection(WHILE Statement)
The format of the WHILE statement is as follows:
@Begin(ProgramExample)

   (@B[WHILE] <condition> @B[DO] <action@-[1]> ... <action@-[n]>)

@End(ProgramExample)
The actions @PE(<action@-[1]>) through @PE(<action@-[n]>) are executed
repeatedly as long as @PE(<condition>) is true.  The keyword @B[DO]
may be omitted.  The value of the expression is NIL.

@Subsection(REPEAT Statement)
The format of the REPEAT statement is as follows:
@Begin(ProgramExample)

   (@B[REPEAT] <action@-[1]> ... <action@-[n]> @B[UNTIL] <condition>)

@End(ProgramExample)
The actions @PE(<action@-[1]>) through @PE(<action@-[n]>) are repeated
(always at least once) until @PE[<condition>] is true.  The value of
the expression is NIL.  The keyword @B[UNTIL] is required.

@Section(Definite Reference to Particular Objects)
In order to simplify reference to particular member(s) of a group,
definite reference may be used.  Such an expression is written using
the word @B[THE] followed by the singular form of the group,
or @B[THOSE] followed by the plural form of the group, and
qualifying phrases (as described for the @B[FOR] statement).
The following examples illustrate these expressions.
@Begin(ProgramExample)
   (THE SLOT WITH SLOTNAME = NAME)
   (THOSE EMPLOYEES WITH JOBTITLE = 'ELECTRICIAN)
@End(ProgramExample)
The value of @B[THE] is a single object (or NIL if no object satisfies
the specified conditions); @B[THOSE] produces a list of all objects
satisfying the conditions.@Foot[In general, nested loops are optimized
so that intermediate lists are not actually constructed.  Therefore,
use of nested THE or THOSE statements is not inefficient.]

@Chapter(Messages)
GLISP supports the @I[Message] metaphor, which has its roots in the
languages SIMULA and SMALLTALK.  These languages provide
@I[Object-Centered Programming], in which objects are thought of as
being active entities which communicate by sending each other
@I[Messages].  The internal structures of objects are hidden; a program
which wishes to access "variables" of an object does so by sending
messages to the object requesting the access desired.  Each object
contains
@Foot[typically by inheritance from some parent in a Class hierarchy]
a list of @I[Selectors], which identify the messages to which the object
can respond.  A @I[Message] specifies the destination object, the
selector, and any arguments associated with the message.  When a
message is executed at runtime, the selector is looked up for the
destination object; associated with the selector is a procedure, which
is executed with the destination object and message arguments as its
arguments.

GLISP treats reference to properties, adjectives, and predicates
associated with an object similarly to the way it treats messages.
The compiler is able to perform much of the lookup of @I[selectors]
at compile time, resulting in efficient code while maintaining the
flexibility of the
message metaphor.  Messages can be defined in such a way that they
compile open, compile as function calls to the function which is
associated with the selector, or compile as messages to be interpreted
at runtime.

Sending of a @I[message] in GLISP is specified using the following syntax:
@Begin(ProgramExample)
@B[(SEND] <object> <selector> <arg@-[1]>@ ...@ <arg@-[n]>@B[)]
@End(ProgramExample)
The keyword "SEND" may be replaced by "@B[@PE(_)]".  The @PE[<selector>]
is assumed to be quoted.  Zero or more arguments may be specified;
the arguments other than @PE[<selector>] are evaluated.
@PE[<object>] is evaluated; if @PE[<object>] is a non-atomic expression,
it must be enclosed in at least one set of parantheses, so that the
@PE[<selector>] will always be the third element of the list.

@SECTION(Compilation of Messages)
When GLISP encounters a message statement, it looks up the <selector>
in the MSG definition of the type of the object to which the message
is sent, or in one of the SUPERS of the type.
@Foot[If an appropriate representation language is provided, the
<selector> and its associated <response>
may be inherited from a parent class in the class hierarchy of the
representation language.]
Each <selector> is paired with the appropriate <response> to the message.
Code is compiled depending on the form
of the <response> associated with the <selector>, as follows:
@Foot[If the type of the destination object is unknown, or if the
<selector> cannot be found, GLISP compiles the (SEND@ ...) statement
as if it is a normal function call.]
@Begin(Enumerate)
If the <response> is an atom, that atom is taken as the name of a
function which is to be called in response to the message.  The code
which is compiled is a direct call to this function,
@Begin(ProgramExample)
(<response> <object> <arg@-[1]> ... <arg@-[n]>)
@End(ProgramExample)

