librep : The language
table of contents
Copying
Introduction
Invocation
The language
The REPL
librep Internals
Reporting bugs
News
Function index
Variable index
Concept index
The language
************
This chapter of the manual is a full guide to the librep Lisp
programming language, including documentation for most of the built-in
functions.
* Menu:
* Intro:: Introduction and Lisp conventions
Fundamental data types.
* Data Types:: Data types and values in Lisp
* Numbers:: Numeric representations and functions
* Sequences:: Ordered sequences of data values
* Symbols:: Symbols are uniquely named objects
The core language.
* Evaluation:: Evaluating expressions
* Variables:: Symbols represent named variables
* Functions:: The building blocks of Lisp programs
* Macros:: User-defined control structures
* Definitions:: Block-structured definitions
* Modules:: Scoping for "global" definitions
* Control Structures:: Conditionals, loops, etc...
* Threads:: Multi-threaded programs
* Loading:: Programs are stored in files
* Compiled Lisp:: Making programs run faster
Data structures and I/O.
* Datums:: Low-level data type definition
* Queues:: FIFO queue type
* Records:: Defining structured data types
* Hash Tables:: Efficient table lookups
* Guardians:: Protecting objects from GC
* Streams:: Data sinks and sources; character streams
* Hooks:: Hooks promote extensibility
* Files:: Manipulating files in the filing system
* Processes:: launch and control subprocesses when
running under Unix
Miscellaneous features.
* Regular Expressions:: Matching regular expressions
* Time and Date:: Manipulating time and date
* i18n:: Internationalisation
* System Information:: Getting details about the host
* User Information:: The name of the user
* Environment Variables:: Reading and writing the environment
* String Functions:: Misc string manipulation
* Sleeping:: Waiting for a period of time
* Beeping:: Making a ding! sound
* Messages:: Writing to the console
* Command Line Options:: Retrieving command line arguments
* Shell Commands:: Executing shell commands
* Timers:: Asynchronous timers
* Debugging:: How to debug Lisp programs
* Tips:: General ideas for `librep' programming
This manual still fails to document the following functions:
default-boundp, default-value, recursive-edit, regexp-cache-control,
sdbm-close, sdbm-delete, sdbm-error, sdbm-fetch, sdbm-firstkey,
sdbm-nextkey, sdbm-open, sdbm-rdonly, sdbm-store, sdbmp, set-default,
setq-default,
back to top
Introduction
============
As you have probably gathered by now, `librep' provides a dialect of
the Lisp programming language--a dialect originally inspired by Emacs
Lisp, but later adapted to include many features from various Scheme
implementations and Common Lisp. The language dialect aims to be
convenient for both extending applications and writing large
stand-alone programs.
All programs written using only the information in this manual should
be compatible with future revisions of `librep'.
This following sections explain some of the most important Lisp
concepts and the conventions I've used in this manual.
* Menu:
* nil and t:: Boolean values in Lisp
* The Lisp Reader:: Basic program structure
* Notation:: Special glyphs used
* Descriptions:: How functions and variables are documented
back to top
nil and t
---------
In the rep Lisp dialect there is a single data value representing
boolean "false"--the empty list, written as `()'. All other values are
considered "not-false", i.e. "true".
By convention the constants `nil' and `t' are used to represent the
canonical boolean values. The constant variable `nil' evaluates to the
empty list (i.e. "false"), while `t' evaluates to itself (i.e.
not-"false", therefore "true").
Reiterating, all of the conditional operations regard _anything_
which is not `()' as being true (i.e. non-false). The actual symbol `t'
should be used where a true boolean value is explicitly stated, to
increase the clarity of the code.
So, `()', and its alias `nil', represent both the empty list and
boolean falsehood. Most Lisp programmers write `()' where its value as
a list should be emphasized, and `nil' where its value as boolean false
is intended. Although neither of these values need be quoted (*note
Quoting::), most programmers will quote the empty list to emphasize
that it is a constant value. However `nil' should not be quoted, doing
so would produce the _symbol_ `nil', not boolean falsehood. For example:
(append '() '()) => () ;Emphasize use of empty lists
(not nil) => t ;Emphasize use as boolean false
(get 'nil 'color) ;Use the symbol `nil'
When a function is said to "return false", it means that it returns
the false boolean value, i.e. the empty list. When a function is said
to "return true", this means that any non-false value is returned.
back to top
The Lisp Reader
---------------
Lisp programs and functions are stored internally as Lisp data
objects, the Lisp Reader is the mechanism that translates from textual
descriptions of Lisp objects to the internal data structures
representing them.
The Lisp Reader is the collection of internal functions accessed by
the `read' Lisp function. It reads a character at a time from an input
stream until a whole Lisp object has been parsed.
*Note Data Types::.
back to top
Notation
--------
Wherever an example of evaluating a Lisp form is shown it will be
formatted like this,
(+ 1 2)
=> 3
The glyph `=>' is used to show the computed value of a form. (1)
When two forms are shown as being exactly equivalent to one another
the glyph `==' is used, for example,
(car some-variable) == (nth 0 some-variable)
Evaluating some forms result in an error being signalled, this is
denoted by the `error-->' glyph.
(open-file "/tmp/foo" 'read)
error--> File error: No such file or directory, /tmp/foo
---------- Footnotes ----------
(1) In this case the list `(+ 1 2)' (i.e. the list containing three
elements, the symbol `+' and, the numbers 1 and 2), represents a
function application. The first element in the list is the name of the
function to be called, all other elements are the arguments to apply to
it. Since the `+' function adds a series of numbers, the above function
call is actually performing the computation `1 + 2'.
back to top
Descriptions
------------
In this document the simplest type of descriptions are those defining
variables (*note Variables::), they look something like:
- Variable: grains-of-sand
This imaginary variable contains the number of grains of sand in a
one-mile long stretch of an averagely sandy beach.
Hooks (*note Hooks::) are also described in this format, the only
difference is that `Variable:' is replaced by `Hook:'.
Functions (*note Functions::) and macros (*note Macros::) have more
complex descriptions; as well as the name of the object being
described, they also have a list of parameters which the object will
accept. Each parameter in the list is named and may be referred to in
the body of the description.
Three keyword parameters may also be used: `#!optional', `#!key' and
`#!rest'. They have the same meaning as when used in the lambda-list of
a function definition (*note Lambda Expressions::). That is,
`#!optional' means that all further parameters are optional, and
`#!rest' means that the following parameter actually receives a list of
any unused argument values.
An example function definition follows.
- Function: useless-function first `#!optional' second `#!rest' tail
This function returns a list consisting of the values SECOND (when
undefined the number 42 is used), all the items in the list TAIL
and FIRST.
(useless-function 'foo 'bar 'xyz 20)
=> (bar xyz 20 foo)
(useless-function '50)
=> (42 50)
Macros and interactive commands are defined in the same way with
`Macro:' or `Command:' replacing `Function:'.
Special forms (*note Special Forms::) are described similarly to
functions except that the argument list is formatted differently, since
special forms are, by definition, more flexible in how they treat their
arguments. Optional values are enclosed in square brackets
(`[OPTIONAL-ARG]') and three dots (`REPEATED-ARG...') indicate where
zero or more arguments are allowed.
back to top
Data Types
==========
The way that data is represented in Lisp is fundamentally different
to languages such as C or Fortran. In Lisp each piece of data (or
"value") has two basic attributes: the data and the _type_ of the data.
This means that type checking is performed at run-time on the actual
data itself, not at compile-time on the "variable" holding the data.
Also, there are no "pointers" in Lisp. As in the Java programming
language, all values are references to data structures, with each
actual data structure (or "Lisp Object") being able to have as many
values referring to it concurrently as necessary. Because of this lack
of pointers, there can be no memory-leakage in Lisp--when an object has
no more extant references, it is automatically deallocated (*note
Garbage Collection::).
Most Lisp objects are a member of one of the primitive types; these
are types built into the Lisp system and can represent things like
strings, numbers, cons cells, vectors, etc... Other primitive types may
be defined at run-time.
More complex objects may be constructed from these primitive types,
for example a vector of three elements could be regarded as a type
`triple' if necessary. In general, each separate type provides a
predicate function which returns true when applied to an object of its
own type.
Finally, one of the most important differences between Lisp and other
languages is that there is no distinction between programs and data.
But this will be explained later.
* Menu:
* Types Summary:: List of the most common types
* Read Syntax:: Some types can be made from source code
* Printed Representation:: All types can be printed
* Equality Predicates:: How to test two objects for equality
* Comparison Predicates:: Comparing two objects as scalars
* Type Predicates:: Each type has a predicate defining it
* Garbage Collection:: Reusing memory from stale objects
back to top
Types Summary
-------------
Each separate data type is documented in its own section, this is a
just a summary of the more common types.
"Numbers"
Numbers: fixnums, bignums, rationals and floats. *Note Numbers::.
"Cons cell"
An object referring to two other Lisp objects. *Note Cons Cells::.
"List"
A sequence of objects, in Lisp lists are not primitive types,
instead they are made by chaining together Cons cells. *Note
Lists::.
"Vector"
A one-dimensional array of objects. *Note Vectors::.
"String"
A vector of characters. *Note Strings::.
"Array"
An ordered sequence of objects which can be accessed in constant
time, either a vector or a string. *Note Sequences::.
"Sequence"
An ordered sequence of objects, either a list or an array. *Note
Sequences::.
"Symbol"
A symbol is a named object; they are used to provide named
variables and functions. *Note Symbols::.
"File"
A link to a notional file in the filing system. This file may be
in the local filing system, or on a FTP server, or wherever. *Note
Files::.
"Process"
An object through which processes may be created and controlled.
*Note Processes::.
"Stream"
Serial data sinks and sources. These may include files, functions,
and processes. *Note Streams::.
"Void"
The empty type, only used in symbols to represent an undefined
value. Note that this is not the same as `()', which is the empty
list, or false truth value.
back to top
Read Syntax
-----------
As previously noted the Lisp reader translates textual descriptions
of Lisp objects into the object they describe (source files are simply
descriptions of objects). However, not all data types can be created in
this way: in fact the only types which can are numbers, strings,
symbols, cons cells (or lists) and vectors, all others have to be
created by calling functions.
Single line comments are introduced by a semi-colon character (`;').
Whenever the Lisp reader encounters a semi-colon where it's looking for
the read syntax of a new Lisp object it will discard the rest of the
line of input. Block comments are also supported, introduced by the
string `#|' and terminated by `|#'. *Note Comment Styles::.
The "read syntax" of an object is the string which when given to the
reader as input will produce the object. The read syntax of each type
of object is documented in that type's main section of this manual but
here is a small summary of how to write each type.
Numbers
A number is number written as an integer--decimal, octal (when the
number is preceded by `#o') or hexadecimal (when the number is
preceded by `#x')--or a decimal rational or floating point value.
An optional minus sign may be the first character in a number. Some
examples are,
42
=> 42
#o177
=> 127
#x-ff
=> -255
3/2
=> 3/2
1.23
=> 1.23
Strings
The read syntax of a string is simply the string with a
double-quote character (`"') at each end, for more details see
*Note Strings::.
"This is a string"
Cons cells
A cons cell is written in what is known as "dotted pair notation",
an opening left-parenthesis, followed by the read syntax of the
first object, followed by a dot, then the second object, and
finally a closing right-parenthesis. For example:
("car" . "cdr")
Lists
The syntax of a list is similar to a cons cell, but the dot is
removed and zero or more objects may be written:
(0 1 2 3)
("foo" ("bar" "baz") 100)
The second example is a list of three elements, a string, an inner
list and a number.
Vectors
The read syntax of a vector is similar to that of a list, but with
square brackets instead of parentheses,
[0 1 2 3]
Symbols
The read syntax of a symbol is its name, for example the read
syntax of the symbol called `my-symbol' is,
my-symbol
back to top
Printed Representation
----------------------
As well as translating textual descriptions to Lisp objects, the
process may be reversed, converting a value back to a textual
description. The resulting text is known as the "printed
representation" of the object, and will usually be very similar to the
read syntax of the object (*note Read Syntax::).
Objects which do not have a read syntax _do_ have a printed
representation, it will normally be of the form,
#
where the RELEVANT-TEXT is object-dependent and usually describes the
object and its contents. The reader will signal an error if it
encounters a description of an object in the format `#<...>'.
back to top
Equality Predicates
-------------------
- Function: eq arg1 arg2
Returns true when ARG1 and ARG2 refer to the same object. Two
objects are the same when they occupy the same place in memory and
hence modifying one object would alter the other. The following
Lisp fragments may illustrate this,
(eq "foo" "foo") ;the objects are distinct
=> ()
(eq t t) ;the same object -- the symbol `t'
=> t
Note that the result of `eq' is undefined when called on two
integer objects with the same value, see `eql'.
- Function: equal arg1 arg2
The function `equal' compares the structure of the two objects
ARG1 and ARG2. If they are considered to be equivalent then
returns true, otherwise returns false.
(equal "foo" "foo")
=> t
(equal 42 42)
=> t
(equal 42 0)
=> ()
(equal '(x . y) '(x . y))
=> t
- Function: eql arg1 arg2
This function is a cross between `eq' and `equal': if ARG1 and
ARG2 are both numbers then the value of these numbers are
compared. Otherwise it behaves in exactly the same manner as `eq'
does.
(eql 3 3)
=> t
(eql 1 2)
=> ()
(eql "foo" "foo")
=> ()
(eql 'x 'x)
=> t
back to top
Comparison Predicates
---------------------
These functions compare their two arguments in a scalar fashion, the
arguments may be of any type but the results are only meaningful for
numbers, strings (ASCII values of each byte compared until a
non-matching pair is found then those two values are compared as
numbers) and cons cells (cars compared before cdrs).
Unlike the `eql' function, inexact and exact numbers will be
compared by first coercing the exact number to be inexact.
- Function: = arg1 arg2 arg3 ... argn
Returns true if all arguments represent the same value.
- Function: /= arg1 arg2 arg3 ... argn
Returns true if no two arguments represent the same value.
- Function: > arg1 arg2 arg3 ... argn
Returns true when ARG1 is `greater than' ARG2, and ARG2 is greater
than ARG3, and so on, upto ARGN.
- Function: >= arg1 arg2 arg3 ... argn
Similar to `>', but for the "greater than or equal to" relation.
- Function: < arg1 arg2 arg3 ... argn
Similar to `>', but for the "less than" relation.
- Function: <= arg1 arg2 arg3 ... argn
Similar to `>', but for the "less than or equal to" relation.
There are two related functions for finding the maximum or minimum
of a sequence of values.
- Function: max #!rest args
Return the maximum value from the list of ARGS. When comparing
numbers, any inexact arguments cause the result to be inexact.
- Function: min #!rest args
Return the minimum value from the list of ARGS. When comparing
numbers, any inexact arguments cause the result to be inexact.
back to top
Type Predicates
---------------
Each type has a corresponding predicate which defines the objects
which are members of that type. Each predicate function has a single
parameter, if that parameter is of the correct type it returns true.
`integerp', `numberp', `null', `consp', `listp', `vectorp', `subrp',
`functionp', `sequencep', `stringp', `symbolp', `processp', `filep'.
The documentation for these functions is with the documentation for
the relevant type.
back to top
Garbage Collection
------------------
In Lisp, data objects are used very freely; a side effect of this is
that it is not possible to (easily) know when an object is "stale",
that is, no references to it exist and it can therefore be reused.
The "garbage collector" is used to overcome this problem; whenever
enough memory has been allocated to make it worthwhile, evaluation
stops and the garbage collector works its way through memory deciding
which objects may still be referenced, and which are stale. The stale
objects are then recorded as being available for reuse and evaluation
continues. (But *note Guardians::)
- Function: garbage-collect
Runs the garbage collector, usually this function doesn't need to
be called manually.
- Variable: garbage-threshold
The number of bytes of data that must have been allocated since the
last garbage collection before evaluation pauses and the garbage
collector is invoked. Its default value is about 100K.
- Variable: idle-garbage-threshold
When the input loop is idle (due to a lack of input), this is the
number of bytes of data that must have been allocated since the
garbage collection, for another collection to be triggered.
This is usually set to a lot less than `garbage-threshold' since
the small delay caused by garbage collection is unnoticeable if the
system is already idle.
- Variable: after-gc-hook
A hook (*note Normal Hooks::) called immediately after each
invocation of the garbage collector.
back to top
Numbers
=======
`Librep' distinguishes between numbers that are represented exactly
and numbers that may not be. This is similar to the Scheme dialect of
Lisp. Quoting from the Scheme standard:
... numbers are either _exact_ or _inexact_. A number is exact if
it was written as an exact constant or was derived from exact
numbers using only exact operations. A number is inexact if it was
written as an inexact constant, if it was derived using inexact
ingredients, or if it was derived using inexact operations. Thus
inexactness is a contagious property of a number.
Exact numbers include both integers and rational numbers, there is no
theoretical limit to the range of the values that may be represented
(1). Inexact numbers are currently implemented using double precision
floating point values.
The read syntax of any number is: `[PFX...][SGN]DATA...', where the
optional SGN is one of the characters `-' or `+', DATA is the
representation of the number, and PFX is zero or more of the following
prefix strings:
`#b'
`#B'
Integers are described in binary,
`#o'
`#O'
Integers are in octal,
`#d'
`#D'
Integers are in decimal (the default),
`#x'
`#X'
Integers are in hexadecimal,
`#e'
`#E'
Coerce the number to an exact representation after parsing it,
`#i'
`#I'
Coerce to an inexact representation.
The representation of an integer is simply the digits representing that
integer, in the radix chosen by any given prefix (defaults to decimal).
Examples of valid integer read syntaxes for the number 42 could be
`42', `#x2a', `#o52', `#o+52', ...
The representation of a rational number is two sequences of digits,
separated by a `/' character. For example, `3/2' represents the
rational number three divided by two.
Inexact numbers are parsed from one of two representations: decimal
point form, which is simply a decimal number containing a decimal
point, and exponential form, which is a decimal number followed by the
letter `e' and a decimal exponent multiplying the first part of the
number by that power of ten. For example, `10.0', `10.' and `1e1' all
read as the inexact number ten. Note that the radix prefixes currently
have no effect when parsing inexact numbers, decimal is used
exclusively.
An integer's printed representation is simply the number printed in
decimal with a preceding minus sign if it is negative. Rational numbers
are printed as two integers separated by a `/' character. Inexact
numbers are printed in their decimal form.
- Function: numberp object
Returns true if OBJECT is a number.
* Menu:
* Arithmetic Functions::
* Integer Functions::
* Rational Functions::
* Real Number Functions::
* Mathematical Functions::
* Bitwise Functions::
* Numeric Predicates::
* Random Numbers::
* Characters::
---------- Footnotes ----------
(1) However, depending on implementation restrictions, very large
integers may be coerced to an inexact representation.
back to top
Arithmetic Functions
--------------------
There are a number of functions which perform arithmetic operations
on numbers, they take a varying number of values as their arguments
returning a new number as their result. When given only exact
arguments, an exact result will be returned.
- Function: + number1 #!rest numbers
This functions adds its arguments then returns their sum.
- Function: - number1 #!rest numbers
If this function is just given one argument (NUMBER1) that number
is negated and returned. Otherwise each of NUMBERS is subtracted
from a running total starting with the value of NUMBER1.
(- 20)
=> -20
(- 20 10 5)
=> 5
- Function: * number1 #!rest numbers
This function multiplies its arguments then returns the result.
- Function: / number1 #!rest numbers
This function performs division, a running-total (initialised from
NUMBER1 is successively divided by each of NUMBERS then the result
is returned.
(/ 100 2)
=> 50
(/ 200 2 5)
=> 20
(/ 3 2)
=> 3/2
(/ 3.0 2)
=> 1.5
- Function: 1+ number
This function returns the result of adding one to NUMBER.
(1+ 42)
=> 43
- Function: 1- number
Returns NUMBER minus one.
back to top
Integer Functions
-----------------
The functions described in this section all operate on, and return,
integer values.
