<!doctype linuxdoc system>
<book>
<title> A Guide to the S-Lang Language
<author> John E. Davis, <tt>davis@space.mit.edu</tt>
<date> Sat Feb 3 01:58:05 2001
<toc>
<chapt>Preface<p>
<bf>S-Lang</bf> is an interpreted language that was designed from the start
to be easily embedded into a program to provide it with a powerful
extension language. Examples of programs that use <bf>S-Lang</bf> as an
extension language include the <bf>jed</bf> text editor, the <bf>slrn</bf>
newsreader, and <bf>sldxe</bf> (unreleased), a numerical computation
program. For this reason, <bf>S-Lang</bf> does not exist as a separate
application and many of the examples in this document are presented
in the context of one of the above applications.
<bf>S-Lang</bf> is also a programmer's library that permits a programmer to
develop sophisticated platform-independent software. In addition to
providing the <bf>S-Lang</bf> extension language, the library provides
facilities for screen management, keymaps, low-level terminal I/O,
etc. However, this document is concerned only with the extension
language and does not address these other features of the <bf>S-Lang</bf>
library. For information about the other components of the library,
the reader is referred to the <bf>The <bf>S-Lang</bf> Library Reference</bf>.
<sect>A Brief History of <bf>S-Lang</bf><p>
I first began working on <bf>S-Lang</bf> sometime during the fall of 1992.
At that time I was writing a text editor (<bf>jed</bf>), which I wanted to
endow with a macro language. It occured to me that an
application-independent language that could be embedded into the
editor would prove more useful because I could envision embedding it
into other programs. As a result, <bf>S-Lang</bf> was born.
<bf>S-Lang</bf> was originally a stack language that supported a
postscript-like syntax. For that reason, I named it <bf>S-Lang</bf>, where
the <em>S</em> was supposed to emphasize its stack-based nature. About
a year later, I began to work on a preparser that would allow one to
write using a more traditional infix syntax making it easier to use
for those unfamiliar with stack based languages. Currently, the
syntax of the language resembles C, nevertheless some
postscript-like features still remain, e.g., the `<tt>%</tt>' character
is still used as a comment delimiter.
<sect>Acknowledgements<p>
Since I first released <bf>S-Lang</bf>, I have received a lot feedback about
the library and the language from many people. This has given me
the opportunity and pleasure to interact with several people to
make the library portable and easy to use. In particular, I would
like to thank the following individuals:
Luchesar Ionkov <tt><lionkov@sf.cit.bg></tt> for his comments and
criticisms of the syntax of the language. He was the person who
made me realize that the low-level byte-code engine should be
totally type-independent. He also improved the tokenizer and
preparser and impressed upon me that the language needed a
grammar.
Mark Olesen <tt><olesen@weber.me.queensu.ca></tt> for his many patches to
various aspects of the library and his support on AIX. He also
contributed a lot to the pre-processing (<tt>SLprep</tt>) routines.
John Burnell <tt><j.burnell@irl.cri.nz></tt> for the OS/2 port of the
video and keyboard routines. He also made value suggestions
regarding the interpreter interface.
Darrel Hankerson <tt><hankedr@mail.auburn.edu></tt> for cleaning up and
unifying some of the code and the makefiles.
Dominik Wujastyk <tt><ucgadkw@ucl.ac.uk></tt> who was always willing to test
new releases of the library.
Michael Elkins <tt><me@muddcs.cs.hmc.edu></tt> for his work on the curses
emulation.
Ulli Horlacher <tt><framstag@belwue.de></tt> and Oezguer Kesim
<tt><kesim@math.fu-berlin.de></tt> for the <bf>S-Lang</bf> newsgroup and mailing list.
Hunter Goatley, Andy Harper <tt><Andy.Harper@kcl.ac.uk></tt>, and Martin
P.J. Zinser <tt><zinser@decus.decus.de></tt> for their VMS support.
Dave Sims <tt><sims@usa.acsys.com></tt> and Chin Huang
<tt><cthuang@vex.net></tt> for Windows 95 and Windows NT support.
Lloyd Zusman <tt><ljz@asfast.com></tt> and Rich Roth <tt><rich@on-the-net.com></tt>
for creating and maintaining <tt>www.s-lang.org</tt>.
I am also grateful to many other people who send in bug-reports and
bug-fixes, for without such community involvement, <bf>S-Lang</bf> would not
be as well-tested and stable as it is. Finally, I would like to
thank my wife for her support and understanding while I spent long
weekend hours developing the library.
<chapt>Introduction<p>
<bf>S-Lang</bf> is a powerful interpreted language that may be embedded into
an application to make the application extensible. This enables
the application to be used in ways not envisioned by the programmer,
thus providing the application with much more flexibility and
power. Examples of applications that take advantage of the
interpreter in this way include the <bf>jed</bf> editor and the <bf>slrn</bf>
newsreader.
<sect>Language Features<p>
The language features both global and local variables, branching
and looping constructs, user-defined functions, structures,
datatypes, and arrays. In addition, there is limited support for
pointer types. The concise array syntax rivals that of commercial
array-based numerical computing environments.
<sect>Data Types and Operators<p>
The language provides built-in support for string, integer (signed
and unsigned long and short), double precision floating point, and
double precision complex numbers. In addition, it supports user
defined structure types, multi-dimensional array types, and
associative arrays. To facilitate the construction of
sophisticated data structures such as linked lists and trees, a
`reference' type was added to the language. The reference type
provides much of the same flexibility as pointers in other
languages. Finally, applications embedding the interpreter may
also provide special application specific types, such as the
<tt>Mark_Type</tt> that the <bf>jed</bf> editor provides.
The language provides standard arithmetic operations such as
addition, subtraction, multiplication, and division. It also
provides support for modulo arithmetic as well as operations at
the bit level, e.g., exclusive-or. Any binary or unary operator
may be extended to work with any data type. For example, the
addition operator (<tt>+</tt>) has been extended to work between
string types to permit string concatenation.
The binary and unary operators work transparently with array types.
For example, if <tt>a</tt> and <tt>b</tt> are arrays, then <tt>a + b</tt>
produces an array whose elements are the result of element by
element addition of <tt>a</tt> and <tt>b</tt>. This permits one to do
vector operations without explicitly looping over the array
indices.
<sect>Statements and Functions<p>
The <bf>S-Lang</bf> language supports several types of looping constructs and
conditional statements. The looping constructs include <tt>while</tt>,
<tt>do...while</tt>, <tt>for</tt>, <tt>forever</tt>, <tt>loop</tt>, <tt>foreach</tt>,
and <tt>_for</tt>. The conditional statements include <tt>if</tt>,
<tt>if-then-else</tt>, and <tt>!if</tt>.
User defined functions may be defined to return zero, one, or more
values. Functions that return zero values are similar to
`procedures' in languages such as PASCAL. The local variables of a
function are always created on a stack allowing one to create
recursive functions. Parameters to a function are always passed by
value and never by reference. However, the language supports a
<em>reference</em> data type that allows one to simulate pass by
reference.
Unlike many interpreted languages, <bf>S-Lang</bf> allows functions to be
dynamically loaded (function autoloading). It also provides
constructs specifically designed for error handling and recovery as
well as debugging aids (e.g., function tracebacks).
Functions and variables may be declared as private belonging to a
namespace associated with the compilation unit that defines the
function or variable. The ideas behind the namespace implementation
stems from the C language and should be quite familiar to any one
familiar with C.
<sect>Error Handling<p>
The <bf>S-Lang</bf> language defines a construct called an <em>error-block</em>
that may be used for error handling and recovery. When a non-fatal
run-time error is encountered, any error blocks that have been
defined are executed as the run-time stack unwinds. An error block
can optionally clear the error and the program will continue
running after the statement that triggered the error. This
mechanism is somewhat similar to try-catch in C++.
<sect>Run-Time Library<p>
Functions that compose the <bf>S-Lang</bf> run-time library are called
<em>intrinsics</em>. Examples of <bf>S-Lang</bf> intrinsic functions available
to every <bf>S-Lang</bf> application include string manipulation functions
such as <tt>strcat</tt>, <tt>strchop</tt>, and <tt>strcmp</tt>. The <bf>S-Lang</bf>
library also provides mathematical functions such as <tt>sin</tt>,
<tt>cos</tt>, and <tt>tan</tt>; however, not all applications enable the
use of these intrinsics. For example, to conserve memory, the 16
bit version of the <bf>jed</bf> editor does not provide support for any
mathematics other than simple integer arithmetic, whereas other
versions of the editor do support these functions.
Most applications embedding the languages will also provide a set of
application specific intrinsic functions. For example, the <bf>jed</bf>
editor adds over 100 application specific intrinsic functions to
the language. Consult your application specific documentation to
see what additional intrinsics are supported.
<sect>Input/Output<p>
The language supports C-like stdio input/output functions such as
<tt>fopen</tt>, <tt>fgets</tt>, <tt>fputs</tt>, and <tt>fclose</tt>. In
addition it provides two functions, <tt>message</tt> and <tt>error</tt>,
for writing to the standard output device and standard error.
Specific applications may provide other I/O mechanisms, e.g.,
the <bf>jed</bf> editor supports I/O to files via the editor's
buffers.
<sect>Obtaining <bf>S-Lang</bf><p>
Comprehensive information about the library may be obtained via the
World Wide Web from <tt>http://www.s-lang.org</tt>.
<bf>S-Lang</bf> as well as some programs that embed it are freely available
via anonymous ftp in the United States from
<itemize>
<item> <htmlurl url="ftp://space.mit.edu/pub/davis" name="ftp://space.mit.edu/pub/davis">.
</itemize>
It is also available outside the United States from the following
mirror sites:
<itemize>
<item> <htmlurl url="ftp://ftp.uni-stuttgart.de/pub/unix/misc/slang/" name="ftp://ftp.uni-stuttgart.de/pub/unix/misc/slang/">
<item> <htmlurl url="ftp://ftp.fu-berlin.de/pub/unix/news/slrn/" name="ftp://ftp.fu-berlin.de/pub/unix/news/slrn/">
<item> <htmlurl url="ftp://ftp.ntua.gr/pub/lang/slang/" name="ftp://ftp.ntua.gr/pub/lang/slang/">
</itemize>
The Usenet newsgroup <tt>alt.lang.s-lang</tt> was created for <bf>S-Lang</bf>
programmers to exchange information and share macros for the various
programs the embed the language. The newsgroup <tt>comp.editors</tt>
can be a useful resource for <bf>S-Lang</bf> macros for the <bf>jed</bf> editor.
Similarly, <bf>slrn</bf> users will find <tt>news.software.readers</tt> to be a
valuable source of information.
Finally, two mailing lists dealing with the <bf>S-Lang</bf> library have been
created:
<itemize>
<item> <tt>slang-announce@babayaga.math.fu-berlin.de</tt>
<item> <tt>slang-workers@babayaga.math.fu-berlin.de</tt>
</itemize>
The first list is for announcements of new releases of the library, while the
second list is intended for those who use the library for their own code
development. To subscribe to the announcement list, send an email to
<tt>slang-announce-subscribe@babayaga.math.fu-berlin.de</tt> and include
the word <tt>subscribe</tt> in the body of the message. To subscribe to
the developers list, use the address
<tt>slang-workers-subscribe@babayaga.math.fu-berlin.de</tt>.
<chapt>Overview of the Language<p>
This purpose of this section is to give the reader a feel for the
<bf>S-Lang</bf> language, its syntax, and its capabilities. The information
and examples presented in this section should be sufficient to
provide the reader with the necessary background to understand the
rest of the document.
<sect>Variables and Functions<p>
<bf>S-Lang</bf> is different from many other interpreted languages in the
sense that all variables and functions must be declared before they
can be used.
Variables are declared using the <tt>variable</tt> keyword, e.g.,
<tscreen><verb>
variable x, y, z;
</verb></tscreen>
declares three variables, <tt>x</tt>, <tt>y</tt>, and <tt>z</tt>. Note the
semicolon at the end of the statement. <em>All <bf>S-Lang</bf> statements must
end in a semi-colon.</em>
Unlike compiled languages such as C, it is not necessary to specify
the data type of a <bf>S-Lang</bf> variable. The data type of a <bf>S-Lang</bf>
variable is determined upon assignment. For example, after
execution of the statements
<tscreen><verb>
x = 3;
y = sin (5.6);
z = "I think, therefore I am.";
</verb></tscreen>
<tt>x</tt> will be an integer, <tt>y</tt> will be a
double, and <tt>z</tt> will be a string. In fact, it is even possible
to re-assign <tt>x</tt> to a string:
<tscreen><verb>
x = "x was an integer, but now is a string";
</verb></tscreen>
Finally, one can combine variable declarations and assignments in
the same statement:
<tscreen><verb>
variable x = 3, y = sin(5.6), z = "I think, therefore I am.";
</verb></tscreen>
Most functions are declared using the <tt>define</tt> keyword. A
simple example is
<tscreen><verb>
define compute_average (x, y)
{
variable s = x + y;
return s / 2.0;
}
</verb></tscreen>
which defines a function that simply computes the average of two
numbers and returns the result. This example shows that a function
consists of three parts: the function name, a parameter list, and
the function body.
The parameter list consists of a comma separated list of variable
names. It is not necessary to declare variables within a parameter
list; they are implicitly declared. However, all other <em>local</em>
variables used in the function must be declared. If the function
takes no parameters, then the parameter list must still be present,
but empty:
<tscreen><verb>
define go_left_5 ()
{
go_left (5);
}
</verb></tscreen>
The last example is a function that takes no arguments and returns
no value. Some languages such as PASCAL distinguish such objects
from functions that return values by calling these objects
<em>procedures</em>. However, <bf>S-Lang</bf>, like C, does not make such a
distinction.
The language permits <em>recursive</em> functions, i.e., functions that
call themselves. The way to do this in <bf>S-Lang</bf> is to first declare
the function using the form:
<tscreen>
define <em>function-name</em> ();
</tscreen>
It is not necessary to declare a parameter list when declaring a
function in this way.
The most famous example of a recursive function is the factorial
function. Here is how to implement it using <bf>S-Lang</bf>:
<tscreen><verb>
define factorial (); % declare it for recursion
define factorial (n)
{
if (n < 2) return 1;
return n * factorial (n - 1);
}
</verb></tscreen>
This example also shows how to mix comments with code. <bf>S-Lang</bf> uses
the `<tt>%</tt>' character to start a comment and all characters from
the comment character to the end of the line are ignored.
<sect>Strings<p>
Perhaps the most appealing feature of any interpreted language is
that it frees the user from the responsibility of memory management.
This is particularly evident when contrasting how
<bf>S-Lang</bf> handles string variables with a lower level language such as
C. Consider a function that concatenates three strings. An
example in <bf>S-Lang</bf> is:
<tscreen><verb>
define concat_3_strings (a, b, c)
{
return strcat (a, strcat (b, c));
}
</verb></tscreen>
This function uses the built-in
<tt>strcat</tt> function for concatenating two strings. In C, the
simplest such function would look like:
<tscreen><verb>
char *concat_3_strings (char *a, char *b, char *c)
{
unsigned int len;
char *result;
len = strlen (a) + strlen (b) + strlen (c);
if (NULL == (result = (char *) malloc (len + 1)))
exit (1);
strcpy (result, a);
strcat (result, b);
strcat (result, c);
return result;
}
</verb></tscreen>
Even this C example is misleading since none of the issues of memory
management of the strings has been dealt with. The <bf>S-Lang</bf> language
hides all these issues from the user.
Binary operators have been defined to work with the string data
type. In particular the <tt>+</tt> operator may be used to perform
string concatenation. That is, one can use the
<tt>+</tt> operator as an alternative to <tt>strcat</tt>:
<tscreen><verb>
define concat_3_strings (a, b, c)
{
return a + b + c;
}
</verb></tscreen>
See section ??? for more information about string variables.
<sect>Referencing and Dereferencing<p>
The unary prefix operator, <tt>&</tt>, may be used to create a
<em>reference</em> to an object, which is similar to a pointer
in other languages. References are commonly used as a mechanism to
pass a function as an argument to another function as the following
example illustrates:
<tscreen><verb>
define compute_functional_sum (funct)
{
variable i, sum;
sum = 0;
for (i = 0; i < 10; i++)
{
sum += @funct (i);
}
return sum;
}
variable sin_sum = compute_functional_sum (&sin);
variable cos_sum = compute_functional_sum (&cos);
</verb></tscreen>
Here, the function <tt>compute_functional_sum</tt> applies the
function specified by the parameter <tt>funct</tt> to the first
<tt>10</tt> integers and returns the sum. The two statements
following the function definition show how the <tt>sin</tt> and
<tt>cos</tt> functions may be used.
Note the <tt>@</tt> operator in the definition of
<tt>compute_functional_sum</tt>. It is known as the <em>dereference</em>
operator and is the inverse of the reference operator.
Another use of the reference operator is in the context of the
<tt>fgets</tt> function. For example,
<tscreen><verb>
define read_nth_line (file, n)
{
variable fp, line;
fp = fopen (file, "r");
while (n > 0)
{
if (-1 == fgets (&line, fp))
return NULL;
n--;
}
return line;
}
</verb></tscreen>
uses the <tt>fgets</tt> function to read the nth line of a file.
In particular, a reference to the local variable <tt>line</tt> is
passed to <tt>fgets</tt>, and upon return <tt>line</tt> will be set to
the character string read by <tt>fgets</tt>.
Finally, references may be used as an alternative to multiple
return values by passing information back via the parameter list.
The example involving <tt>fgets</tt> presented above provided an
illustration of this. Another example is
<tscreen><verb>
define set_xyz (x, y, z)
{
@x = 1;
@y = 2;
@z = 3;
}
variable X, Y, Z;
set_xyz (&X, &Y, &Z);
</verb></tscreen>
which, after execution, results in <tt>X</tt> set to <tt>1</tt>, <tt>Y</tt>
set to <tt>2</tt>, and <tt>Z</tt> set to <tt>3</tt>. A C programmer will
note the similarity of <tt>set_xyz</tt> to the following C
implementation:
<tscreen><verb>
void set_xyz (int *x, int *y, int *z)
{
*x = 1;
*y = 2;
*z = 3;
}
</verb></tscreen>
<sect>Arrays<p>
The <bf>S-Lang</bf> language supports multi-dimensional arrays of all
datatypes. For example, one can define arrays of references to
functions as well as arrays of arrays. Here are a few examples of
creating arrays:
<tscreen><verb>
variable A = Integer_Type [10];
variable B = Integer_Type [10, 3];
variable C = [1, 3, 5, 7, 9];
</verb></tscreen>
The first example creates an array of <tt>10</tt> integers and assigns
it to the variable <tt>A</tt>. The second example creates a 2-d array
of <tt>30</tt> integers arranged in <tt>10</tt> rows and <tt>3</tt> columns
and assigns the result to <tt>B</tt>. In the last example, an array
of <tt>5</tt> integers is assigned to the variable <tt>C</tt>. However,
in this case the elements of the array are initialized to the
values specified. This is known as an <em>inline-array</em>.
<bf>S-Lang</bf> also supports something called an
<em>range-array</em>. An example of such an array is
<tscreen><verb>
variable C = [1:9:2];
</verb></tscreen>
This will produce an array of 5 integers running from <tt>1</tt>
through <tt>9</tt> in increments of <tt>2</tt>.