If the <response> is a list, the contents of the list are recursively
compiled in-line as GLISP code, with the name "@PE[self]" artificially
"bound" to the <object> to which the message was sent.  Because the
compilation is recursive, a message may be defined in terms of other
messages, substructures, or properties, which may themselves be defined
as messages.
@Foot[Such recursive definitions must of course be acyclic.]
The outer pair of parentheses of the <response> serves only to bound
its contents; thus, if the <response> is a function call, the function
call must be enclosed in an additional set of parentheses.
@End(Enumerate)

The following examples illustrate the various ways of defining message
responses.
@Begin(ProgramExample)

(EDIT         EDITV)

(SUCCESSOR    (self + 1))

(MAGNITUDE    ((SQRT X*X + Y*Y)))

@End(ProgramExample)
In the first example, a message with <selector> EDIT is
compiled as a direct call to the function EDITV.  In the
second example, the SUCCESSOR message is compiled as the sum of
the object receiving the message (represented by "@PE[self]") and the
constant 1; if the object receiving the message is the value of the
variable J and has the type INTEGER, the code generated
for the SUCCESSOR would be @PE[(ADD1 J)].  The third example illustrates
a call to a function, SQRT, with arguments containing definite
references to X and Y (which presumably are defined as part of the
object whose MAGNITUDE is sought).  Note that since MAGNITUDE is
defined by a function call, an "extra" pair of parentheses is
required around the function call to distinguish it from in-line code.

The user can determine whether a message is to be compiled open,
compiled as a function call, or compiled as a message which is to
be executed at runtime.
When a GLISP expression is specified as a <response>, the <response>
is always compiled open; open compilation can be requested by using
the OPEN property when the <response> is a function name.
Open compilation operates like
macro expansion; since the "macro" is a GLISP expression, it is easy
to define messages and properties in terms of other messages and
properties.  The combined capabilities of open compilation, message
inheritance, conditional compilation, and flexible assignment provide
a great deal of power.
The ability to use definite reference in GLISP makes
the definition and use of the "macros" simple and natural.

@Section(Compilation of Properties and Adjectives)
Properties, Adjectives, and ISA-adjectives are compiled in the
same way as Messages.  Since the syntax of use of properties and
adjectives does not permit specification of any arguments, the only
argument available to code or a function which implements the
@PE[<response>] for a property or adjective is the @PE[ self ]
argument, which denotes the object to which the property or adjective
applies.  A @PE[<response>] which is written directly as GLISP code
may use the name @PE[ self ] directly
@Foot[The name @PE< self > is "declared" by the compiler, and does
not have to be specified in the Structure Description.], as in the
SUCCESSOR example above; a function which is specified as the
@PE[<response>] will be called with the @PE[self]
object as its single argument.

@Section(Declarations for Message Compilation)
Declarations which affect compilation of Messages, Adjectives, or
Properties may be specified following the <response> for a given
message; such declarations are in (Interlisp) property-list format,
@PE[<prop@-[1]><value@-[1]>@ ...@ <prop@-[n]><value@-[n]>].  The
following declarations may be specified:
@Begin(Enumerate)
@B[RESULT]@PE[ <type>]
@*
This declaration specifies the @I[type] of the result of the
message or other property.  Specification of result types helps the
compiler to perform type inference, thus reducing the number of type
declarations needed in user programs.
The RESULT type for simple GLISP expressions will be inferred by the
compiler; the RESULT declaration should be used if the @PE[<response>]
is a complex GLISP expression or a function name.
@Foot[Alternatively, the result of a function may be specified by the
RESULT declaration within the function itself.]@ 

@B[OPEN@ @ T]
@*
This declaration specifies that the function which is specified as the
<response> is to be compiled open at each reference.  A <response>
which is a list of GLISP code is always compiled open; however, such
a <response> can have only the @PE[self] argument.  If it is desired to
compile open a Message <response> which has arguments besides @PE[self],
the <response> must be coded as a function (in order to bind the
arguments) and the OPEN declaration must be used.
Functions which are compiled open may not be recursive via any chain
of open-compiled functions.