- Function: quotient dividend divisor
Return the integer part of dividing DIVIDEND by DIVISOR.
- Function: remainder dividend divisor
Returns the integer remainder from dividing the DIVIDEND by
DIVISOR. The remainder is either zero or has the same sign as
DIVIDEND.
- Function: modulo dividend divisor
- Function: mod dividend divisor
Return the value of DIVIDEND modulo DIVISOR. Unlike the
`remainder' function the `modulo' function always has the sign of
the DIVISOR, not of the DIVIDEND
- Function: gcd args...
Returns the greatest common divisor of the integers ARGS... If no
arguments are given, returns zero.
- Function: lcm args...
Return the lowest common multiple of the integers ARGS... If no
arguments are given, returns one.
back to top
Rational Functions
------------------
These functions operate on rational numbers.
- Function: numerator x
Returns the exact numerator of X.
- Function: denominator x
Returns the exact denominator of X.
- Function: exact->inexact x
Returns an inexact version of rational number X.
back to top
Real Number Functions
---------------------
- Function: abs x
Returns the magnitude of X.
- Function: floor x
Round X downwards to the nearest integer less than or equal to X.
- Function: ceiling x
Round X upwards to the nearest integer less than or equal to X.
- Function: truncate x
Round X to the nearest integer between X and zero.
- Function: round x
Round X to the nearest integer. Halfway cases are rounded to the
nearest even integer.
- Function: inexact->exact x
Returns an exact representation of X. This may involve a loss of
accuracy.
back to top
Mathematical Functions
----------------------
- Function: exp x
Return `e' (the base of natural logarithms) raised to the power X.
- Function: log x
Return the natural logarithm of X. An arithmetic error is
signalled if X is less than zero.
- Function: sin x
Return the sine of angle X; x is in terms of radians.
- Function: cos x
Return the cosine of angle X.
- Function: tan x
Return the tangent of angle X.
- Function: asin x
Return the arc sine of X (the value whose sine is X), in radians.
- Function: acos x
Return the arc cosine of X.
- Function: atan x
Return the arc tangent of X.
- Function: sqrt x
Return the non-negative square root of X. Currently, if X is
negative, an arithmetic error is signalled.
- Function: expt x y
Returns X raised to the power Y.
If X is negative and Y is a non-integer, then an arithmetic error
is signalled (mathematically should return a complex number).
back to top
Bitwise Functions
-----------------
These functions operate on the bit string which an integer
represents, assuming a two's complement representation.
- Function: lsh number count
This function shifts the integer NUMBER COUNT bits to the left, if
COUNT is negative NUMBER is shifted to the right instead.
(lsh 1 8)
=> 256
(lsh 256 -8)
=> 1
- Function: logand number1 #!rest numbers
This function uses a bit-wise logical `and' operation to combine
all its arguments (there must be at least one argument).
(logand 15 8)
=> 8
(logand 15 7 20)
=> 4
- Function: logior number1 #!rest numbers
Uses a bit-wise logical `inclusive-or' to combine all its
arguments (there must always be at least one argument).
(logior 1 2 4)
=> 7
- Function: logxor number1 #!rest numbers
Uses a bitwise logical `exclusive-or' to combine all its arguments
(there must be at least one).
(logxor 7 3)
=> 4
- Function: lognot number
This function inverts all the bits in NUMBER.
(lognot 0)
=> -1
(lognot 2)
=> -3
(lognot -1)
=> 0
back to top
Numeric Predicates
------------------
For the documentation of the functions `=', `/=', `>', `<', `>=',
`<=', `max' and `min', see *Note Comparison Predicates::.
- Function: exactp object
Returns true when OBJECT is an exact number.
- Function: inexactp object
Returns true when OBJECT is an inexact number.
- Function: integerp object
Returns true when OBJECT is an integer.
- Function: rationalp object
Returns true when OBJECT is a rational number (including integers).
- Function: realp object
Returns true when OBJECT is a real number.
- Function: oddp x
Return true if X is an odd number.
- Function: evenp x
Return true if X is an even number.
- Function: positivep x
Return true if X is a number greater than zero.
- Function: negativep x
Return true if X is a number less than zero.
- Function: zerop x
Returns true if X is equal to zero.
back to top
Pseudo-Random Numbers
---------------------
The `random' function allows pseudo-random numbers to be generated.
- Function: random #!optional limit
Return a pseudo-random number between zero and LIMIT-1 inclusive.
If LIMIT is undefined, it is taken as being the largest positive
integer representable in a fixnum.
Calling `random' with LIMIT equal to the symbol `t' seeds the
generator with the current time of day.
back to top
Characters
----------
In `librep' characters are stored in integers. Their read syntax is
a question mark followed by the character itself, which may be an
escape sequence introduced by a backslash. For details of the available
escape sequences see *Note Strings::.
?a
=> 97
?\n
=> 10
?\177
=> 127
- Function: alpha-char-p character
This function returns true when CHARACTER is one of the alphabetic
characters.
(alpha-char-p ?a)
=> t
- Function: upper-case-p character
When CHARACTER is one of the upper-case characters this function
returns true.
- Function: lower-case-p character
Returns true when CHARACTER is lower-case.
- Function: digit-char-p character
This function returns true when CHARACTER is one of the decimal
digit characters.
- Function: alphanumericp character
This function returns true when CHARACTER is either an alphabetic
character or a decimal digit character.
- Function: space-char-p character
Returns true when CHARACTER is a white-space character (space, tab,
newline or form feed).
- Function: char-upcase character
This function returns the upper-case equivalent of CHARACTER. If
CHARACTER is already upper-case or has no upper-case equivalent it
is returned unchanged.
(char-upcase ?a)
=> 65 ;`A'
(char-upcase ?A)
=> 65 ;`A'
(char-upcase ?!)
=> 33 ;`!'
- Function: char-downcase character
Returns the lower-case equivalent of the character CHARACTER.
back to top
Sequences
=========
Sequences are ordered groups of objects, there are several primitive
types which can be considered sequences, each with their pros and cons.
A sequence is either an array or a list, where an array is either a
vector or a string.
- Function: sequencep object
This function returns true if OBJECT is a sequence.
* Menu:
* Cons Cells:: An ordered pair of two objects
* Lists:: Chains of cons cells
* Vectors:: A chunk of memory holding a number of objects
* Strings:: Strings are efficiently-stored vectors
* Array Functions:: Accessing elements in vectors and strings
* Sequence Functions:: These work on any type of sequence
back to top
Cons Cells
----------
A "cons cell" is an ordered pair of two objects, the "car" and the
"cdr".
The read syntax of a cons cell is an opening parenthesis followed by
the read syntax of the car, a dot, the read syntax of the cdr and a
closing parenthesis. For example a cons cell with a car of 10 and a cdr
of the string `foo' would be written as,
(10 . "foo")
- Function: cons car cdr
This function creates a new cons cell. It will have a car of CAR
and a cdr of CDR.
(cons 10 "foo")
=> (10 . "foo")
- Function: consp object
This function returns true if OBJECT is a cons cell.
(consp '(1 . 2))
=> t
(consp '())
=> ()
(consp (cons 1 2))
=> t
The strange syntax `'(1 . 2)' is known as "quoting" (*note
Quoting::), it tells the evaluator that the object following the
quote-mark is a constant, and therefore should not be evaluated. This
will be explained in more detail later.
In Lisp an "atom" is any object which is not a cons cell (and is,
therefore, atomic).
- Function: atom object
Returns true if OBJECT is an atom (not a cons cell).
Given a cons cell there are a number of operations which can be
performed on it.
- Function: car cons-cell
This function returns the object which is the car (first element)
of the cons cell CONS-CELL.
(car (cons 1 2))
=> 1
(car '(1 . 2))
=> 1
- Function: cdr cons-cell
This function returns the cdr (second element) of the cons cell
CONS-CELL.
(cdr (cons 1 2))
=> 2
(cdr '(1 . 2))
=> 2
- Function: rplaca cons-cell new-car
This function sets the value of the car (first element) in the cons
cell CONS-CELL to NEW-CAR. The value returned is CONS-CELL.
(setq x (cons 1 2))
=> (1 . 2)
(rplaca x 3)
=> (3 . 2)
x
=> (3 . 2)
- Function: rplacd cons-cell new-cdr
This function is similar to `rplacd' except that the cdr slot
(second element) of CONS-CELL is modified.
back to top
Lists
-----
A list is a sequence of zero or more objects, the main difference
between lists and vectors is that lists are more dynamic: they can
change size, be split, reversed, concatenated, etc... very easily.
In Lisp lists are not a primitive type; instead singly-linked lists
are formed by chaining cons cells together (*note Cons Cells::). The
empty list is represented by the special value `()'.
- Function: listp arg
This functions returns true when its argument, ARG, is a list
(i.e. either a cons cell or `()').
- Function: null arg
Returns a true value if ARG is the empty list.
* Menu:
* List Structure:: How lists are built from cons cells
* Building Lists:: Dynamically creating lists
* Accessing List Elements:: Getting at the elements which make the list
* Modifying Lists:: How to alter the contents of a list
* Association Lists:: Lists can represent relations
* Infinite Lists:: Circular data structures in Lisp
back to top
List Structure
..............
Each element in a list is given its own cons cell and stored in the
car of that cell. The list is then constructed by having the cdr of a
cell point to the cons cell containing the next element (and hence the
entire rest of the list). The cdr of the cell containing the last
element in the list is `()'. A list of zero elements is represented by
`()'.
The read syntax of a list is an opening parenthesis, followed by the
read syntax of zero or more space-separated objects, followed by a
closing parenthesis. Alternatively, lists can be constructed `manually'
using dotted-pair notation.
All of the following examples result in the same list of five
elements: the numbers from zero to four.
(0 1 2 3 4)
(0 . (1 . (2 . (3 . (4 . ())))))
(0 1 2 . (3 4))
An easy way to visualise lists and how they are constructed is to see
each cons cell in the list as a separate "box" with pointers to its car
and cdr,
+-----+-----+
| o | o----> cdr
+--|--+-----+
|
--> car
Complex box-diagrams can now be drawn to represent lists. For
example the following diagram represents the list `(1 2 3 4)'.
+-----+-----+ +-----+-----+ +-----+-----+ +-----+-----+
| o | o----> | o | o----> | o | o----> | o | o----> ()
+--|--+-----+ +--|--+-----+ +--|--+-----+ +--|--+-----+
| | | |
--> 1 --> 2 --> 3 --> 4
A more complex example, the list `((1 2) (foo bar))' can be drawn as,
+-----+-----+ +-----+-----+
| o | o---------------------------> | o | o----> ()
+--|--+-----+ +--|--+-----+
| |
+-----+-----+ +-----+-----+ +-----+-----+ +-----+-----+
| o | o----> | o | o----> () | o | o----> | o | o----> ()
+--|--+-----+ +--|--+-----+ +--|--+-----+ +--|--+-----+
| | | |
--> 1 --> 2 --> foo --> bar
back to top
Building Lists
..............
It has already been described how you can create lists using the Lisp
reader; this method does have a drawback though: the list created is
effectively static. If you modify the contents of the list and that
list was created when a function was defined the list will remain
modified for all future invocations of that function. This is not
usually a good idea, consider the following function definition,
(defun bogus-function (x)
"Return a list whose first element is nil and whose second element is X."
(let
((result '(nil nil))) ;Static list which is filled in each time
(rplaca (cdr result) x) ; the function is called
result))
This function does in fact do what its documentation claims, but a
problem arises when it is called more than once,
(setq x (bogus-function 'foo))
=> (nil foo)
(setq y (bogus-function 'bar))
=> (nil bar) ;The first result has been destroyed
x
=> (nil bar) ;See!
This example is totally contrived--no one would ever write a
function like the one in the example but it does demonstrate the need
for a dynamic method of creating lists.
- Function: list #!rest elements
This function creates a list out of its arguments, if zero
arguments are given the empty list, `()', is returned.
(list 1 2 3)
=> (1 2 3)
(list (major-version-number) (minor-version-number))
=> (3 2)
(list)
=> ()
- Function: list* arg1 arg2 ... argn-1 argn
Creates a new list `(ARG1 ARG2 ... ARGN-1 . ARGN)'.
(list* 1 2 '(3 4))
=> (1 2 3 4)
- Function: make-list length #!optional initial-value
This function creates a list LENGTH elements long. If the
INITIAL-VALUE argument is given it defines the value of all
elements in the list, if it is not defined they are all `()'.
(make-list 2)
=> (() ())
(make-list 3 t)
=> (t t t)
(make-list 0)
=> ()
- Function: append #!rest lists
This function creates a new list with the elements of each of its
arguments (which must be lists). Unlike the function `nconc' this
function preserves the structure of all its arguments.
(append '(1 2 3) '(4 5))
=> (1 2 3 4 5)
(append)
=> ()
What actually happens is that all arguments but the last are
copied, then the last argument is linked on to the end of the list
(uncopied).
(setq foo '(1 2))
=> (1 2)
(setq bar '(3 4))
=> (3 4)
(setq baz (append foo bar))
=> (1 2 3 4)
(eq (nthcdr 2 baz) bar)
=> t
The following diagram shows the final state of the three variables
more clearly,
foo--> +-----+-----+ +-----+-----+
| o | o----> | o | |
+--|--+-----+ +--|--+-----+
| |
o--> 1 o--> 2 bar
| | ->
baz--> +--|--+-----+ +--|--+-----+ +-----+-----+ +-----+-----+
| o | o----> | o | o----> | o | o----> | o | |
+-----+-----+ +-----+-----+ +--|--+-----+ +--|--+-----+
| |
--> 3 --> 4
Note how `foo' and the first half of `baz' use the _same_ objects
for their elements--copying a list only copies its cons cells, its
elements are reused. Also note how the variable `bar' actually
references the mid-point of `baz' since the last list in an
`append' call is not copied.
- Function: remove elt list
Return a copy of LIST, with all elements the same as ELT discarded
(using the `equal' function to compare).
- Function: remq elt list
Similar to the `remove' function, except that comparisons are made
using `eq'.
- Function: reverse list
This function returns a new list; it is made from the elements of
the list LIST in reverse order. Note that this function does not
alter its argument.
(reverse '(1 2 3 4))
=> (4 3 2 1)
As a postscript to this section, the function used as an example at
the beginning could now be written as,
(defun not-so-bogus-function (x)
(list nil x))
Also note that the `cons' function can be used to create lists by
hand and to add new elements onto the front of a list. For example:
(setq x (list 1 2 3))
=> (1 2 3)
(setq x (cons 0 x))
=> (0 1 2 3)
back to top
Accessing List Elements
.......................
The most flexible method of accessing an element in a list is via a
combination of the `car' and `cdr' functions. There are other functions
which provide an easier way to get at the elements in a flat list.
These will usually be faster than a string of `car' and `cdr'
operations.
- Function: nth count list
This function returns the element COUNT elements down the list,
therefore to access the first element use a COUNT of zero (or even
better the `car' function). If there are too few elements in the
list and no element number COUNT can be found `()' is returned.
(nth 3 '(0 1 2 3 4 5))
=> 3
(nth 0 '(foo bar)
=> foo
- Function: nthcdr count list
This function takes the cdr of the list LIST COUNT times,
returning the last cdr taken.
(nthcdr 3 '(0 1 2 3 4 5))
=> (3 4 5)
(nthcdr 0 '(foo bar))
=> (foo bar)
- Function: last list
This function returns the last element in the list LIST. If the
list has zero elements `()' is returned.
(last '(1 2 3))
=> 3
(last '())
=> ()
- Function: member object list
This function scans through the list LIST until it finds an element
which is `equal' to OBJECT. The tail of the list (the cons cell
whose car is the matched object) is then returned. If no elements
match OBJECT then the empty list `()' is returned.
(member 'c '(a b c d e))
=> (c d e)
(member 20 '(1 2))
=> ()
- Function: memq object list
This function is similar to `member' except that comparisons are
performed by the `eq' function not `equal'.
back to top
Modifying Lists
...............
The `nthcdr' function can be used in conjunction with the `rplaca'
function to modify an arbitrary element in a list. For example,
(rplaca (nthcdr 2 '(0 1 2 3 4 5)) 'foo)
=> foo
sets the third element of the list `(0 1 2 3 4 5)' to the symbol called
`foo'.
There are also functions which modify the structure of a whole list.
These are called "destructive" operations because they modify the actual
structure of a list--no copy is made. This can lead to unpleasant side
effects if care is not taken.
- Function: nconc #!rest lists
This function is the destructive equivalent of the function
`append', it modifies its arguments so that it can return a list
which is the concatenation of the elements in its arguments lists.
Like all the destructive functions this means that the lists given
as arguments are modified (specifically, the cdr of their last
cons cell is made to point to the next list). This can be seen
with the following example (similar to the example in the `append'
documentation).
(setq foo '(1 2))
=> (1 2)
(setq bar '(3 4))
=> (3 4)
(setq baz (nconc foo bar))
=> (1 2 3 4)
foo
=> (1 2 3 4) ;`foo' has been altered!
(eq (nthcdr 2 baz) bar)
=> t
The following diagram shows the final state of the three variables
more clearly,
foo--> bar-->
baz--> +-----+-----+ +-----+-----+ +-----+-----+ +-----+-----+
| o | o----> | o | o----> | o | o----> | o | |
+--|--+-----+ +--|--+-----+ +--|--+-----+ +--|--+-----+
| | | |
--> 1 --> 2 --> 3 --> 4
- Function: nreverse list
This function rearranges the cons cells constituting the list LIST
so that the elements are in the reverse order to what they were.
(setq foo '(1 2 3))
=> (1 2 3)
(nreverse foo)
=> (3 2 1)
foo
=> (1) ;`foo' wasn't updated when the list
; was altered.
- Function: delete object list
This function destructively removes all elements of the list LIST
which are `equal' to OBJECT then returns the modified list.
(delete 1 '(0 1 0 1 0))
=> (0 0 0)
When this function is used to remove an element from a list which
is stored in a variable that variable must be set to the return
value of the `delete' function. Otherwise, if the first element of
the list has to be deleted (because it is `equal' to OBJECT) the
value of the variable will not change.
(setq foo '(1 2 3))
=> (1 2 3)
(delete 1 foo)
=> (2 3)
foo
=> (1 2 3)
(setq foo (delete 1 foo))
=> (2 3)
- Function: delq object list
This function is similar to the `delete' function, the only
difference is that the `eq' function is used to compare OBJECT
with each of the elements in LIST, instead of the `equal' function
which is used by `delete'.
- Function: sort list #!optional predicate
Destructively sorts (i.e. by modifying cdrs) the list of values
LIST, to satisfy the function PREDICATE, returning the sorted
list. If PREDICATE is undefined, the `<' function is used, sorting
the list into ascending order.
PREDICATE is called with two values, it should return true if the
first is considered less than the second.
(sort '(5 3 7 4))
=> (3 4 5 7)
The sort is stable, in that elements in the list which are equal
will preserve their original positions in relation to each other.
back to top
Association Lists
.................
An "association list" (or "alist") is a list mapping keys to to.
Each element of the alist is a cons cell, the car of which is the
"key", the cdr the value that it associates to. For example an alist
could look like,
((fred . 20)
(bill . 30))
this alist has two keys, `fred' and `bill' which both associate to an
integer (20 and 30 respectively).
It is possible to make the associated values lists, this looks like,
((fred 20 male)
(bill 30 male)
(sue 25 female))
in this alist the symbol `fred' is associated with the list `(20 male)'.
There are a number of functions which let you interrogate an alist
with a given key for its association.
- Function: assoc key alist
This function scans the association list ALIST for the first
element whose car is `equal' to KEY, this element is then
returned. If no match of KEY is found false is returned.
(assoc 'two '((one . 1) (two . 2) (three . 3)))
=> (two . 2)
- Function: assq key alist
Similar to the function `assoc' except that the function `eq' is
used to compare elements instead of `equal'.
It is not usually wise to use `assq' when the keys of the alist
may not be symbols--`eq' won't think two objects are equivalent
unless they are the _same_ object!