Arrays are passed by reference to functions and never by value.
This permits one to write functions which can initialize arrays.
For example,
<tscreen><verb>
define init_array (a)
{
variable i, imax;
imax = length (a);
for (i = 0; i < imax; i++)
{
a[i] = 7;
}
}
variable A = Integer_Type [10];
init_array (A);
</verb></tscreen>
creates an array of <tt>10</tt> integers and initializes all its
elements to <tt>7</tt>.
There are more concise ways of accomplishing the result of the
previous example. These include:
<tscreen><verb>
variable A = [7, 7, 7, 7, 7, 7, 7, 7, 7, 7];
variable A = Integer_Type [10]; A[[0:9]] = 7;
variable A = Integer_Type [10]; A[*] = 7;
</verb></tscreen>
The second and third methods use an array of indices to index the array
<tt>A</tt>. In the second, the range of indices has been explicitly
specified, whereas the third example uses a wildcard form. See
section ??? for more information about array indexing.
Although the examples have pertained to integer arrays, the fact is
that <bf>S-Lang</bf> arrays can be of any type, e.g.,
<tscreen><verb>
variable A = Double_Type [10];
variable B = Complex_Type [10];
variable C = String_Type [10];
variable D = Ref_Type [10];
</verb></tscreen>
create <tt>10</tt> element arrays of double, complex, string, and
reference types, respectively. The last example may be used to
create an array of functions, e.g.,
<tscreen><verb>
D[0] = &sin;
D[1] = &cos;
</verb></tscreen>
The language also defines unary, binary, and mathematical
operations on arrays. For example, if <tt>A</tt> and <tt>B</tt> are
integer arrays, then <tt>A + B</tt> is an array whose elements are
the sum of the elements of <tt>A</tt> and <tt>B</tt>. A trivial example
that illustrates the power of this capability is
<tscreen><verb>
variable X, Y;
X = [0:2*PI:0.01];
Y = 20 * sin (X);
</verb></tscreen>
which is equivalent to the highly simplified C code:
<tscreen><verb>
double *X, *Y;
unsigned int i, n;
n = (2 * PI) / 0.01 + 1;
X = (double *) malloc (n * sizeof (double));
Y = (double *) malloc (n * sizeof (double));
for (i = 0; i < n; i++)
{
X[i] = i * 0.01;
Y[i] = 20 * sin (X[i]);
}
</verb></tscreen>
<sect>Structures and User-Defined Types<p>
A <em>structure</em> is similar to an array in the sense that it is a
container object. However, the elements of an array must all be of
the same type (or of <tt>Any_Type</tt>), whereas a structure is
heterogeneous. As an example, consider
<tscreen><verb>
variable person = struct
{
first_name, last_name, age
};
variable bill = @person;
bill.first_name = "Bill";
bill.last_name = "Clinton";
bill.age = 51;
</verb></tscreen>
In this example a structure consisting of the three fields has been
created and assigned to the variable <tt>person</tt>. Then an
<em>instance</em> of this structure has been created using the
dereference operator and assigned to <tt>bill</tt>. Finally, the
individual fields of <tt>bill</tt> were initialized. This is an
example of an <em>anonymous</em> structure.
A <em>named</em> structure is really a new data type and may be created
using the <tt>typedef</tt> keyword:
<tscreen><verb>
typedef struct
{
first_name, last_name, age
}
Person_Type;
variable bill = @Person_Type;
bill.first_name = "Bill";
bill.last_name = "Clinton";
bill.age = 51;
</verb></tscreen>
The big advantage of creating a new type is that one can go on to
create arrays of the data type
<tscreen><verb>
variable People = Person_Type [100];
People[0].first_name = "Bill";
People[1].first_name = "Hillary";
</verb></tscreen>
The creation and initialization of a structure may be facilitated
by a function such as
<tscreen><verb>
define create_person (first, last, age)
{
variable person = @Person_Type;
person.first_name = first;
person.last_name = last;
person.age = age;
return person;
}
variable Bill = create_person ("Bill", "Clinton", 51);
</verb></tscreen>
Other common uses of structures is the creation of linked lists,
binary trees, etc. For more information about these and other
features of structures, see section ???.
<sect>Namespaces<p>
In addition to the global namespace, each compilation unit (e.g., a
file) is given a private namespace. A variable or function name
that is declared using the <tt>static</tt> keyword will be placed in
the private namespace associated with compilation unit. For
example,
<tscreen><verb>
variable i;
static variable i;
</verb></tscreen>
defines two variables called <tt>i</tt>. The first declaration
defines <tt>i</tt> in the global namespace, but the second declaration
defines <tt>i</tt> in the private namespace.
The <tt>-></tt> operator may be used in conjunction with the name of
the namespace to access objects in the name space. In the above
example, to access the variable <tt>i</tt> in the global namespace,
one would use <tt>Global->i</tt>. Unless otherwise specified, a
private namespace has no name and its objects may not be accessed
from outside the compilation unit. However, the <tt>implements</tt>
function may be used give the private namespace a name, allowing
access to its objects. For example, if the file <tt>t.sl</tt> contains
<tscreen><verb>
implements ("A");
static variable i;
</verb></tscreen>
then another file may access the variable <tt>i</tt> via <tt>A->i</tt>.
<chapt>Data Types and Literal Constants<p>
The current implementation of the <bf>S-Lang</bf> language permits up to 256
distinct data types, including predefined data types such as integer and
floating point, as well as specialized applications specific data
types. It is also possible to create new data types in the
language using the <tt>typedef</tt> mechanism.
Literal constants are objects such as the integer <tt>3</tt> or the
string <tt>"hello"</tt>. The actual data type given to a literal
constant depends upon the syntax of the constant. The following
sections describe the syntax of literals of specific data types.
<sect>Predefined Data Types<p>
The current version of <bf>S-Lang</bf> defines integer, floating point,
complex, and string types. It also defines special purpose data
types such as <tt>Null_Type</tt>, <tt>DataType_Type</tt>, and
<tt>Ref_Type</tt>. These types are discussed below.
<sect1>Integers<p>
The <bf>S-Lang</bf> language supports both signed and unsigned characters,
short integer, long integer, and plain integer types. On most 32
bit systems, there is no difference between an integer and a long
integer; however, they may differ on 16 and 64 bit systems.
Generally speaking, on a 16 bit system, plain integers are 16 bit
quantities with a range of -32767 to 32767. On a 32 bit system,
plain integers range from -2147483648 to 2147483647.
An plain integer <em>literal</em> can be specified in one of several ways:
<itemize>
<item> As a decimal (base 10) integer consisting of the characters
<tt>0</tt> through <tt>9</tt>, e.g., <tt>127</tt>. An integer specified
this way cannot begin with a leading <tt>0</tt>. That is,
<tt>0127</tt> is <em>not</em> the same as <tt>127</tt>.
<item> Using hexadecimal (base 16) notation consisting of the characters
<tt>0</tt> to <tt>9</tt> and <tt>A</tt> through <tt>F</tt>. The hexadecimal
number must be preceded by the characters <tt>0x</tt>. For example,
<tt>0x7F</tt> specifies an integer using hexadecimal notation and has
the same value as decimal <tt>127</tt>.
<item> In Octal notation using characters <tt>0</tt> through <tt>7</tt>. The Octal
number must begin with a leading <tt>0</tt>. For example,
<tt>0177</tt> and <tt>127</tt> represent the same integer.
Short, long, and unsigned types may be specified by using the
proper suffixes: <tt>L</tt> indicates that the integer is a long
integer, <tt>h</tt> indicates that the integer is a short integer, and
<tt>U</tt> indicates that it is unsigned. For example, <tt>1UL</tt>
specifies an unsigned long integer.
Finally, a character literal may be specified using a notation
containing a character enclosed in single quotes as <tt>'a'</tt>.
The value of the character specified this way will lie in the
range 0 to 256 and will be determined by the ASCII value of the
character in quotes. For example,
<tscreen><verb>
i = '0';
</verb></tscreen>
assigns to <tt>i</tt> the character 48 since the <tt>'0'</tt> character
has an ASCII value of 48.
</itemize>
Any integer may be preceded by a minus sign to indicate that it is a
negative integer.
<sect1>Floating Point Numbers<p>
Single and double precision floating point literals must contain either a
decimal point or an exponent (or both). Here are examples of
specifying the same double precision point number:
<tscreen><verb>
12. 12.0 12e0 1.2e1 120e-1 .12e2 0.12e2
</verb></tscreen>
Note that <tt>12</tt> is <em>not</em> a floating point number since it
contains neither a decimal point nor an exponent. In fact,
<tt>12</tt> is an integer.
One may append the <tt>f</tt> character to the end of the number to
indicate that the number is a single precision literal.
<sect1>Complex Numbers<p>
The language implements complex numbers as a pair of double
precision floating point numbers. The first number in the pair
forms the <em>real</em> part, while the second number forms the
<em>imaginary</em> part. That is, a complex number may be regarded as the
sum of a real number and an imaginary number.
Strictly speaking, the current implementation of the <bf>S-Lang</bf> does
not support generic complex literals. However, it does support
imaginary literals and a more generic complex number with a non-zero
real part may be constructed from the imaginary literal via
addition of a real number.
An imaginary literal is specified in the same way as a floating
point literal except that <tt>i</tt> or <tt>j</tt> is appended. For
example,
<tscreen><verb>
12i 12.0i 12e0j
</verb></tscreen>
all represent the same imaginary number. Actually, <tt>12i</tt> is
really an imaginary integer except that <bf>S-Lang</bf> automatically
promotes it to a double precision imaginary number.
A more generic complex number may be constructed from an imaginary
literal via addition, e.g.,
<tscreen><verb>
3.0 + 4.0i
</verb></tscreen>
produces a complex number whose real part is <tt>3.0</tt> and whose
imaginary part is <tt>4.0</tt>.
The intrinsic functions <tt>Real</tt> and <tt>Imag</tt> may be used to
retrieve the real and imaginary parts of a complex number,
respectively.
<sect1>Strings<p>
A string literal must be enclosed in double quotes as in:
<tscreen><verb>
"This is a string".
</verb></tscreen>
Although there is no imposed limit on the length of a string,
string literals must be less than 256 characters in length. It is
possible to go beyond this limit by string concatenation, e.g.,
<tscreen><verb>
"This is the first part of a long string"
+ "and this is the second half"
</verb></tscreen>
Any character except a newline (ASCII 10) or the null character
(ASCII 0) may appear explicitly in a string literal. However,
these characters may be used implicitly using the mechanism
described below.
The backslash character is a special character and is used to
include other special characters (such as a newline character) in
the string. The special characters recognized are:
<tscreen><verb>
\" -- double quote
\' -- single quote
\\ -- backslash
\a -- bell character (ASCII 7)
\t -- tab character (ASCII 9)
\n -- newline character (ASCII 10)
\e -- escape character (ASCII 27)
\xhhh -- character expressed in HEXADECIMAL notation
\ooo -- character expressed in OCTAL notation
\dnnn -- character expressed in DECIMAL
</verb></tscreen>
For example, to include the double quote character as part of the
string, it must be preceded by a backslash character, e.g.,
<tscreen><verb>
"This is a \"quote\""
</verb></tscreen>
Similarly, the next illustrates how a newline character may be
included:
<tscreen><verb>
"This is the first line\nand this is the second"
</verb></tscreen>
<sect1>Null_Type<p>
Objects of type <tt>Null_Type</tt> can have only one value:
<tt>NULL</tt>. About the only thing that you can do with this data
type is to assign it to variables and test for equality with
other objects. Nevertheless, <tt>Null_Type</tt> is an important and
extremely useful data type. Its main use stems from the fact that
since it can be compared for equality with any other data type, it
is ideal to represent the value of an object which does not yet
have a value, or has an illegal value.
As a trivial example of its use, consider
<tscreen><verb>
define add_numbers (a, b)
{
if (a == NULL) a = 0;
if (b == NULL) b = 0;
return a + b;
}
variable c = add_numbers (1, 2);
variable d = add_numbers (1, NULL);
variable e = add_numbers (1,);
variable f = add_numbers (,);
</verb></tscreen>
It should be clear that after these statements have been executed,
<tt>c</tt> will have a value of <tt>3</tt>. It should also be clear
that <tt>d</tt> will have a value of <tt>1</tt> because <tt>NULL</tt> has
been passed as the second parameter. One feature of the language
is that if a parameter has been omitted from a function call, the
variable associated with that parameter will be set to <tt>NULL</tt>.
Hence, <tt>e</tt> and <tt>f</tt> will be set to <tt>1</tt> and <tt>0</tt>,
respectively.
The <tt>Null_Type</tt> data type also plays an important role in the
context of <em>structures</em>.
<sect1>Ref_Type<p>
Objects of <tt>Ref_Type</tt> are created using the unary
<em>reference</em> operator <tt>&</tt>. Such objects may be
<em>dereferenced</em> using the dereference operator <tt>@</tt>. For
example,
<tscreen><verb>
variable sin_ref = &sin;
variable y = @sin_ref (1.0);
</verb></tscreen>
creates a reference to the <tt>sin</tt> function and assigns it to
<tt>sin_ref</tt>. The second statement uses the dereference operator
to call the function that <tt>sin_ref</tt> references.
The <tt>Ref_Type</tt> is useful for passing functions as arguments to
other functions, or for returning information from a function via
its parameter list. The dereference operator is also used to create
an instance of a structure. For these reasons, further discussion
of this important type can be found in section ??? and section ???.
<sect1>Array_Type and Struct_Type<p>
Variables of type <tt>Array_Type</tt> and <tt>Struct_Type</tt> are known
as <em>container objects</em>. They are much more complicated than the
simple data types discussed so far and each obeys a special syntax.
For these reasons they are discussed in a separate chapters.
See ???.
<sect1>DataType_Type Type<p>
<bf>S-Lang</bf> defines a type called <tt>DataType_Type</tt>. Objects of
this type have values that are type names. For example, an integer
is an object of type <tt>Integer_Type</tt>. The literals of
<tt>DataType_Type</tt> include:
<tscreen><verb>
Char_Type (signed character)
UChar_Type (unsigned character)
Short_Type (short integer)
UShort_Type (unsigned short integer)
Integer_Type (plain integer)
UInteger_Type (plain unsigned integer)
Long_Type (long integer)
ULong_Type (unsigned long integer)
Float_Type (single precision real)
Double_Type (double precision real)
Complex_Type (complex numbers)
String_Type (strings, C strings)
BString_Type (binary strings)
Struct_Type (structures)
Ref_Type (references)
Null_Type (NULL)
Array_Type (arrays)
DataType_Type (data types)
</verb></tscreen>
as well as the names of any other types that an application
defines.
The built-in function <tt>typeof</tt> returns the data type of
its argument, i.e., a <tt>DataType_Type</tt>. For instance
<tt>typeof(7)</tt> returns <tt>Integer_Type</tt> and
<tt>typeof(Integer_Type)</tt> returns <tt>DataType_Type</tt>. One can use this
function as in the following example:
<tscreen><verb>
if (Integer_Type == typeof (x)) message ("x is an integer");
</verb></tscreen>
The literals of <tt>DataType_Type</tt> have other uses as well. One
of the most common uses of these literals is to create arrays, e.g.,
<tscreen><verb>
x = Complex_Type [100];
</verb></tscreen>
creates an array of <tt>100</tt> complex numbers and assigns it to
<tt>x</tt>.
<sect>Typecasting: Converting from one Type to Another<p>
Occasionally, it is necessary to convert from one data type to
another. For example, if you need to print an object as a string,
it may be necessary to convert it to a <tt>String_Type</tt>. The
<tt>typecast</tt> function may be used to perform such conversions.
For example, consider
<tscreen><verb>
variable x = 10, y;
y = typecast (x, Double_Type);
</verb></tscreen>
After execution of these statements, <tt>x</tt> will have the integer
value <tt>10</tt> and <tt>y</tt> will have the double precision floating
point value <tt>10.0</tt>. If the object to be converted is an
array, the <tt>typecast</tt> function will act upon all elements of
the array. For example,
<tscreen><verb>
variable x = [1:10]; % Array of integers
variable y = typecast (x, Double_Type);
</verb></tscreen>
will create an array of <tt>10</tt> double precision values and
assign it to <tt>y</tt>. One should also realize that it is not
always possible to perform a typecast. For example, any attempt to
convert an <tt>Integer_Type</tt> to a <tt>Null_Type</tt> will result in a
run-time error.
Often the interpreter will perform implicit type conversions as necessary
to complete calculations. For example, when multiplying an
<tt>Integer_Type</tt> with a <tt>Double_Type</tt>, it will convert the
<tt>Integer_Type</tt> to a <tt>Double_Type</tt> for the purpose of the
calculation. Thus, the example involving the conversion of an
array of integers to an array of doubles could have been performed
by multiplication by <tt>1.0</tt>, i.e.,
<tscreen><verb>
variable x = [1:10]; % Array of integers
variable y = 1.0 * x;
</verb></tscreen>
The <tt>string</tt> intrinsic function is similar to the typecast
function except that it converts an object to a string
representation. It is important to understand that a typecast from
some type to <tt>String_Type</tt> is <em>not</em> the same as converting
an object to its string operation. That is,
<tt>typecast(x,String_Type)</tt> is not equivalent to
<tt>string(x)</tt>. The reason for this is that when given an array,
the <tt>typecast</tt> function acts on each element of the array to
produce another array, whereas the <tt>string</tt> function produces a
a string.
The <tt>string</tt> function is useful for printing the value of an
object. This use is illustrated in the following simple example:
<tscreen><verb>
define print_object (x)
{
message (string (x));
}
</verb></tscreen>
Here, the <tt>message</tt> function has been used because it writes a
string to the display. If the <tt>string</tt> function was not used
and the <tt>message</tt> function was passed an integer, a
type-mismatch error would have resulted.
<chapt>Identifiers<p>
The names given to variables, functions, and data types are called
<em>identifiers</em>. There are some restrictions upon the actual
characters that make up an identifier. An identifier name must
start with a letter (<tt>[A-Za-z]</tt>), an underscore character, or a
dollar sign. The rest of the characters in the name can be any
combination of letters, digits, dollar signs, or underscore
characters. However, all identifiers whose name begins with two
underscore characters are reserved for internal use by the
interpreter and declarations of objects with such names should be
avoided.
Examples of valid identifiers include:
<tscreen><verb>
mary _3 _this_is_ok
a7e1 $44 _44$_Three
</verb></tscreen>
However, the following are not legal:
<tscreen><verb>
7abc 2e0 #xx
</verb></tscreen>
In fact, <tt>2e0</tt> actually specifies the real number
<tt>2.0</tt>.
Although the maximum length of identifiers is unspecified by the
language, the length should be kept below <tt>64</tt> characters.
The following identifiers are reserved by the language for use as
keywords:
<tscreen><verb>
!if _for do mod sign xor
ERROR_BLOCK abs do_while mul2 sqr public
EXIT_BLOCK and else not static private
USER_BLOCK0 andelse exch or struct
USER_BLOCK1 break for orelse switch
USER_BLOCK2 case foreach pop typedef
USER_BLOCK3 chs forever return using
USER_BLOCK4 continue if shl variable
__tmp define loop shr while
</verb></tscreen>
<chapt>Variables<p>
A variable must be declared before it can be used, otherwise an
undefined name error will be generated. A variable is declared
using the <tt>variable</tt> keyword, e.g,
<tscreen><verb>
variable x, y, z;
</verb></tscreen>
declares three variables, <tt>x</tt>, <tt>y</tt>, and <tt>z</tt>. This
is an example of a variable declaration statement, and like all
statements, it must end in a semi-colon.