@B[MESSAGE@ @ T]
@*
This declaration specifies that a runtime message should be generated
for messages with this <selector> sent to objects of this Class.
Typically, such a declaration would be used in a higher-level Class
whose subclasses have different responses to the same message
<selector>.
@End(Enumerate)

@Section(Operator Overloading)
GLISP provides operator overloading for user-defined objects using
the Message facility.  If an arithmetic operator is defined as the
@I[selector] of a message for a user datatype, an arithmetic
subexpression using that operator will be compiled as if it were
a message call with two arguments.  For example, the type VECTOR
might have the declaration and function definitions below:
@Begin(ProgramExample)

(GLISPOBJECTS
   (VECTOR  (CONS (X INTEGER) (Y INTEGER))
      MSG  ((+  VECTORPLUS OPEN T)
            (_+ VECTORINCR OPEN T)) )    )

(DEFINEQ

   (VECTORPLUS (GLAMBDA (U,V:VECTOR)
       (A VECTOR WITH X = U:X + V:X , Y = U:Y + V:Y) ))

   (VECTORINCR (GLAMBDA (U,V:VECTOR)
       (U:X _+ V:X)
       (U:Y _+ V:Y) ))    )

@End(ProgramExample)
With these definitions, an expression involving the operators @PE[+]
or @PE[_+] will be compiled by open compilation of the respective
functions.

The compound operators (@PE[_+ +_ _- -_]) are conventionally thought of as
"destructive replacement" operators; thus, the expression
@PE[(U@ _@ U@ +@ V)] will create a new VECTOR structure and assign
the new structure to U, while the expression @PE[(U@ _+@ V)] will
smash the existing structure U, given the definitions above.
The convention of letting the compound operators specify "destructive
replacement" allows the user to specify both the destructive and
non-destructive cases.  However, if the compound operators are not
overloaded but the arithmetic operators @PE[+] and @PE[-] are
overloaded, the compound operators are compiled using the definitions
of @PE[+] for @PE[_+] and @PE[+_], and @PE[-] for @PE[_-] and @PE[-_].
Thus, if only the @PE[+] operator were overloaded for VECTOR, the
expression @PE[(U@ _+@ V)] would be compiled as if it were
@PE[(U@ _@ U@ +@ V)].

@Section(Runtime Interpretation of Messages)
In some cases, the type of the object which will receive a given message
is not known at compile time; in such cases, the message must be
executed interpretively, at runtime.  Interpretive
execution is provided for all types of GLISP messages.

An interpretive message call (i.e., a call to the function @PE[SEND])
is generated by the GLISP compiler in response to a message call in
a GLISP program when the specified message selector cannot be found
for the declared type of the object receiving the message, or when
the MESSAGE flag is set for that selector.  Alternatively, a call to
SEND may be entered interactively by the user or may be contained in
a function which has not been compiled by GLISP.

Messages can be interpreted only for those objects which are represented
as one of the OBJECT types, since it is necessary that the object
contain a pointer to its CLASS.  The <selector> of the message is
looked up in the MSG declarations of the CLASS; if it is not found
there, the SUPERS of the CLASS are examined (depth-first) until the
selector is found.  The <response> associated with the <selector> is
then examined.  If the <response> is a function name, that function is
simply called with the specified arguments.@Foot{The object to which
the message is sent is always inserted as the first argument, followed
by the other arguments specified in the message call.}  If the
<response> is a GLISP expression, the expression is compiled as a
LAMBDA form and cached for future use.

Interpretive execution is available for other property types (PROP,
ADJ, and ISA) using the call:
@Programexample[

(SENDPROP <object> <selector> <proptype>)

]
where @PE[<proptype>] is PROP, ADJ, or ISA.  @PE[<proptype>] is not
evaluated.

@Chapter(Context Rules and Reference)
The ability to use definite reference to features of objects which
are in @I[Context] is the key to much of GLISP's power.  At the
same time, definite reference introduces the possibility of ambiguity,
i.e., there could be more than one object in Context which has
a feature with a specified name.  In this chapter, guidelines are
presented for use of definite reference to allow the user to avoid
ambiguity.