(assq "foo" '(("bar" . 1) ("foo" . 2)))
=> ()
(assoc "foo" '(("bar" . 1) ("foo" . 2)))
=> ("foo" . 2)
- Function: rassoc association alist
This function searches through ALIST until it finds an element
whose cdr is `equal' to ASSOCIATION, that element is then returned.
false will be returned if no elements match.
(rassoc 2 '((one . 1) (two . 2) (three . 3)))
=> (two . 2)
- Function: rassq association alist
This function is equivalent to `rassoc' except that it uses `eq'
to make comparisons.
back to top
Infinite Lists
..............
Sometimes it is useful to be able to create `infinite' lists--that
is, lists which appear to have no last element--this can easily be done
in Lisp by linking the cdr of the last cons cell in the list structure
back to the beginning of the list.
-----------------------------------
| |
--> +-----+-----+ +-----+-----+ |
| o | o----> | o | o-----
+--|--+-----+ +--|--+-----+
| |
--> 1 --> 2
The diagram above represents the infinite list `(1 2 1 2 1 2 ...)'.
Infinite lists have a major drawback though, many of the standard
list manipulation functions can not be used on them. These functions
work by moving through the list until they reach the end. If the list
has _no_ end the function may never terminate and the only option is to
send the interpreter an interrupt signal.
The only functions which may be used on circular lists are: the cons
cell primitives (`cons', `car', `cdr', `rplaca', `rplacd'), `nth' and
`nthcdr'.
Also note that infinite lists can't be printed. But note the
`print-length' and `print-level' variables, see *Note Output
Functions::.
back to top
Vectors
-------
A vector is a fixed-size sequence of Lisp objects, each element may
be accessed in constant time--unlike lists where the time taken to
access an element is proportional to the position of the element.
The read syntax of a vector is an opening square bracket, followed
by zero or more space-separated objects, followed by a closing square
bracket. For example,
[zero one two three]
In general it is best to use vectors when the number of elements to
be stored is known and lists when the sequence may grow or shrink.
- Function: vectorp object
This function returns true if its argument, OBJECT, is a vector.
- Function: vector #!rest elements
This function creates a new vector containing the arguments given
to the function.
(vector 1 2 3)
=> [1 2 3]
(vector)
=> []
- Function: make-vector size #!optional initial-value
Returns a new vector, SIZE elements big. If INITIAL-VALUE is
defined each element of the new vector is set to INITIAL-VALUE,
otherwise they are all `()'.
(make-vector 4)
=> [() () () ()]
(make-vector 2 t)
=> [t t]
back to top
Strings
-------
A string is a vector of characters (*note Characters::), they are
generally used for storing and manipulating pieces of text. `librep'
puts no restrictions on the values which may be stored in a
string--specifically, the null character (`^@') may be stored with no
problems.
The read syntax of a string is a double quote character, followed by
the contents of the string, the object is terminated by a second double
quote character. For example, `"abc"' is the read syntax of the string
`abc'.
Any backslash characters in the string's read syntax introduce an
escape sequence; one or more of the following characters are treated
specially to produce the next _actual_ character in the string.
The following escape sequences are supported (all are shown without
their leading backslash `\' character).
`n'
A newline character.
`r'
A carriage return character.
`f'
A form feed character.
`t'
A TAB character.
`a'
A `bell' character (this is Ctrl-g).
`\'
A backslash character.
`^C'
The `control' code of the character C. This is calculated by
toggling the seventh bit of the _upper-case_ version of C.
For example,
\^C ;A Ctrl-c character (ASCII value 3)
\^@ ;The NUL character (ASCII value 0)
`012'
The character whose ASCII value is the octal value `012'. After the
backslash character the Lisp reader reads up to three octal digits
and combines them into one character.
`x12'
The character whose ASCII value is the hexadecimal value `12', i.e.
an `x' character followed by one or two hex digits.
- Function: stringp object
This function returns true if its argument is a string.
- Function: make-string length #!optional initial-character
Creates a new string containing LENGTH characters, each character
is initialised to INITIAL-CHARACTER (or to spaces if
INITIAL-CHARACTER is not defined).
(make-string 3)
=> " "
(make-string 2 ?$)
=> "$$"
- Function: concat #!rest args
This function concatenates all of its arguments, ARGS, into a
single string which is returned. If no arguments are given then
the null string (`') results.
Each of the ARGS may be a string, a character or a list or vector
of characters. Characters are stored in strings modulo 256.
(concat "foo" "bar")
=> "foobar"
(concat "a" ?b)
=> "ab"
(concat "foo" [?b ?a ?r])
=> "foobar"
(concat)
=> ""
- Function: substring string start #!optional end
This function creates a new string which is a partial copy of the
string STRING. The first character copied is START characters from
the beginning of the string. If the END argument is defined it is
the index of the character to stop copying at, if it is not defined
all characters until the end of the string are copied.
(substring "xxyfoozwx" 3 6)
=> "foo"
(substring "xyzfoobar" 3)
=> "foobar"
- Function: string= string1 string2
This function compares the two strings STRING1 and STRING2--if
they are made from the same characters in the same order then true
is returned.
(string= "one" "one")
=> t
(string= "one" "two")
=> ()
Note that an alternate way to compare strings (or anything!) is to
use the `equal' function.
- Function: string-equal string1 string2
Returns true if STRING1 and STRING2 are the same, ignoring
differences in character case.
- Function: string< string1 string2
This function returns true if STRING1 is `less' than `string2'.
This is determined by comparing the two strings a character at a
time, the first pair of characters which do not match each other
are then compared with a normal `less-than' function.
In `librep' the standard `<' function understands strings so
`string<' is just a macro calling that function.
(string< "abc" "abd")
=> t
(string< "abc" "abb")
=> ()
- Function: string-lessp string1 string2
Similar to `string<' but ignores character case in comparisons.
See *Note String Functions:: for a few more string manipulating
functions, and *Note Regular Expressions:: for a method of pattern
matching in strings.
back to top
Array Functions
---------------
- Function: arrayp object
This function returns true if OBJECT is an array.
- Function: aref array position
Returns the element of the array (vector or string) ARRAY POSITION
elements from the first element (i.e. the first element is
numbered zero). If no element exists at POSITION in ARRAY, false
is returned.
(aref [0 1 2 3] 2)
=> 2
(aref "abcdef" 3)
=> 100 ;`d'
- Function: aset array position value
This function sets the element of the array ARRAY with an index of
POSITION (counting from zero) to VALUE. An error is signalled if
element POSITION does not exist. The result of the function is
VALUE.
(setq x [0 1 2 3])
=> [0 1 2 3]
(aset x 2 'foo)
=> foo
x
=> [0 1 foo 3]
back to top
Sequence Functions
------------------
- Function: sequencep arg
Returns true if ARG is a sequence, i.e. a list or an array.
- Function: length sequence
This function returns the length (an integer) of the sequence
SEQUENCE.
(length "abc")
=> 3
(length '(1 2 3 4))
=> 4
(length [x y])
=> 2
- Function: copy-sequence sequence
Returns a new copy of the sequence SEQUENCE. Where possible (in
lists and vectors) only the `structure' of the sequence is newly
allocated: the same objects are used for the elements in both
sequences.
(copy-sequence "xy")
=> "xy"
(setq x '("one" "two"))
=> ("one" "two")
(setq y (copy-sequence x))
=> ("one" "two")
(eq x y)
=> ()
(eq (car x) (car y))
=> t
- Function: elt sequence position
This function returns the element of SEQUENCE POSITION elements
from the beginning of the sequence.
This function is a combination of the `nth' and `aref' functions.
(elt [0 1 2 3] 1)
=> 1
(elt '(foo bar) 0)
=> foo
back to top
Symbols
=======
Symbols are objects with a name (almost always a unique name). They
are one of the most important data types in Lisp since they are used to
provided named variables (*note Variables::) and functions (*note
Functions::).
- Function: symbolp arg
This function returns true when its argument is a symbol.
* Menu:
* Symbol Syntax:: The read syntax of symbols
* Symbol Attributes:: The objects stored in a symbol
* Obarrays:: Vectors used to store symbols
* Creating Symbols:: Allocating new symbols
* Interning:: Putting a symbol into an obarray
* Property Lists:: Each symbol has a set of properties
* Keyword Symbols:: Self-evaluating keywords
back to top
Symbol Syntax
-------------
The read syntax of a symbol is usually its name; however, if the name
contains any meta-characters (whitespace or any from `()[]'";|\') they
will have to be entered specially. There are two ways to tell the
reader that a meta-character is actually part of the symbol's name:
1. Precede the meta-character by a backslash character (`\'), for
example:
xy\(z\) ;the symbol whose name is `xy(z)'
2. Enclose part of the name in vertical bars (two `|' characters).
All characters after the starting vertical line are copied as-is
until the closing vertical line is encountered. For example:
xy|(z)| ;the symbol `xy(z)'
Here are some example read syntaxes.
setq ; `setq'
|setq| ; `setq'
\s\e\t\q ; `setq'
1 ; the _number_ 1
\1 ; the _symbol_ `1'
|!$%zf78&| ; `!$%zf78&'
foo|(bar)| ; `foo(bar)'
foo\(bar\) ; `foo(bar)'
back to top
Symbol Attributes
-----------------
All symbols have two basic attributes: print name and property list.
Most important is the "print name" of the symbol. This is a string
naming the symbol, after it has been defined (when the symbol is first
created) it may not be changed.
- Function: symbol-name symbol
This function returns the print name of the symbol SYMBOL.
(symbol-name 'unwind-protect)
=> "unwind-protect"
The symbol's "property list" (or plist) is similar to an alist
(*note Association Lists::), though stored differently, and provides a
method of storing arbitrary extra values in each symbol. *Note Property
Lists::.
Although not strictly an attribute of the symbol, symbols also
provide a means of associating values with names (i.e. variables).
Within a defined context, a symbol may have a "binding", this binding
associates the symbol with a memory location within which a value may
be stored. When writing Lisp programs, the value of a symbol's current
binding is accessed by writing the print name of the symbol. Similarly
the binding may be modified by using the `setq' special form. *Note
Variables::.
back to top
Obarrays
--------
An "obarray" is the structure used to ensure that no two symbols
have the same name and to provide quick access to a symbol given its
name. An obarray is a vector, each element of the vector is a chain of
symbols whose names share the same hash-code (a "bucket"). These
symbols are chained together through links which are invisible to Lisp
programs: if you examine an obarray you will see that each bucket looks
as though it has at most one symbol stored in it.
The normal way to reference a symbol is simply to type its name in
the program, when the Lisp reader encounters a name of a symbol it looks
in the default obarray for a symbol of that name. If the named symbol
doesn't exist it is created and hashed into the obarray--this process
is known as "interning" the symbol, for more details see *Note
Interning::.
- Variable: obarray
This variable contains the obarray that the `read' function uses
when interning symbols.
- Function: make-obarray size
This function creates a new obarray with SIZE hash buckets (this
should probably be a prime number for the fewest hash collisions).
This is the only way of creating an obarray. `make-vector' is _not
suitable_.
- Function: find-symbol symbol-name #!optional obarray
This function scans the specified obarray (OBARRAY or the value of
the variable `obarray' if OBARRAY is undefined) for a symbol whose
name is the string SYMBOL-NAME. The value returned is the symbol
if it can be found or false otherwise.
(find-symbol "setq")
=> setq
- Function: apropos regexp #!optional predicate obarray
Returns a list of symbols from the obarray OBARRAY (or the
default) whose print name matches the regular expression REGEXP
(*note Regular Expressions::). If PREDICATE is true, each symbol
which matches REGEXP is applied to the function PREDICATE, if the
value is true it is considered a match.
The PREDICATE argument is useful for restricting matches to a
certain type of symbol, for example only commands.
(apropos "^yank" 'commandp)
=> (yank-rectangle yank yank-to-mouse)
back to top
Creating Symbols
----------------
It is possible to allocate symbols dynamically, this is normally only
necessary when the symbol is to be interned in a non-default obarray or
the symbol is a temporary object which should not be interned (for
example: labels in a compiler).
- Function: make-symbol print-name
This function creates and returns a new, uninterned, symbol whose
print name is the string PRINT-NAME. Its value cell is void
(undefined) and it will have an empty property list.
(make-symbol "foo")
=> foo
- Function: gensym
This function returns a new, uninterned, symbol that has a unique
print name.
(gensym)
=> G0001
(gensym)
=> G0002
back to top
Interning
---------
"Interning" a symbol means to store it in an obarray so that it can
be found in the future: all variables and named-functions are found
through interned symbols.
When a symbol is interned a hash function is applied to its print
name to determine which bucket in the obarray it should be stored in.
Then it is simply pushed onto the front of that bucket's chain of
symbols.
Normally all interning is done automatically by the Lisp reader. When
it encounters the name of a symbol which it can't find in the default
obarray (the value of the variable `obarray') it creates a new symbol
of that name and interns it. This means that no two symbols can have
the same print name, and that the read syntax of a particular symbol
always produces the same object (unless the value of `obarray' is
altered).
(eq 'some-symbol 'some-symbol)
=> t
- Function: intern symbol-name #!optional obarray
This function uses `find-symbol' to search the OBARRAY (or the
standard obarray) for a symbol called SYMBOL-NAME. If a symbol of
that name is found it is returned, otherwise a new symbol of that
name is created, interned into the obarray, and returned.
(intern "setq")
=> setq
(intern "my-symbol" my-obarray)
=> my-symbol
- Function: intern-symbol symbol #!optional obarray
Interns the symbol SYMBOL into the obarray OBARRAY (or the
standard one) then returns the symbol. If SYMBOL is currently
interned in an obarray an error is signalled.
(intern-symbol (make-symbol "foo"))
=> foo
(intern-symbol 'foo)
error--> Error: Symbol is already interned, foo
- Function: unintern symbol #!optional obarray
This function removes the symbol SYMBOL from the obarray OBARRAY
then returns the symbol.
Beware! this function should be used with _extreme_ caution--once
you unintern a symbol there may be no way to recover it.
(unintern 'setq) ;This is extremely stupid
=> setq
back to top
Property Lists
--------------
Each symbol has a property list (or "plist"), this is a structure
which associates an arbitrary Lisp object with a key (usually a
symbol). The keys in a plist may not have any duplications (so that
each property is only defined once).
The concept of a property list is very similar to an association list
(*note Association Lists::) but there are two main differences:
1. Structure; each element of an alist represents one key/association
pair. In a plist each pair of elements represents an association:
the first is the key, the second the property. For example, where
an alist may be,
((one . 1) (two . 2) (three . 3))
a property list would be,
(one 1 two 2 three 3)
2. Plists have their own set of functions to modify the list. This is
done destructively, altering the property list (since the plist is
stored in only one location, the symbol, this is quite safe).
- Function: get symbol property
This function searches the property list of the symbol SYMBOL for
a property `equal' to PROPERTY. If such a property is found it is
returned, otherwise false is returned.
(get 'if 'lisp-indent)
=> 2
(get 'set 'lisp-indent)
=> ()
- Function: put symbol property new-value
`put' sets the value of the property PROPERTY to NEW-VALUE in the
property list of the symbol SYMBOL. If there is an existing value
for this property (using `equal' to compare keys) it is
overwritten. The value returned is NEW-VALUE.
(put 'foo 'prop 200)
=> 200
- Function: symbol-plist symbol
Returns the property list of the symbol SYMBOL.
(symbol-plist 'if)
=> (lisp-indent 2)
- Function: setplist symbol plist
This function sets the property list of the symbol SYMBOL to PLIST.
(setplist 'foo '(zombie yes))
=> (zombie yes)
back to top
Keyword Symbols
---------------
Keywords are a special class of symbols. They evaluate to themselves,
and have the read syntax `#:SYMBOL', where SYMBOL is anything
satisfying the usual symbol syntax. These objects are normally used to
mark keyword parameters in function applications (*note Lambda
Expressions::).
- Function: make-keyword symbol
Return the keyword symbol that could be used to mark an argument
value for the keyword parameter SYMBOL.
(make-keyword 'x)
=> #:x
- Function: keywordp arg
Returns true if ARG is a keyword symbol.
back to top
Evaluation
==========
So far only the primitive data types have been discussed, and how the
Lisp reader converts textual descriptions of these types into Lisp
objects. Obviously there has to be a way of actually computing
something--it would be difficult to write a useful program otherwise.
What sets Lisp apart from other languages is that in Lisp there is no
difference between programs and data: a Lisp program is just a sequence
of Lisp objects which will be evaluated as a program when required.
The subsystem which does this evaluation is called the "Lisp
evaluator" and each expression to be evaluated is called a "form". The
evaluator (the function `eval') examines the structure of the form that
is applied to it and computes the value of that form within the current
Lisp environment.
A form can be any type of data object; the only types which the
evaluator treats specially are symbols (which describe variables) and
lists (subroutine applications), anything else is returned as-is (and
is called a "self-evaluating form").
- Function: eval form
This function computes and returns the value of FORM within the
current module and dynamic environment, and a null lexical
environment.
However, `eval' is rarely explicitly invoked, except in the
read-eval-print loop. Lisp provides many other methods of evaluation
that are usually much more suitable within a program.
- Variable: max-lisp-depth
This variable limits the number of nested calls to `eval'. If more
than this many nested calls to `eval' exist, an error is
signalled. The intention is to detect infinite recursion before
hitting the stack size limit (causing a segmentation fault).
* Menu:
* Symbol Forms:: How variables are accessed
* List Forms:: Subroutine calls
* Self-Evaluating Forms:: Forms which don't get evaluated
* Quoting:: How to prevent evaluation of forms
back to top
Symbol Forms
------------
When the evaluator is applied to a symbol the computed value of the
form is the value associated with the symbol in the current
environment. Basically this means that to get the value of a variable
you simply write its name. For example,
rep-version
=> "1.0"
this extract from a Lisp session shows the read syntax of a form to get
the value of the variable `rep-version' and the result when this form
is evaluated.
Since forms are evaluated within the current environment the value
of a variable is its most-recent extant binding (with slight
differences for lexical and special variables). *Note Variables::.
If an evaluated symbol has no current binding, an error is signalled.
back to top
List Forms
----------
Forms which are lists are used to invoke a subroutine. The first
element of the list defines the subroutine to be called; all further
elements are arguments to be applied to that subroutine invocation.
There are several different types of subroutines available:
functions, macros, special forms and autoloads. When the evaluator
finds a form which is a list it tries to classify the form into one of
these four types.
First of all it evaluates the first element of the list; the computed
value of this element decides how the rest of the elements in the list
are treated. For example, if the first element is a symbol whose value
is a function, then that function is called with the other values in
the list.
* Menu:
* Function Call Forms:: `Normal' subroutines
* Macro Call Forms:: Source code expansions
* Special Forms:: Abnormal control structures
* Autoload Forms:: Loading subroutines from files on the fly
back to top
Function Call Forms
...................
When the first element of a list form evaluates to a function object
(either a primitive subroutine or a closure), all other elements in the
list are evaluated sequentially from left-to-right, then these values
are applied to the function definition. The result returned by the
function is then taken as the value of the whole list form.
For example, consider the form `(/ 100 (1+ 4))'. This is a function
call to the function stored in the variable `/'. First the `/' form is
evaluated, it is a variable containing a data value representing the
primitive subroutine for integer division. Then the `100' form is
evaluated: it is a number, so self-evaluates to the value `100'. Next
the form `(1+ 4)' is evaluated. This is also a function call and
computes to a value of `5' which becomes the second argument to the `/'
function. Now the `/' function is applied to its evaluated arguments of
`100' and `5', and returns the value `20'. This then becomes the value
of the form `(/ 100 (1+ 4))'.
(/ 100 (1+ 4))
== (/ 100 5)
=> 20
Or another example,
(+ (- 10 (1- 7)) (* (1+ 2) 4)
== (+ (- 10 6) (* (1+ 2) 4)
== (+ 4 (* (1+ 2) 4)
== (+ 4 (* 3 4))
== (+ 4 12)
=> 16
The system is also capable of eliminating tail calls where possible,
allowing tail-recursive function definitions to run with bounded space
requirements.