Variables declared this way are untyped and inherit a type upon
assignment. The actual type checking is performed at run-time. For
example,
<tscreen><verb>
x = "This is a string";
x = 1.2;
x = 3;
x = 2i;
</verb></tscreen>
results in x being set successively to a string, a float, an
integer, and to a complex number (<tt>0+2i</tt>). Any attempt to use
a variable before it has acquired a type will result in an
uninitialized variable error.
It is legal to put executable code in a variable declaration list.
That is,
<tscreen><verb>
variable x = 1, y = sin (x);
</verb></tscreen>
are legal variable declarations. This also provides a convenient way
of initializing a variable.
Variables are classified as either <em>global</em> or <em>local</em>. A
variable declared inside a function is said to be local and has no
meaning outside the function. A variable is said to be global if
it was declared outside a function. Global variables are further
classified as being <tt>public</tt>, <tt>static</tt>, or <tt>private</tt>,
according to the name space where they were defined.
See chapter ??? for more information about name spaces.
The following global variables are predefined by the language and
are mainly used as convenience variables:
<tscreen><verb>
$0 $1 $2 $3 $4 $5 $6 $7 $8 $9
</verb></tscreen>
An <em>intrinsic</em> variable is another type of global variable.
Such variables have a definite type which cannot be altered.
Variables of this type may also be defined to be read-only, or
constant variables. An example of an intrinsic variable is
<tt>PI</tt> which is a read-only double precision variable with a value
of approximately <tt>3.14159265358979323846</tt>.
<chapt>Operators<p>
<bf>S-Lang</bf> supports a variety of operators that are grouped into three
classes: assignment operators, binary operators, and unary operators.
An assignment operator is used to assign a value to a variable.
They will be discussed more fully in the context of the assignment
statement in section ???.
An unary operator acts only upon a single quantity while a binary
operation is an operation between two quantities. The boolean
operator <tt>not</tt> is an example of an unary operator. Examples of
binary operators include the usual arithmetic operators
<tt>+</tt>, <tt>-</tt>, <tt>*</tt>, and <tt>/</tt>. The operator given by
<tt>-</tt> can be either an unary operator (negation) or a binary operator
(subtraction); the actual operation is determined from the context
in which it is used.
Binary operators are used in algebraic forms, e.g., <tt>a + b</tt>.
Unary operators fall in one of two classes: postfix-unary or
prefix-unary. For example, in the expression <tt>-x</tt>, the minus
sign is a prefix-unary operator.
Not all data types have binary or unary operations defined. For
example, while <tt>String_Type</tt> objects support the <tt>+</tt>
operator, they do not admit the <tt>*</tt> operator.
<sect>Unary Operators<p>
The <bf>unary</bf> operators operate only upon a single operand. They
include: <tt>not</tt>, <tt>~</tt>, <tt>-</tt>, <tt>@</tt>, <tt>&</tt>, as well as the
increment and decrement operators <tt>++</tt> and <tt>--</tt>,
respectively.
The boolean operator <tt>not</tt> acts only upon integers and produces
<tt>0</tt> if its operand is non-zero, otherwise it produces <tt>1</tt>.
The bit-level not operator <tt>~</tt> performs a similar function,
except that it operates on the individual bits of its integer
operand.
The arithmetic negation operator <tt>-</tt> is the most well-known
unary operator. It simply reverses the sign of its operand.
The reference (<tt>&</tt>) and dereference (<tt>@</tt>) operators will be
discussed in greater detail in section ???. Similarly, the
increment (<tt>++</tt>) and decrement (<tt>--</tt>) operators will be
discussed in the context of the assignment operator.
<sect>Binary Operators<p>
The binary operators may be grouped according to several classes:
arithmetic operators, relational operators, boolean operators, and
bitwise operators.
All binary and unary operators may be overloaded. For example, the
arithmetic plus operator has been overloaded by the
<tt>String_Type</tt> data type to permit concatenation between strings.
<sect1>Arithmetic Operators<p>
The arithmetic operators include <tt>+</tt>, <tt>-</tt>, <tt>*</tt>, <tt>/</tt>,
which perform addition, subtraction, multiplication, and division,
respectively. In addition to these, <bf>S-Lang</bf> supports the <tt>mod</tt>
operator as well as the power operator <tt>^</tt>.
The data type of the result produced by the use of one of these
operators depends upon the data types of the binary participants.
If they are both integers, the result will be an integer. However,
if the operands are not of the same type, they will be converted to
a common type before the operation is performed. For example, if
one is a floating point value and the other is an integer, the
integer will be converted to a float. In general, the promotion
from one type to another is such that no information is lost, if
possible. As an example, consider the expression <tt>8/5</tt> which
indicates division of the integer <tt>8</tt> by the integer <tt>5</tt>.
The result will be the integer <tt>1</tt> and <em>not</em> the floating
point value <tt>1.6</tt>. However, <tt>8/5.0</tt> will produce
<tt>1.6</tt> because <tt>5.0</tt> is a floating point number.
<sect1>Relational Operators<p>
The relational operators are <tt>></tt>, <tt>>=</tt>, <tt><</tt>, <tt><=</tt>,
<tt>==</tt>, and <tt>!=</tt>. These perform the comparisons greater
than, greater than or equal, less than, less than or equal, equal,
and not equal, respectively. The result of one of these
comparisons is the integer <tt>1</tt> if the comparison is true, or
<tt>0</tt> if the comparison is false. For example, <tt>6 >= 5</tt>
returns <tt>1</tt>, but <tt>6 == 5</tt> produces
<tt>0</tt>.
<sect1>Boolean Operators<p>
There are only two boolean binary operators: <tt>or</tt> and
<tt>and</tt>. These operators are defined only for integers and
produce an integer result. The <tt>or</tt> operator returns <tt>1</tt>
if either of its operands are non-zero, otherwise it produces
<tt>0</tt>. The <tt>and</tt> operator produces <tt>1</tt> if and only if
both its operands are non-zero, otherwise it produces <tt>0</tt>.
Neither of these operators perform the so-called boolean
short-circuit evaluation. For example, consider the expression:
<tscreen><verb>
(x != 0) and (1/x > 10)
</verb></tscreen>
Here, if <tt>x</tt> were to have a value of zero, a division by zero error
would occur because even though <tt>x!=0</tt> evaluates to zero, the
<tt>and</tt> operator is not short-circuited and the <tt>1/x</tt> expression
would still be evaluated. Although these operators are not
short-circuited, <bf>S-Lang</bf> does have another mechanism of performing
short-circuit boolean evaluation via the <tt>orelse</tt> and
<tt>andelse</tt> expressions. See below for information about these
constructs.
<sect1>Bitwise Operators<p>
The bitwise binary operators are defined only with integer operands
and are used for bit-level operations. Operators that fall in this
class include <tt>&</tt>, <tt>|</tt>, <tt>shl</tt>, <tt>shr</tt>, and
<tt>xor</tt>. The <tt>&</tt> operator performs a boolean AND operation
between the corresponding bits of the operands. Similarly, the
<tt>|</tt> operator performs the boolean OR operation on the bits.
The bit-shifting operators <tt>shl</tt> and <tt>shr</tt> shift the bits
of the first operand by the number given by the second operand to
the left or right, respectively. Finally, the <tt>xor</tt> performs
an EXCLUSIVE-OR operation.
These operators are commonly used to manipulate variables whose
individual bits have distinct meanings. In particular, <tt>&</tt> is
usually used to test bits, <tt>|</tt> can be used to set bits, and
<tt>xor</tt> may be used to flip a bit.
As an example of using <tt>&</tt> to perform tests on bits, consider
the following: The <bf>jed</bf> text editor stores some of the information
about a buffer in a bitmapped integer variable. The value of this
variable may be retrieved using the <bf>jed</bf> intrinsic function
<tt>getbuf_info</tt>, which actually returns four quantities: the
buffer flags, the name of the buffer, directory name, and file
name. For the purposes of this section, only the buffer flags are
of interest and can be retrieved via a function such as
<tscreen><verb>
define get_buffer_flags ()
{
variable flags;
(,,,flags) = getbuf_info ();
return flags;
}
</verb></tscreen>
The buffer flags is a bitmapped quantity where the 0th bit
indicates whether or not the buffer has been modified, the first
bit indicates whether or not autosave has been enabled for the
buffer, and so on. Consider for the moment the task of determining
if the buffer has been modified. This can be
determined by looking at the zeroth bit, if it is <tt>0</tt> the
buffer has not been modified, otherwise it has. Thus we can create
the function,
<tscreen><verb>
define is_buffer_modified ()
{
variable flags = get_buffer_flags ();
return (flags & 1);
}
</verb></tscreen>
where the integer <tt>1</tt> has been used since it has all of its
bits set to <tt>0</tt>, except for the zeroth one, which is set to
<tt>1</tt>. (At this point, it should also be apparent that bits are
numbered from zero, thus an <tt>8</tt> bit integer consists of bits
<tt>0</tt> to <tt>7</tt>, where <tt>0</tt> is the least significant bit and
<tt>7</tt> is the most significant one.) Similarly, we can create another
function
<tscreen><verb>
define is_autosave_on ()
{
variable flags = get_buffer_flags ();
return (flags & 2);
}
</verb></tscreen>
to determine whether or not autosave has been turned on for the
buffer.
The <tt>shl</tt> operator may be used to form the integer with only
the <em>nth</em> bit set. For example, <tt>1 shl 6</tt> produces an
integer with all bits set to zero except the sixth bit, which is
set to one. The following example exploits this fact:
<tscreen><verb>
define test_nth_bit (flags, nth)
{
return flags & (1 shl nth);
}
</verb></tscreen>
<sect1>Namespace operator<p>
The operator <tt>-></tt> is used to in conjunction with the name of a
namespace to access an object within the namespace. For example,
if <tt>A</tt> is the name of a namespace containing the variable
<tt>v</tt>, then <tt>A->v</tt> refers to that variable.
<sect1>Operator Precedence<p>
<sect1>Binary Operators and Functions Returning Multiple Values<p>
Care must be exercised when using binary operators with an operand
the returns multiple values. In fact, the current implementation
of the <bf>S-Lang</bf> language will produce incorrect results if both
operands of a binary expression return multiple values. <em>At
most, only one of operands of a binary expression can return
multiple values, and that operand must be the first one, not the
second.</em> For example,
<tscreen><verb>
define read_line (fp)
{
variable line, status;
status = fgets (&line, fp);
if (status == -1)
return -1;
return (line, status);
}
</verb></tscreen>
defines a function, <tt>read_line</tt> that takes a single argument, a
handle to an open file, and returns one or two values, depending
upon the return value of <tt>fgets</tt>. Now consider
<tscreen><verb>
while (read_line (fp) > 0)
{
text = ();
% Do something with text
.
.
}
</verb></tscreen>
Here the relational binary operator <tt>></tt> forms a comparison
between one of the return values (the one at the top of the stack)
and <tt>0</tt>. In accordance with the above rule, since <tt>read_line</tt>
returns multiple values, it occurs as the left binary operand.
Putting it on the right as in
<tscreen><verb>
while (0 < read_line (fp)) % Incorrect
{
text = ();
% Do something with text
.
.
}
</verb></tscreen>
violates the rule and will result in the wrong answer.
<sect>Mixing Integer and Floating Point Arithmetic<p>
If a binary operation (<tt>+</tt>, <tt>-</tt>, <tt>*</tt> , <tt>/</tt>) is
performed on two integers, the result is an integer. If at least
one of the operands is a float, the other is converted to float and
the result is float. For example:
<tscreen><verb>
11 / 2 --> 5 (integer)
11 / 2.0 --> 5.5 (float)
11.0 / 2 --> 5.5 (float)
11.0 / 2.0 --> 5.5 (float)
</verb></tscreen>
Finally note that only integers may be used as array indices,
loop control variables, and bit operations. The conversion
functions, <tt>int</tt> and <tt>float</tt>, may be used convert between
floats and ints where appropriate, e.g.,
<tscreen><verb>
int (1.5) --> 1 (integer)
float(1.5) --> 1.5 (float)
float (1) --> 1.0 (float)
</verb></tscreen>
<sect>Short Circuit Boolean Evaluation<p>
The boolean operators <tt>or</tt> and <tt>and</tt> <em>are not short
circuited</em> as they are in some languages. <bf>S-Lang</bf> uses
<tt>orelse</tt> and <tt>andelse</tt> expressions for short circuit boolean
evaluation. However, these are not binary operators. Expressions
of the form:
<tscreen>
<em>expr-1</em> and <em>expr-2</em> and ... <em>expr-n</em>
</tscreen>
can be replaced by the short circuited version using <tt>andelse</tt>:
<tscreen>
andelse {<em>expr-1</em>} {<em>expr-2</em>} ... {<em>expr-n</em>}
</tscreen>
A similar syntax holds for the <tt>orelse</tt> operator. For example, consider
the statement:
<tscreen><verb>
if ((x != 0) and (1/x > 10)) do_something ();
</verb></tscreen>
Here, if <tt>x</tt> were to have a value of zero, a division by zero error
would occur because even though <tt>x!=0</tt> evaluates to zero, the
<tt>and</tt> operator is not short circuited and the <tt>1/x</tt> expression
would be evaluated causing division by zero. For this case, the
<tt>andelse</tt> expression could be used to avoid the problem:
<tscreen><verb>
if (andelse
{x != 0}
{1 / x > 10}) do_something ();
</verb></tscreen>
<chapt>Statements<p>
Loosely speaking, a <em>statement</em> is composed of <em>expressions</em>
that are grouped according to the syntax or grammar of the language
to express a complete computation. Statements are analogous to
sentences in a human language and expressions are like phrases.
All statements in the <bf>S-Lang</bf> language must end in a semi-colon.
A statement that occurs within a function is executed only during
execution of the function. However, statements that occur outside
the context of a function are evaluated immediately.
The language supports several different types of statements such as
assignment statements, conditional statements, and so forth. These
are described in detail in the following sections.
<sect>Variable Declaration Statements<p>
Variable declarations were already discussed in chapter ???. For
the sake of completeness, a variable declaration is a statement of
the form
<tscreen>
variable <em>variable-declaration-list</em> ;
</tscreen>
where the <em>variable-declaration-list</em> is a comma separated list
of one or more variable names with optional initializations, e.g.,
<tscreen><verb>
variable x, y = 2, z;
</verb></tscreen>
<sect>Assignment Statements<p>
Perhaps the most well known form of statement is the <em>assignment
statement</em>. Statements of this type consist of a left-hand side,
an assignment operator, and a right-hand side. The left-hand side
must be something to which an assignment can be performed. Such
an object is called an <em>lvalue</em>.
The most common assignment operator is the simple assignment
operator <tt>=</tt>. Simple of its use include
<tscreen><verb>
x = 3;
x = some_function (10);
x = 34 + 27/y + some_function (z);
x = x + 3;
</verb></tscreen>
In addition to the simple assignment operator, <bf>S-Lang</bf>
also supports the assignment operators <tt>+=</tt> and <tt>-=</tt>.
Internally, <bf>S-Lang</bf> transforms
<tscreen><verb>
a += b;
</verb></tscreen>
to
<tscreen><verb>
a = a + b;
</verb></tscreen>
Similarly, <tt>a -= b</tt> is transformed to <tt>a = a - b</tt>. It is
extremely important to realize that, in general, <tt>a+b</tt> is not
equal to <tt>b+a</tt>. This means that <tt>a+=b</tt> is not the same
as <tt>a=b+a</tt>. As an example consider
<tscreen><verb>
a = "hello"; a += "world";
</verb></tscreen>
After execution of these two statements, <tt>a</tt> will have the
value <tt>"helloworld"</tt> and not <tt>"worldhello"</tt>.
Since adding or subtracting <tt>1</tt> from a variable is quite
common, <bf>S-Lang</bf> also supports the unary increment and decrement
operators <tt>++</tt>, and <tt>--</tt>, respectively. That is, for
numeric data types,
<tscreen><verb>
x = x + 1;
x += 1;
x++;
</verb></tscreen>
are all equivalent. Similarly,
<tscreen><verb>
x = x - 1;
x -= 1;
x--;
</verb></tscreen>
are also equivalent.
Strictly speaking, <tt>++</tt> and <tt>--</tt> are unary operators. When
used as <tt>x++</tt>, the <tt>++</tt> operator is said to be a
<em>postfix-unary</em> operator. However, when used as <tt>++x</tt> it is
said to be a <em>prefix-unary</em> operator. The current
implementation does not distinguish between the two forms, thus
<tt>x++</tt> and <tt>++x</tt> are equivalent. The reason for this
equivalence is <em>that assignment expressions do not return a value in
the <bf>S-Lang</bf> language</em> as they do in C. Thus one should exercise care
and not try to write C-like code such as
<tscreen><verb>
x = 10;
while (--x) do_something (x); % Ok in C, but not in S-Lang
</verb></tscreen>
The closest valid <bf>S-Lang</bf> form involves a <em>comma-expression</em>:
<tscreen><verb>
x = 10;
while (x--, x) do_something (x); % Ok in S-Lang and in C
</verb></tscreen>
<bf>S-Lang</bf> also supports a <em>multiple-assignment</em> statement. It is
discussed in detail in section ???.
<sect>Conditional and Looping Statements<p>
<bf>S-Lang</bf> supports a wide variety of conditional and looping
statements. These constructs operate on statements grouped together
in <em>blocks</em>. A block is a sequence of <bf>S-Lang</bf> statements enclosed
in braces and may contain other blocks. However, a block cannot
include function declarations. In the following,
<em>statement-or-block</em> refers to either a single
<bf>S-Lang</bf> statement or to a block of statements, and
<em>integer-expression</em> is an integer-valued expression.
<em>next-statement</em> represents the statement following the form
under discussion.
<sect1>Conditional Forms<p>
<sect2>if<p>
The simplest condition statement is the <tt>if</tt> statement. It
follows the syntax
<tscreen>
if (<em>integer-expression</em>) <em>statement-or-block</em>
<em>next-statement</em>
</tscreen>
If <em>integer-expression</em> evaluates to a non-zero result, then the
statement or group of statements implied <em>statement-or-block</em>
will get executed. Otherwise, control will proceed to
<em>next-statement</em>.
An example of the use of this type of conditional statement is
<tscreen><verb>
if (x != 0)
{
y = 1.0 / x;
if (x > 0) z = log (x);
}
</verb></tscreen>
This example illustrates two <tt>if</tt> statements where the second
<tt>if</tt> statement is part of the block of statements that belong to
the first.
<sect2>if-else<p>
Another form of <tt>if</tt> statement is the <em>if-else</em> statement.