@Section(Organization of Context)
The Context maintained by the compiler is organized in levels, each
of which may have multiple entries; the sequence of
levels is a stack.  Searching of the Context
proceeds from the top (nearest) level of the stack to the bottom
(farthest) level.  The bottom level of the stack is composed of the
LAMBDA variables of the function being compiled.  New levels
are added to the Context in the following cases:
@Begin(Enumerate)
When a PROG is compiled.  The PROG variables are added to the new
level.

When a @B[For] loop is compiled.  The "loop index" variable (which may
be either a user variable or a compiler variable) is added to the
new level, so that it is in context during the loop.

When a @B[While] loop is compiled.

When a new clause of an @B[If] statement is compiled.
@End(Enumerate)

When a Message, Property, or Adjective is compiled, that compilation
takes place in a @I[ new ] context consisting only of the @PE[ self ]
argument and other message arguments.

@Section(Rules for Using Definite Reference)
The possibility of referential ambiguity is easily controlled in practice.
First, it should be noted that the traditional methods of unique
naming and complete path specification ("PASCAL style")
are available, and should be
used whenever there is any possibility of ambiguity.  Second, there
are several cases which are guaranteed to be unambiguous:
@Begin(Enumerate)
In compiling GLISP code which implements a Message, Property, or
Adjective, only the @PE[@ self@ ] argument is in context initially;
definite reference to any substructure or property of the object
is therefore unambiguous.
@Foot[Unless there are duplicated names in the object definition.
However, if the same name is used as both a Property and an Adjective,
for example, it is not considered a duplicate since Properties and
Adjectives are specified by different source language constructs.]@ 

Within a @B[For] loop, the loop variable is the closest thing in
context.

In many cases, a function will only have a single structured argument;
in such cases, definite reference is unambiguous.
@End(Enumerate)
If "PASCAL" syntax (or the equivalent English-like form) is used for
references other than the above cases, no ambiguities will occur.
@Section(Type Inference)
In order to interpret definite references to features of objects,
the compiler must know the @I[ types ] of the objects.  However,
explicit type specification can be burdensome, and makes it difficult
to change types without rewriting existing type declarations.
The GLISP compiler performs type inference in many cases, relieving
the programmer of the burden of specifying types explicitly.  The
following rules enable the programmer to know when types will be
inferred by the compiler.
@Begin(Enumerate)
Whenever a variable is set to a value whose type is known,
the type of the variable
is inferred to be the type of the value to which it was set.

If a variable whose initial type was NIL (e.g., an untyped PROG variable)
appears on the left-hand side of the @PE[@ _+@ ] operator, its type
is inferred to be @PE[(LISTOF@ <type>)], where @PE[@ <type>@ ] is
the type of the right-hand side of the @PE[@ _+@ ] expression.

Whenever a substructure of a structured object is retrieved, the type
of the substructure is retrieved also.

Types of infix expressions are inferred.

Types of Properties, Adjectives, and Messages are inferred if:
@Begin(Enumerate)
The @PE[ <response> ] is GLISP code whose type can be inferred.

The @PE[ <response> ] has a RESULT declaration associated with it.

The @PE[ <response> ] is a function whose definition includes a
RESULT declaration, or whose property list contains a GLRESULTTYPE
declaration.
@End(Enumerate)

The type of the "loop variable" in a @B[For] loop is inferred and is
added to a new level of Context by the compiler.

If an @B[If] statement tests the type of a variable using a @PE[@ self@ ]
adjective, the variable is inferred to be of that type if the test is
satisfied.  Similar type inference is performed if the test of the type
of the variable is the condition of a @B[While] statement.

When possible, GLISP infers the type of the function it is compiling
and adds the type of the result to the property list of the function
name under the indicator GLRESULTTYPE.

The types returned by many standard Lisp functions are known by the
compiler.
@End(Enumerate)

@Chapter(GLISP and Knowledge Representation Languages)
GLISP provides a convenient @I[Access Language] which allows uniform
specification of access to objects, without regard to the way in
which the objects are actually stored; in addition, GLISP provides
a basic @I[Representation Language], in which the structures and
properties of objects can be declared.  The field of Artificial
Intelligence has spawned a number of powerful Representation
Languages, which provide power in describing large numbers of object
classes by allowing hierarchies of @I[Class] descriptions, in which
instances of Classes can inherit properties and procedures from
parent Classes.  The @I[Access Languages] provided for these Representation
Languages, however, have typically been rudimentary, often being no
more than variations of LISP's GETPROP and PUTPROP.  In addition,
by performing inheritance of procedures and data values at runtime,
these Representation Languages have often been computationally costly.