A "tail-call" is a function call that occurs immediately before
exiting the containing function. Since the containing function need not
receive the result of the function call, it is possible to, in effect,
exit from the containing function before invoking the called function.
Note however, that this is only possible where none of the dynamic
features of the language (i.e. bindings to special variables,
`unwind-protect', `condition-case', `catch', etc...) are currently
active in the containing function.
Consider, for example, the following function:
(defun print-list (l)
(unless (null l)
(format standard-output "%s\n" (car l))
(print-list (cdr l))))
the call to `print-list' occurs in the "tail-position" of the function.
This means that the call may be made after removing the previous call
to `print-list' from the interpreter's stack of active functions.
[ XXX currently the interpreter is incapable of eliminating tail
calls to subrs, i.e. Lisp functions implemented in C ]
back to top
Macro Call Forms
................
Macros are source code expansions, the general idea is that a macro
is a function which using the unevaluated arguments applied to it,
computes another form (the expansion of the macro and its arguments)
which is then evaluated to provide the value of the form.
Macros are generally used to implement control-flow operations, where
not all arguments may be evaluated, or evaluated in an unusual order.
For more details see *Note Macros::.
back to top
Special Forms
.............
Special forms are built-in subroutines which the evaluator knows must
be handled specially. The main difference between a special form and a
function is that the arguments applied to a special form are _not_
automatically evaluated--if necessary the special form will evaluate
arguments itself. This will be noted in the documentation of the
special form.
Special forms are generally used to provide control structures, for
example, the primitive conditional constructs are special forms (if all
of their arguments, including the forms to be conditionally evaluated,
were evaluated automatically this would defeat the object of being
conditional!).
The special forms supported by `librep' are: `cond', `defvar',
`progn', `quote', `setq'.
- Function: special-form-p arg
Returns true if ARG is a special form.
(special-form-p quote)
=> t
back to top
Autoload Forms
..............
Not all parts of `librep' are needed at once, autoload forms provide
a means of marking that a function (or macro) is contained by a
specific Lisp library. The first time that the function is accessed the
autoload form will be evaluated; this loads the file containing the
function, then re-evaluates the original form. By then the autoload
form will have been overwritten in the symbol's function slot by the
true function (when it was loaded) so the form will execute properly.
For more details see *Note Autoloading::.
back to top
Self-Evaluating Forms
---------------------
The computed value of any form which is not a symbol or a list will
simply be the form itself and the form is said to be a "self-evaluating
form".
Usually the only forms to be evaluated in this way will be numbers,
strings and vectors (since they are the only other data types which
have read syntaxes) but the effect is the same for other types of data.
This means that forms you know are self-evaluating do not have to be
quoted to be used as constants (like lists and symbols do).
"foo"
=> "foo"
back to top
Quoting
-------
As the above sections explain some types of Lisp object have special
meaning to the Lisp evaluator (namely the symbol and list types) this
means that if you want to refer to a symbol or a list in a program you
can't because the evaluator will treat the form as either a variable
reference or a function call respectively.
To get around this Lisp uses an idea called "quoting". The special
form `quote' simply returns its argument without evaluating it. For
example,
(quote my-symbol)
=> my-symbol
the `quote' form prevents the `my-symbol' being treated as a
variable--it is effectively `hidden' from the evaluator.
Writing `quote' all the time would be a bit time-consuming so there
is a shortcut: the Lisp reader treats any form X preceded by a single
quote character (`'') as the form `(quote X)'. So the example above
would normally be written as,
'my-symbol
=> my-symbol
The general way to prevent evaluation of a form is to simply precede
it by a single quote-mark.
- Special Form: quote form
This special form returns its single argument without evaluating
it. This is used to "quote" constant objects to prevent them from
being evaluated.
For another form of quoting, see *Note Backquoting::.
back to top
Variables
=========
In Lisp, symbols are used to represent variables. Each symbol
contains a "value" slot that is used to contain the value of the symbol
when it used as a variable.
The normal way to obtain the current value of a variable is simply to
evaluate the symbol of the same name (i.e. write the name of the
variable in your program). The `symbol-value' function can be used to
evaluate variables whose names not known statically.
- Function: symbol-value variable
This function returns the value of the symbol VARIABLE in the
current environment.
* Menu:
* Local Variables:: Creating temporary variables
* Setting Variables:: Altering a variable's value
* Scope and Extent:: Technical jargon
* Void Variables:: Some variables have no values
* Defining Variables:: How to define a variable before
using it
* Fluid Variables:: Another dynamic bindingd methodb
back to top
Local Variables
---------------
A "local variable" is a variable which has a temporary value. For
example, when a function is called the variables which are the names of
its arguments are temporarily bound to the values of the arguments
passed to the function. When the function call exits its arguments are
unbound and the previous definitions of the variables come back into
view.
A "binding" is a particular instance of a local variable. Even if a
variable has more than one binding currently in place, only the most
recent is available--there is no way the previous binding can be
accessed until the previous binding is removed.
One way of visualising variable binding is to think of each variable
as a stack. When the variable is bound to, a new value is pushed onto
the stack, when it is unbound the top of the stack is popped. Similarly
when the stack is empty the value of the variable is void (*note Void
Variables::). Assigning a value to the variable (*note Setting
Variables::) overwrites the top value on the stack with a new value.
When the value of the variable is required it is simply read from the
top of the stack.
Apart from function applications there are two special forms which
perform variable binding (i.e. creating local variables), `let' and
`let*'.
- Macro: let bindings body-forms...
`let' creates new variable bindings as specified by the BINDINGS
argument, then evaluates the BODY-FORMS in order. The bindings
are then removed, returning all variables to their state before
the `let' statement was entered. The value of the statement is the
value of the implicit `progn'.
The BINDINGS argument is a list of the bindings to perform. Each
binding is either a symbol, in which case that variable is bound to
`()', or a list whose car is a symbol. The cdr of this list is a
list of forms which, when evaluated as a `progn', gives the value
to bind to that variable.
(setq foo 42)
=> 42
(let
((foo (+ 1 2))
bar)
;; Body forms
(setq foo (1+ foo)) ;This sets the new binding
(cons foo bar))
=> (4 . ())
foo
=> 42 ;The original values is back
No bindings are made until all new values have been computed. For
example:
(setq foo 42)
=> 42
(let
((foo 100)
(bar foo))
(cons foo bar))
=> (100 . 42)
Although `foo' is given a new binding this is not actually done
until all the new values have been computed, hence `bar' is bound
to the _old_ value of `foo'.
- Macro: let* bindings body-forms...
This special form is exactly the same as `let' except for one
important difference: the new bindings are installed _as they are
computed_.
You can see the difference by comparing the following example with
the last example in the `let' documentation (above),
(setq foo 42)
=> 42
(let* ;Using `let*' this time
((foo 100)
(bar foo))
(cons foo bar))
=> (100 . 100)
By the time the new value of `bar' is computed the new binding of
`foo' is already active.
- Macro: letrec bindings body-forms...
`letrec' is similar to `let' and `let*', with the differerence
being that the values of bindings are evaluated with all other
bindings in scope. This means that recursive functions may be
defined with `letrec'. For example, a local factorial function
(from SICP):
(letrec ((fact
(lambda (n)
(if (= n 1)
1
(* n (fact (1- n)))))))
(fact 10))
Note also that letrec allows groups of mutually recursive
functions to be defined, as in the following example (also from
SICP):
(defun f (x)
(letrec ((evenp
(lambda (n)
(if (= n 0)
t
(oddp (1- n)))))
(oddp
(lambda (n)
(if (= n 0)
nil
(evenp (1- n))))))
...
back to top
Setting Variables
-----------------
"Setting" a variable means to overwrite its current value (that is,
the value of its most recent active binding) with a new one. In the
variable-as-stack analogy, this is analogous to overwriting the top of
the stack. The old value is irretrievably lost (unlike when a new value
is bound to a variable, *note Local Variables::).
The `setq' special form is the usual method of altering the value of
a variable.
- Special Form: setq variable form ...
Each VARIABLE is set to the result of evaluating its corresponding
FORM. The last value assigned becomes the value of the `setq' form.
(setq x 20 y (+ 2 3))
=> 5
In the above example the variable `x' is set to `20' and `y' is
set to the value of the form `(+ 2 3)' (5).
- Function: set variable new-value
The value of the variable VARIABLE (a symbol) is set to NEW-VALUE
and the NEW-VALUE is returned.
This function is used when the VARIABLE is unknown until run-time,
and therefore has to be computed from a form.
(set 'foo 20)
==
(setq foo 20) ;`setq' means `set-quoted'
=> 20
_Note:_ currently the `set' function may be used to set any type
of variable (i.e. lexical or special). However this likely to
change in the future, such that only special variables will be
allowed to be modified using the `set' function. It is strongly
advised to avoid using this function on lexical bindings!
(Moreover the compiler may generate incorrect code in certain
circumstances.)
back to top
Scope and Extent
----------------
In the `librep' dialect of Lisp by default variables have "lexical
scope". This means that bindings are associated with textual regions of
programs, and may be accessed by any forms within this associated
textual region. Moreover, the bindings are persistent, even when the
flow of control is currently outside the associated region.
Consider the following example:
(let
((counter 0))
(defun count ()
(setq counter (1+ counter))
counter))
the value of the `counter' variable persists, and is incremented each
time the `count' function is called. The `counter' variable is
accessible from nowhere but the forms written inside the `let'
statement declaring it.
(count)
=> 1
(count)
=> 2
An alternative method of scoping variables is also available. Any
variables declared using the `defvar' special form are said to be
"special" variables, they have "indefinite scope" and "dynamic extent",
often simplified to "dynamic scope". What this means is that references
to these variables may occur anywhere in a program (i.e. bindings
established in one function are visible within functions called from
the original function) and that references may occur at any point in
time between the binding being created and it being unbound.
Dynamic scoping is easy to abuse, making programs hard to understand
and debug. A quick example of the use of dynamic scope,
(defvar *foo-var* nil)
(defun foo (x)
(let
;; a dynamically-scoped binding
((*foo-var* (* x 20)))
(bar x)
...
(defun bar (y)
;; Since this function is called from
;; the function `foo' it can refer
;; to `*foo-var*'
(setq y (+ y *foo-var*))
...
As shown in the previous example, a common convention is to mark
special variables by enclosing their names within asterisks.
back to top
Void Variables
--------------
A variable which has no value is said to be "void", attempting to
reference the value of such a symbol will result in an error. It is
possible for the most recent binding of a variable to be void even
though the inactive bindings may have values.
- Function: boundp variable
Returns true if the symbol VARIABLE has a value.
- Function: makunbound variable
This function makes the current binding of the symbol VARIABLE be
void, then returns VARIABLE.
(setq foo 42)
=> 42
foo
=> 42
(boundp 'foo)
=> t
(makunbound 'foo)
=> foo
(boundp 'foo)
=> ()
foo
error--> Value as variable is void: foo
back to top
Defining Variables
------------------
The special forms `define', `defvar' and `defconst' allow you to
define the global variables that will be used by a program.
- Macro: define variable form
Defines a lexically scoped global variable called VARIABLE. It
will have the result of evaluating FORM assigned to it.
Note that the `define' special form may also be used to declare
block-structured functions, *Note Definitions::.
- Special Form: defvar variable [form [doc-string]]
This special form defines a special (i.e. dynamically scoped)
variable, the symbol VARIABLE. If the value of VARIABLE is void the
FORM is evaluated and its value is stored as the value of VARIABLE
(note that only the default value is modified, never a
buffer-local value). If no FORM is given the assigned value
defaults to false.
If the DOC-STRING argument is defined it is a string documenting
VARIABLE. This string is then stored as the symbol's
`variable-documentation' property and can be accessed by the
`describe-variable' function.
(defvar *my-variable* '(x y)
"This variable is an example showing the usage of the `defvar'
special form.")
=> *my-variable*
- Macro: defconst constant form [doc-string]
`defconst' defines a global constant, the symbol CONSTANT. Its
value is set to the result of evaluating FORM. Note that unlike
`defvar' the value of the symbol is _always_ set, even if it
already has a value.
The DOC-STRING argument, if defined, is the documentation string
for the constant.
(defconst the-answer 42
"An example constant.")
=> the-answer
back to top
Fluid Variables
---------------
Special variables have a number of drawbacks, especially when used in
conjunction with the module system (*note Modules and Special
Variables::). As a consequence of these drawbacks, `rep' provides a
second method of implementing dynamically scoped variables, known as
"fluid variables", or just "fluids".
A fluid is a first class Lisp object that may be passed around like
any other Lisp object. Its sole function is to provide a location from
which dynamic bindings may be created. Fluids are anonymous objects,
they are usually named by being stored in lexically scoped variables.
- Function: make-fluid #!optional value
Create and return a new fluid, it will have an initial binding of
VALUE (or false if VALUE is undefined).
- Function: fluid fluid
Return the value of the most recently created binding of the fluid
variable object FLUID.
- Function: fluid-set fluid value
Set the value of the most recently created binding of the fluid
variable object FLUID to VALUE.
- Function: with-fluids fluids values thunk
Call the zero parameter function THUNK (and return the value that
it returns) with new bindings created for each of the fluid
variables specified in the list FLUIDS.
For each member of FLUIDS the corresponding member of the VALUES
list provides the initial value of the new binding.
If the lists FLUIDS and VALUES are not of the same length, an
error is signalled.
- Macro: let-fluids bindings body ...
A convenient wrapper around `with-fluids', similar to the `let'
syntax.
The list BINDINGS associates the names of lexical variables
containing fluid objects, with the values to bind to those fluid
objects. Once the bindings have been installed, the BODY ...
forms are evaluated, and the bindings removed. The value of the
last of the BODY ... forms is returned.
Here is an example code fragment using fluid variables and
`let-fluids':
(define a (make-fluid))
(define b (make-fluid))
(let-fluids ((a 1)
(b 2))
(+ (fluid a) (fluid b))) => 3
back to top
Functions
=========
A "function" is a Lisp object which, when applied to a sequence of
argument values, produces another value--the function's "result". It
may also induce side-effects (e.g. changing the environment of the
calling function). All Lisp functions return results -- there is
nothing like a procedure in Pascal.
Note that special forms (*note Special Forms::) and macros (*note
Macros::) are _not_ functions since they do not guarantee to evaluate
all of their arguments.
Functions are the main building-block in Lisp programs, each program
is usually a system of interrelated functions.
There are two types of function: "primitive functions" are functions
written in the C language, these are sometimes called built-in
functions, the object containing the C code itself is called a "subr".
All other functions are defined in Lisp.
- Function: functionp object
Returns true if OBJECT is a function (i.e. it can be used as the
function argument of `funcall'.
(functionp set)
=> t
(functionp setq)
=> ()
(functionp (lambda (x) (+ x 2)))
=> t
- Function: subrp arg
Returns true is ARG is a primitive subroutine object.
- Function: subr-name subr
Returns a string naming the primitive subroutine SUBR.
* Menu:
* Lambda Expressions:: Structure of a function object
* Defining Functions:: How to write a function definition
* Anonymous Functions:: Or they can be un-named
* Predicate Functions:: Functions which return boolean values
* Local Functions:: Binding functions temporarily
* Calling Functions:: Functions can be called by hand
* Mapping Functions:: Map a function to the elements of a list
back to top
Lambda Expressions
------------------
"Lambda expressions" are used to create functions from other Lisp
objects. A lambda expression is a list whose first element is the
symbol `lambda'. All functions written in Lisp (as opposed to the
primitive functions in C) are defined using lambda expressions.
The general format of a lambda expression is:
(lambda LAMBDA-LIST [DOC] [INTERACTIVE-DECLARATION] BODY-FORMS... )
Where LAMBDA-LIST is a list defining the formal parameters of the
function, DOC is an optional documentation string,
INTERACTIVE-DECLARATION is only required by interactive commands (1)
and BODY-FORMS is the sequence of forms making up the function body,
evaluated using an implicit `progn'.
The LAMBDA-LIST is a list, it defines how the values applied to the
function are bound to local variables which represent the parameters of
the function. At its simplest it is simply a list of symbols, each
symbol will have the corresponding argument value bound to it. For
example, the lambda list `(x y)' defines two parameters, `x' and `y'.
When called with two arguments the first will be bound to the variable
`x', the second to `y'. When used in a full lambda expression this
looks like:
(lambda (x y) (+ x y))
this evaluates to an anonymous function with two parameters, `x' and
`y', which when called evaluates to their sum.
Note that a lambda expression itself is _not_ a function, it must be
associated with a lexical environment, this conjunction is usually
called a closure; it is the closure that may be called as a function.
However, to confuse matters, a lambda expression _evaluates_ to the
closure of itself and the current environment. Consider the following
example:
(lambda (x) (1+ x))
=> #
(functionp (lambda (x) (1+ x)))
=> t
(functionp '(lambda (x) (1+ x)))
=> ()
There are several "lambda-list keywords" which modify the meaning of
symbols in the lambda-list. The syntax of the lambda list is:
([REQUIRED-PARAMETERS...]
[#!optional OPTIONAL-PARAMETERS...]
[#!key KEYWORD-PARAMETERS...]
[#!rest REST-PARAMETER | . REST-PARAMETER])
Each lambda list keyword is a symbol whose name begins `#!', they are
interpreted as follows:
`#!optional'
All variables following this keyword are considered "optional" (all
variables before the first keyword are "required": an error will be
signalled if a required argument is undefined in a function call).
OPTIONAL-PARAMETERS may either be of the form `SYMBOL' or of the
form `(SYMBOL DEFAULT)'. If no argument is supplied for this
parameter the DEFAULT form is evaluated to give the bound
value(2). If no DEFAULT form is given, then the variable is bound
to a false value.
Note that optional parameters must be specified if a later
parameter is also specified.
((lambda (#!optional a b) (list a b)))
=> (() ())
((lambda (#!optional a b) (list a b)) 1)
=> (1 ())
((lambda (#!optional a b) (list a b)) nil 1)
=> (() 1)
((lambda (#!optional (a 1)) (list a)))
=> (1)
((lambda (#!optional (a 1)) (list a)) 2)
=> (2)
`#!key'
This object marks that the parameters up to the next lambda list
keyword are keyword parameters. The values bound to these
parameters when the function is called are determined not by
position (as with normal parameters), but by being marked by a
preceding keyword symbol. Keyword symbols have the syntax
`#:SYMBOL'.
As with optional parameters, default values may be supplied
through the use of the `(SYMBOL DEFAULT)' syntax. If no default
value is given and no keyword argument of the specified kind is
available, the variable is bound to a false value.
For example, the lambda list `(a #!key b c)' accepts one required
argument, and two optional keyword arguments. The variable `a'
would be bound to the first supplied argument; the variable `b'
would be bound to the argument preceded by the keyword `#:b', or
`()' if no such argument exists. (After extracting required and
optional arguments, each remaining pair of values is checked for
associating a value with each keyword.)
((lambda (a #!key b c) (list a b c)) 1 2 3)
=> (1 () ())
((lambda (a #!key b c) (list a b c)) 1 #:b 2 3)
=> (1 2 ())
((lambda (a #!key b c) (list a b c)) 1 #:b 2 #:c 3)
=> (1 2 3)
((lambda (a #!key b c) (list a b c)) 1 #:c 3 #:b 2)
=> (1 2 3)
`#!rest'
The `#!rest' keyword allows a variable number of arguments to be
applied to a function, all the argument values which have not been
bound to argument variables (or used to mark keyword arguments) are
made into a list and bound to the variable following the `#!rest'
keyword. For example, in
(lambda (x #!rest y) ...)
the first argument, `x', is required. Any other arguments applied
to this function are made into a list and this list is bound to the
variable `y'.
Variable argument functions may also be defined through the Scheme
method of using an improper lambda-list. The previous example is
exactly equivalent to:
(lambda (x . y) ...)
When a function represented by a lambda-list is called the first
action is to bind the argument values to the formal parameters. The
LAMBDA-LIST and the list of argument values applied to the function are
worked through in parallel. Any required arguments which are left
undefined when the end of the argument values has been reached causes
an error.