It follows the syntax:
<tscreen>
if (<em>integer-expression</em>) <em>statement-or-block-1</em>
else <em>statement-or-block-2</em>
<em>next-statement</em>
</tscreen>
Here, if <em>expression</em> returns non-zero,
<em>statement-or-block-1</em> will get executed and control will pass
on to <em>next-statement</em>. However, if <em>expression</em> returns zero,
<em>statement-or-block-2</em> will get executed before continuing with
<em>next-statement</em>. A simple example of this form is
<tscreen><verb>
if (x > 0) z = log (x); else error ("x must be positive");
</verb></tscreen>
Consider the more complex example:
<tscreen><verb>
if (city == "Boston")
if (street == "Beacon") found = 1;
else if (city == "Madrid")
if (street == "Calle Mayor") found = 1;
else found = 0;
</verb></tscreen>
This example illustrates a problem that beginners have with
<em>if-else</em> statements. The grammar presented above shows that
the this example is equivalent to
<tscreen><verb>
if (city == "Boston")
{
if (street == "Beacon") found = 1;
else if (city == "Madrid")
{
if (street == "Calle Mayor") found = 1;
else found = 0;
}
}
</verb></tscreen>
It is important to understand the grammar and not be seduced by the
indentation!
<sect2>!if<p>
One often encounters <tt>if</tt> statements similar to
<tscreen>
if (<em>integer-expression</em> == 0) <em>statement-or-block</em>
</tscreen>
or equivalently,
<tscreen>
if (not(<em>integer-expression</em>)) <em>statement-or-block</em>
</tscreen>
The <tt>!if</tt> statement was added to the language to simplify the
handling of such statements. It obeys the syntax
<tscreen>
!if (<em>integer-expression</em>) <em>statement-or-block</em>
</tscreen>
and is functionally equivalent to
<tscreen>
if (not (<em>expression</em>)) <em>statement-or-block</em>
</tscreen>
<sect2>orelse, andelse<p>
These constructs were discussed earlier. The syntax for the
<tt>orelse</tt> statement is:
<tscreen>
orelse {<em>integer-expression-1</em>} ... {<em>integer-expression-n</em>}
</tscreen>
This causes each of the blocks to be executed in turn until one of
them returns a non-zero integer value. The result of this statement
is the integer value returned by the last block executed. For
example,
<tscreen><verb>
orelse { 0 } { 6 } { 2 } { 3 }
</verb></tscreen>
returns <tt>6</tt> since the second block is the first to return a
non-zero result. The last two block will not get executed.
The syntax for the <tt>andelse</tt> statement is:
<tscreen>
andelse {<em>integer-expression-1</em>} ... {<em>integer-expression-n</em>}
</tscreen>
Each of the blocks will be executed in turn until one of
them returns a zero value. The result of this statement is the
integer value returned by the last block executed. For example,
<tscreen><verb>
andelse { 6 } { 2 } { 0 } { 4 }
</verb></tscreen>
returns <tt>0</tt> since the third block will be the last to execute.
<sect2>switch<p>
The switch statement deviates the most from its C counterpart. The
syntax is:
<tscreen><verb>
switch (x)
{ ... : ...}
.
.
{ ... : ...}
</verb></tscreen>
The `<tt>:</tt>' operator is a special symbol which means to test
the top item on the stack, and if it is non-zero, the rest of the block
will get executed and control will pass out of the switch statement.
Otherwise, the execution of the block will be terminated and the process
will be repeated for the next block. If a block contains no
<tt>:</tt> operator, the entire block is executed and control will
pass onto the next statement following the <tt>switch</tt> statement.
Such a block is known as the <em>default</em> case.
As a simple example, consider the following:
<tscreen><verb>
switch (x)
{ x == 1 : print("Number is one.");}
{ x == 2 : print("Number is two.");}
{ x == 3 : print("Number is three.");}
{ x == 4 : print("Number is four.");}
{ x == 5 : print("Number is five.");}
{ print ("Number is greater than five.");}
</verb></tscreen>
Suppose <tt>x</tt> has an integer value of <tt>3</tt>. The first two
blocks will terminate at the `<tt>:</tt>' character because each of the
comparisons with <tt>x</tt> will produce zero. However, the third
block will execute to completion. Similarly, if <tt>x</tt> is
<tt>7</tt>, only the last block will execute in full.
A more familiar way to write the previous example used the
<tt>case</tt> keyword:
<tscreen><verb>
switch (x)
{ case 1 : print("Number is one.");}
{ case 2 : print("Number is two.");}
{ case 3 : print("Number is three.");}
{ case 4 : print("Number is four.");}
{ case 5 : print("Number is five.");}
{ print ("Number is greater than five.");}
</verb></tscreen>
The <tt>case</tt> keyword is a more useful comparison operator because
it can perform a comparison between different data types while
using <tt>==</tt> may result in a type-mismatch error. For example,
<tscreen><verb>
switch (x)
{ (x == 1) or (x == "one") : print("Number is one.");}
{ (x == 2) or (x == "two") : print("Number is two.");}
{ (x == 3) or (x == "three") : print("Number is three.");}
{ (x == 4) or (x == "four") : print("Number is four.");}
{ (x == 5) or (x == "five") : print("Number is five.");}
{ print ("Number is greater than five.");}
</verb></tscreen>
will fail because the <tt>==</tt> operation is not defined between
strings and integers. The correct way to write this to use the
<tt>case</tt> keyword:
<tscreen><verb>
switch (x)
{ case 1 or case "one" : print("Number is one.");}
{ case 2 or case "two" : print("Number is two.");}
{ case 3 or case "three" : print("Number is three.");}
{ case 4 or case "four" : print("Number is four.");}
{ case 5 or case "five" : print("Number is five.");}
{ print ("Number is greater than five.");}
</verb></tscreen>
<sect1>Looping Forms<p>
<sect2>while<p>
The <tt>while</tt> statement follows the syntax
<tscreen>
while (<em>integer-expression</em>) <em>statement-or-block</em>
<em>next-statement</em>
</tscreen>
It simply causes <em>statement-or-block</em> to get executed as long as
<em>integer-expression</em> evaluates to a non-zero result. For
example,
<tscreen><verb>
i = 10;
while (i)
{
i--;
newline ();
}
</verb></tscreen>
will cause the <tt>newline</tt> function to get called 10 times.
However,
<tscreen><verb>
i = -10;
while (i)
{
i--;
newline ();
}
</verb></tscreen>
would loop forever (or until <tt>i</tt> wraps from the most negative
integer value to the most positive and then decrements to zero).
If you are a C programmer, do not let the syntax of the language
seduce you into writing this example as you would in C:
<tscreen><verb>
i = 10;
while (i--) newline ();
</verb></tscreen>
The fact is that expressions such as <tt>i--</tt> do not return a
value in <bf>S-Lang</bf> as they do in C. If you must write this way, use
the comma operator as in
<tscreen><verb>
i = 10;
while (i, i--) newline ();
</verb></tscreen>
<sect2>do...while<p>
The <tt>do...while</tt> statement follows the syntax
<tscreen>
do
<em>statement-or-block</em>
while (<em>integer-expression</em>);
</tscreen>
The main difference between this statement and the <tt>while</tt>
statement is that the <tt>do...while</tt> form performs the test
involving <em>integer-expression</em> after each execution
of <em>statement-or-block</em> rather than before. This guarantees that
<em>statement-or-block</em> will get executed at least once.
A simple example from the <bf>jed</bf> editor follows:
<tscreen><verb>
bob (); % Move to beginning of buffer
do
{
indent_line ();
}
while (down (1));
</verb></tscreen>
This will cause all lines in the buffer to get indented via the
<bf>jed</bf> intrinsic function <tt>indent_line</tt>.
<sect2>for<p>
Perhaps the most complex looping statement is the <tt>for</tt>
statement; nevertheless, it is a favorite of many programmers.
This statement obeys the syntax
<tscreen>
for (<em>init-expression</em>; <em>integer-expression</em>; <em>end-expression</em>)
<em>statement-or-block</em>
<em>next-statement</em>
</tscreen>
In addition to <em>statement-or-block</em>, its specification requires
three other expressions. When executed, the <tt>for</tt> statement
evaluates <em>init-expression</em>, then it tests
<em>integer-expression</em>. If <em>integer-expression</em> returns zero,
control passes to <em>next-statement</em>. Otherwise, it executes
<em>statement-or-block</em> as long as <em>integer-expression</em>
evaluates to a non-zero result. After every execution of
<em>statement-or-block</em>, <em>end-expression</em> will get evaluated.
This statement is <em>almost</em> equivalent to
<tscreen>
<em>init-expression</em>;
while (<em>integer-expression</em>)
{
<em>statement-or-block</em>
<em>end-expression</em>;
}
</tscreen>
The reason that they are not fully equivalent involves what happens
when <em>statement-or-block</em> contains a <tt>continue</tt> statement.
Despite the apparent complexity of the <tt>for</tt> statement, it is
very easy to use. As an example, consider
<tscreen><verb>
sum = 0;
for (i = 1; i <= 10; i++) sum += i;
</verb></tscreen>
which computes the sum of the first 10 integers.
<sect2>loop<p>
The <tt>loop</tt> statement simply executes a block of code a fixed
number of times. It follows the syntax
<tscreen>
loop (<em>integer-expression</em>) <em>statement-or-block</em>
<em>next-statement</em>
</tscreen>
If the <em>integer-expression</em> evaluates to a positive integer,
<em>statement-or-block</em> will get executed that many times.
Otherwise, control will pass to <em>next-statement</em>.
For example,
<tscreen><verb>
loop (10) newline ();
</verb></tscreen>
will cause the function <tt>newline</tt> to get called 10 times.
<sect2>forever<p>
The <tt>forever</tt> statement is similar to the <tt>loop</tt> statement
except that it loops forever, or until a <tt>break</tt> or a
<tt>return</tt> statement is executed. It obeys the syntax
<tscreen>
forever <em>statement-or-block</em>
</tscreen>
A trivial example of this statement is
<tscreen><verb>
n = 10;
forever
{
if (n == 0) break;
newline ();
n--;
}
</verb></tscreen>
<sect2>foreach<p>
The <tt>foreach</tt> statement is used to loop over one or more
statements for every element in a container object. A container
object is a data type that consists of other types. Examples
include both ordinary and associative arrays, structures, and
strings. Every time through the loop the current member of the
object is pushed onto the stack.
The simple type of <tt>foreach</tt> statement obeys the syntax
<tscreen>
foreach (<em>container-object</em>) <em>statement-or-block</em>
</tscreen>
Here <em>container-object</em> can be an expression that returns a
container object. A simple example is
<tscreen><verb>
foreach (["apple", "peach", "pear"])
{
fruit = ();
process_fruit (fruit);
}
</verb></tscreen>
This example shows that if the container object is an array, then
successive elements of the array are pushed onto the stack prior to
each execution cycle. If the container object is a string, then
successive characters of the string are pushed onto the stack.
What actually gets pushed onto the stack may be controlled via the
<tt>using</tt> form of the <tt>foreach</tt> statement. This more complex
type of <tt>foreach</tt> statement follows the syntax
<tscreen>
foreach ( <em>container-object</em> ) using ( <em>control-list</em> )
<em>statement-or-block</em>
</tscreen>
The allowed values of <em>control-list</em> will depend upon the type
of container object. For associative arrays (<tt>Assoc_Type</tt>),
<em>control-list</em> specified whether <em>keys</em>, <em>values</em>, or both
are pushed onto the stack. For example,
<tscreen><verb>
foreach (a) using ("keys")
{
k = ();
.
.
}
</verb></tscreen>
results in the keys of the associative array <tt>a</tt> being pushed
on the list. However,
<tscreen><verb>
foreach (a) using ("values")
{
v = ();
.
.
}
</verb></tscreen>
will cause the values to be used, and
<tscreen><verb>
foreach (a) using ("keys", "values")
{
(k,v) = ();
.
.
}
</verb></tscreen>
will use both the keys and values of the array.
Similarly, for linked-lists of structures, one may walk the list via
code like
<tscreen><verb>
foreach (linked_list) using ("next")
{
s = ();
.
.
}
</verb></tscreen>
This <tt>foreach</tt> statement is equivalent
<tscreen><verb>
s = linked_list;
while (s != NULL)
{
.
.
s = s.next;
}
</verb></tscreen>
Consult the type-specific documentation for a discussion of the
<tt>using</tt> control words, if any, appropriate for a given type.
<sect>break, return, continue<p>
<bf>S-Lang</bf> also includes the non-local transfer functions <tt>return</tt>, <tt>break</tt>,
and <tt>continue</tt>. The <tt>return</tt> statement causes control to return to the
calling function while the <tt>break</tt> and <tt>continue</tt> statements are used in
the context of loop structures. Consider:
<tscreen><verb>
define fun ()
{
forever
{
s1;
s2;
..
if (condition_1) break;
if (condition_2) return;
if (condition_3) continue;
..
s3;
}
s4;
..
}
</verb></tscreen>
Here, a function <tt>fun</tt> has been defined that contains a <tt>forever</tt>
loop consisting of statements <tt>s1</tt>, <tt>s2</tt>,...,<tt>s3</tt>, and
three <tt>if</tt> statements. As long as the expressions <tt>condition_1</tt>,
<tt>condition_2</tt>, and <tt>condition_3</tt> evaluate to zero, the statements
<tt>s1</tt>, <tt>s2</tt>,...,<tt>s3</tt> will be repeatedly executed. However,
if <tt>condition_1</tt> returns a non-zero value, the <tt>break</tt> statement
will get executed, and control will pass out of the <tt>forever</tt> loop to
the statement immediately following the loop which in this case is
<tt>s4</tt>. Similarly, if <tt>condition_2</tt> returns a non-zero number,
the <tt>return</tt> statement will cause control to pass back to the
caller of <tt>fun</tt>. Finally, the <tt>continue</tt> statement will
cause control to pass back to the start of the loop, skipping the
statement <tt>s3</tt> altogether.
<chapt>Functions<p>
A function may be thought of as a group of statements that work
together to perform a computation. While there are no imposed
limits upon the number statements that may occur within a function,
it is considered poor programming practice if a function contains
many statements. This notion stems from the belief that a function
should have a simple, well defined purpose.
<sect>Declaring Functions<p>
Like variables, functions must be declared before they can be used. The
<tt>define</tt> keyword is used for this purpose. For example,
<tscreen><verb>
define factorial ();
</verb></tscreen>
is sufficient to declare a function named <tt>factorial</tt>. Unlike
the <tt>variable</tt> keyword used for declaring variables, the
<tt>define</tt> keyword does not accept a list of names.
Usually, the above form is used only for recursive functions. In
most cases, the function name is almost always followed by a
parameter list and the body of the function:
<tscreen>
define <em>function-name</em> (<em>parameter-list</em>)
{
<em>statement-list</em>
}
</tscreen>
The <em>function-name</em> is an identifier and must conform to the
naming scheme for identifiers discussed in chapter ???.
The <em>parameter-list</em> is a comma-separated list of variable names
that represent parameters passed to the function, and
may be empty if no parameters are to be passed.
The body of the function is enclosed in braces and consists of zero
or more statements (<em>statement-list</em>).
The variables in the <em>parameter-list</em> are implicitly declared,
thus, there is no need to declare them via a variable declaration
statement. In fact any attempt to do so will result in a syntax
error.
<sect>Parameter Passing Mechanism<p>
Parameters to a function are always passed by value and never by
reference. To see what this means, consider
<tscreen><verb>
define add_10 (a)
{
a = a + 10;
}
variable b = 0;
add_10 (b);
</verb></tscreen>
Here a function <tt>add_10</tt> has been defined, which when executed,
adds <tt>10</tt> to its parameter. A variable <tt>b</tt> has also been
declared and initialized to zero before it is passed to
<tt>add_10</tt>. What will be the value of <tt>b</tt> after the call to
<tt>add_10</tt>? If <bf>S-Lang</bf> were a language that passed parameters by
reference, the value of <tt>b</tt> would be changed to
<tt>10</tt>. However, <bf>S-Lang</bf> always passes by value, which means that
<tt>b</tt> would retain its value of zero after the function call.
<bf>S-Lang</bf> does provide a mechanism for simulating pass by reference
via the reference operator. See the next section for more details.
If a function is called with a parameter in the parameter list
omitted, the corresponding variable in the function will be set to
<tt>NULL</tt>. To make this clear, consider the function
<tscreen><verb>
define add_two_numbers (a, b)
{
if (a == NULL) a = 0;
if (b == NULL) b = 0;
return a + b;
}
</verb></tscreen>
This function must be called with two parameters. However, we can
omit one or both of the parameters by calling it in one of the
following ways:
<tscreen><verb>
variable s = add_two_numbers (2,3);
variable s = add_two_numbers (2,);
variable s = add_two_numbers (,3);
variable s = add_two_numbers (,);
</verb></tscreen>
The first example calls the function using both parameters;
however, at least one of the parameters was omitted in the other
examples. The interpreter will implicitly convert the last three
examples to
<tscreen><verb>
variable s = add_two_numbers (2, NULL);
variable s = add_two_numbers (NULL, 3);
variable s = add_two_numbers (NULL, NULL);
</verb></tscreen>
It is important to note that this mechanism is available only for
function calls that specify more than one parameter. That is,
<tscreen><verb>
variable s = add_10 ();
</verb></tscreen>
is <em>not</em> equivalent to <tt>add_10(NULL)</tt>. The reason for this
is simple: the parser can only tell whether or not <tt>NULL</tt> should
be substituted by looking at the position of the comma character in
the parameter list, and only function calls that indicate more than
one parameter will use a comma. A mechanism for handling single
parameter function calls is described in the next section.
<sect>Referencing Variables<p>
One can achieve the effect of passing by reference by using the
reference (<tt>&</tt>) and dereference (<tt>@</tt>) operators. Consider
again the <tt>add_10</tt> function presented in the previous section.
This time we write it as
<tscreen><verb>
define add_10 (a)
{
@a = @a + 10;
}
variable b = 0;
add_10 (&b);
</verb></tscreen>
The expression <tt>&b</tt> creates a <em>reference</em> to the variable
<tt>b</tt> and it is the reference that gets passed to <tt>add_10</tt>.
When the function <tt>add_10</tt> is called, the value of <tt>a</tt> will
be a reference to <tt>b</tt>. It is only by <em>dereferencing</em> this
value that <tt>b</tt> can be accessed and changed. So, the statement
<tt>@a=@a+10;</tt> should be read `add <tt>10</tt>' to the value of the
object that <tt>a</tt> references and assign the result to the object
that <tt>a</tt> references.
The reader familiar with C will note the similarity between
<em>references</em> in <bf>S-Lang</bf> and <em>pointers</em> in C.
One of the main purposes for references is that this mechanism
allows reference to functions to be passed to other functions. As
a simple example from elementary calculus, consider the following
function which returns an approximation to the derivative of another
function at a specified point:
<tscreen><verb>
define derivative (f, x)
{
variable h = 1e-6;
return (@f(x+h) - @f(x)) / h;
}
</verb></tscreen>
It can be used to differentiate the function
<tscreen><verb>
define x_squared (x)
{
return x^2;
}
</verb></tscreen>
at the point <tt>x = 3</tt> via the expression
<tt>derivative(&x_squared,3)</tt>.
<sect>Functions with a Variable Number of Arguments<p>
<bf>S-Lang</bf> functions may be defined to take a variable number of
arguments. The reason for this is that the calling routine pushes
the arguments onto the stack before making a function call, and it
is up to the called function to pop the values off the stack and
make assignments to the variables in the parameter list. These
details are, for the most part, hidden from the programmer.
However, they are important when a variable number of arguments are
passed.