Facilities are provided for interfacing GLISP with representation
languages of the user's choice.  When this is done,
GLISP provides a convenient and uniform language for
accessing both objects in the Representation Language and LISP objects.
In addition, GLISP can greatly improve the efficiency of programs which
access the representations by performing lookup of procedures and data
in the Class hierarchy @I[at compile time].  Finally, a LISP structure
can be specified @I[as the way of implementing] instances of a Class
in the Representation Language, so that while the objects in such a
class appear the same as other objects in the Representation Language
and are accessed in the same way, they are actually implemented as
LISP objects which are efficient in both time and storage.

A clean
@Foot[Cleanliness is in the eye of the beholder and, being next to
Godliness, difficult to attain.  However, it's @I(relatively) clean.]
interface between GLISP and a Representation Language is provided.
With such an interface, each @I[Class] in the Representation Language
is acceptable as a GLISP @I[type].  When the program which is being
compiled specifies an access to an object which is known to be a
member of some Class, the interface module for the Representation
Language is called to generate code to perform the access.  The
interface module can perform inheritance within the Class hierarchy,
and can call GLISP compiler functions to compile code for
subexpressions.  Properties, Adjectives, and Messages in GLISP format
can be added to Class definitions, and can be inherited by subclasses
at compile time.  In an Object-Centered representation language or
other representation language which relies heavily on procedural
inheritance, substantial improvements in execution speed can be
achieved by performing the inheritance lookup at compile time and
compiling direct procedure calls to inherited procedures when the
procedures are static and the type of the object which inherits the
procedure is known at compile time.

Specifications for an interface module for GLISP are contained in a
separate document@Foot[to be written.].  To date, GLISP has been
interfaced to our own GIRL representation language, and to LOOPS.
@Foot[LOOPS, a LISP Object Oriented Programming System, is being
developed at Xerox Palo Alto Research Center by Dan Bobrow and
yMark Stefik.]

@Chapter(Obtaining and Using GLISP)
GLISP and its documentation are available free of charge over the
ARPANET.  The host computers involved will accept the login
"ANONYMOUS GUEST" for transferring files with FTP.
@Section(Documentation)
This user's manual, in line-printer format, is contained in
@PE([UTEXAS-20]<CS.NOVAK>GLUSER.LPT) .  The SCRIBE source file is
@PE([SU-SCORE]<CSD.NOVAK>GLUSER.MSS) .  Printed copies of this manual
can be ordered from Publications Coordinator, Computer Science
Department, Stanford University, Stanford, CA 94305, as technical report
STAN-CS-82-895 ($3.15 prepaid); the printed version may not be as
up-to-date as the on-line version.
@Section(Compiler Files)
There are two files, GLISP (the compiler itself) and GLTEST (a file
of examples).  The files for the different Lisp dialects are:
@Tabset(1.4 inch)
@Begin(Format)
Interlisp:@\@PE([SU-SCORE]<CSD.NOVAK>GLISP.LSP) and @PE(GLTEST.LSP)
Maclisp:@\@PE([SU-SCORE]<CSD.NOVAK>GLISP.MAC) and @PE(GLTEST.MAC)
UCI Lisp:@\@PE([UTEXAS-20]<CS.NOVAK>GLISP.UCI) and @PE(GLTEST.UCI)
ELISP:@\the UCI version plus @PE([UTEXAS-20]<CS.NOVAK>ELISP.FIX)
Franz Lisp:@\@PE([SUMEX-AIM]<NOVAK>GLISP.FRANZ) and @PE(GLTEST.FRANZ)
PSL:@\@PE([SU-SCORE]<CSD.NOVAK>GLISP.PSL) and @PE(GLTEST.PSL)
@End(Format)
@Section(Getting Started)
Useful functions for invoking GLISP are:
@Begin(Format)
@PE[(GLCC 'FN)]@\Compile FN.

@PE[(GLCP 'FN)]@\Compile FN and prettyprint result.

@PE[(GLP 'FN)]@\Prettyprint GLISP-compiled version of FN.

@PE[(GLED 'NAME)]@\Edit the property list of NAME.