After the arguments have been processed the BODY-FORMS are evaluated
by an implicit progn, the value of which becomes the value of the
function call. Finally, all parameters are unbound and control passes
back to the caller.
---------- Footnotes ----------
(1) Only used when `librep' is embedded within another application.
(2) The DEFAULT form is evaluated in the environment of the closure
being called, but without any of the bindings created by the lambda
expression.
back to top
Defining Functions
------------------
Globally accessible functions are usually defined by the `defun'
special form.
- Macro: defun name lambda-list body-forms...
`defun' initialises the function definition of the symbol NAME to
the lambda expression resulting from the concatenation of the
symbol `lambda', LAMBDA-LIST and the BODY-FORMS.
The BODY-FORMS may contain a documentation string for the function
as its first form and an interactive calling specification as its
first (if there is no doc-string) or second form if the function
may be called interactively by the user (*note Lambda
Expressions::).
An example function definition taken from the `librep' source code
is:
(defun load-all (file)
"Try to load files called FILE (or FILE.jl, etc) from all
directories in the Lisp load path."
(mapc (lambda (dir)
(let
((full-name (expand-file-name file dir)))
(when (or (file-exists-p full-name)
(file-exists-p (concat full-name ".jl"))
(file-exists-p (concat full-name ".jlc")))
(load full-name nil t))))
load-path))
back to top
Anonymous Functions
-------------------
When supplying functions as arguments to other functions it is often
useful to give an actual function _definition_ (i.e. an enclosed lambda
expression) instead of the name of a function.
In Lisp, unlike most other programming languages, functions have no
inherent name. As seen in the last section named-functions are created
by storing a function object in a variable, if you want, a function can
have many different names: simply store the function in many different
variables!
So, when you want to pass a function as an argument there is the
option of just writing down its definition. This is especially useful
with functions like `mapc' and `delete-if'. For example, the following
form removes all elements from the LIST which are even and greater than
20.
(setq LIST (delete-if (lambda (x)
(and (zerop (% x 2)) (> x 20)))
LIST))
The above lambda expression combines two predicates applied to its
argument.
In certain cases it may be necessary to create a non-constant
function, for example by using backquoting (*note Backquoting::). In
these cases the `make-closure' function may be used to create a function
object from a lambda expression.
- Function: make-closure arg
Return the closure of ARG and the current lexical environment.
- Function: closurep arg
Returns true if ARG is a closure.
- Function: closure-function closure
Returns the function object associated with the lexical closure
CLOSURE.
back to top
Predicate Functions
-------------------
In Lisp, a function which returns a boolean `true' or boolean `false'
value is called a "predicate". As is the convention in Lisp a value of
`()' means false, anything else means true. The symbols `nil' and `t'
are often used to represent constant false and true values (*note nil
and t::).
Another Lisp convention is that the names of predicate functions
should name the quality that the predicate is testing followed by
either a `p' or `-p' string. The `p' variant is used when the first
string does not contain any hyphens.
For example, the predicate to test for the quality "const-variable"
(a variable which has a constant value, *note Defining Variables::) is
called `const-variable-p'. On the other hand the predicate to test for
the quality "cons" (a Cons cell) is called `consp'.
back to top
Local Functions
---------------
The `defun' special form allows globally-accessible functions to be
defined. It is often desirable to declare functions local to the
current lexical environment. The `let' and `let*' special form that
were introduced earlier allow this since named functions are simply
functional values stored in variables.
For example,
(let
((temporary-function (lambda (x)
(+ x 42))))
...
(temporary-function 20)
...
back to top
Calling Functions
-----------------
Most of the time function applications are made by the evaluator when
it finds a functional value after evaluating the first element of a
list form. However two functions are available for manually calling
functions.
- Function: funcall function #!rest args
Applies the argument values ARGS to the function FUNCTION, then
returns its result.
- Function: apply function #!rest args
Similar to `funcall' except that the last of its arguments is a
_list_ of arguments which are appended to the other members of
ARGS to form the list of argument values to apply to the function
FUNCTION.
(apply + 1 '(2 3))
=> 6
(apply + (make-list 1000000 1))
=> 1000000
back to top
Mapping Functions
-----------------
A "mapping function" applies a function to each of a collection of
objects. `librep' currently has two mapping functions, `mapcar' and
`mapc'.
- Function: mapcar function list
Each element of LIST is individually applied to the function
FUNCTION. The values returned are made into a new list which is
returned.
The FUNCTION must accept a single argument value.
(mapcar 1+ '(1 2 3 4 5))
=> (2 3 4 5 6)
- Function: mapc function list
Similar to `mapcar' except that the values returned when each
element is applied to the function FUNCTION are discarded. The
value returned is undefined.
This function is generally used where the side effects of calling
the function are the important thing, not the results. It is often
the most efficient way of traversing all items in a list, for
example:
(mapc (lambda (x)
(print x standard-error)) list)
The two following functions are also mapping functions of a sort.
They are variants of the `delete' function (*note Modifying Lists::)
and use predicate functions to classify the elements of the list which
are to be deleted.
- Function: delete-if predicate list
This function is a variant of the `delete' function. Instead of
comparing each element of LIST with a specified object, each
element of LIST is applied to the predicate function PREDICATE. If
it returns true then the element is destructively removed from
LIST.
(delete-if stringp '(1 "foo" 2 "bar" 3 "baz"))
=> (1 2 3)
- Function: delete-if-not predicate list
This function does the inverse of `delete-if'. It applies
PREDICATE to each element of LIST, if it returns false then the
element is destructively removed from the list.
(delete-if-not stringp '(1 "foo" 2 "bar" 3 "baz"))
=> ("foo" "bar" "baz")
The `filter' function is similar to `delete-if-not', except that the
original list isn't modified, a new list is created.
- Function: filter predicate list
Return a new list, consisting of the elements in LIST which the
function PREDICATE returns true when applied to. This function is
equivalent to:
(mapcar nconc (mapcar (lambda (x)
(and (PREDICATE x) (list x)))
LIST))
back to top
Macros
======
"Macros" are used to extend the Lisp language. They consist of a
function which instead of returning a computed value, transform their
unevaluated arguments into a new form that, when evaluated, produces
the actual value of the original form.
For example, the `when' macro (*note Conditional Structures::)
implements a new conditional operation by transforming its arguments
into a `cond' statement. That is,
(when CONDITION FORM ...)
==> (cond (CONDITION FORM ...))
Since macros do not evaluate their arguments, instead just
transforming them, they may be expanded at _compile-time_. The
resulting form is then compiled as usual.
- Function: macrop arg
Returns true if ARG is a macro object.
* Menu:
* Defining Macros:: Macros are defined like functions
* Backquoting:: Creating macros from templates
* Macro Expansion:: How macros are used by the evaluator
* Compiling Macros:: The compiler expands macros at compile-
time.
back to top
Defining Macros
---------------
Macros are defined in the same style as functions, the only
difference is the name of the special form used to define them.
A macro object is a list whose car is the symbol `macro', its cdr is
the function which creates the expansion of the macro when applied to
the macro calls unevaluated arguments.
- Macro: defmacro name lambda-list body-forms...
Defines the macro stored in the function cell of the symbol NAME.
LAMBDA-LIST is the lambda-list specifying the arguments to the
macro (*note Lambda Expressions::) and BODY-FORMS are the forms
evaluated when the macro is expanded. The first of BODY-FORMS may
be a documentation string describing the macro's use.
Here is a simple macro definition, it is the definition of the
`when' macro shown in the previous section.
(defmacro when (condition #!rest body)
"Evaluates CONDITION, if it's true evaluates the BODY
forms."
(list 'cond (list* condition body)))
When a form of the type `(when C B ...)' is evaluated the macro
definition of `when' expands to the form `(cond (C (progn B ...)))'
which is then evaluated to perform the `when'-construct.
When you define a macro ensure that the forms which produce the
expansion have no side effects; otherwise undefined effects will occur
when programs using the macro are compiled.
back to top
Backquoting
-----------
As seen in the previous sections, macros are a very powerful
mechanism of defining new control structures. However due to the need
to create the expansion, i.e. the form that will be actually evaluated,
they can often be complex to write and understand.
We have already seen that constants may be produced through the use
of the quote-mark (*note Quoting::), here another form of quoting is
described, where only some of the quoted object is actually constant.
This is known as "backquoting", since it is introduced by the backquote
character ``', a shortcut for the `backquote' macro.
- Macro: backquote arg
Constructs a new version of ARG (a list). All parts of LIST are
preserved except for expressions introduced by comma (`,')
characters, which are evaluated and spliced into the list. For
example:
`(1 2 ,(+ 1 2))
=> (1 2 3)
Also, the `,@' prefix will splice the following _list_ into the
output list, at the same level:
`(1 2 ,@(list 3))
=> (1 2 3)
Backquoting allows macros expansions to be created from static
templates. For example the `when' macro shown in the previous sections
can be rewritten as:
(defmacro when (condition #!rest body)
`(cond (,condition ,@body)))
which is easier to read, since it is a lot closer to the actual
expansion.
back to top
Macro Expansion
---------------
When a macro call is detected (*note List Forms::) the function which
is the cdr of the macro's definition (*note Defining Macros::) is
applied to the macro call's arguments. Unlike in a function call, the
arguments are _not evaluated_, the actual forms are the arguments to
the macro's expansion function. This is to allow these forms to be
rearranged by the macro's expansion function, creating the form that
will finally be evaluated.
There is a function which performs macro expansion, its main use is
to let the Lisp compiler expand macro calls at compile time.
- Function: macroexpand form #!optional environment
If FORM is a macro call `macroexpand' will expand that call by
calling the macro's expansion function (the cdr of the macro
definition). If this expansion is another macro call the process
is repeated until an expansion is obtained which is not a macro
call, this form is then returned.
The optional ENVIRONMENT argument is a function to call to do the
actual expansion.
(defmacro when (condition #!rest body)
"Evaluates CONDITION, if it's true evaluates the BODY
forms."
(list 'if condition (cons 'progn body)))
=> when
(macroexpand '(when x (setq foo bar)))
=> (cond (x (progn (setq foo bar))))
While a macro is being expanded, the special variable
`macro-environment' is bound to value of the ENVIRONMENT parameter
in the containing call to `macroexpand'. This allows macros to
expand inner macros correctly.
- Function: macroexpand-1 form #!optional environment
Similar to `macroexpand', but only a single macro expansion is
ever performed, i.e. if FORM is a macro call the result of
expanding that call will be returned, otherwise FORM is returned.
(macroexpand-1 '(when x (setq foo bar)))
=> (if x (progn (setq foo bar)))
back to top
Compiling Macros
----------------
Although it may seem odd that macros return a form to produce a
result and not simply the result itself, this is actually their most
important feature. It allows the expansion and the evaluation of the
expansion to occur at different times.
The Lisp compiler makes use of this; when it comes across a macro
call in a form it is compiling it uses the `macroexpand' function to
produce the expansion of that form. This expansion is then compiled
straight into the object code. Obviously this is good for performance
(why evaluate the expansion every time it is needed when once will do?).
Some rules do need to be observed to make this work properly:
* The macro expansion function (i.e. the definition of the macro)
should not have any side effects or evaluate its arguments (the
value of a symbol at compile-time probably won't be the same as
its value at run-time).
* Macros which are defined by another file must be loaded so they
can be recognised. Use the `require' function, the compiler will
evaluate any `require' forms it sees loading any macro definitions
used.
Note however, that the `librep' compiler does allow macros to be
used before they are defined (two passes are made through the source
file).
back to top
Block-Structured Definitions
============================
Previous sections of this document have described several special
forms and macros for defining top-level functions and variables.
`librep' also provides a higher-level method of creating these
definitions, the `define' statement. `define' originates in the Scheme
dialect of Lisp, it allows block-structured programs to be defined
intuitively.
The most basic use of `define' is very similar to `defun', e.g. the
two following forms have exactly the same effect:
(defun foo (x) (1+ x))
(define (foo x) (1+ x))
But note the different position of the parentheses. This is because
`define' may also be used to define (lexical) variables. Hence the
following is also equivalent:
(define foo (lambda (x) (1+ x)))
However this is the most uninteresting aspect of `define'. More
interesting is that it allows "internal definitions".
Within a `define' form, any inner calls to `define' (that occur in a
contiguous block at the start of the body of a `let', `let*', `letrec',
`lambda', or `define' form) are also used to create definitions, but
definitions that are local to the containing scope. For example:
(define (foo x)
(define (bar)
(* x 42))
(1+ (bar)))
This defines a top-level function called `foo'. However it also
contains an inner function named `bar', that is only accessible within
`foo'. Since `bar' is defined inside `foo', and librep uses lexical
scope by default, the variable `x' defined by `foo' may also be
accessed by `bar'.
- Macro: define name form
- Macro: define (name . args) body-forms...
Define a global lexical variable called NAME, whose value will be
set to FORM.
If the first argument to the macro is a list, then a function is
defined whose name is NAME and whose formal parameters are
specified by ARGS. The body of the function is defined by the
BODY-FORMS. The body forms have any macros expanded, and are
scanned for internal definitions (at the start of the body of
`let', `let*', `lambda' special forms)
- Macro: define-macro name form
- Macro: define-macro (name . args) body-forms...
Similar to `define', except that it creates a macro definition
(*note Macros::).
- Macro: with-internal-definitions body-forms
Recursively expand macros in BODY-FORMS, while scanning out any
internal definitions into `letrec' statements.
back to top
Modules
=======
When creating large programs from many separate components, it is
important to be able to encapsulate these components, such that the
interfaces they present to other components are well defined, and the
implementations of these interfaces may be modified without affecting
any other components. To this end `rep' provides a "module system" for
managing the scope of global definitions. This module system was
inspired by the Scheme48, Xerox Scheme and Standard ML module systems.
Modules are known as "structures" and may be anonymous or named.
Each structure specifies and implements an "interface", essentially a
list of names listing the definitions within that module that may be
accessed by other modules. Each structure is a separate global
namespace, with a number of variable bindings. Each closure contains a
reference to the structure it was instantiated in, for accessing the
bindings of any free variables referenced by that closure.
As well as specifying its name and interface, each module also lists
the other modules whose bindings it may reference. Structures may
either "open" or "access" other structures; when opening a structure
all its exported bindings are immediately referenceable from the
importing module. Exported bindings from accessed structures are
referenced using the `structure-ref' form.
* Menu:
* Module Interfaces::
* Module Definition::
* Module Loading::
* Modules and Special Variables::
back to top
Module Interfaces
-----------------
Each module implements an interface--the set of bindings (i.e.
functions, macros or variables) that it exports to other modules.
Interfaces may either be defined and then referenced by name, written
literally, or combined from a number of sources.
The syntax of interface definitions is as follows:
INTERFACE -> (export ID ...)
| NAME
| (compound-interface INTERFACE ...)
| (structure-interface MODULE-NAME)
where each ID is the name of a binding to export, and each NAME is the
name of an interface previously defined using `define-interface'.
- Macro: define-interface name interface
Associate the symbol NAME with the module interface INTERFACE
(using one of the forms listed above.
Here is an example defining an interface called `foo':
(define-interface foo (compound-interface bar (export baz quux)))
It includes the interface called `bar' and adds two extra exported
symbols: `baz' and `quux'.
back to top
Module Definition
-----------------
Two special forms are used to define modules, one for anonymous
modules, one for named modules. When storing modules in files, each
file often contains a single instance of one of these forms.
- Macro: structure interface config body...
- Macro: define-structure name interface config body...
These special forms each create a new module with interface
INTERFACE (using the syntax described in the previous section),
and configuration CONFIG.
After configuring the module as specified, the sequence of forms
BODY... is evaluated; it should include the definitions required
by the interface that the module has promised to implement.
The CONFIG form is either a list of configuration clauses, or a
single configuration clause. Each such clause must be of the
following syntax:
CLAUSE -> (open NAME ...)
| (access NAME ...)
Each NAME specifies the name of a module, in the case of `open'
clauses, the named module(s) will be loaded such that their
exported bindings may be referenced from within the current module
with no qualification (i.e. as if they had been defined within the
module itself).
Alternatively, if an `access' clause was used, the named module(s)
will be loaded, but their exported bindings will only be accessible
from within the current module using the `structure-ref' form.
E.g. if a module `foo' has been accessed and it exports a binding
named `bar', then the following form could be used to access its
value:
(structure-ref foo bar)
Since this form is used so often, the reader allows the
abbreviation `foo#bar' to be used instead, it is expanded to the
form above when read. Note that no whitespace is allowed between
the three tokens.
Note that to access the standard features of the `rep' language
described in this manual, modules need to import the `rep' module.
Alternatively, they may import the `scheme' module to load a minimal
R4RS Scheme environment.
Here is an example module definition, defining a module named `test'
that exports two functions `foo' and `bar'.
(define-structure test (export foo bar)
(open rep)
(define (foo x) (* x 42))
(define (bar x y) (+ (foo x) (1+ y))))
It is also possible to export multiple views of a single underlying
set of bindings, by using the `define-structures' form to create a
number of modules.
- Macro: define-structures ((name interface) ...) config body...
Create a module for each `(NAME INTERFACE)' pair. The module is
called NAME and exports the interface defined by INTERFACE.
The CONFIG and BODY... forms are as in `define-structure'.
Here is a trivial example:
(define-structures ((foo (export foo both))
(bar (export bar both)))
(open rep)
(define both 1)
(define foo 2)
(define bar 3))
the underlying environment has three bindings. Each created module
exports two of these.
back to top
Module Loading
--------------
As described above, the common way of loading modules is to use the
`open' and `access' clauses of the configuration language.
If the modules named by these clauses are not currently loaded into
the interpreter, then the system will attempt to load them from the
filing system, using the standard `load-path' variable to define the
directories to search.
To allow modules names to be hierarchical, any dot characters in a
module's name are replaced by the operating system's directory
separator string (i.e. on unix, all `.' characters are simply replaced
by `/' characters).
When searching for files to load, the standard filename suffixes are
used to differentiate Lisp files from other types of files (*note Load
Function::). This file should contain a `define-structure' form (as
described in the previous section) as the last top-level form in the
file.
For backwards compatibility, the `require' function can also be used
to import modules. If a module of the same name as the requested
feature has already been loaded, then it is imported into the current
module. Otherwise if a file is loaded that contains a module definition
as its last top-level form, this module is imported into the current
module. *Note Features::.
back to top
Modules and Special Variables
-----------------------------
As described earlier, the `defvar' special form may be used to
create variables that are scoped dynamically, known as special
variables, see *Note Defining Variables::. Due to their dynamic scope,
special variables do not fit well with the lexically scoped module
system described here.
As a result of this mismatch, special variables are stored in a
separate namespace. This means that modules defining special variables
must take the necessary steps to avoid the names of these variables
clashing with those declared in other modules(1).
In fact, it is often advisable to avoid using special variables as
much as possible, especially when writing modules of Lisp code. An
alternative method of creating dynamically scoped variables is to use
fluid variable objects. These use first class Lisp objects to represent
anonymous dynamically scoped variables. Since they are just Lisp
objects, they may be stored in lexically scoped variables--this gives
the benefits of both lexical (i.e. encapsulation) and dynamic scoping.
*Note Fluid Variables::.
---------- Footnotes ----------
(1) The usual convention is to prefix the variable name with a
unique string derived from the module name.
back to top
Control Structures
==================
Control structures are special forms or macros that control _which_
forms get evaluated, _when_ they get evaluated and the _number_ of
times to evaluate them. This includes conditional structures, loops,
etc...
The simplest control structures are the sequencing structures; they
are used to evaluate a list of forms in left to right order.
* Menu:
* Sequencing Structures:: Evaluating several forms in sequence
* Conditional Structures:: Making decisions based on truth values
* Looping Structures:: `while' loops
* Non-Local Exits:: Exiting from several levels of evaluation
* Continuations:: Capturing the call stack
back to top
Sequencing Structures
---------------------
Each of the special forms in this section simply evaluates its
arguments in left-to-right order. The only difference is the result
returned.