Consider the <tt>add_10</tt> example presented earlier. This time it
is written
<tscreen><verb>
define add_10 ()
{
variable x;
x = ();
return x + 10;
}
variable s = add_10 (12); % ==> s = 22;
</verb></tscreen>
For the uninitiated, this example looks as if it
is destined for disaster. The <tt>add_10</tt> function looks like it
accepts zero arguments, yet it was called with a single argument.
On top of that, the assignment to <tt>x</tt> looks strange. The truth
is, the code presented in this example makes perfect sense, once you
realize what is happening.
First, consider what happened when <tt>add_10</tt> is called with the
the parameter <tt>12</tt>. Internally, <tt>12</tt> is
pushed onto the stack and then the function called. Now,
consider the function itself. <tt>x</tt> is a variable local to the
function. The strange looking assignment `<tt>x=()</tt>' simply
takes whatever is on the stack and assigns it to <tt>x</tt>. In
other words, after this statement, the value of <tt>x</tt> will be
<tt>12</tt>, since <tt>12</tt> will be at the top of the stack.
A generic function of the form
<tscreen><verb>
define function_name (x, y, ..., z)
{
.
.
}
</verb></tscreen>
is internally transformed by the interpreter to
<tscreen><verb>
define function_name ()
{
variable x, y, ..., z;
z = ();
.
.
y = ();
x = ();
.
.
}
</verb></tscreen>
before further parsing. (The <tt>add_10</tt> function, as defined above, is
already in this form.) With this knowledge in hand, one can write a
function that accepts a variable number of arguments. Consider the
function:
<tscreen><verb>
define average_n (n)
{
variable x, y;
variable sum;
if (n == 1)
{
x = ();
sum = x;
}
else if (n == 2)
{
y = ();
x = ();
sum = x + y;
}
else error ("average_n: only one or two values supported");
return sum / n;
}
variable ave1 = average_n (3.0, 1); % ==> 3.0
variable ave2 = average_n (3.0, 5.0, 2); % ==> 4.0
</verb></tscreen>
Here, the last argument passed to <tt>average_n</tt> is an integer
reflecting the number of quantities to be averaged. Although this
example works fine, its principal limitation is obvious: it only
supports one or two values. Extending it to three or more values
by adding more <tt>else if</tt> constructs is rather straightforward but
hardly worth the effort. There must be a better way, and there is:
<tscreen><verb>
define average_n (n)
{
variable sum, x;
sum = 0;
loop (n)
{
x = (); % get next value from stack
sum += x;
}
return sum / n;
}
</verb></tscreen>
The principal limitation of this approach is that one must still
pass an integer that specifies how many values are to be averaged.
Fortunately, a special variable exists that is local to every function
and contains the number of values that were passed to the function.
That variable has the name <tt>_NARGS</tt> and may be used as follows:
<tscreen><verb>
define average_n ()
{
variable x, sum = 0;
if (_NARGS == 0) error ("Usage: ave = average_n (x, ...);");
loop (_NARGS)
{
x = ();
sum += x;
}
return sum / _NARGS;
}
</verb></tscreen>
Here, if no arguments are passed to the function, a simple message
that indicates how it is to be used is printed out.
<sect>Returning Values<p>
As stated earlier, the usual way to return values from a function
is via the <tt>return</tt> statement. This statement has the
simple syntax
<tscreen>
return <em>expression-list</em> ;
</tscreen>
where <em>expression-list</em> is a comma separated list of expressions.
If the function does not return any values, the expression list
will be empty. As an example of a function that can return
multiple values, consider
<tscreen><verb>
define sum_and_diff (x, y)
{
variable sum, diff;
sum = x + y; diff = x - y;
return sum, diff;
}
</verb></tscreen>
which is a function returning two values.
It is extremely important to note that <em>the calling routine must
explicitly handle all values returned by a function</em>. Although
some languages such as C do not have this restriction, <bf>S-Lang</bf> does
and it is a direct result of a <bf>S-Lang</bf> function's ability to return
many values and accept a variable number of parameters. Examples
of properly handling the above function include
<tscreen><verb>
variable sum, diff;
(sum, diff) = sum_and_diff (5, 4); % ignore neither
(sum, ) = sum_and_diff (5, 4); % ignore diff
(,) = sum_and_diff (5, 4); % ignore both sum and diff
</verb></tscreen>
See the section below on assignment statements for more information
about this important point.
<sect>Multiple Assignment Statement<p>
<bf>S-Lang</bf> functions can return more than one value, e.g.,
<tscreen><verb>
define sum_and_diff (x, y)
{
return x + y, x - y;
}
</verb></tscreen>
returns two values. It accomplishes this by placing both values on
the stack before returning. If you understand how <bf>S-Lang</bf> functions
handle a variable number of parameters (section ???), then it
should be rather obvious that one assigns such values to variables.
One way is to use, e.g.,
<tscreen><verb>
sum_and_diff (9, 4);
d = ();
s = ();
</verb></tscreen>
However, the most convenient way to accomplish this is to use a
<em>multiple assignment statement</em> such as
<tscreen><verb>
(s, d) = sum_and_diff (9, 4);
</verb></tscreen>
The most general form of the multiple assignment statement is
<tscreen><verb>
( var_1, var_2, ..., var_n ) = expression;
</verb></tscreen>
In fact, internally the interpreter transforms this statement into
the form
<tscreen><verb>
expression; var_n = (); ... var_2 = (); var_1 = ();
</verb></tscreen>
for further processing.
If you do not care about one of return values, simply omit the
variable name from the list. For example,
<tscreen><verb>
(s, ) = sum_and_diff (9, 4);
</verb></tscreen>
assigns the sum of <tt>9</tt> and <tt>4</tt> to <tt>s</tt> and the
difference (<tt>9-4</tt>) will be removed from the stack.
As another example, the <bf>jed</bf> editor provides a function called
<tt>down</tt> that takes an integer argument and returns an integer.
It is used to move the current editing position down the number of
lines specified by the argument passed to it. It returns the number
of lines it successfully moved the editing position. Often one does
not care about the return value from this function. Although it is
always possible to handle the return value via
<tscreen><verb>
variable dummy = down (10);
</verb></tscreen>
it is more convenient to use a multiple assignment expression and
omit the variable name, e.g.,
<tscreen><verb>
() = down (10);
</verb></tscreen>
Some functions return a <em>variable number</em> of values instead of a
<em>fixed number</em>. Usually, the value at the top of the stack will
indicate the actual number of return values. For such functions,
the multiple assignment statement cannot directly be used. To see
how such functions can be dealt with, consider the following
function:
<tscreen><verb>
define read_line (fp)
{
variable line;
if (-1 == fgets (&line, fp))
return -1;
return (line, 0);
}
</verb></tscreen>
This function returns either one or two values, depending upon the
return value of <tt>fgets</tt>. Such a function may be handled as in
the following example:
<tscreen><verb>
status = read_line (fp);
if (status != -1)
{
s = ();
.
.
}
</verb></tscreen>
In this example, the <em>last</em> value returned by <tt>read_line</tt> is
assigned to <tt>status</tt> and then tested. If it is non-zero, the
second return value is assigned to <tt>s</tt>. In particular note the
empty set of parenthesis in the assignment to <tt>s</tt>. This simply
indicates that whatever is on the top of the stack when the
statement is executed will be assigned to <tt>s</tt>.
Before leaving this section it is important to reiterate the fact
that if a function returns a value, the caller must deal with that
return value. Otherwise, the value will continue to live onto the
stack and may eventually lead to a stack overflow error.
Failing to handle the return value of a function is the
most common mistake that inexperienced <bf>S-Lang</bf> programmers make.
For example, the <tt>fflush</tt> function returns a value that many C
programmer's never check. Instead of writing
<tscreen><verb>
fflush (fp);
</verb></tscreen>
as one could in C, a <bf>S-Lang</bf> programmer should write
<tscreen><verb>
() = fflush (fp);
</verb></tscreen>
in <bf>S-Lang</bf>. (Many good C programmer's write <tt>(void)fflush(fp)</tt>
to indicate that the return value is being ignored).
<sect>Exit-Blocks<p>
An <em>exit-block</em> is a set of statements that get executed when a
functions returns. They are very useful for cleaning up when a
function returns via an explicit call to <tt>return</tt> from deep
within a function.
An exit-block is created by using the <tt>EXIT_BLOCK</tt> keyword
according to the syntax
<tscreen>
EXIT_BLOCK { <em>statement-list</em> }
</tscreen>
where <em>statement-list</em> represents the list of statements that
comprise the exit-block. The following example illustrates the use
of an exit-block:
<tscreen><verb>
define simple_demo ()
{
variable n = 0;
EXIT_BLOCK { message ("Exit block called."); }
forever
{
if (n == 10) return;
n++;
}
}
</verb></tscreen>
Here, the function contains an exit-block and a <tt>forever</tt> loop.
The loop will terminate via the <tt>return</tt> statement when <tt>n</tt>
is 10. Before it returns, the exit-block will get executed.
A function can contain multiple exit-blocks, but only the last
one encountered during execution will actually get executed. For
example,
<tscreen><verb>
define simple_demo (n)
{
EXIT_BLOCK { return 1; }
if (n != 1)
{
EXIT_BLOCK { return 2; }
}
return;
}
</verb></tscreen>
If <tt>1</tt> is passed to this function, the first exit-block will
get executed because the second one would not have been encountered
during the execution. However, if some other value is passed, the
second exit-block would get executed. This example also
illustrates that it is possible to explicitly return from an
exit-block, although nested exit-blocks are illegal.
<chapt>Name Spaces<p>
By default, all global variables and functions are defined in the
global namespace. In addition to the global namespace, every
compilation unit (e.g., a file containing <bf>S-Lang</bf> code) has an
anonymous namespace. Objects may be defined in the anonymous
namespace via the <tt>static</tt> declaration keyword. For example,
<tscreen><verb>
static variable x;
static define hello () { message ("hello"); }
</verb></tscreen>
defines a variable <tt>x</tt> and a function <tt>hello</tt> in the
anonymous namespace. This is useful when one wants to define
functions and variables that are only to be used within the file, or
more precisely the compilation unit, that defines them.
The <tt>implements</tt> function may be used to give the anonymous
namespace a name to allow access to its objects from outside the
compilation unit that defines them. For example,
<tscreen><verb>
implements ("foo");
static variable x;
</verb></tscreen>
allows the variable <tt>x</tt> to be accessed via <tt>foo->x</tt>, e.g.,
<tscreen><verb>
if (foo->x == 1) foo->x = 2;
</verb></tscreen>
The <tt>implements</tt> function does more than simply giving the
anonymous namespace a name. It also changes the default variable
and function declaration mode from <tt>public</tt> to <tt>static</tt>.
That is,
<tscreen><verb>
implements ("foo");
variable x;
</verb></tscreen>
and
<tscreen><verb>
implements ("foo");
static variable x;
</verb></tscreen>
are equivalent. Then to create a public object within the
namespace, one must explicitly use the <tt>public</tt> keyword.
Finally, the <tt>private</tt> keyword may be used to create an object
that is truly private within the compilation unit. For example,
<tscreen><verb>
implements ("foo");
variable x;
private variable y;
</verb></tscreen>
allows <tt>x</tt> to be accessed from outside the namespace via
<tt>foo->x</tt>, however <tt>y</tt> cannot be accessed.
<chapt>Arrays<p>
An array is a container object that can contain many values of one
data type. Arrays are very useful objects and are indispensable
for certain types of programming. The purpose of this chapter is
to describe how arrays are defined and used in the <bf>S-Lang</bf> language.
<sect>Creating Arrays<p>
The <bf>S-Lang</bf> language supports multi-dimensional arrays of all data
types. Since the <tt>Array_Type</tt> is a data type, one can even
have arrays of arrays. To create a multi-dimensional array of
<em>SomeType</em> use the syntax
<tscreen><verb>
SomeType [dim0, dim1, ..., dimN]
</verb></tscreen>
Here <em>dim0</em>, <em>dim1</em>, ... <em>dimN</em> specify the size of
the individual dimensions of the array. The current implementation
permits arrays consist of up to <tt>7</tt> dimensions. When a
numeric array is created, all its elements are initialized to zero.
The initialization of other array types depend upon the data type,
e.g., <tt>String_Type</tt> and <tt>Struct_Type</tt> arrays are
initialized to <tt>NULL</tt>.
As a concrete example, consider
<tscreen><verb>
a = Integer_Type [10];
</verb></tscreen>
which creates a one-dimensional array of <tt>10</tt> integers and
assigns it to <tt>a</tt>.
Similarly,
<tscreen><verb>
b = Double_Type [10, 3];
</verb></tscreen>
creates a <tt>30</tt> element array of double precision numbers
arranged in <tt>10</tt> rows and <tt>3</tt> columns, and assigns it to
<tt>b</tt>.
<sect1>Range Arrays<p>
There is a more convenient syntax for creating and initializing a
1-d arrays. For example, to create an array of ten
integers whose elements run from <tt>1</tt> through <tt>10</tt>, one
may simply use:
<tscreen><verb>
a = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10];
</verb></tscreen>
Similarly,
<tscreen><verb>
b = [1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0];
</verb></tscreen>
specifies an array of ten doubles.
An even more compact way of specifying a numeric array is to use a
<em>range-array</em>. For example,
<tscreen><verb>
a = [0:9];
</verb></tscreen>
specifies an array of 10 integers whose elements range from <tt>0</tt>
through <tt>9</tt>. The most general form of a range array is
<tscreen><verb>
[first-value : last-value : increment]
</verb></tscreen>
where the <em>increment</em> is optional and defaults to <tt>1</tt>. This
creates an array whose first element is <em>first-value</em> and whose
successive values differ by <em>increment</em>. <em>last-value</em> sets
an upper limit upon the last value of the array as described below.
If the range array <tt>[a:b:c]</tt> is integer valued, then the
interval specified by <tt>a</tt> and <tt>b</tt> is closed. That is, the
kth element of the array x_k is given by x_k=a+ck and
must satisfy a<=x_k<=b. Hence, the number of elements in an
integer range array is given by the expression 1 + (b-a)/c.
The situation is somewhat more complicated for floating point range
arrays. The interval specified by a floating point range array
<tt>[a:b:c]</tt> is semi-open such that <tt>b</tt> is not contained in
the interval. In particular, the kth element of <tt>[a:b:c]</tt> is
given by x_k=a+kc such that a<=x_k<b when
c>=0, and b<x_k<=a otherwise. The number of elements
in the array is one greater than the largest k that
satisfies the open interval constraint.
Here are a few examples that illustrate the above comments:
<tscreen><verb>
[1:5:1] ==> [1,2,3,4,5]
[1.0:5.0:1.0] ==> [1.0, 2.0, 3.0, 4.0]
[5:1:-1] ==> [5,4,3,2,1]
[5.0:1.0:-1.0] ==> [5.0, 4.0, 3.0, 2.0];
[1:1] ==> [1]
[1.0:1.0] ==> []
[1:-3] ==> []
</verb></tscreen>
<sect1>Creating arrays via the dereference operator<p>
Another way to create an array is apply the dereference operator
<tt>@</tt> to the <tt>DataType_Type</tt> literal <tt>Array_Type</tt>. The
actual syntax for this operation resembles a function call
<tscreen>
variable a = @Array_Type (<em>data-type</em>, <em>integer-array</em>);
</tscreen>
where <em>data-type</em> is of type <tt>DataType_Type</tt> and
<em>integer-array</em> is a 1-d array of integers that specify the size
of each dimension. For example,
<tscreen><verb>
variable a = @Array_Type (Double_Type, [10, 20]);
</verb></tscreen>
will create a <tt>10</tt> by <tt>20</tt> array of doubles and assign it
to <tt>a</tt>. This method of creating arrays derives its power from
the fact that it is more flexible than the methods discussed in this
section. We shall encounter it again in section ??? in the context
of the <tt>array_info</tt> function.
<sect>Reshaping Arrays<p>
It is sometimes possible to change the `shape' of an array using
the <tt>reshape</tt> function. For example, a 1-d 10 element array
may be reshaped into a 2-d array consisting of 5 rows and 2
columns. The only restriction on the operation is that the arrays
must be commensurate. The <tt>reshape</tt> function follows the
syntax
<tscreen>
reshape (<em>array-name</em>, <em>integer-array</em>);
</tscreen>
where <em>array-name</em> specifies the array to be reshaped to have
the dimensions given by <tt>integer-array</tt>, a 1-dimensional array of
integers. It is important to note that this does <em>not</em> create a
new array, it simply reshapes the existing array. Thus,
<tscreen><verb>
variable a = Double_Type [100];
reshape (a, [10, 10]);
</verb></tscreen>
turns <tt>a</tt> into a <tt>10</tt> by <tt>10</tt> array.
<sect>Indexing Arrays<p>
An individual element of an array may be referred to by its
<em>index</em>. For example, <tt>a[0]</tt> specifies the zeroth element
of the one dimensional array <tt>a</tt>, and <tt>b[3,2]</tt> specifies
the element in the third row and second column of the two
dimensional array <tt>b</tt>. As in C array indices are numbered from
<tt>0</tt>. Thus if <tt>a</tt> is a one-dimensional array of ten
integers, the last element of the array is given by <tt>a[9]</tt>.
Using <tt>a[10]</tt> would result in a range error.
A negative index may be used to index from the end of the array,
with <tt>a[-1]</tt> referring to the last element of <tt>a</tt>,
<tt>a[-2]</tt> referring to the next to the last element, and so on.
One may use the indexed value like any other variable. For
example, to set the third element of an integer array to <tt>6</tt>, use
<tscreen><verb>
a[2] = 6;
</verb></tscreen>
Similarly, that element may be used in an expression, such as
<tscreen><verb>
y = a[2] + 7;
</verb></tscreen>
Unlike other <bf>S-Lang</bf> variables which inherit a type upon assignment,
array elements already have a type. For example, an attempt to
assign a string value to an element of an integer array will result
in a type-mismatch error.
One may use any integer expression to index an array. A simple
example that computes the sum of the elements of 10 element 1-d
array is
<tscreen><verb>
variable i, sum;
sum = 0;
for (i = 0; i < 10; i++) sum += a[i];
</verb></tscreen>
Unlike many other languages, <bf>S-Lang</bf> permits arrays to be indexed by
other integer arrays. Suppose that <tt>a</tt> is a 1-d array of 10
doubles. Now consider:
<tscreen><verb>
i = [6:8];
b = a[i];
</verb></tscreen>
Here, <tt>i</tt> is a 1-dimensional range array of three integers with
<tt>i[0]</tt> equal to <tt>6</tt>, <tt>i[1]</tt> equal to <tt>7</tt>,
and <tt>i[2]</tt> equal to <tt>8</tt>. The statement <tt>b = a[i];</tt>
will create a 1-d array of three doubles and assign it to <tt>b</tt>.
The zeroth element of <tt>b</tt>, <tt>b[0]</tt> will be set to the sixth
element of <tt>a</tt>, or <tt>a[6]</tt>, and so on. In fact, these two simple
statements are equivalent to
<tscreen><verb>
b = Double_Type [3];
b[0] = a[6];
b[1] = a[7];
b[2] = a[8];
</verb></tscreen>
except that using an array of indices is not only much more
convenient, but executes much faster.