@PE[(GLEDF 'FN)]@\Edit the original (GLISP) definition of FN.
@\(The original definition is saved under the property
@\"GLORIGINALEXPR" when the function is compiled, and
@\the compiled version replaces the function
@\definition.)

@PE[(GLEDS 'STR)]@\Edit the structure declarations of STR.
@End(Format)
The editing functions call the "BBN/Interlisp" structure editor.

To try out GLISP, load the GLTEST file and use GLCP to compile the
functions CURRENTDATE, GIVE-RAISE, TESTFN1, TESTFN2, DRAWRECT,
TP, GROWCIRCLE, and SQUASH.  To run compiled functions on test data,
do:
@Begin(ProgramExample)
(GIVE-RAISE 'COMPANY1)
(TP '(((A (B (C D (E (G H (I J (K))))))))))
(GROWCIRCLE MYCIRCLE)
@END(ProgramExample)

@Section(Reserved Words and Characters)
GLISP contains ordinary lisp as a sublanguage.  However, in order to
avoid having code which was intended as "ordinary lisp" interpreted
as GLISP code, it is necessary to follow certain conventions when
writing "ordinary lisp" code.

@Subsection(Reserved Characters)
The colon and the characters which represent the arithmetic operators
should not be used within atom names, since GLISP splits apart "atoms"
which contain operators.  The set of characters to be avoided within
atom names is:
@Programexample{

+ * / ^ _ ~ = < > : ' ,

}
The character "minus" (@PE[ - ]) is permitted within atom names unless
the flag @PE[GLSEPMINUS] is set.

Some GLISP constructs permit (but do
not require) use of the character "comma" (@PE[ , ]); since the comma
is used as a "backquote" character in some Lisp dialects, the user may
wish to avoid its use.  When used in Lisp dialects which use comma as
a backquote character, all commas must be "escaped" or "slashified";
this makes porting of GLISP code containing commas more difficult.

@Subsection(Reserved Function Names)
Most GLISP function, variable, and property names begin with "@PE[GL]"
to avoid conflict with user names.  Those "function" names which are
used in GLISP constructs or in interpretive functions should be
avoided.  This set includes the following names:
@Programexample{

A           AN          CASE         FOR         IF
REPEAT      SEND        SENDPROP     THE         WHILE

}

@SUBSECTION(Other Reserved Names)
Words which are used within GLISP constructs should be avoided as
variable names.  This set of names includes:
@ProgramExample{

A           AN          DO           ELSE        ELSEIF
IS          OF          THE          THEN        UNTIL
}

@SECTION(Lisp Dialect Idiosyncrasies)

GLISP code passes through the Lisp reader before it is seen by GLISP.
For this reason, operators in expressions may need to be set off from
operands by blanks; the operator "@PE[-]" should always be surrounded
by blanks, and the operator "@PE[+]" should be separated from numbers
by blanks.

@Subsection(Interlisp)
GLISP compilation happens automatically, and usually does not need
to be invoked explicitly.  GLISP declarations are integrated with the
file package.
@Subsection(UCI Lisp)
The following command is needed before loading to make room for GLISP:
@ProgramExample[(REALLOC 3000 1000 1000 1000 35000)]
The compiler file modifies the syntax of the character @B[~] to be
"alphabetic" so it can be used as a GLISP operator.
The character "@PE[/]" must be "slashified" to "@PE[//]".
@Subsection(ELISP)
For ELISP, the UCI Lisp version of the compiler is used, together with
a small compatibility file.  The above comments about UCI lisp do not
apply to ELISP.
The characters "@PE[/]" and "@PE[,]" must be "slashified" to "@PE[//]"
and "@PE[/,]".
@Subsection(Maclisp)
The characters "@PE[/]" and "@PE[,]" must be "slashified" to "@PE[//]"
and "@PE[/,]".
@Subsection(Franz Lisp)
Automatic compilation is implemented for Franz Lisp.
The character "@PE[,]" and the operators "@PE[+_]" and "@PE[-_]"
must be "slashified" to "@PE[\,]", "@PE[+\_]", and "@PE[-\_]",
respectively.  Before loading GLISP, edit something to cause the
editor files to be loaded@Foot[Some versions of the "CMU editor"
contain function definitions which may conflict with those of
GLISP; if the editor is loaded first, the GLISP versions override.].
The Franz Lisp version of GLISP has been tested
on Opus 38 Franz Lisp; users with earlier versions of Franz might
encounter difficulties.