The most widely used sequencing special form is `progn': it
evaluates all its argument forms and returns the computed value of the
last one. Many other control structures are said to perform an
"implicit progn", this means that internally they call `progn' with a
list of forms.
`progn' in Lisp is nearly analogous to a `begin...end' block in
Pascal; it is used in much the same places--to allow you to evaluate a
sequence of form where only one form was allowed (for example the
"true" clause of an `if' structure).
- Special Form: progn forms...
All of the FORMS are evaluated sequentially (from left-to-right),
the result of the last evaluated FORM is the return value of the
special form. If no arguments are given to `progn' it returns
false.
(progn 'one (+ 1 1) "three")
=> "three"
(progn)
=> ()
- Macro: prog1 first forms...
This special form evaluates its FIRST form then performs an
implicit progn on the rest of its arguments. The result of this
structure is the computed value of the FIRST form.
(prog1 'one (+ 1 1) "three")
=> one
- Macro: prog2 first second forms...
This is similar to `prog1' except that the evaluation of its
SECOND form is returned.
The FIRST form is evaluated, then its SECOND, then it performs an
implicit progn on the remaining arguments.
(prog2 'one (+ 1 1) "three")
=> 2
back to top
Conditional Structures
----------------------
Lisp provides a number of conditional constructs, the most complex of
which (`cond') takes a list of conditions, the first of which evaluates
to true has its associated list of forms evaluated. Theoretically this
is the only conditional special form necessary--all others can be
implemented as macros.
- Macro: if condition true-form else-forms...
The `if' form is the nearest thing in Lisp to the "if-then-else"
construct found in most programming languages.
First the CONDITION form is evaluated, if it returns true the
TRUE-FORM is evaluated and its result returned. Otherwise the
result of an implicit progn on the ELSE-FORMS is returned. If
there are no ELSE-FORMS false is returned.
Note that one of the TRUE-FORM or the ELSE-FORMS is completely
ignored--it is not evaluated.
(if (special-form-p if)
"`if' is a special form"
"`if' is not a special form")
=> "`if' is not a special form"
- Macro: when condition true-forms...
CONDITION is evaluated, if it is true the result of an implicit
progn on the TRUE-FORMS is returned, otherwise false is returned.
(when t
(message "Pointless")
'foo)
=> foo
- Macro: unless condition else-forms...
This special form evaluates CONDITION, if its computed value is
true, `()' is returned. Otherwise the ELSE-FORMS are evaluated
sequentially, the value of the last is returned.
- Special Form: cond clause...
The `cond' special form is used to choose between an arbitrary
number of conditions. Each CLAUSE is a list; the car of which is a
CONDITION, the cdr is a list of forms to evaluate (in an implicit
`progn') if the CONDITION evaluates to true. This means that each
CLAUSE looks something like:
(CONDITION BODY-FORMS...)
and a whole `cond' form looks like:
(cond
(CONDITION-1 BODY-FORMS-1...)
(CONDITION-2 BODY-FORMS-2...)
...)
The CONDITION in each CLAUSE is evaluated in sequence
(CONDITION-1, then CONDITION-2, ...), the first one which
evaluates to a true value has an implicit `progn' performed on its
BODY-FORMS. The value of this `progn' is also the value of the
`cond' statement.
If the true CONDITION has no BODY-FORMS the value returned is the
value of the CONDITION. If none of the clauses has a true
CONDITION the value of the `cond' statement is false.
Often you want a "default" clause which has its BODY-FORMS
evaluated when none of the other clauses are true. The way to do
this is to add a clause with a CONDITION of `t' and BODY-FORMS of
whatever you want the default action to be.
(cond
((stringp buffer-list)) ;Clause with no BODY-FORMS
((consp buffer-list)
(setq x buffer-list) ;Two BODY-FORMS
t)
(t ;Default clause
(error "`buffer-list' is corrupted!")))
=> t
- Macro: case key clauses...
This special form is similar to `cond', but switches on the result
of evaluating a single form KEY, checking for equality with a
number of other values, defined by the CLAUSES. If any of these
other values is the same as the result of evaluating KEY, then a
sequence of forms associated with the value is evaluated.
Each element of the CLAUSES list has the format:
((VALUE-1 VALUE-2 ... VALUE-N) FORMS...)
Each of the values in the car of the clause is tested for equality
with KEY, using the `eql' function. If any test positively, then
the associated FORMS are evaluated and the resulting value becomes
the result of the special form.
Instead of supplying a list of possible values, it is also
possible to just specify the symbol `t'. If such a clause is
encountered, and no other clauses have matched the value of KEY,
then this clause is assumed to match by default.
If any of the values in the CLAUSES appear multiply, then the
behaviour of the construct is undefined.
Here is an example use of `case':
(case foo
((bar baz)
(print "It was either bar or baz"))
((quux)
(print "It was quux"))
(t
(print "I've no idea what it was...")))
There are also a number of special forms which combine conditions
together by the normal logical rules.
- Macro: or forms...
The first of the FORMS is evaluated, if it is true its value is
the value of the `or' form and no more of `forms' are evaluated.
Otherwise this step is repeated for the next member of FORMS.
If all of the FORMS have been evaluated and none have a true value
the `or' form evaluates to false.
(or nil 1 nil (beep)) ;`(beep)' won't be evaluated
=> 1
- Macro: and forms...
The first of the FORMS is evaluated. If it is false no more of the
FORMS are evaluated and false is the value of the `and' statement.
Otherwise the next member of FORMS is evaluated and its value
tested. If none of the FORMS are false the computed value of the
last member of FORMS is returned from the `and' form.
(and 1 2 nil (beep)) ;`(beep)' won't be evaluated
=> ()
(and 1 2 3) ;All forms are evaluated
=> 3
- Function: not object
This function inverts the truth value of its argument. If OBJECT
is true, false is returned, otherwise true is returned.
(not nil)
=> t
(not t)
=> ()
(not (not 42))
=> t
back to top
Looping Structures
------------------
The `librep' Lisp dialect only has one method of creating looping
control structures--recursion. Any looping construct found in an
imperative language can be represented as a recursive function. For
example the common `while' statement:
(while CONDITION BODY...)
==
(letrec ((loop (lambda ()
(when CONDITION
BODY
(loop)))))
(loop))
Each successive iteration of the loop is simply another call to the
function. Also note that the recursive call to the `(loop)' function
occurs in the tail-position of the function. When combined with the
system's ability to eliminate tail-calls (*note Function Call Forms::)
the above example loop has bounded space requirements. This is
important when loops make a large number of iterations.
Although tail-recursion is the only primitive method of looping, the
language offers a number of looping forms for convenience.
- Macro: do vars (test expr...) body...
`do' is an iteration construct; VARS specifies a set of variable
bindings to be created, how they are initialized and how they are
updated on each iteration. TEST specifies the termination
condition of the loop, any EXPR... forms are evaluated immediately
prior to exiting the `do' construct. The BODY... forms specify
the side effecting body of the loop.
VARS is a list of variable clauses, each of which has the
structure `(VARIABLE INIT STEP)' where VARIABLE is the name of a
variable, INIT defines the initial value of its binding, and STEP
defines how the next value of the binding is computed. An
alternative form is `(VARIABLE INIT)', in this case the value of
the binding does not change across loop iterations.
Each iteration begins by evaluating TEST, if the result is false,
then the BODY... expressions are evaluated, and the variables
bound to new locations initialized to the results of evaluating
the associated STEP forms.
If the result of evaluating TEST is true then the EXPR... forms
are evaluated, and the `do' construct returns the value of the
last EXPR form evaluated.
(do ((vec (make-vector 5))
(i 0 (1+ i)))
((= i 5) vec)
(aset vec i i))
=> [0 1 2 3 4]
The "named-let" variant of the `let' form also provides a convenient
looping construct.
- Macro: let variable bindings body...
This is the same as the `(let BINDINGS BODY...)' form described in
*Note Local Variables::, but within the BODY... forms, the symbol
VARIABLE is bound to a function whose parameters are the bound
variables defined by BINDINGS and whose body is the sequence of
forms BODY...
This means that the body of the `let' may be repeated by invoking
the function VARIABLE with suitable parameters.
(let loop ((rest '(1 2 3))
(total 0))
(if (null rest)
total
(loop (cdr rest) (+ total (car rest)))))
=> 6
Finally, the imperative `while' form shown at the start of the
section is also provided:
- Macro: while condition body...
The CONDITION form is evaluated. If it is true an implicit progn
is performed on the BODY forms and the whole procedure is repeated.
This continues until the CONDITION form evaluates to false. The
value of every `while' structure that terminates is false.
back to top
Non-Local Exits
---------------
A "non-local exit" is a transfer of control from the current point
of evaluation to a different point (somewhat similar to the
much-maligned `goto' statement in imperative languages).
Non-local exits can either be used explicitly (`catch' and `throw')
or implicitly (errors).
* Menu:
* Catch and Throw:: Programmed non-local exits
* Function Exits:: Returning values from a function
* Cleanup Forms:: Forms which will always be evaluated
* Errors:: Signalling that an error occurred
back to top
Catch and Throw
...............
The `catch' and `throw' structures are used to perform explicit
transfers of control. First a `catch' form is used to setup a "tag";
this acts like a label for a `goto' statement. To transfer control a
`throw' form is then used to transfer to the named tag. The tag is
destroyed and the `catch' form exits with the value provided by the
`throw'.
In a program this looks like,
(catch 'TAG
;; Forms which may `throw' back to TAG
...
(throw 'TAG VALUE)
;; Control has now passed to the `catch',
;; no more forms in this progn will be evaluated.
...)
=> VALUE
where TAG is the tag to be used (this is normally a symbol) and VALUE
is the result of the `catch' form.
When a throw actually happens all catches in scope are searched for
one with a tag which is `eq' to the tag in the throw. If more than one
exists the innermost is selected. Now that the catch has been located
the environment is `wound-back' to the catch's position (i.e. local
variables are unbound, cleanup forms executed, unused catches removed,
etc...) and all Lisp constructs between the current point of control
and the catch are immediately exited.
For example,
(let
((test 'outer))
(cons (catch 'foo
(let
((test 'inner))
(throw 'foo test)
(setq test 'unreachable))) ;Never reached
test))
=> (inner . outer)
when the throw executes the second binding of `test' is unwound and the
first binding comes back into effect. For more details on variable
binding see *Note Local Variables::.
Note that catch tags are _dynamically_ scoped, the thrower does not
have to be within the same lexical scope (this means that you can
`throw' through functions).
- Macro: catch tag body-forms...
This special form defines a catch tag which will be accessible
while the BODY-FORMS are evaluated.
TAG is evaluated and recorded as the tag for this catch. Next the
BODY-FORMS are evaluated as an implicit `progn'. The value of the
`catch' form is either the value of the `progn', or, if a `throw'
happened, the value specified in the THROW form.
Before exiting, the tag installed by this form is removed.
- Function: throw tag #!optional catch-value
This function transfers the point of control to the catch form
with a tag which is `eq' to TAG. The value returned by this catch
form is either CATCH-VALUE or false if CATCH-VALUE is undefined.
If there is no catch with a tag of TAG an error is signalled and
the interpreter returns to the top-level of evaluation.
There are a number of pre-defined throw tags:
`quit'
Terminate the `librep' interpreter, returning the value of the
throw (if a number).
`exit'
Exit the innermost event loop, unless currently in the outermost
event loop, when control just passes back to the event loop.
`user-interrupt'
As if a `SIGINT' or `C-c' signal has been received. Control passes
back to the top-level event loop.
`term-interrupt'
Triggered when a `SIGTERM' or `SIGHUP' signal is received. Tries
to clean up any existing state, then terminates the interpreter.
Note that it is the event loop that catches these tags. If no event
loop is active (i.e. just in read-eval-print on the console mode), any
uncaught throws will result in termination.
back to top
Function Exits
..............
`librep' has no explicit `return' statement, as found in most other
languages. Where a value has to returned from a function before the
function would normally exit, a `catch'/`throw' pair may be used.
For example:
(defun foo (x y)
(catch 'return
(when (= x 2)
(throw 'return nil))
...
back to top
Cleanup Forms
.............
It is sometimes necessary ensure that a certain form is _always_
evaluated, even when a non-local exit would normally bypass that form.
The `unwind-protect' special form is used in this case.
- Macro: unwind-protect body-form cleanup-forms...
The BODY-FORM is evaluated, if it exits normally the CLEANUP-FORMS
are evaluated sequentially then the value which the BODY-FORM
returned becomes the value of the `unwind-protect' form. If the
BODY-FORM exits abnormally though (i.e. a non-local exit happened)
the CLEANUP-FORMS are evaluated anyway and the non-local exit
continues.
One use of this is to ensure that an opened file is always closed,
for example,
(catch 'foo
(unwind-protect
(let
((temporary-file (open-file (make-temp-name) 'write)))
;; Use `temporary-file'
(write temporary-file "A test\n")
;; Now force a non-local exit
(throw 'foo))
;; This is the CLEANUP-FORM it will _always_
;; be evaluated, despite the `throw'.
(close temporary-file)))
=> ()
back to top
Errors
......
Errors are a type of non-local exit; when a form can not be evaluated
for some reason an error is normally "signalled". If an error-handler
has been installed for that type of error, control is passed to the
handler for that error, and evaluation continues. If there is no
suitable handler, control is passed back to the innermost input loop
and a suitable error message is printed.
- Function: signal error-symbol data
Signals that an error has happened. ERROR-SYMBOL is a symbol
classifying the type of error, it should have a property
`error-message' (a string) which is the error message to be
printed.
DATA is a list of objects which are relevant to the error -- they
will be made available to any error-handler or printed with the
error message otherwise.
(signal 'void-value '(some-symbol))
error--> Value as variable is void: some-symbol
- Variable: debug-on-error
This variable is consulted by the function `signal'. If its value
is either `t' or a list containing the ERROR-SYMBOL to `signal' as
one of its elements, the Lisp debugger is entered. When the
debugger exits the error is signalled as normal.
- Variable: backtrace-on-error
Similar to `debug-on-error', but if an error is matched, the
current backtrace is printed to the standard error stream, and
control continues.
When you expect an error to occur and need to be able to regain
control afterwards the `condition-case' special form may be used.
- Macro: condition-case symbol body-form error-handlers...
`condition-case' evaluates the BODY-FORM with the ERROR-HANDLERS
in place. If an error occurs and one of the handles matches the
error, then it is evaluated with the value of SYMBOL set to the
error information.
Each of the ERROR-HANDLERS is a list whose car is a symbol
defining the type of error which this handler catches. The cdr of
the list is a list of forms to be evaluated in a `progn' if the
handler is invoked.
While the forms of the error handler are being evaluated the
variable SYMBOL is bound to the value `(ERROR-SYMBOL . DATA)'
(these were the arguments to the `signal' form which caused the
error). If SYMBOL is the symbol `nil' (or the empty list `()'),
then the error information is not available to the handler.
The special value, the symbol `error', in the car of one of the
ERROR-HANDLERS will catch _all_ types of errors.
(condition-case data
(signal 'file-error '("File not found" "/tmp/foo"))
(file-error
data)
(error
(setq x z))) ;Default handler
=> (file-error "File not found" "/tmp/foo")
back to top
Continuations
=============
Whenever a function is called, there is a control path waiting to
receive the result of the function, e.g. often the form following the
function invocation. This waiting control path is called the
"continuation" of the function, since control will continue down this
path when the called function exits.
These continuations are usually not paid much thought, but in some
cases it may be useful to be able to directly manipulate the
continuation of a function. For this purpose rep provides the
`call-with-current-continuation' function (often shortened to
`call/cc') that is standard in the Scheme dialect of Lisp.
- Function: call/cc function
FUNCTION is a function with a single parameter; it will be
immediately invoked with this parameter bound to an object
representing the current continuation (i.e. the control path that
would be taken after FUNCTION exits).
The continuation object passed to FUNCTION is itself a function
accepting a single argument, when called it transfers control to
the continuation of FUNCTION, as if FUNCTION had returned the
argument applied to the continuation object.
- Function: call-with-current-continuation function
This is an alias for `call/cc'.
In its simplest form, `call/cc' can mimic the `catch' and `throw'
procedures (*note Catch and Throw::), for example:
(defun foo (bar)
(call/cc (lambda (esc)
(when (null bar)
;; throws out of the call/cc
(esc nil))
;; do something with bar
...
this is roughly equivalent to:
(defun foo (bar)
(catch 'tag
(when (null bar)
(throw 'tag nil))
;; do something with bar
...
This is only half the story--the most powerful feature of `call/cc'
is that since continuations have dynamic extent (that is, no object is
freed until no references to it exist) it is possible to return control
to scopes that have already exited.
For example, consider the following fragment of a lisp interaction:
(prog1 (call/cc (lambda (esc)
(setq cont esc)))
(message "foo!"))
-| foo!
=> #
cont
=> #
(cont 10)
-| foo!
=> 10
The continuation of the `prog1' form is saved into the variable `cont'.
When subsequently called with a single argument, it has exactly the
same effect as the first time that the second form in the `prog1'
construct was evaluated.
Implementation Notes
--------------------
`call/cc' works by making a copy of the process' entire call stack.
For this reason, it is likely to be less efficient than using the
control structures described in the previous parts of this section. Of
course, it is much more powerful than the other constructs, so this
often outweighs the slight inefficiency.
Also note that currently no attempt is made to save or restore the
dynamic state of the Lisp system, apart from variable bindings (both
lexical and special). This means that any `unwind-protect',
`condition-case' or `catch' forms that are active when invoking a
continuation are all ignored.
Another restriction is that invoking a continuation may not cause
control to pass across a dynamic root (*note Threads::).
back to top
Threads
=======
`librep' supports a simple model of multi-threaded programming.
Multiple threads of execution may be created, with control preemptively
being switched between them.
Unless otherwise noted, all definitions described in this section are
provided by the `rep.threads' module.
* Menu:
* Thread Contexts::
* Creating Threads::
* Deleting Threads::
* Manipulating Threads::
* Mutexes::
* Thread Implementation Notes::
back to top
Thread Contexts
---------------
Every thread created by rep is a member of a "thread context", this
context is defined by the current position in the lisp call stack. At
any point in time, only threads which are members of the current
context may be executing.
- Function: call-with-dynamic-root thunk
Call the function of zero-parameters THUNK in a new thread
context. The new context will contain a single thread, that
executing THUNK.
The call to `call-with-dynamic-root' will only return once all
threads in the newly created context have been deleted, or a
non-local exit causes control to leave forcibly.
back to top
Creating Threads
----------------
The `make-thread' function may be used to create threads that
execute within the current thread context (dynamic root). Each thread
is represented by a lisp object.
- Function: threadp arg
Return true if lisp object ARG represents a thread of execution in
the lisp environment.
- Function: make-thread thunk #!optional name
Create and return a new thread of execution; it will initially
invoke the zero-parameter function THUNK. If the call to THUNK
returns the thread is automatically deleted.
If NAME is defined, it is a string naming the current thread.
- Function: make-suspended-thread #!optional name
Similar to `make-thread', except that the newly created thread
will be immediately suspended from running.
- Function: current-thread
Returns the currently executing thread. If no threads have been
created yet in the current dynamic root (i.e. there is a single
"implicit" thread) then false is returned.
- Function: all-threads
Returns a newly-created list containing all threads in the current
dynamic root. If no threads have been created yet, returns a null
list.
back to top
Deleting Threads
----------------
A thread may be deleted by either returning from the function
specified when it was created, or by explicit deletion. Also, the
implicit thread created by the `call-with-dynamic-root' function may be
deleted by exiting from the function called in the new context.
- Function: thread-delete #!optional thread
Mark THREAD (or the current thread), as being deleted. It will not
be switched to in the future. If the current thread is deleted,
control will be passed to the next runnable thread. Deleting the
last runnable thread results forces the containing dynamic root to
be closed.