More generally, one may use an index array to specify which
elements are to participate in a calculation. For example, consider
<tscreen><verb>
a = Double_Type [1000];
i = [0:499];
j = [500:999];
a[i] = -1.0;
a[j] = 1.0;
</verb></tscreen>
This creates an array of <tt>1000</tt> doubles and sets the first
<tt>500</tt> elements to <tt>-1.0</tt> and the last <tt>500</tt> to
<tt>1.0</tt>. Actually, one may do away with the <tt>i</tt> and <tt>j</tt>
variables altogether and use
<tscreen><verb>
a = Double_Type [1000];
a [[0:499]] = -1.0;
a [[500:999]] = 1.0;
</verb></tscreen>
It is important to understand the syntax used and, in particular,
to note that <tt>a[[0:499]]</tt> is <em>not</em> the same as
<tt>a[0:499]</tt>. In fact, the latter will generate a syntax error.
Often, it is convenient to use a <em>rubber</em> range to specify
indices. For example, <tt>a[[500:]]</tt> specifies all elements of
<tt>a</tt> whose index is greater than or equal to <tt>500</tt>. Similarly,
<tt>a[[:499]]</tt> specifies the first 500 elements of <tt>a</tt>.
Finally, <tt>a[[:]]</tt> specifies all the elements of <tt>a</tt>;
however, using <tt>a[*]</tt> is more convenient.
Now consider a multi-dimensional array. For simplicity, suppose
that <tt>a</tt> is a <tt>100</tt> by <tt>100</tt> array of doubles. Then
the expression <tt>a[0, *]</tt> specifies all elements in the zeroth
row. Similarly, <tt>a[*, 7]</tt> specifies all elements in the
seventh column. Finally, <tt>a[[3:5][6:12]]</tt> specifies the
<tt>3</tt> by <tt>7</tt> region consisting of rows <tt>3</tt>, <tt>4</tt>,
and <tt>5</tt>, and columns <tt>6</tt> through <tt>12</tt> of <tt>a</tt>.
We conclude this section with a few examples.
Here is a function that computes the trace (sum of the diagonal
elements) of a square 2 dimensional <tt>n</tt> by <tt>n</tt> array:
<tscreen><verb>
define array_trace (a, n)
{
variable sum = 0, i;
for (i = 0; i < n; i++) sum += a[i, i];
return sum;
}
</verb></tscreen>
This fragment creates a <tt>10</tt> by <tt>10</tt> integer array, sets
its diagonal elements to <tt>5</tt>, and then computes the trace of
the array:
<tscreen><verb>
a = Integer_Type [10, 10];
for (j = 0; j < 10; j++) a[j, j] = 5;
the_trace = array_trace(a, 10);
</verb></tscreen>
We can get rid of the <tt>for</tt> loop as follows:
<tscreen><verb>
j = Integer_Type [10, 2];
j[*,0] = [0:9];
j[*,1] = [0:9];
a[j] = 5;
</verb></tscreen>
Here, the goal was to construct a 2-d array of indices that
correspond to the diagonal elements of <tt>a</tt>, and then use that
array to index <tt>a</tt>. To understand how
this works, consider the middle statements. They are equivalent
to the following <tt>for</tt> loops:
<tscreen><verb>
variable i;
for (i = 0; i < 10; i++) j[i, 0] = i;
for (i = 0; i < 10; i++) j[i, 1] = i;
</verb></tscreen>
Thus, row <tt>n</tt> of <tt>j</tt> will have the value <tt>(n,n)</tt>,
which is precisely what was sought.
Another example of this technique is the function:
<tscreen><verb>
define unit_matrix (n)
{
variable a = Integer_Type [n, n];
variable j = Integer_Type [n, 2];
j[*,0] = [0:n - 1];
j[*,1] = [0:n - 1];
a[j] = 1;
return a;
}
</verb></tscreen>
This function creates an <tt>n</tt> by <tt>n</tt> unit matrix,
that is a 2-d <tt>n</tt> by <tt>n</tt> array whose elements are all zero
except on the diagonal where they have a value of <tt>1</tt>.
<sect>Arrays and Variables<p>
When an array is created and assigned to a variable, the
interpreter allocates the proper amount of space for the array,
initializes it, and then assigns to the variable a <em>reference</em>
to the array. So, a variable that represents an array has a value
that is really a reference to the array. This has several
consequences, some good and some bad. It is believed that the
advantages of this representation outweigh the disadvantages.
First, we shall look at the positive aspects.
When a variable is passed to a function, it is always the value of
the variable that gets passed. Since the value of a variable
representing an array is a reference, a reference to the array gets
passed. One major advantage of this is rather obvious: it is a
fast and efficient way to pass the array. This also has another
consequence that is illustrated by the function
<tscreen><verb>
define init_array (a, n)
{
variable i;
for (i = 0; i < n; i++) a[i] = some_function (i);
}
</verb></tscreen>
where <tt>some_function</tt> is a function that generates a scalar
value to initialize the <em>ith</em> element. This function can be
used in the following way:
<tscreen><verb>
variable X = Double_Type [100000];
init_array (X, 100000);
</verb></tscreen>
Since the array is passed to the function by reference, there is no
need to make a separate copy of the <tt>100000</tt> element array. As
pointed out above, this saves both execution time and memory. The
other salient feature to note is that any changes made to the
elements of the array within the function will be manifested in the
array outside the function. Of course, in this case, this is a
desirable side-effect.
To see the downside of this representation, consider:
<tscreen><verb>
variable a, b;
a = Double_Type [10];
b = a;
a[0] = 7;
</verb></tscreen>
What will be the value of <tt>b[0]</tt>? Since the value of <tt>a</tt>
is really a reference to the array of ten doubles, and that
reference was assigned to <tt>b</tt>, <tt>b</tt> also refers to the same
array. Thus any changes made to the elements of <tt>a</tt>, will also
be made implicitly to <tt>b</tt>.
This begs the question: If the assignment of one variable which
represents an array, to another variable results in the assignment
of a reference to the array, then how does one make separate copies
of the array? There are several answers including using an index
array, e.g., <tt>b = a[*]</tt>; however, the most natural method is
to use the dereference operator:
<tscreen><verb>
variable a, b;
a = Double_Type [10];
b = @a;
a[0] = 7;
</verb></tscreen>
In this example, a separate copy of <tt>a</tt> will be created and
assigned to <tt>b</tt>. It is very important to note that <bf>S-Lang</bf>
never implicitly dereferences an object. So, one must explicitly use
the dereference operator. This means that the elements of a
dereferenced array are not themselves dereferenced. For example,
consider dereferencing an array of arrays, e.g.,
<tscreen><verb>
variable a, b;
a = Array_Type [2];
a[0] = Double_Type [10];
a[1] = Double_Type [10];
b = @a;
</verb></tscreen>
In this example, <tt>b[0]</tt> will be a reference to the array that
<tt>a[0]</tt> references because <tt>a[0]</tt> was not explicitly
dereferenced.
<sect>Using Arrays in Computations<p>
Many functions and operations work transparently with arrays.
For example, if <tt>a</tt> and <tt>b</tt> are arrays, then the sum
<tt>a + b</tt> is an array whose elements are formed from the sum of
the corresponding elements of <tt>a</tt> and <tt>b</tt>. A similar
statement holds for all other binary and unary operations.
Let's consider a simple example. Suppose, that we wish to solve a
set of <tt>n</tt> quadratic equations whose coefficients are given by
the 1-d arrays <tt>a</tt>, <tt>b</tt>, and <tt>c</tt>. In general, the
solution of a quadratic equation will be two complex numbers. For
simplicity, suppose that all we really want is to know what subset of
the coefficients, <tt>a</tt>, <tt>b</tt>, <tt>c</tt>, correspond to
real-valued solutions. In terms of <tt>for</tt> loops, we can write:
<tscreen><verb>
variable i, d, index_array;
index_array = Integer_Type [n];
for (i = 0; i < n; i++)
{
d = b[i]^2 - 4 * a[i] * c[i];
index_array [i] = (d >= 0.0);
}
</verb></tscreen>
In this example, the array <tt>index_array</tt> will contain a
non-zero value if the corresponding set of coefficients has a
real-valued solution. This code may be written much more compactly
and with more clarity as follows:
<tscreen><verb>
variable index_array = ((b^2 - 4 * a * c) >= 0.0);
</verb></tscreen>
<bf>S-Lang</bf> has a powerful built-in function called <tt>where</tt>. This
function takes an array of integers and returns a 2-d array of
indices that correspond to where the elements of the input array
are non-zero. This simple operation is extremely useful. For
example, suppose <tt>a</tt> is a 1-d array of <tt>n</tt> doubles, and it
is desired to set to zero all elements of the array whose value is
less than zero. One way is to use a <tt>for</tt> loop:
<tscreen><verb>
for (i = 0; i < n; i++)
if (a[i] < 0.0) a[i] = 0.0;
</verb></tscreen>
If <tt>n</tt> is a large number, this statement can take some time to
execute. The optimal way to achieve the same result is to use the
<tt>where</tt> function:
<tscreen><verb>
a[where (a < 0.0)] = 0;
</verb></tscreen>
Here, the expression <tt>(a < 0.0)</tt> returns an array whose
dimensions are the same size as <tt>a</tt> but whose elements are
either <tt>1</tt> or <tt>0</tt>, according to whether or not the
corresponding element of <tt>a</tt> is less than zero. This array of
zeros and ones is then passed to <tt>where</tt> which returns a 2-d
integer array of indices that indicate where the elements of
<tt>a</tt> are less than zero. Finally, those elements of <tt>a</tt> are
set to zero.
As a final example, consider once more the example involving the set of
<tt>n</tt> quadratic equations presented above. Suppose that we wish
to get rid of the coefficients of the previous example that
generated non-real solutions. Using an explicit <tt>for</tt> loop requires
code such as:
<tscreen><verb>
variable i, j, nn, tmp_a, tmp_b, tmp_c;
nn = 0;
for (i = 0; i < n; i++)
if (index_array [i]) nn++;
tmp_a = Double_Type [nn];
tmp_b = Double_Type [nn];
tmp_c = Double_Type [nn];
j = 0;
for (i = 0; i < n; i++)
{
if (index_array [i])
{
tmp_a [j] = a[i];
tmp_b [j] = b[i];
tmp_c [j] = c[i];
j++;
}
}
a = tmp_a;
b = tmp_b;
c = tmp_c;
</verb></tscreen>
Not only is this a lot of code, it is also clumsy and error-prone.
Using the <tt>where</tt> function, this task is trivial:
<tscreen><verb>
variable i;
i = where (index_array != 0);
a = a[i];
b = b[i];
c = c[i];
</verb></tscreen>
All the examples up to now assumed that the dimensions of the array
were known. Although the intrinsic function <tt>length</tt> may be
used to get the total number of elements of an array, it cannot be
used to get the individual dimensions of a multi-dimensional array.
However, the function <tt>array_info</tt> may be used to
get information about an array, such as its data type and size.
The function returns three values: the data type, the number of
dimensions, and an integer array containing the size
of each dimension. It may be used to determine the number of rows
of an array as follows:
<tscreen><verb>
define num_rows (a)
{
variable dims, type, num_dims;
(dims, num_dims, type) = array_info (a);
return dims[0];
}
</verb></tscreen>
The number of columns may be obtained in a similar manner:
<tscreen><verb>
define num_cols (a)
{
variable dims, type, num_dims;
(dims, num_dims, type) = array_info (a);
if (num_dims > 1) return dims[1];
return 1;
}
</verb></tscreen>
Another use of <tt>array_info</tt> is to create an array that has the
same number of dimensions as another array:
<tscreen><verb>
define make_int_array (a)
{
variable dims, num_dims, type;
(dims, num_dims, type) = array_info (a);
return @Array_Type (Integer_Type, dims);
}
</verb></tscreen>
<chapt>Associative Arrays<p>
An associative array differs from an ordinary array in the sense
that its size is not fixed and that is indexed by a string, called
the <em>key</em>. For example, consider:
<tscreen><verb>
variable A = Assoc_Type [Integer_Type];
A["alpha"] = 1;
A["beta"] = 2;
A["gamma"] = 3;
</verb></tscreen>
Here, <tt>A</tt> represents an associative array of integers
(<tt>Integer_Type</tt>) and three keys have been added to the array.
As the example suggests, an associative array may be created using
one of the following forms:
<tscreen>
Assoc_Type [<em>type</em>]
Assoc_Type [<em>type</em>, <em>default-value</em>]
Assoc_Type []
</tscreen>
The last form returns an associative array of <tt>Any_Type</tt>
objects allowing any type of object to may be stored in
the array.
The form involving a <em>default-value</em> is useful for associating a
default value for non-existent array members. This feature is
explained in more detail below.
There are several functions that are specially designed to work
with associative arrays. These include:
<itemize>
<item> <tt>assoc_get_keys</tt>, which returns an ordinary array of strings
containing the keys in the array.
<item> <tt>assoc_get_values</tt>, which returns an ordinary array of the
values of the associative array.
<item> <tt>assoc_key_exists</tt>, which can be used to determine whether
or not a key exists in the array.
<item> <tt>assoc_delete_key</tt>, which may be used to remove a key (and
its value) from the array.
</itemize>
To illustrate the use of an associative array, consider the problem
of counting the number of repeated occurrences of words in a list.
Let the word list be represented as an array of strings given by
<tt>word_list</tt>. The number of occurrences of each word may be
stored in an associative array as follows:
<tscreen><verb>
variable a, word;
a = Assoc_Type [Integer_Type];
foreach (word_list)
{
word = ();
if (0 == assoc_key_exists (a, word))
a[word] = 0;
a[word]++; % same as a[word] = a[word] + 1;
}
</verb></tscreen>
Note that <tt>assoc_key_exists</tt> was necessary to determine whether
or not a word was already added to the array in order to properly
initialize it. However, by creating the associative array with a
default value of <tt>0</tt>, the above code may be simplified to
<tscreen><verb>
variable a, word;
a = Assoc_Type [Integer_Type, 0];
foreach (word_list)
{
word = ();
a[word]++;
}
</verb></tscreen>
<chapt>Structures and User-Defined Types<p>
A <em>structure</em> is a heterogeneous container object, i.e., it is
an object with elements whose values do not have to be of the same
data type. The elements or fields of a structure are named, and
one accesses a particular field of the structure via the field
name. This should be contrasted with an array whose values are of
the same type, and whose elements are accessed via array indices.
A <em>user-defined</em> data type is a structure with a fixed set of
fields defined by the user.
<sect>Defining a Structure<p>
The <tt>struct</tt> keyword is used to define a structure. The syntax
for this operation is:
<tscreen>
struct {<em>field-name-1</em>, <em>field-name-2</em>, ... <em>field-name-N</em>};
</tscreen>
This creates and returns a structure with <em>N</em> fields whose names
are specified by <em>field-name-1</em>, <em>field-name-2</em>, ...,
<em>field-name-N</em>. When a structure is created, all its fields are
initialized to <tt>NULL</tt>.
For example,
<tscreen><verb>
variable t = struct { city_name, population, next };
</verb></tscreen>
creates a structure with three fields and assigns it to the
variable <tt>t</tt>.
Alternatively, a structure may be created by dereferencing
<tt>Struct_Type</tt>. For example, the above structure may also be
created using one of the two forms:
<tscreen><verb>
t = @Struct_Type ("city_name", "population", "next");
t = @Struct_Type (["city_name", "population", "next"]);
</verb></tscreen>
These are useful when creating structures dynamically where one does
not know the name of the fields until run-time.
Like arrays, structures are passed around via a references. Thus,
in the above example, the value of <tt>t</tt> is a reference to the
structure. This means that after execution of
<tscreen><verb>
variable u = t;
</verb></tscreen>
<em>both</em> <tt>t</tt> and <tt>u</tt> refer to the <em>same</em> structure,
since only the reference was used in the assignment. To actually
create a new copy of the structure, use the <em>dereference</em>
operator, e.g.,
<tscreen><verb>
variable u = @t;
</verb></tscreen>
<sect>Accessing the Fields of a Structure<p>
The dot (<tt>.</tt>) operator is used to specify the a particular
field of structure. If <tt>s</tt> is a structure and <tt>field_name</tt>
is a field of the structure, then <tt>s.field_name</tt> specifies
that field of <tt>s</tt>. This specification can be used in
expressions just like ordinary variables. Again, consider
<tscreen><verb>
variable t = struct { city_name, population, next };
</verb></tscreen>
described in the last section. Then,
<tscreen><verb>
t.city_name = "New York";
t.population = 13000000;
if (t.population > 200) t = t.next;
</verb></tscreen>
are all valid statements involving the fields of <tt>t</tt>.
<sect>Linked Lists<p>
One of the most important uses of structures is to create a
<em>dynamic</em> data structure such as a <em>linked-list</em>. A
linked-list is simply a chain of structures that are linked together
such that one structure in the chain is the value of a field of the
previous structure in the chain. To be concrete, consider the
structure discussed earlier:
<tscreen><verb>
variable t = struct { city_name, population, next };
</verb></tscreen>
and suppose that we desire to create a list of such structures.
The purpose of the <tt>next</tt> field is to provide the link to the
next structure in the chain. Suppose that there exists a function,
<tt>read_next_city</tt>, that reads city names and populations from a
file. Then we can create the list via:
<tscreen><verb>
define create_population_list ()
{
variable city_name, population, list_root, list_tail;
variable next;
list_root = NULL;
while (read_next_city (&city_name, &population))
{
next = struct {city_name, population, next };
next.city_name = city_name;
next.population = population;
next.next = NULL;
if (list_root == NULL)
list_root = next;
else
list_tail.next = next;
list_tail = next;
}
return list_root;
}
</verb></tscreen>
In this function, the variables <tt>list_root</tt> and <tt>list_tail</tt>
represent the beginning and end of the list, respectively. As long
as <tt>read_next_city</tt> returns a non-zero value, a new structure is
created, initialized, and then appended to the list via the
<tt>next</tt> field of the <tt>list_tail</tt> structure. On the first
time through the loop, the list is created via the assignment to the
<tt>list_root</tt> variable.
This function may be used as follows:
<tscreen><verb>
variable Population_List = create_population_list ();
if (Population_List == NULL) error ("List is empty");
</verb></tscreen>
We can create other functions that manipulate the list. An example is
a function that finds the city with the largest population:
<tscreen><verb>
define get_largest_city (list)
{
variable largest;
largest = list;
while (list != NULL)
{
if (list.population > largest.population)
largest = list;
list = list.next;
}
return largest.city_name;
}
vmessage ("%s is the largest city in the list",
get_largest_city (Population_List)));
</verb></tscreen>
The <tt>get_largest_city</tt> is a typical example of how one traverses
a linear linked-list by starting at the head of the list and
successively moves to the next element of the list via the
<tt>next</tt> field.
In the previous example, a <tt>while</tt> loop was used to traverse the
linked list. It is faster to use a <tt>foreach</tt> loop for this:
<tscreen><verb>
define get_largest_city (list)
{
variable largest, elem;
largest = list;
foreach (list)
{
elem = ();
if (item.population > largest.population)
largest = item;
}
return largest.city_name;
}
</verb></tscreen>
Here a <tt>foreach</tt> loop has been used to walk the list via its
<tt>next</tt> field. If the field name was not <tt>next</tt>, then it
would have been necessary to use the <tt>using</tt> form of the
<tt>foreach</tt> statement. For example, if the field name implementing the
linked list was <tt>next_item</tt>, then
<tscreen><verb>
foreach (list) using ("next_item")
{
elem = ();
.