@Section(Bug Reports and Mailing List)
To get on the GLISP mailing list or to report bugs, send mail to
CSD.NOVAK@@SU-SCORE.


@Chapter(GLISP Hacks)
This chapter discusses some ways of doing things in GLISP which might
not be entirely obvious at first glance.
@Section(Overloading Basic Types)
GLISP provides the ability to define properties of structures described
in the Structure Description language; since the elementary LISP types
are structures in this language, objects whose storage representation
is an elementary type can be "overloaded" by specifying properties
and operators for them.  The following examples illustrate how this
can be done.
@Begin(ProgramExample)

(GLDEFSTRQ


(ArithmeticOperator  (self ATOM)

   PROP ((Precedence OperatorPrecedenceFn  RESULT INTEGER)
         (PrintForm  ((GETPROP self 'PRINTFORM) or self)) )

   MSG  ((PRIN1      ((PRIN1 the PrintForm)))) )


(IntegerMod7         (self INTEGER)

   PROP ((Modulus    (7))
         (Inverse    ((If self is ZERO then 0
                            else (Modulus - self))) ))

   ADJ  ((Even       ((ZEROP (LOGAND self 1))))
         (Odd        (NOT Even)))

   ISA  ((Prime      PrimeTestFn))

   MSG  ((+          IMod7Plus  OPEN T  RESULT IntegerMod7)
         (_          IMod7Store OPEN T  RESULT IntegerMod7)) )

)
(DEFINEQ

(IMod7Store  (GLAMBDA (LHS:IntegerMod7 RHS:INTEGER)
         (LHS:self __ (IREMAINDER RHS Modulus)) ))

(IMod7Plus   (GLAMBDA (X,Y:IntegerMod7)
         (IREMAINDER (X:self + Y:self) X:Modulus) ))
)
@End(ProgramExample)
A few subtleties of the function IMod7Store are worth noting.
First, the left-hand-side expression used in storing the result is
LHS:self rather than simply LHS.  LHS and LHS:self of course refer
to the same actual structure; however, the @I[type] of LHS is
IntegerMod7, while the type of LHS:self is INTEGER.  If LHS were
used on the left-hand side, since the @PE[ _ ] operator is
overloaded for IntegerMod7, the function IMod7Store would be invoked
again to perform its own function; since the function is compiled
OPEN, this would be an infinite loop.  A second subtlety is that the
assignment to LHS:self must use the self-assignment operator, @PE[@ __@ ],
since it is desired to perform assignment as seen "outside" the
function IMod7Store, i.e., in the environment in which the original
assignment operation was specified.
@Section(Disjunctive Types)
LISP programming often involves objects which may in fact be of
different types, but which are for some purposes treated alike.
For example, LISP data structures are typically constructed of
CONS cells whose fields may point to other CONS cells or to ATOMs.
The GLISP Structure Description language does not permit the user
to specify that a certain field of a structure is a CONS cell @P[or]
an ATOM.  However, it is possible to create a GLISP datatype which
encompasses both.  Typically, this is done by declaring the structure
of the object to be the complex structure, and testing for the
simpler structure explicitly.  This is illustrated for the case of
the LISP tree below.
@Begin(ProgramExample)

   (LISPTREE  (CONS (CAR LISPTREE) (CDR LISPTREE))

      ADJ    ((EMPTY     (@R<~>self)))

      PROP   ((LEFTSON   ((If self is ATOMIC then NIL else CAR)))
              (RIGHTSON  ((If self is ATOMIC then NIL else CDR)))))

@End(ProgramExample)
@Section(Generators)
Often, one would like to define such properties of an object as the
way of enumerating its parts in some order.  Such things
cannot be specified directly as properties of the object because they
depend on the previous state of the enumeration.  However, it is
possible to define an object, associated with the original datatype,
which contains the state of the enumeration and responds to Messages.
This is illustrated below by an object which searches a tree in Preorder.
@Begin(ProgramExample)

(PreorderSearchRecord  (CONS (Node LISPTREE)
                             (PreviousNodes (LISTOF LISPTREE)))