- Function: thread-deleted-p thread
Returns true if THREAD has been deleted.
back to top
Manipulating Threads
--------------------
- Function: thread-yield
This function may be used to pass control away from the current
thread if other threads are waiting to run. There is usually no
need to call this function since running threads will be preempted
after a period of time.
- Function: thread-suspend #!optional thread milliseconds
Mark THREAD (or the current thread) as being suspended. It will
not be selected until either it has had this status removed, or
MILLISECONDS milliseconds time has passed.
Suspending the current thread will pass control to the next
runnable thread in the same dynamic root. If there are no runnable
threads, then the interpreter will sleep until the next thread
becomes runnable.
- Function: thread-join thread #!optional timeout default-value
Suspends the current thread until either THREAD has exited, or
TIMEOUT milliseconds have passed.
If THREAD exits normally, then the value of the last form it
evaluated is returned; otherwise DEFAULT-VALUE is returned.
It is an error to call `thread-join' on a THREAD that is not a
member of the current dynamic root.
- Function: thread-wake thread
Remove the suspended state from thread THREAD. It will then be
scheduled for execution sometime subsequently, if its dynamic root
is active.
- Function: thread-suspended-p thread
Returns true if THREAD is currently suspended.
Thread preemption may be forbidden at times, to allow atomic
operations to take place. Each dynamic root has its own "forbid
counter". Only when this counter is zero may the current thread be
preempted.
- Function: thread-forbid
Increment the forbid count.
- Function: thread-permit
Decrement the forbid count.
- Macro: without-interrupts #!rest forms
Evaluate the list of forms FORMS with thread preemption
temporarily disabled.
back to top
Mutual Exclusion Devices
------------------------
"Mutexes" are lisp objects used to coordinate access to data shared
across multiple threads (where interleaved access would be bad). These
functions are exported by the `rep.threads.mutex' module (*note
Modules::).
- Function: make-mutex
Create and return a mutex object. No thread will own the new mutex.
- Function: mutexp arg
Return true if ARG is a mutex object.
- Function: obtain-mutex mutex
Obtain the mutex MUTEX for the current thread. Will suspend the
current thread until the mutex is exclusively available.
- Function: maybe-obtain-mutex mutex
Attempt to obtain mutex MUTEX for the current thread without
blocking. Returns true if able to obtain the mutex, false
otherwise.
- Function: release-mutex mutex
Release the mutex object MUTEX (which must have previously been
obtained by the current thread). Returns true if the mutex has no
new owner.
back to top
Thread Implementation Notes
---------------------------
The threads used by `librep' are _software threads_. This means that
they are currently implemented by manually switching in and out thread
context (i.e. the call stack) as required. There are a number of
disadvantages to this method:
* blocking I/O blocks _all_ threads, not just the thread doing the
I/O,
* only a single processor is used, thereby avoiding any true
parallelism on multi-processor systems.
The main advantage is the ease of implementation, especially when
retrofitting threads into the previously single-threaded interpreter.
back to top
Loading
=======
In Lisp, programs (also called "modules", or "libraries") are stored
in files. Each file is a sequence of Lisp forms (known as "top-level
forms"). Most of the top-level forms in a program will be definitions
(i.e. function, macro or variable definitions) since generally each
library is a system of related functions and variables.
Before the program can be used it has to be "loaded" into the
editor's workspace; this involves reading and evaluating each top-level
form in the file, i.e. instantiating all function definitions, or
whatever.
* Menu:
* Load Function:: The function which loads programs
* Autoloading:: Functions can be loaded on reference
* Features:: Module management functions
back to top
Load Function
-------------
- Function: load program #!optional no-error no-path no-suffix
This function loads the file containing the program called PROGRAM;
first the file is located then each top-level form contained by
the file is read and evaluated in order.
Each directory named by the variable `load-path' is searched until
the file containing PROGRAM is found. In each directory three
different file names are tried,
1. PROGRAM with `.jlc' appended to it. Files with a `.jlc'
suffix are usually compiled Lisp files. *Note Compiled Lisp::.
2. PROGRAM with `.jl' appended, most uncompiled Lisp programs are
stored in files with names like this.
3. PROGRAM with no modifications.
If none of these gives a result the next directory is searched in
the same way, when all directories in `load-path' have been
exhausted and the file still has not been found an error is
signalled.
Next the file is opened for reading and Lisp forms are read from it
one at a time, each form is evaluated before the next form is
read. When the end of the file is reached the file has been loaded
and this function returns true.
The optional arguments to this function are used to modify its
behaviour,
NO-ERROR
When this argument is true no error is signalled if the file
can not be located. Instead the function returns false.
NO-PATH
The variable `load-path' is not used, PROGRAM must point to
the file from the current working directory.
NO-SUFFIX
When true no `.jlc' or `.jl' suffixes are applied to the
PROGRAM argument when locating the file.
If a version of the program whose name ends in `.jlc' is older than
a `.jl' version of the same file (i.e. the source code is newer
than the compiled version) a warning is displayed and the `.jl'
version is used.
If no Lisp file can be found matching PROGRAM, then each directory
in the variable `dl-load-path' is searched for a `libtool' shared
library called `PROGRAM.la' (*note Shared Libraries::).
- Variable: load-filename
Whilst loading a Lisp library, this variable is bound to the name
of the file being loaded.
- Variable: load-path
A list of strings, each element is the name of a directory which is
prefixed to the name of a program when Lisp program files are being
searched for.
load-path
=> ("/usr/local/lib/rep/1.0/lisp/"
"/usr/local/lib/rep/site-lisp/" "")
The element `""' refers to the current directory, note that
directory names should have an ending `/' (or whatever) so that
when concatenated with the name of the file they make a meaningful
filename.
- Variable: dl-load-path
A list of strings defining all directories to search for shared
libraries.
- Variable: lisp-lib-directory
The name of the directory in which the standard Lisp files are
stored.
lisp-lib-dir
=> "/usr/local/lib/rep/1.0/lisp/"
- Variable: after-load-alist
An association list of elements of the format `(FILE FORMS ...)'.
When the library FILE is loaded, all FORMS are executed. However,
note that FILE must _exactly_ match the PROGRAM argument to the
`load' function.
- Function: eval-after-load library form
Arrange for FORM to be evaluated immediately after the Lisp
library of LIBRARY has been read by the `load' function. Note
that LIBRARY must exactly match the PROGRAM argument to `load'.
back to top
Autoloading
-----------
Obviously, not all features of the `librep' environment are always
used. "Autoloading" allows libraries to only be loaded when they are
first required. This speeds up the initialisation process and may save
memory.
Functions which may be autoloaded have a special form in their
symbol's function cell--an "autoload form". This is a special kind of
closure. When the function call dispatcher finds one of these forms it
loads the program file specified in the form then re-evaluates the
function call. The true function definition will then have been loaded
and therefore the call may proceed as normal.
Autoload stubs may be created through the `autoload' function.
- Function: autoload symbol file #!optional is-command
Installs an autoload form into the symbol SYMBOL. It marks that
when SYMBOL is called as a function the lisp library FILE should
be loaded to provided the actual definition of SYMBOL.
It is not necessary to call the `autoload' function manually.
Simply prefix the definitions of all the functions that may be
autoloaded (i.e. the entry points to your module; _not_ all the
internal functions.) with the magic comment `;;;###autoload'. Then load
the file into the Jade editor and invoke the `add-autoloads' command,
creating all the necessary calls to the autoload function in the
`autoloads.jl' Lisp file (this file which lives in the Lisp library
directory is loaded when the environment is initialised).
`Meta-x add-autoloads'
Scans the current buffer for any autoload definitions. Functions
with the comment `;;;###autoload' preceding them have autoload
forms inserted into the `autoloads.jl' file. Simply save this
file's buffer and the new autoloads will be used the next time
Jade is initialised.
It is also possible to mark arbitrary forms for inclusion in the
`autoloads.jl' file: put them on a single line which starts with
the comment `;;;###autoload' call the command.
The unsaved `autoloads.jl' buffer will become the current buffer.
;;;###autoload
(defun foo (bar) ;`foo' is to be autoloaded
...
;;;###autoload (setq x y) ;Form to eval on initialisation
`Meta-x remove-autoloads'
Remove all autoload forms from the `autoloads.jl' file which are
marked by the `;;;###autoload' comment in the current buffer.
The unsaved `autoloads.jl' buffer will become the current buffer.
XXX these editor commands don't really belong here, but they'll do
for now...
back to top
Features
--------
"Features" correspond to libraries of Lisp code. Each feature is
loaded separately. Each feature has a name, when a certain feature is
required its user asks for it to be present (with the `require'
function), the feature may then be used as normal.
When a feature is loaded one of the top-level forms evaluated is a
call to the `provide' function. This names the feature and installs it
into the list of present features.
- Variable: features
A list of the features currently present (that is, loaded) in the
current module. Each feature is represented by a symbol. Usually
the print name of the symbol (the name of the feature) is the same
as the name of the file it was loaded from, minus any `.jl' or
`.jlc' suffix.
features
=> (info isearch fill-mode texinfo-mode lisp-mode xc)
- Function: featurep feature
Returns true if the feature FEATURE has been loaded into the
current module.
- Function: provide feature
Adds FEATURE (a symbol) to the list of loaded features. A call to
this function is normally one of the top-level forms in a file.
;;;; maths.jl -- the `maths' library
(provide 'maths)
...
- Function: require feature #!optional file
Show that the caller is planning to use the feature FEATURE (a
symbol). This function will check the `features' variable to see
if FEATURE is already loaded, if so it will return immediately.
If FEATURE is not present it will be loaded. If FILE is given it
specifies the first argument to the `load' function, else the
print name of the symbol FEATURE is used, with any `.' characters
replaced by the operating system's directory separator (*note
Module Loading::).
;;;; physics.jl -- the `physics' library
(require 'maths) ;Need the `maths' library
(provide 'physics)
...
When called interactively the symbol FEATURE is prompted for.
Features may also be provided by modules, for more details *Note
Module Loading::.
back to top
Compiled Lisp
=============
`librep' contains a Lisp compiler as well as an interpreter; this
takes a Lisp form or program and compiles it into a "byte-code" object.
This byte-code object is a string of characters representing virtual
machine instructions, a vector of constants and some other
meta-information. The system also contains a byte-code interpreter;
this takes the compiled byte-codes and executes them by simulating the
virtual machine. This simulation will have exactly the same effect as
interpreting the original form or program.
One of the main reasons for compiling programs is to increase their
efficiency. Compiled functions are likely to be more efficient than
interpreted counterparts in all areas (space and time). For example:
user> (define (fib n) (if (<= n 2) 1 (+ (fib (- n 1)) (fib (- n 2)))))
user> ,time (fib 30)
832040
Elapsed: 17.05572 seconds
user> ,compile
user> ,time (fib 30)
832040
Elapsed: 1.479007 seconds
--the compiled function is over an order of magnitude faster than the
interpreted version.
* Menu:
* Compilation Functions:: How to compile Lisp programs
* Compiler Declarations:: Hinting to the compiler
* Compilation Tips:: Getting the most out of the compiler
* Disassembly:: Examining compiled functions
back to top
Compilation Functions
---------------------
- Function: compile-form form
This function compiles the Lisp form FORM into a byte-code form
which is returned.
(compile-form '(setq foo bar))
=> (run-byte-code "F!" [bar foo] 2)
- Command: compile-function function
This function replaces the uncompiled body of the function FUNCTION
(a symbol) with a compiled version, then returns FUNCTION.
- Command: compile-file file-name
This function compiles the file called FILE-NAME into a file of
compiled Lisp forms whose name is FILE-NAME with `c' appended to
it (i.e. if FILE-NAME is `foo.jl' it will be compiled to
`foo.jlc').
If an error occurs while the file is being compiled any
semi-written file will be deleted.
When called interactively this function will ask for the value of
FILE-NAME.
- Command: compile-directory directory #!optional force exclude
Compiles all the Lisp files in the directory called DIRECTORY which
either haven't been compiled or whose compiled version is older
than the source file (Lisp files are those ending in `.jl').
If the optional argument FORCE is true _all_ Lisp files will be
recompiled whatever the status of their compiled version.
The EXCLUDE argument may be a list of filenames, these files will
_not_ be compiled.
When this function is called interactively it prompts for the
directory.
- Command: compile-module module-name
Compiles all uncompiled function definitions in the module named
MODULE-NAME (a symbol).
When called interactively the module name will be prompted for.
- Function: run-byte-code byte-codes constants stack
Interprets the string of byte instructions BYTE-CODES with the
vector of constants CONSTANTS.
This function should _never_ be called by hand. The compiler will
produce calls to this function when it compiles a form or a
function.
There is a second form that byte-code objects can take: a vector
whose read syntax includes a preceding `#' character is a "byte-code
subr". These objects represent compiled Lisp functions and macros.
- Function: bytecodep arg
Returns true if ARG is a byte-code subroutine.
back to top
Compiler Declarations
---------------------
It is often useful to be able to give the compiler extra knowledge
about the program forms that it is compiling. The language includes
special declaration forms that have no effect when interpreted, but are
meaningful to the compiler as it traverses the program.
- Macro: declare clause...
Offer the information contained in the CLAUSE... forms to the
compiler, which it may or may not use when compiling the program.
Each CLAUSE is a list, the first element of each clause is a
symbol defining the type of declaration, the interpretation of any
other elements in the clause depends upon the declaration type.
The following table lists the syntax of all currently supported
declaration types:
`(bound VARIABLES...)'
This declaration tells the compiler that all symbols VARIABLES
have lexical bindings for the extent of the current lexical
scope. This is often useful to prevent spurious compiler
warnings.
`(special VARIABLES...)'
This tells the compiler that all symbols VARIABLES have
special (dynamic) bindings for the extent of the current
lexical scope.
(It is important that the compiler is able to distinguish
special bindings from lexical bindings, since different
instruction sequences must be generated to access the
different types of binding.)
`(unused VARIABLES...)'
Directs the compiler not to warn about bindings for
VARIABLES... being unreferenced.
`(inline NAMES...)'
Tells the compiler that it should consider inlining calls to
the functions called NAMES.... Inlining will only occur if
these functions are declared in the same module as, and
after, the declaration itself.
`(in-module MODULE-NAME)'
This declaration should occur at the top-level of a program;
it tells the compiler that the forms in the program will be
evaluated within the context of the module called MODULE-NAME
(a symbol).
`(language MODULE)'
Explicitly specifies the particular language dialect that the
current module or file body is written for. Language dialects
included with the librep distribution include `rep', `scheme'
and `unscheme'. These are also the names of the modules that
should be imported to use a particular dialect.
By default, the `rep' dialect is assumed for code outside
module definitions. For code inside a module definition the
list of imported modules is scanned for a known language
dialect (i.e. if the module imports `rep', then the rep
language dialect is compiled for).
`(unsafe-for-call/cc)'
Tell the compiler that it may register-allocate variables,
even if it can't prove that doing so wouldn't produce
incorrect results if `call/cc' causes a function call to
return more than once (*note Continuations::). This
declaration applies to the entire file that it occurs in.
Without this declaration, the compiler will only
register-allocate bindings if the following conditions are
met:
* the binding is not accessed from any inner closures, and,
* the binding is never modified after being initialized
(actually, the binding may be modified between being
intialized and the next function call)
this declaration is often useful where `call/cc' isn't used,
and there is a lot of side effecting of local variables.
Declaration forms always evaluate to false.
A second type of declaration is the `eval-when-compile' form, it
allows Lisp forms to be evaluated only at compile-time.
- Macro: eval-when-compile form
This form tells the system that FORM should only be evaluated when
the containing code is being compiled.
The compiler knows to recognize FORMs of the pattern
`(eval-when-compile (require 'FEATURE))' as marking that FEATURE
should be imported at compile-time. Any other FORMs are simply
evaluated in an unspecified environment.
When interpreted, `eval-when-compile' forms alway evaluate to
false, when compiled they evaluate to the result of evaluating the
FORM at compile-time.
back to top
Compilation Tips
----------------
Here are some tips for making compiled code run fast(er):
* Instead of using `while' loops to traverse lists, use `mapc' or
tail recursion.
For example you might code a function to scan a list using
iteration through a `while' loop:
(defun scan-list (lst elt)
"Search the LST for an element similar to ELT.
Return it if one is found."
(catch 'return
(while (consp lst)
(when (equal (car lst) elt)
(throw 'return elt))
(setq lst (cdr lst)))))
As well as obscuring what is actually happening, this will
probably be fairly slow to execute. A more elegant solution is to
use tail-recursion:
(defun scan-list (lst elt)
(if (equal (car lst) elt)
elt
(scan-list (cdr lst) elt)))
An alternative idiom is to map an anonymous function over the list
using the `mapc' function:
(defun scan-list (lst elt)
(catch 'return
(mapc (lambda (x)
(when (equal x elt)
(throw 'return elt)))
lst)
nil))
In fact, the compiler knows that calls to `mapc' with a constant
lambda expression can be open-coded, so it will code the list
traversal directly using the virtual machine stack.
However, in most cases the execution time differences are likely to
negligible.
* In some cases the functions `member', `memq', `assoc', etc... can
be used to search lists. Since these are primitives written in C
they will probably execute several times faster than an equivalent
Lisp function.
So the above `scan-list' example can again be rewritten, this time
as:
(defun scan-list (lst elt)
(car (member elt lst)))
* All conditional structures are equivalent when compiled (they are
all translated to `cond' statements), so use whichever is the
easiest to understand.
* A certain amount of constant folding is performed. If a function is
known to be free of side effects, and all its arguments are
constants, then it is evaluated at compile-time, and the result
folded into the program in its place. For example
(logor (lsh 1 6) x)
==> (logor 32 x)
* Careful use of named constants (*note Defining Variables::) can
increase the speed of some programs. For example, in the Lisp
compiler itself all the opcode values (small integers) are defined
as constants.
It must be stressed that in some cases constants may _not_ be
suitable; they may drastically increase the size of the compiled
program (when the constants are `big' objects, i.e. long lists) or
even introduce subtle bugs (since two references to the same
constant may not be `eq' whereas two references to the same
variable are always `eq').
* Many primitives have corresponding byte-code instructions; these
primitives will be quicker to call than those that don't (and
incur a normal function call). Currently, the functions which have
byte-code instructions (apart from all the special forms) are:
`cons', `car', `cdr', `rplaca', `rplacd', `nth', `nthcdr', `aset',
`aref', `length', `eval', `+', `*', `/', `%', `mod', `lognot',
`not', `logior', `logand', `logxor', `equal', `eq', `=', `/=',
`>', `<', `>=', `<=', `1+', `1-', `-', `set', `lsh', `zerop',
`null', `atom', `consp', `listp', `numberp', `stringp', `vectorp',
`throw', `boundp', `symbolp', `get', `put', `signal', `return',
`reverse', `nreverse', `assoc', `assq', `rassoc', `rassq', `last',
`mapcar', `mapc', `member', `memq', `delete', `delq', `delete-if',
`delete-if-not', `copy-sequence', `sequencep', `functionp',
`special-form-p', `subrp', `eql', `max', `min', `filter',
`macrop', `bytecodep', `bind-object'.
* When a file is being compiled each top-level form it contains is
inspected to see if it should be compiled into a byte-code form.
Different types of form are processed in different ways:
* Function and macro definitions have their body forms compiled
into a single byte-code form. The doc-string and interactive
declaration are not compiled.
* If the form is a list form (*note List Forms::) and the
symbol which is the car of the list is one of:
`if', `cond', `when', `unless', `let', `let*', `catch',
`unwind-protect', `error-protect', `with-buffer',
`with-window', `progn', `prog1', `prog2', `while', `and',
`or', `case'.
then the form is compiled. Otherwise it is just written to
the output file in its uncompiled state.
If your program contains a lot of top-level forms which you know
will not be compiled automatically, consider putting them in a
`progn' block to make the compiler coalesce them into one
byte-code form.
back to top
Disassembly
-----------
It is possible to disassemble byte-code forms; originally this was
so I could figure out why the compiler wasn't working but if you're
curious about how the compiler compiles a form it may be of use to you.