.
}
</verb></tscreen>
would have been used. In other words, unless otherwise indicated
via the <tt>using</tt> clause, <tt>foreach</tt> walks the list using a field
named <tt>next</tt>.
Now consider a function that sorts the list according to population.
To illustrate the technique, a <em>bubble-sort</em> will be used, not
because it is efficient, it is not, but because it is simple and
intuitive.
<tscreen><verb>
define sort_population_list (list)
{
variable changed;
variable node, next_node, last_node;
do
{
changed = 0;
node = list;
next_node = node.next;
last_node = NULL;
while (next_node != NULL)
{
if (node.population < next_node.population)
{
% swap node and next_node
node.next = next_node.next;
next_node.next = node;
if (last_node != NULL)
last_node.next = next_node;
if (list == node) list = next_node;
node = next_node;
next_node = node.next;
changed++;
}
last_node = node;
node = next_node;
next_node = next_node.next;
}
}
while (changed);
return list;
}
</verb></tscreen>
Note the test for equality between <tt>list</tt> and <tt>node</tt>, i.e.,
<tscreen><verb>
if (list == node) list = next_node;
</verb></tscreen>
It is important to appreciate the fact that the values of these
variables are references to structures, and that the
comparison only compares the references and <em>not</em> the actual
structures they reference. If it were not for this, the algorithm
would fail.
<sect>Defining New Types<p>
A user-defined data type may be defined using the <tt>typedef</tt>
keyword. In the current implementation, a user-defined data type
is essentially a structure with a user-defined set of fields. For
example, in the previous section a structure was used to represent
a city/population pair. We can define a data type called
<tt>Population_Type</tt> to represent the same information:
<tscreen><verb>
typedef struct
{
city_name,
population
} Population_Type;
</verb></tscreen>
This data type can be used like all other data types. For example,
an array of Population_Type types can be created,
<tscreen><verb>
variable a = Population_Type[10];
</verb></tscreen>
and `populated' via expressions such as
<tscreen><verb>
a[0].city_name = "Boston";
a[0].population = 2500000;
</verb></tscreen>
The new type <tt>Population_Type</tt> may also be used with the
<tt>typeof</tt> function:
<tscreen><verb>
if (Population_Type = typeof (a)) city = a.city_name;
</verb></tscreen>
The dereference <tt>@</tt> may be used to create an instance of the
new type:
<tscreen><verb>
a = @Population_Type;
a.city_name = "Calcutta";
a.population = 13000000;
</verb></tscreen>
<chapt>Error Handling<p>
Many intrinsic functions signal errors in the event of failure.
User defined functions may also generate an error condition via the
<tt>error</tt> function. Depending upon the severity of the error, it
can be caught and cleared using a construct called an
<em>error-block</em>.
<sect>Error-Blocks<p>
When the interpreter encounters a recoverable run-time error, it
will return to top-level by <em>unwinding</em> its function call
stack. Any error-blocks that it encounters as part of this
unwinding process will get executed. Errors such as syntax errors
and memory allocation errors are not recoverable, and error-blocks
will not get executed when such errors are encountered.
An error-block is defined using the syntax
<tscreen><verb>
ERROR_BLOCK { statement-list }
</verb></tscreen>
where <em>statement-list</em> represents a list of statements that
comprise the error-block. A simple example of an error-block is
<tscreen><verb>
define simple (a)
{
ERROR_BLOCK { message ("error-block executed"); }
if (a) error ("Triggering Error");
message ("hello");
}
</verb></tscreen>
Executing this function via <tt>simple(0)</tt> will result in the
message <tt>"hello"</tt>. However, calling it using <tt>simple(1)</tt>
will generate an error that will be caught, but not cleared, by
the error-block and the <tt>"error-block executed"</tt> message will
result.
Error-blocks are never executed unless triggered by an error. The
only exception to this is when the user explicitly indicates that
the error-block in scope should execute. This is indicated by the
special keyword <tt>EXECUTE_ERROR_BLOCK</tt>. For example,
<tt>simple</tt> could be recoded as
<tscreen><verb>
define simple (a)
{
variable err_string = "error-block executed";
ERROR_BLOCK { message (err_string); }
if (a) error ("Triggering Error");
err_string = "hello";
EXECUTE_ERROR_BLOCK;
}
</verb></tscreen>
Please note that <tt>EXECUTE_ERROR_BLOCK</tt> does not initiate an
error condition; it simply causes the error-block to be executed
and control will pass onto the next statement following the
<tt>EXECUTE_ERROR_BLOCK</tt> statement.
<sect>Clearing Errors<p>
Once an error has been caught by an error-block, the error can be cleared
by the <tt>_clear_error</tt> function. After the error has been cleared,
execution will resume at the next statement at the level of the error block
following the statement that generated the error. For example, consider:
<tscreen><verb>
define make_error ()
{
error ("Error condition created.");
message ("This statement is not executed.");
}
define test ()
{
ERROR_BLOCK
{
_clear_error ();
}
make_error ();
message ("error cleared.");
}
</verb></tscreen>
Calling <tt>test</tt> will trigger an error in the <tt>make_error</tt>
function, but will get cleared in the <tt>test</tt> function. The
call-stack will unwind from <tt>make_error</tt> back into <tt>test</tt>
where the error-block will get executed. As a result, execution
resumes after the statement that makes the call to <tt>make_error</tt>
since this statement is at the same level as the error-block that
cleared the error.
Here is another example that illustrates how multiple error-blocks
work:
<tscreen><verb>
define example ()
{
variable n = 0, s = "";
variable str;
ERROR_BLOCK {
str = sprintf ("s=%s,n=%d", s, n);
_clear_error ();
}
forever
{
ERROR_BLOCK {
s += "0";
_clear_error ();
}
if (n == 0) error ("");
ERROR_BLOCK {
s += "1";
}
if (n == 1) error ("");
n++;
}
return str;
}
</verb></tscreen>
Here, three error-blocks have been declared. One has been declared
outside the <tt>forever</tt> loop and the other two have been declared
inside the <tt>forever</tt> loop. Each time through the loop, the variable
<tt>n</tt> is incremented and a different error-block is triggered. The
error-block that gets triggered is the last one encountered, since
that will be the one in scope. On the first time through the loop,
<tt>n</tt> will be zero and the first error-block in the loop will get
executed. This error block clears the error and execution resumes
following the <tt>if</tt> statement that triggered the error. The
variable <tt>n</tt> will get incremented to <tt>1</tt> and, on the
second cycle through the loop the second <tt>if</tt> statement
will trigger an error causing the second error-block to execute.
This time, the error is not cleared and the call-stack unwinds out
of the <tt>forever</tt> loop, at which point the error-block outside
the loop is in scope, causing it to execute. This error-block
prints out the values of the variables <tt>s</tt> and <tt>n</tt>. It
will clear the error and execution resumes on the statement
<em>following</em> the <tt>forever</tt> loop. The result of this
complicated series of events is that the function will return the
string <tt>"s=01,n=1"</tt>.
<chapt>Loading Files: evalfile and autoload<p>
<chapt>File Input/Output<p>
<bf>S-Lang</bf> provides built-in supports for two different I/O facilities.
The simplest interface is modeled upon the C language <tt>stdio</tt>
streams interface and consists of functions such as <tt>fopen</tt>,
<tt>fgets</tt>, etc. The other interface is modeled on a lower level
POSIX interface consisting of functions such as <tt>open</tt>,
<tt>read</tt>, etc. In addition to permitting more control, the lower
level interface permits one to access network objects as well as disk
files.
<sect>Input/Output via stdio<p>
<sect1>Stdio Overview<p>
The <tt>stdio</tt> interface consists of the following functions:
<itemize>
<item> <tt>fopen</tt>, which opens a file for read or writing.
<item> <tt>fclose</tt>, which closes a file opened by <tt>fopen</tt>.
<item> <tt>fgets</tt>, used to read a line from the file.
<item> <tt>fputs</tt>, which writes text to the file.
<item> <tt>fprintf</tt>, used to write formatted text to the file.
<item> <tt>fwrite</tt>, which may be used to write objects to the
file.
<item> <tt>fread</tt>, which reads a specified number of objects from
the file.
<item> <tt>feof</tt>, which is used to test whether the file pointer is at the
end of the file.
<item> <tt>ferror</tt>, which is used to see whether or not the stream
associated with the file has an error.
<item> <tt>clearerr</tt>, which clears the end-of-file and error
indicators for the stream.
<item> <tt>fflush</tt>, used to force all buffered data associated with
the stream to be written out.
<item> <tt>ftell</tt>, which is used to query the file position indicator
of the stream.
<item> <tt>fseek</tt>, which is used to set the position of the file
position indicator of the stream.
<item> <tt>fgetslines</tt>, which reads all the lines in a text file and
returns them as an array of strings.
</itemize>
In addition, the interface supports the <tt>popen</tt> and <tt>pclose</tt>
functions on systems where the corresponding C functions are available.
Before reading or writing to a file, it must first be opened using
the <tt>fopen</tt> function. The only exceptions to this rule involves
use of the pre-opened streams: <tt>stdin</tt>, <tt>stdout</tt>, and
<tt>stderr</tt>. <tt>fopen</tt> accepts two arguments: a file name and a
string argument that indicates how the file is to be opened, e.g.,
for reading, writing, update, etc. It returns a <tt>File_Type</tt>
stream object that is used as an argument to all other functions of
the <tt>stdio</tt> interface. Upon failure, it returns <tt>NULL</tt>. See the
reference manual for more information about <tt>fopen</tt>.
<sect1>Stdio Examples<p>
In this section, some simple examples of the use of the <tt>stdio</tt>
interface is presented. It is important to realize that all the
functions of the interface return something, and that return value
must be dealt with.
The first example involves writing a function to count the number of
lines in a text file. To do this, we shall read in the lines, one by
one, and count them:
<tscreen><verb>
define count_lines_in_file (file)
{
variable fp, line, count;
fp = fopen (file, "r"); % Open the file for reading
if (fp == NULL)
verror ("%s failed to open", file);
count = 0;
while (-1 != fgets (&line, fp))
count++;
() = fclose (fp);
return count;
}
</verb></tscreen>
Note that <tt>&line</tt> was passed to the <tt>fgets</tt> function. When
<tt>fgets</tt> returns, <tt>line</tt> will contain the line of text read in
from the file. Also note how the return value from <tt>fclose</tt> was
handled.
Although the preceding example closed the file via <tt>fclose</tt>,
there is no need to explicitly close a file because <bf>S-Lang</bf> will
automatically close the file when it is no longer referenced. Since
the only variable to reference the file is <tt>fp</tt>, it would have
automatically been closed when the function returned.
Suppose that it is desired to count the number of characters in the
file instead of the number of lines. To do this, the <tt>while</tt>
loop could be modified to count the characters as follows:
<tscreen><verb>
while (-1 != fgets (&line, fp))
count += strlen (line);
</verb></tscreen>
The main difficulty with this approach is that it will not work for
binary files, i.e., files that contain null characters. For such
files, the file should be opened in <em>binary</em> mode via
<tscreen><verb>
fp = fopen (file, "rb");
</verb></tscreen>
and then the data read in using the <tt>fread</tt> function:
<tscreen><verb>
while (-1 != fread (&line, Char_Type, 1024, fp))
count += bstrlen (line);
</verb></tscreen>
The <tt>fread</tt> function requires two additional arguments: the type
of object to read (<tt>Char_Type</tt> in the case), and the number of
such objects to read. The function returns the number of objects
actually read, or -1 upon failure. The <tt>bstrlen</tt> function was
used to compute the length of <tt>line</tt> because for <tt>Char_Type</tt>
or <tt>UChar_Type</tt> objects, the <tt>fread</tt> function assigns a
<em>binary</em> string (<tt>BString_Type</tt>) to <tt>line</tt>.
The <tt>foreach</tt> construct also works with <tt>File_Type</tt> objects.
For example, the number of characters in a file may be counted via
<tscreen><verb>
foreach (fp) using ("char")
{
ch = ();
count++;
}
</verb></tscreen>
To count the number of lines, one can use:
<tscreen><verb>
foreach (fp) using ("line")
{
line = ();
num_lines++;
count += strlen (line);
}
</verb></tscreen>
Finally, it should be mentioned that neither of these examples should
be used to count the number of characters in a file when that
information is more readily accessible by another means. For
example, it is preferable to get this information via the
<tt>stat_file</tt> function:
<tscreen><verb>
define count_chars_in_file (file)
{
variable st;
st = stat_file (file);
if (st == NULL)
error ("stat_file failed.");
return st.st_size;
}
</verb></tscreen>
<sect>POSIX I/O<p>
<sect>Advanced I/O techniques<p>
The previous examples illustrate how to read and write objects of a
single data-type from a file, e.g.,
<tscreen><verb>
num = fread (&a, Double_Type, 20, fp);
</verb></tscreen>
would result in a <tt>Double_Type[num]</tt> array being assigned to
<tt>a</tt> if successful. However, suppose that the binary data file
consists of numbers in a specified byte-order. How can one read
such objects with the proper byte swapping? The answer is to use
the <tt>fread</tt> function to read the objects as <tt>Char_Type</tt> and
then <em>unpack</em> the resulting string into the specified data type,
or types. This process is facilitated using the <tt>pack</tt> and
<tt>unpack</tt> functions.
The <tt>pack</tt> function follows the syntax
<tscreen>
BString_Type pack (<em>format-string</em>, <em>item-list</em>);
</tscreen>
and combines the objects in the <em>item-list</em> according to
<em>format-string</em> into a binary string and returns the result.
Likewise, the <tt>unpack</tt> function may be used to convert a binary
string into separate data objects:
<tscreen>
(<em>variable-list</em>) = unpack (<em>format-string</em>, <em>binary-string</em>);
</tscreen>
The format string consists of one or more data-type specification
characters, and each may be followed by an optional decimal length
specifier. Specifically, the data-types are specified according to
the following table:
<tscreen><verb>
c char
C unsigned char
h short
H unsigned short
i int
I unsigned int
l long
L unsigned long
j 16 bit int
J 16 unsigned int
k 32 bit int
K 32 bit unsigned int
f float
d double
F 32 bit float
D 64 bit float
s character string, null padded
S character string, space padded
x a null pad character
</verb></tscreen>
A decimal length specifier may follow the data-type specifier. With
the exception of the <tt>s</tt> and <tt>S</tt> specifiers, the length
specifier indicates how many objects of that data type are to be
packed or unpacked from the string. When used with the <tt>s</tt> or
<tt>S</tt> specifiers, it indicates the field width to be used. If the
length specifier is not present, the length defaults to one.
With the exception of <tt>c</tt>, <tt>C</tt>, <tt>s</tt>, <tt>S</tt>, and
<tt>x</tt>, each of these may be prefixed by a character that indicates
the byte-order of the object:
<tscreen><verb>
> big-endian order (network order)
< little-endian order
= native byte-order
</verb></tscreen>
The default is native byte order.
Here are a few examples that should make this more clear:
<tscreen><verb>
a = pack ("cc", 'A', 'B'); % ==> a = "AB";
a = pack ("c2", 'A', 'B'); % ==> a = "AB";
a = pack ("xxcxxc", 'A', 'B'); % ==> a = "\0\0A\0\0B";
a = pack ("h2", 'A', 'B'); % ==> a = "\0A\0B" or "\0B\0A"
a = pack (">h2", 'A', 'B'); % ==> a = "\0\xA\0\xB"
a = pack ("<h2", 'A', 'B'); % ==> a = "\0B\0A"
a = pack ("s4", "AB", "CD"); % ==> a = "AB\0\0"
a = pack ("s4s2", "AB", "CD"); % ==> a = "AB\0\0CD"
a = pack ("S4", "AB", "CD"); % ==> a = "AB "
a = pack ("S4S2", "AB", "CD"); % ==> a = "AB CD"
</verb></tscreen>
When unpacking, if the length specifier is greater than one, then an
array of that length will be returned. In addition, trailing
whitespace and null character are stripped when unpacking an object
given by the <tt>S</tt> specifier. Here are a few examples:
<tscreen><verb>
(x,y) = unpack ("cc", "AB"); % ==> x = 'A', y = 'B'
x = unpack ("c2", "AB"); % ==> x = ['A', 'B']
x = unpack ("x<H", "\0\xAB\xCD"); % ==> x = 0xCDABuh
x = unpack ("xxs4", "a b c\0d e f"); % ==> x = "b c\0"
x = unpack ("xxS4", "a b c\0d e f"); % ==> x = "b c"
</verb></tscreen>
<sect1>Example: Reading /var/log/wtmp<p>
Consider the task of reading the Unix system file
<tt>/var/log/utmp</tt>, which contains login records about who logged
onto the system. This file format is documented in section 5 of the
online Unix man pages, and consists of a sequence of entries
formatted according to the C structure <tt>utmp</tt> defined in the
<tt>utmp.h</tt> C header file. The actual details of the structure
may vary from one version of Unix to the other. For the purposes of
this example, consider its definition under the Linux operating
system running on an Intel processor:
<tscreen><verb>
struct utmp {
short ut_type; /* type of login */
pid_t ut_pid; /* pid of process */
char ut_line[12]; /* device name of tty - "/dev/" */
char ut_id[2]; /* init id or abbrev. ttyname */
time_t ut_time; /* login time */
char ut_user[8]; /* user name */
char ut_host[16]; /* host name for remote login */
long ut_addr; /* IP addr of remote host */
};
</verb></tscreen>
On this system, <tt>pid_t</tt> is defined to be an <tt>int</tt> and
<tt>time_t</tt> is a <tt>long</tt>. Hence, a format specifier for the
<tt>pack</tt> and <tt>unpack</tt> functions is easily constructed to be:
<tscreen><verb>
"h i S12 S2 l S8 S16 l"
</verb></tscreen>
However, this particular definition is naive because it does not
allow for structure padding performed by the C compiler in order to
align the data types on suitable word boundaries. Fortunately, the
intrinsic function <tt>pad_pack_format</tt> may be used to modify a
format by adding the correct amount of padding in the right places.
In fact, <tt>pad_pack_format</tt> applied to the above format on an
Intel-based Linux system produces the result:
<tscreen><verb>
"h x2 i S12 S2 x2 l S8 S16 l"
</verb></tscreen>
Here we see that 4 bytes of padding were added.
The other missing piece of information is the size of the structure.
This is useful because we would like to read in one structure at a
time using the <tt>fread</tt> function. Knowing the size of the
various data types makes this easy; however it is even easier to use
the <tt>sizeof_pack</tt> intrinsic function, which returns the size (in
bytes) of the structure described by the pack format.