   MSG  ((NEXT  ((PROG (TMP)
                    (If TMP_Node:LEFTSON
                        then (If Node:RIGHTSON
                                 then PreviousNodes+_Node)
                             Node_TMP
                        else TMP-_PreviousNodes
                             Node_TMP:RIGHTSON) ))))


(TP (GLAMBDA ((A LISPTREE))
      (PROG (PSR)
         (PSR _ (A PreorderSearchRecord
                   with Node = (the LISPTREE)))
         (While Node (If Node is ATOMIC (PRINT Node))
                     (_ PSR NEXT)) )))

@End(ProgramExample)
The object class PreorderSearchRecord serves two purposes: it holds
the state of the enumeration, and it responds to messages to step
through the enumeration.  With these definitions, it is easy to write
a program involving enumeration of a LISPTREE, as illustrated by
the example function TP above.  By being open-compiled, messages to
an object can be as efficient as in-line hand coding; yet, the code
for the messages only has to be written once, and can easily be
changed without changing the programs which use the messages.
@Chapter(Program Examples)
In this chapter, examples of GLISP object declarations and programs
are presented.  Each example is discussed as a section of this
chapter; the code for the examples and the code produced by the
compiler are shown for each example at the end of the chapter.
@Section(GLTST1 File)
The GLTST1 file illustrates the use of several types of LISP
structures, and the use of fairly complex Property definitions
for objects.  SENIORITY of an EMPLOYEE, for example, is defined
in terms of the YEAR of DATE-HIRED, which is a substructure of
EMPLOYEE, and the YEAR of the function (CURRENTDATE).
@Foot[The @I<type> of (CURRENTDATE) must be known to the compiler,
either by compiling it first, or by including a RESULT declaration
in the function definition of CURRENTDATE, or by specifying the
GLRESULTTYPE property for the function name.]
@Section(GLTST2 File)
The GLTST2 file illustrates the use of Messages for ordinary LISP
objects.  By defining the arithmetic operators as Message selectors
for the object VECTOR, use of vectors in arithmetic expressions
is enabled; OPEN compilation is specified for these messages.

The definition of GRAPHICSOBJECT uses VECTORs as components.
While the actual structure of a GRAPHICSOBJECT is simple,
numerous properties are defined for user convenience.
The definition of CENTER is easily stated as a VECTOR expression.

The Messages of GRAPHICSOBJECT illustrate how different responses
to a message for different types of objects can be achieved, even
though for GLISP compilation of messages to LISP objects the code
for a message must be resolved at compile time.
@Foot[For objects in a Representation Language, messages may be
compiled directly as LISP code or as messages to be interpreted at
runtime, depending on how much is known about the object to which the
message is sent and the compilation declarations in effect.]
The DRAW and
ERASE messages get the function to be used from the property list
of the SHAPE name of the GRAPHICSOBJECT and APPLY it to draw the
desired object.

MOVINGGRAPHICSOBJECT contains a GRAPHICSOBJECT as a TRANSPARENT
component, so that it inherits the properties of a GRAPHICSOBJECT;
a MOVINGGRAPHICSOBJECT is a GRAPHICSOBJECT which has a VELOCITY,
and will move itself by the amount of its velocity upon the message
command STEP.@Foot[This example is adapted from the MovingPoint
example written by Dan Bobrow for LOOPS.]
The compilation of the message
@PE[(_@ MGO@ STEP)] in the function TESTFN1 is of particular
interest.  This message is expanded
into the sending of the message @PE[(_@ self@ MOVE@ VELOCITY)]
to the MOVINGGRAPHICSOBJECT.  The MOVINGGRAPHICSOBJECT cannot respond
to such a message; however, since it contains a GRAPHICSOBJECT as a
TRANSPARENT component, its GRAPHICSOBJECT responds to the message.
@Foot[TRANSPARENT substructures thus permit procedural inheritance by
LISP objects.]
A GRAPHICSOBJECT responds to a MOVE message by
erasing itself, increasing its START point by the (vector) distance
to be moved, and
then redrawing itself.  All of the messages are specified as being
compiled open, so that the short original message actually generates
a large amount of code.

A rectangle is drawn by the function DRAWRECT.  Note how the use of
the properties defined for a GRAPHICSOBJECT allows an easy interface
to the system functions MOVETO and DRAWTO in terms of the properties
LEFT, RIGHT, TOP, and BOTTOM.