So, with all the pieces in place, it is rather straightforward to
write the code:
<tscreen><verb>
variable format, size, fp, buf;
typedef struct
{
ut_type, ut_pid, ut_line, ut_id,
ut_time, ut_user, ut_host, ut_addr
} UTMP_Type;
format = pad_pack_format ("h i S12 S2 l S8 S16 l");
size = sizeof_pack (format);
define print_utmp (u)
{
() = fprintf (stdout, "%-16s %-12s %-16s %s\n",
u.ut_user, u.ut_line, u.ut_host, ctime (u.ut_time));
}
fp = fopen ("/var/log/utmp", "rb");
if (fp == NULL)
error ("Unable to open utmp file");
() = fprintf (stdout, "%-16s %-12s %-16s %s\n",
"USER", "TTY", "FROM", "LOGIN@");
variable U = @UTMP_Type;
while (-1 != fread (&buf, Char_Type, size, fp))
{
set_struct_fields (U, unpack (format, buf));
print_utmp (U);
}
() = fclose (fp);
</verb></tscreen>
A few comments about this example are in order. First of all, note
that a new data type called <tt>UTMP_Type</tt> was created, although
this was not really necessary. We also opened the file in binary
mode, but this too is optional under a Unix system where there is no
distinction between binary and text modes. The <tt>print_utmp</tt>
function does not print all of the structure fields. Finally, last
but not least, the return values from <tt>fprintf</tt> and <tt>fclose</tt>
were dealt with.
<chapt>Debugging<p>
The current implementation provides no support for an interactive
debugger, although a future version will. Nevertheless, <bf>S-Lang</bf> has
several features that aid the programmer in tracking down problems,
including function call tracebacks and the tracing of function calls.
However, the biggest debugging aid stems from the fact that the
language is interpreted permitting one to easily add debugging
statements to the code.
To enable debugging information, add the lines
<tscreen><verb>
_debug_info = 1;
_traceback = 1;
</verb></tscreen>
to the top of the source file of the code containing the bug and the
reload the file. Setting the <tt>_debug_info</tt> variable to
<tt>1</tt> causes line number information to be compiled into the
functions when the file is loaded. The <tt>_traceback</tt> variable
controls whether or not traceback information should be generated.
If it is set to <tt>1</tt>, the values of local variables will be
dumped when the traceback is generated. Setting this variable
to <tt>-1</tt> will cause only function names to be reported in the
traceback.
Here is an example of a traceback report:
<tscreen><verb>
S-Lang Traceback: error
S-Lang Traceback: verror
S-Lang Traceback: (Error occurred on line 65)
S-Lang Traceback: search_generic_search
Local Variables:
$0: Type: String_Type, Value: "Search forward:"
$1: Type: Integer_Type, Value: 1
$2: Type: Ref_Type, Value: _function_return_1
$3: Type: String_Type, Value: "abcdefg"
$4: Type: Integer_Type, Value: 1
S-Lang Traceback: (Error occurred on line 72)
S-Lang Traceback: search_forward
</verb></tscreen>
There are several ways to read this report; perhaps the simplest is
to read it from the bottom. This report says that on line <tt>72</tt>,
the <tt>search_forward</tt> function called the
<tt>search_generic_search</tt> function. On line <tt>65</tt> it called the
<tt>verror</tt> function, which called <tt>error</tt>. The
<tt>search_generic_search</tt> function contains <tt>5</tt> local variables
and are represented symbolically as <tt>$0</tt> through <tt>$4</tt>.
<chapt>Regular Expressions<p>
The S-Lang library includes a regular expression (RE) package that
may be used by an application embedding the library. The RE syntax
should be familiar to anyone acquainted with regular expressions. In
this section the syntax of the <bf>S-Lang</bf> regular expressions is
discussed.
<sect><bf>S-Lang</bf> RE Syntax<p>
A regular expression specifies a pattern to be matched against a
string, and has the property that the contcatenation of two REs is
also a RE.
The <bf>S-Lang</bf> library supports the following standard regular
expressions:
<tscreen><verb>
. match any character except newline
* matches zero or more occurences of previous RE
+ matches one or more occurences of previous RE
? matches zero or one occurence of previous RE
^ matches beginning of a line
$ matches end of line
[ ... ] matches any single character between brackets.
For example, [-02468] matches `-' or any even digit.
and [-0-9a-z] matches `-' and any digit between 0 and 9
as well as letters a through z.
\< Match the beginning of a word.
\> Match the end of a word.
\( ... \)
\1, \2, ..., \9 Matches the match specified by nth \( ... \)
expression.
</verb></tscreen>
In addition the following extensions are also supported:
<tscreen><verb>
\c turn on case-sensitivity (default)
\C turn off case-sensitivity
\d match any digit
\e match ESC char
</verb></tscreen>
Here are some simple examples:
<tt>"^int "</tt> matches the <tt>"int "</tt> at the beginning of a line.
<tt>"\<money\>"</tt> matches <tt>"money"</tt> but only if it appears
as a separate word.
<tt>"^$"</tt> matches an empty line.
A more complex pattern is
<tscreen><verb>
"\(\<[a-zA-Z]+\>\)[ ]+\1\>"
</verb></tscreen>
which matches any word repeated consecutively. Note how the grouping
operators <tt>\(</tt> and <tt>\)</tt> are used to define the text
matched by the enclosed regular expression, and then subsequently
referred to <tt>\1</tt>.
Finally, remember that when used in string literals either in the
<bf>S-Lang</bf> language or in the C language, care must be taken to
"double-up" the <tt>'\'</tt> character since both languages treat it
as as an escape character.
<sect>Differences between <bf>S-Lang</bf> and egrep REs<p>
There are several differences between <bf>S-Lang</bf> regular expressions and,
e.g., <bf>egrep</bf> regular expressions.
The most notable difference is that the <bf>S-Lang</bf> regular expressions do
not support the <bf>OR</bf> operator <tt>|</tt> in expressions. This means
that <tt>"a|b"</tt> or <tt>"a\|b"</tt> do not have the meaning that they
have in regular expression packages that support egrep-style
expressions.
The other main difference is that while <bf>S-Lang</bf> regular expressions
support the grouping operators <tt>\(</tt> and <tt>\)</tt>, they are
only used as a means of specifying the text that is matched. That
is, the expression
<tscreen><verb>
"@\([a-z]*\)@.*@\1@"
</verb></tscreen>
matches <tt>"xxx@abc@silly@abc@yyy"</tt>, where the pattern <tt>\1</tt>
matches the text enclosed by the <tt>\(</tt> and <tt>\)</tt>
expressions. However, in the current implementation, the grouping
operators are not used to group regular expressions to form a single
regular expression. Thus expression such as <tt>"\(hello\)*"</tt> is
<em>not</em> a pattern to match zero or more occurances of <tt>"hello"</tt>
as it is in e.g., <bf>egrep</bf>.
One question that comes up from time to time is why doesn't <bf>S-Lang</bf>
simply employ some posix-compatible regular expression library. The
simple answer is that, at the time of this writing, none exists that
is is available across all the platforms that the <bf>S-Lang</bf> library
supports (Unix, VMS, OS/2, win32, win16, BEOS, MSDOS, and QNX) and
can be distributed under both the GNU and Artistic licenses. It is
particularly important that the library and the interpreter support a
common set of regular expressions in a platform independent manner.
<chapt>Future Directions<p>
Several new features or enhancements to the <bf>S-Lang</bf> language are
planned for the next major release. In no particular order, these
include:
<itemize>
<item> An interactive debugging facility.
<item> Function qualifiers. These entities should already be
familiar to VMS users or to those who are familiar with the IDL
language. Basically, a qualifier is an optional argument that is
passed to a function, e.g., <tt>plot(X,Y,/logx)</tt>. Here
<tt>/logx</tt> is a qualifier that specifies that the plot function
should use a log scale for <tt>x</tt>.
<item> File local variables and functions. A file local variable or
function is an object that is global to the file that defines it.
<item> Multi-threading. Currently the language does not support
multiple threads.
</itemize>
<appendix>
<chapt>Copyright<p>
The <bf>S-Lang</bf> library is distributed under two copyrights: the GNU
Genral Public License, and the Artistic License. Any program
that uses the interpreter must adhere to rules of one of these
licenses.
<sect>The GNU Public License<p>
<tscreen><verb>
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
Copyright (C) 1989, 1991 Free Software Foundation, Inc.
675 Mass Ave, Cambridge, MA 02139, USA
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
</verb></tscreen>
The licenses for most software are designed to take away your
freedom to share and change it. By contrast, the GNU General Public
License is intended to guarantee your freedom to share and change free
software--to make sure the software is free for all its users. This
General Public License applies to most of the Free Software
Foundation's software and to any other program whose authors commit to
using it. (Some other Free Software Foundation software is covered by
the GNU Library General Public License instead.) You can apply it to
your programs, too.
When we speak of free software, we are referring to freedom, not
price. Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
this service if you wish), that you receive source code or can get it
if you want it, that you can change the software or use pieces of it
in new free programs; and that you know you can do these things.
To protect your rights, we need to make restrictions that forbid
anyone to deny you these rights or to ask you to surrender the rights.
These restrictions translate to certain responsibilities for you if you
distribute copies of the software, or if you modify it.
For example, if you distribute copies of such a program, whether
gratis or for a fee, you must give the recipients all the rights that
you have. You must make sure that they, too, receive or can get the
source code. And you must show them these terms so they know their
rights.
We protect your rights with two steps: (1) copyright the software, and
(2) offer you this license which gives you legal permission to copy,
distribute and/or modify the software.
Also, for each author's protection and ours, we want to make certain
that everyone understands that there is no warranty for this free
software. If the software is modified by someone else and passed on, we
want its recipients to know that what they have is not the original, so
that any problems introduced by others will not reflect on the original
authors' reputations.
Finally, any free program is threatened constantly by software
patents. We wish to avoid the danger that redistributors of a free
program will individually obtain patent licenses, in effect making the
program proprietary. To prevent this, we have made it clear that any
patent must be licensed for everyone's free use or not licensed at all.
The precise terms and conditions for copying, distribution and
modification follow.
<tscreen><verb>
GNU GENERAL PUBLIC LICENSE
TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
</verb></tscreen>
0. This License applies to any program or other work which contains
a notice placed by the copyright holder saying it may be distributed
under the terms of this General Public License. The "Program", below,
refers to any such program or work, and a "work based on the Program"
means either the Program or any derivative work under copyright law:
that is to say, a work containing the Program or a portion of it,
either verbatim or with modifications and/or translated into another
language. (Hereinafter, translation is included without limitation in
the term "modification".) Each licensee is addressed as "you".
Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope. The act of
running the Program is not restricted, and the output from the Program
is covered only if its contents constitute a work based on the
Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does.
1. You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any warranty;
and give any other recipients of the Program a copy of this License
along with the Program.
You may charge a fee for the physical act of transferring a copy, and
you may at your option offer warranty protection in exchange for a fee.
2. You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:
<tscreen><verb>
a) You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.
b) You must cause any work that you distribute or publish, that in
whole or in part contains or is derived from the Program or any
part thereof, to be licensed as a whole at no charge to all third
parties under the terms of this License.
c) If the modified program normally reads commands interactively
when run, you must cause it, when started running for such
interactive use in the most ordinary way, to print or display an
announcement including an appropriate copyright notice and a
notice that there is no warranty (or else, saying that you provide
a warranty) and that users may redistribute the program under
these conditions, and telling the user how to view a copy of this
License. (Exception: if the Program itself is interactive but
does not normally print such an announcement, your work based on
the Program is not required to print an announcement.)
</verb></tscreen>
These requirements apply to the modified work as a whole. If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works. But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it.
Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Program.
In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.
3. You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:
<tscreen><verb>
a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three
years, to give any third party, for a charge no more than your
cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer
to distribute corresponding source code. (This alternative is
allowed only for noncommercial distribution and only if you
received the program in object code or executable form with such
an offer, in accord with Subsection b above.)
</verb></tscreen>
The source code for a work means the preferred form of the work for
making modifications to it. For an executable work, complete source
code means all the source code for all modules it contains, plus any
associated interface definition files, plus the scripts used to
control compilation and installation of the executable. However, as a
special exception, the source code distributed need not include
anything that is normally distributed (in either source or binary
form) with the major components (compiler, kernel, and so on) of the
operating system on which the executable runs, unless that component
itself accompanies the executable.
If distribution of executable or object code is made by offering
access to copy from a designated place, then offering equivalent
access to copy the source code from the same place counts as
distribution of the source code, even though third parties are not
compelled to copy the source along with the object code.
4. You may not copy, modify, sublicense, or distribute the Program
except as expressly provided under this License. Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License.
However, parties who have received copies, or rights, from you under
this License will not have their licenses terminated so long as such
parties remain in full compliance.
5. You are not required to accept this License, since you have not
signed it. However, nothing else grants you permission to modify or
distribute the Program or its derivative works. These actions are
prohibited by law if you do not accept this License. Therefore, by
modifying or distributing the Program (or any work based on the
Program), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Program or works based on it.
6. Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License.
7. If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all. For example, if a patent
license would not permit royalty-free redistribution of the Program by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Program.
If any portion of this section is held invalid or unenforceable under
any particular circumstance, the balance of the section is intended to
apply and the section as a whole is intended to apply in other
circumstances.
It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system, which is
implemented by public license practices. Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.
This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.
8. If the distribution and/or use of the Program is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Program under this License
may add an explicit geographical distribution limitation excluding
those countries, so that distribution is permitted only in or among
countries not thus excluded. In such case, this License incorporates
the limitation as if written in the body of this License.
9. The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
Each version is given a distinguishing version number. If the Program
specifies a version number of this License which applies to it and "any
later version", you have the option of following the terms and conditions
either of that version or of any later version published by the Free
Software Foundation. If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation.
10. If you wish to incorporate parts of the Program into other free
programs whose distribution conditions are different, write to the author
to ask for permission. For software which is copyrighted by the Free
Software Foundation, write to the Free Software Foundation; we sometimes
make exceptions for this. Our decision will be guided by the two goals
of preserving the free status of all derivatives of our free software and
of promoting the sharing and reuse of software generally.
<tscreen><verb>
NO WARRANTY
11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS
TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE
PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
REPAIR OR CORRECTION.
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
POSSIBILITY OF SUCH DAMAGES.
END OF TERMS AND CONDITIONS
</verb></tscreen>
Appendix: How to Apply These Terms to Your New Programs
If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
<tscreen><verb>
<one line to give the program's name and a brief idea of what it does.>
Copyright (C) 19yy <name of author>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
</verb></tscreen>
Also add information on how to contact you by electronic and paper mail.
If the program is interactive, make it output a short notice like this
when it starts in an interactive mode:
<tscreen><verb>
Gnomovision version 69, Copyright (C) 19yy name of author
Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
</verb></tscreen>
The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, the commands you use may
be called something other than `show w' and `show c'; they could even be
mouse-clicks or menu items--whatever suits your program.
You should also get your employer (if you work as a programmer) or your
school, if any, to sign a "copyright disclaimer" for the program, if
necessary. Here is a sample; alter the names:
<tscreen><verb>
Yoyodyne, Inc., hereby disclaims all copyright interest in the program
`Gnomovision' (which makes passes at compilers) written by James Hacker.
<signature of Ty Coon>, 1 April 1989
Ty Coon, President of Vice
</verb></tscreen>
This General Public License does not permit incorporating your program into
proprietary programs. If your program is a subroutine library, you may
consider it more useful to permit linking proprietary applications with the
library. If this is what you want to do, use the GNU Library General
Public License instead of this License.
<sect>The Artistic License<p>
<tscreen><verb>
The "Artistic License"
Preamble
</verb></tscreen>
The intent of this document is to state the conditions under which a
Package may be copied, such that the Copyright Holder maintains some
semblance of artistic control over the development of the package,
while giving the users of the package the right to use and distribute
the Package in a more-or-less customary fashion, plus the right to make
reasonable modifications.
Definitions:
<tscreen><verb>
"Package" refers to the collection of files distributed by the
Copyright Holder, and derivatives of that collection of files
created through textual modification.
"Standard Version" refers to such a Package if it has not been
modified, or has been modified in accordance with the wishes
of the Copyright Holder as specified below.
"Copyright Holder" is whoever is named in the copyright or
copyrights for the package.
"You" is you, if you're thinking about copying or distributing
this Package.
"Reasonable copying fee" is whatever you can justify on the
basis of media cost, duplication charges, time of people involved,
and so on. (You will not be required to justify it to the
Copyright Holder, but only to the computing community at large
as a market that must bear the fee.)
"Freely Available" means that no fee is charged for the item
itself, though there may be fees involved in handling the item.
It also means that recipients of the item may redistribute it
under the same conditions they received it.
</verb></tscreen>
1. You may make and give away verbatim copies of the source form of the
Standard Version of this Package without restriction, provided that you
duplicate all of the original copyright notices and associated disclaimers.
2. You may apply bug fixes, portability fixes and other modifications
derived from the Public Domain or from the Copyright Holder. A Package
modified in such a way shall still be considered the Standard Version.
3. You may otherwise modify your copy of this Package in any way, provided
that you insert a prominent notice in each changed file stating how and
when you changed that file, and provided that you do at least ONE of the
following:
<tscreen><verb>
a) place your modifications in the Public Domain or otherwise make them
Freely Available, such as by posting said modifications to Usenet or
an equivalent medium, or placing the modifications on a major archive
site such as uunet.uu.net, or by allowing the Copyright Holder to include
your modifications in the Standard Version of the Package.
b) use the modified Package only within your corporation or organization.
c) rename any non-standard executables so the names do not conflict
with standard executables, which must also be provided, and provide
a separate manual page for each non-standard executable that clearly
documents how it differs from the Standard Version.
d) make other distribution arrangements with the Copyright Holder.
</verb></tscreen>
4. You may distribute the programs of this Package in object code or
executable form, provided that you do at least ONE of the following:
<tscreen><verb>
a) distribute a Standard Version of the executables and library files,
together with instructions (in the manual page or equivalent) on where
to get the Standard Version.
b) accompany the distribution with the machine-readable source of
the Package with your modifications.
c) give non-standard executables non-standard names, and clearly
document the differences in manual pages (or equivalent), together
with instructions on where to get the Standard Version.
d) make other distribution arrangements with the Copyright Holder.
</verb></tscreen>
5. You may charge a reasonable copying fee for any distribution of this
Package. You may charge any fee you choose for support of this
Package. You may not charge a fee for this Package itself. However,
you may distribute this Package in aggregate with other (possibly
commercial) programs as part of a larger (possibly commercial) software
distribution provided that you do not advertise this Package as a
product of your own. You may embed this Package's interpreter within
an executable of yours (by linking); this shall be construed as a mere
form of aggregation, provided that the complete Standard Version of the
interpreter is so embedded.
6. The scripts and library files supplied as input to or produced as
output from the programs of this Package do not automatically fall
under the copyright of this Package, but belong to whomever generated
them, and may be sold commercially, and may be aggregated with this
Package. If such scripts or library files are aggregated with this
Package via the so-called "undump" or "unexec" methods of producing a
binary executable image, then distribution of such an image shall
neither be construed as a distribution of this Package nor shall it
fall under the restrictions of Paragraphs 3 and 4, provided that you do
not represent such an executable image as a Standard Version of this
Package.
7. C subroutines (or comparably compiled subroutines in other
languages) supplied by you and linked into this Package in order to
emulate subroutines and variables of the language defined by this
Package shall not be considered part of this Package, but are the
equivalent of input as in Paragraph 6, provided these subroutines do
not change the language in any way that would cause it to fail the
regression tests for the language.
8. Aggregation of this Package with a commercial distribution is always
permitted provided that the use of this Package is embedded; that is,
when no overt attempt is made to make this Package's interfaces visible
to the end user of the commercial distribution. Such use shall not be
construed as a distribution of this Package.
9. The name of the Copyright Holder may not be used to endorse or promote
products derived from this software without specific prior written permission.
10. THIS PACKAGE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR
IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.
</book>