Naturally, the output of the disassembler is a listing in the
assembly language of the `librep' virtual machine--it won't take a
byte-code form and produce the equivalent Lisp code!
- Command: disassemble-fun function #!optional stream
This function disassembles the compile Lisp function FUNCTION. It
writes a listing to the output stream STREAM (normally the value
of the `standard-output' variable).
When called interactively it will prompt for a function to
disassemble.
When reading the output of the disassembler bear in mind that
`librep' simulates a stack machine for the code to run on. All
calculations are performed on the stack, the value left on the stack
when the piece of code ends is the value of the byte-code form.
Here is a small example. Consider the `fib' function given at the
start of this section:
(define (fib n)
(if (<= n 2)
1
(+ (fib (- n 1))
(fib (- n 2)))))
After compilation and disassembly, the following is produced (but
without the annotations):
Disassembly of #:
21 bytes, 1 constants, and (5,0,1) stack slots
0 required-arg ;requires a single parameter
1 dup
2 slot-set #0 ;store it in register 0 (r0)
3 pushi 2
4 le
5 jn 10 ;unless r0 <= 2, goto 10
8 pushi 1
9 return ;else, return 1
10 refg [0] fib
11 slot-ref #0
12 dec
13 call #1 ;push result of (fib (1- n))
14 refg [0] fib
15 slot-ref #0
16 pushi 2
17 sub
18 call #1 ;push (fib (- n 2))
19 add
20 return ;return the sum of the two calls
back to top
Datums
======
"Datums" are the mechanism by which `librep' allows Lisp programs to
define new data types, such that these types are completely distinct
from the built-in data types (i.e. they match none of the standard type
predicates).
They also provide encapsulation, in that the data objects they
provide are completely opaque, unless a pre-defined value is known
(which was specified when the object was created, and is typically
known only by the object's creator).
- Function: make-datum value key
Create and return a new datum object. It has the value VALUE
associated with it, and has type KEY.
- Function: datum-ref arg key
If ARG has type KEY, then return the value associated with it.
Otherwise, an error is signalled.
- Function: datum-set arg key value
If ARG has type KEY, then set the value associated with it to be
VALUE. Otherwise, an error is signalled.
- Function: has-type-p arg key
Return true if ARG has type KEY.
- Function: define-datum-printer key printer
Associate the function PRINTER with all datum objects of type KEY.
When any such object is printed, PRINTER is applied to two
arguments, the datum and the stream to which it should be printed
(*note Output Streams::).
back to top
Queues
======
A "queue" is an ordered set of objects, such that objects enter at
one end of the queue (the "tail"), and leave from the other end of the
queue (the "head"). The acts of entering and leaving a queue are often
called "enqueing" and "dequeueing".
`librep' provides a straightforward queue implementation,
implemented by the `rep.data.queues' module (*note Modules::).
- Function: make-queue
Create and return a new queue object. The queue will initially be
empty.
- Function: enqueue q arg
Add the object ARG to the tail of the queue Q.
- Function: dequeue q
Remove the object at the head of the queue Q, and return it. If Q
is empty, an error is signalled.
- Function: queue-empty-p q
Return true if the queue Q is not empty.
- Function: queuep arg
Return true if the object ARG is a queue.
- Function: queue->list q
Return a list of objects representing the contents of the queue Q,
with objects ordered from head to tail. Modifying the list
structure causes undefined effects to the queue itself.
- Function: queue-length q
Return the number of objects stored in the queue Q.
- Function: delete-from-queue q arg
Removes any occurrences of the object ARG from the queue Q.
back to top
Records
=======
`librep' provides a convenient means of defining structured data
types, these types are known as "records". Each record is a distinct
data type, meaning that there will only be a single type-predicate
matching objects of any individual record type.
All definitions documented in this section are provided by the
`rep.data.records' module (*note Modules::).
Record types are defined using the `define-record-type' macro, this
in turn defines a number of functions implementing the type. These
functions include a constructor, a type predicate, and a user-defined
set of field-accessor and -modifier functions.
- Macro: define-record-type type (constructor fields...) [predicate]
(field accessor [modifier])...
This macro creates a new record type storing an opaque object
identifying the type in the variable named TYPE.
It then defines a function CONSTRUCTOR with parameter list as
specified by the FIELDS..., and a predicate function called
PREDICATE if PREDICATE is given.
The fields of the record are defined by the sequence of `(FIELD
ACCESSOR [MODIFIER])' forms, each form describes a single field
(named FIELD, which may match one of the constructor arguments).
For each field a function ACCESSOR will be defined that when
applied to an argument of the record type, returns the value
stored in the associated FIELD. If the MODIFIER name is defined a
function will be defined of that name, that when applied to a
record and an object, stores the object into the associated field
of the record.
Note that the FIELDS... may include all the standard lambda-list
features (*note Lambda Expressions::), including keyword
parameters and default values.
Here is an example record definition:
(define-record-type :pare
(kons x y) ; constructor
pare? ; predicate
(x kar set-kar!) ; fields w/ optional accessors
(y kdr)) ;and modifiers
the variable `:pare' is bound to the record type. Following this
definition, the record type could be used as follows:
(define x (kons 1 2))
(pare? x)
=> t
(kar x)
=> 1
(set-kar! x 42)
(kar x)
=> 42
By default record objects print as the name of their type in angle
brackets, e.g. for the above `pare' type, each object would print as
the string `#<:pare>'. This may be redefined using the
`define-record-discloser' function.
- Function: define-record-discloser type discloser
Associate the function DISCLOSER with the record type TYPE. When
any record of this type is printed, DISCLOSER is applied to the
object, it should return the value that will actually be printed.
For the above example, the following could be used:
(define-record-discloser :pare (lambda (x) `(pare ,(kar x) ,(kdr x))))
(kons 'a 'b)
=> (pare a b)
Constructors for records with large numbers of fields often benefit
from using keyword parameters. For example the `kons' record above
could be defined as follows (though this would make more sense if it
had more than two fields):
(define-record-type :pare
(kons #!key (kar 1) (kdr 2))
pare?
(kar kar set-kar!)
(kdr kdr set-kdr!))
(kons #:kar 42) => (pare 42 2)
(kons #:kdr 42) => (pare 1 42)
back to top
Hash Tables
===========
The `rep.data.tables' module provides a flexible hash table
implementation (*note Modules::). Each hash table is represented by a
lisp object satisfying the `tablep' predicate:
- Function: tablep arg
Return true if ARG is a hash table.
Hash tables may be created by using the `make-table' and
`make-weak-table' functions:
- Function: make-table hash-fun compare-fun
Create and return a new hash table. When storing and referencing
keys it will use the function HASH-FUN to map keys to hash codes
(positive fixnums), and the predicate function COMPARE-FUN to
compare two keys (should return true if the keys are considered
equal).
- Function: make-weak-table hash-fun compare-fun
Similar to `make-table', except that key-value pairs stored in the
table are said to be "weakly keyed". That is, they are only
retained in the table as long the key has not been garbage
collected.
Unlike with tables created by the `make-table' function, the fact
that the key is stored in the table is not considered good enough
to prevent it being garbage collected.
- Function: table-ref table key
Return the value stored in hash table TABLE indexed by object KEY.
Returns false if no such value exists.
- Function: table-bound-p table key
Returns true if the hash table TABLE contains a value associated
with KEY.
- Function: table-set table key value
Associate the value VALUE with KEY in hash table TABLE. Returns
`value'.
- Function: table-unset table key
Remove any value stored in TABLE associated with KEY.
- Function: table-walk function table
Call function FUNCTION for every key-value pair stored in hash
table TABLE. For each pair, the function is called with arguments
`(KEY VALUE)'.
Several hash functions are also provided:
- Function: string-hash string
Return an integer representing the string STRING.
- Function: symbol-hash symbol
Call `(string-hash (symbol-name SYMBOL))'.
- Function: eq-hash arg
Return a hash value representing object ARG. The hash is generated
from the _address_ of the object.
- Function: equal-hash arg
Return a hash value representing object ARG. The hash is generated
from the _contents_ of the object.
back to top
Guardians
=========
A "guardian" is a lisp object used to control when other data
objects are recycled by the garbage collector (*note Garbage
Collection::).(1) The usual behaviour of the collector is to recycle
objects as soon as they have no remaining references.
Guardians allow the programmer to detect when a specified object
would be freed by the garbage collector, and to implement their own
allocation policy. This can be useful, for example, with objects that
have a high creation-overhead, and thus need to be cached for
performance reasons.
- Function: make-guardian
This function allocates and returns a new guardian. Each guardian
has a list of data objects associated with it; some of which may
have been proved to have no remaining references to them (except
from the guardian system).
Calling the guardian object with a single argument, adds that
value to the list of objects associated with that guardian.
Calling the guardian with no arguments has one of two effects:
* If objects are associated with the guardian that have been
proved to be inaccessible, then return one of those objects,
and remove it from the list of objects associated with the
guardian.
* If none of the associated objects have been proved to be
inaccessible, then return the value false.
Note the use of the word "prove" in the above description, objects
are only moved into a guardian's inaccessible set by the garbage
collector.
Here is an example use of the guardian system:
;; create a new guardian object
(setq G (make-guardian))
;; create a lisp object
(setq x (cons 'a 'b))
=> (a . b)
;; protect the object using the guardian
(G x)
;; remove the sole reference to the object
(setq x nil)
=> ()
;; invoke the garbage collector, this will
;; prove that the value added to the
;; guardian is no longer accessible
(garbage-collect)
;; call the guardian to retrieve the
;; inaccessible value
(G)
=> (a . b)
;; no more inaccessible values available
(G)
=> ()
---------- Footnotes ----------
(1) Guardians were first described in a paper by R. Kent Dybvig,
Carl Bruggeman, and David Eby: `"Guardians in a Generation-Based
Garbage Collector", ACM SIGPLAN Conference on Programming Language
Design and Implementation, June 1993.'
back to top
Streams
=======
A "stream" is a Lisp object which is either a data sink (an "output
stream") or a data source (an "input stream"). All streams produce or
consume sequences of 8-bit characters.
Streams are very flexible, functions using streams for their input
and output do not need to know the type of stream being accessed. For
example the Lisp reader (the `read' function) takes an input stream as
its sole argument, it then reads characters from this stream until it
has parsed a whole object. This stream could be a file, a function, or
even a string; the `read' function does not need to differentiate.
- Function: streamp arg
Return true if ARG is a stream.
- Function: input-stream-p arg
Return true if ARG is an input stream.
- Function: output-stream-p arg
Return true if ARG is an output stream.
* Menu:
* Input Streams:: Types of input stream
* Output Streams:: Types of output stream
* Input Functions:: Functions to read from streams
* Output Functions:: How to output to a stream
* Formatted Output:: Output by template
back to top
Input Streams
-------------
These are the possible types of input stream, for the functions which
use them see *Note Input Functions::.
`FILE'
Characters are read from the file object FILE, for the functions
which manipulate file objects see *Note Files::.
`FUNCTION'
Each time an input character is required the FUNCTION is called
with no arguments. It should return the character read (an
integer) or false if for some reason no character is available.
FUNCTION should also be able to `unread' one character. When this
happens the function will be called with one argument--the value of
the last character read. The function should arrange it so that the
next time it is called it returns this character. A possible
implementation could be,
(defvar ms-unread-char nil
"If true the character which was pushed back.")
(defun my-stream (#!optional unread-char)
(if unread-char
(setq ms-unread-char unread-char)
(if ms-unread-char
(prog1
ms-unread-char
(setq ms-unread-char nil))
;; Normal case -- read and return a character from somewhere
...
`nil'
Read from the stream stored in the variable `standard-input'.
It is also possible to use a string as an input stream. The string to
be read from must be applied to the `make-string-input-stream' function
and the result from this function used as the input stream.
- Function: make-string-input-stream string #!optional start
Returns an input stream which will supply the characters of the
string STRING in order starting with the character at position
START (or from position zero if this argument is undefined).
(read (make-string-input-stream "(1 . 2)"))
=> (1 . 2)
- Variable: standard-input
The input stream which is used when no other is specified or is
false.
Applications that embed `librep', or dynamically loaded extensions,
may provide further input stream types.
back to top
Output Streams
--------------
These are the different types of output stream, for the functions
which use them see *Note Output Functions::.
`FILE'
Writes to the file object FILE. *Note Files::.
`FUNCTION'
The function FUNCTION is called with one argument, either a string
or a character. This should be used as the circumstances dictate.
If the function returns a number it is the number of characters
actually used, otherwise it is assumed that all the characters
were successful.
`PROCESS'
Writes to the standard input of the process object PROCESS. If
PROCESS isn't running an error is signalled. *Note Processes::.
`t'
Appends the character(s) to the end of the status line message.
`()'
Write to the stream stored in the variable `standard-output'.
It is also possible to store the characters sent to an output stream
in a string.
- Function: make-string-output-stream
Returns an output stream. It accumulates the text sent to it for
the benefit of the `get-output-stream-string' function.
- Function: get-output-stream-string string-output-stream
Returns a string consisting of the text sent to the
STRING-OUTPUT-STREAM since the last call to
GET-OUTPUT-STREAM-STRING (or since this stream was created by
`make-string-output-stream').
(setq stream (make-string-output-stream))
=> ("" . 0)
(prin1 keymap-path stream)
=> ("(lisp-mode-keymap global-keymap)" . 64)
(get-output-stream-string stream)
=> "(lisp-mode-keymap global-keymap)"
- Variable: standard-output
This variable contains the output stream which is used when no
other is specified (or when the given output stream is false).
- Variable: standard-error
This variable contains the output stream which is used when an
error message is being reported.
Applications that embed `librep', or dynamically loaded extensions,
may provide further output stream types.
back to top
Input Functions
---------------
- Function: read-char stream
Read and return the next character from the input stream STREAM. If
the end of the stream is reached false is returned.
- Function: read-line stream
This function reads one line of text from the input stream STREAM,
a string containing the line (including the newline character
which terminates the line).
If the end of stream is reached before any characters can be read
false is returned, if the end of stream is reached but some
characters have been read (but not the newline) these characters
are made into a string and returned.
Note that unlike the Common Lisp function of the same name, the
newline character is not removed from the returned string.
- Function: read stream
This function is the function which encapsulates the Lisp reader
(*note The Lisp Reader::). It reads as many characters from the
input stream STREAM as required to form the read syntax of a single
Lisp object (*note Read Syntax::), this object is then returned.
- Function: read-from-string string #!optional start
Reads one Lisp object from the string STRING, the first character
is read from position START (or position zero).
(read-from-string STRING START)
==
(read (make-string-input-stream STRING START))
back to top
Output Functions
----------------
- Function: write stream data #!optional length
Writes the specified character(s) to the output stream STREAM.
DATA is either the character or the string to be written. If DATA
is a string the optional argument LENGTH may specify how many
characters are to be written. The value returned is the number of
characters successfully written.
(write standard-output "Testing 1.. 2.. 3..")
-| Testing 1.. 2.. 3..
=> 19
- Function: copy-stream input-stream output-stream
This function copies all characters which may be read from
INPUT-STREAM to OUTPUT-STREAM. The copying process is not stopped
until the end of the input stream is read. Returns the number of
characters copied.
Be warned, if you don't choose the streams carefully you may get a
deadlock which only an interrupt signal can break!
- Function: print object #!optional stream
Outputs a newline character to the output stream STREAM, then
writes a textual representation of OBJECT to the stream.
If possible, this representation will be the read syntax of OBJECT.
OBJECT is returned.
(print '(1 2 3))
-|
-| (1 2 3)
=> (1 2 3)
- Function: prin1 object #!optional stream
Similar to `print' but no initial newline is output.
(prin1 '(1 2 3))
-| (1 2 3)
=> (1 2 3)
(prin1 '|(xy((z]|) ;A strange symbol
-| \(xy\(\(z\]
=> \(xy\(\(z\]
- Function: prin1-to-string object
Returns a string containing the characters that `prin1' would
output when it prints OBJECT.
(prin1-to-string '(1 2 3))
=> "(1 2 3)"
- Function: princ object #!optional stream
Prints a textual representation of OBJECT to the output stream
STREAM. No steps are taken to create output that `read' can parse;
in particular, no double-quote characters surround strings.
(princ "foo")
-| foo
=> "foo"
(princ '|(xy((z]|)
-| (xy((z]
=> \(xy\(\(z\]
Several variables may be used to control how objects are printed.
- Variable: print-escape
This defines which control characters `print' and `prin1' will
escape (using backslashes). Possible values are:
`()'
Only escape double-quote and backslash characters.
`newlines'
Only escape double-quote, backslash, newline, `TAB', and
formfeed characters.
`t'
Escape double-quote, backslash, and all control characters
(anything with a numeric value less than 32, or greater than
126).
- Variable: print-length
This variable, if true, limits the number of elements printed from
lists.
- Variable: print-level
This variable, if true, limits the recursion depth when printing
lists.
back to top
Formatted Output
----------------
- Function: format stream template #!rest values
Writes to a stream, STREAM, a string constructed from the format
string, TEMPLATE, and list of arguments VALUES.
If STREAM is false the resulting string will be returned, not
written to a stream.
TEMPLATE is a template for the output string, any `%' characters
introduce a substitution, using the next unused argument. The
substitutions have the following syntax,
%[INDEX$][FLAGS][FIELD-WIDTH]CONVERSION
INDEX is an optional decimal number specifying exactly which of
the VALUES this conversion refers to (with the first at position
one), and is usually used when translating messages; by default the
next value is used.
FIELD-WIDTH is a positive decimal integer, defining the size in
characters of the substitution output.
CONVERSION is a character defining how to convert the
corresponding argument value to text. The default options are:
`s'
Write the printed representation of the value without quoting
(as if from the `princ' function).
`S'
Write the printed representation _with_ quoting enabled (like
the `prin1' function).
`d'
Output the value as a decimal number.
`o'
Write the value in octal.
`x'
`X'
In hexadecimal.
`c'
Write the character specified by the value.
`%'
Print a literal percent character. None of the VALUES are
used.
FLAGS is a sequence of zero or more of the following characters,
`_'
Left justify the substitution within the field.
`^'
Truncate the substitution at the size of the field.
`0'
Pad the field with zeros instead of spaces.
`+'
For `d', `x', and `o' conversions, output a leading plus sign
if the argument is positive.
` ' (a space)
For `d', `x', and `o' conversions, if the result doesn't
start with a plus or minus sign, output a leading space.
The list of CONVERSIONS can be extended through the
`format-hooks-alist' variable; the strings created by these extra
conversions are formatted as if by the `s' conversion.
Note that the FIELD-WIDTH and all flags currently have no effect
on the `S' conversion, (or the `s' conversion when the argument
isn't a string).
If STREAM isn't false (in which case the created string is
returned) the value of STREAM is returned.
(format nil "foo %S bar 0x%x" '(x . y) 255)
=> "foo (x . y) bar 0xff"
(format standard-output "The %2$s is %1$s!" "purple" "dog")
-| The dog is purple!
=> #
- Variable: format-hooks-alist
This variable is an association-list, each element being `(CHAR .
FUNCTION)', defining extra conversions for the `format' function.
If a conversion `%X' is given, and the alist contains an element
whose car is the character X, the the associated function is
called with one value, the next argument to be formatted. It should
return the string to be inserted.
back to top
Hooks
=====
A "hook" allows you to wedge your own pieces of Lisp code into the
operation of other functions, enable the extension of that
functionality. These pieces of code are evaluated via the hook and the
result is available to the hook's caller. One hook has already been
encountered, the `format-hooks-alist' variable (*note Formatted
Output::).
* Menu:
* Functions As Hooks:: Some hooks are a single function,
* Normal Hooks:: Others may be a list of pieces of code
to evaluate.
back to top
Functions As Hooks
------------------
Some hooks only allow a single piece of code to be hooked in.
Usually a normally-undefined function is used; to install your hook
defined a function with the name of the hook. When the hook is to be
evaluated the function is called.
Generally the name of the hook's function will end in `-function'.
An alternative scheme is to use a variable to store the hook, its
value should be the function to call.
back to top
Normal Hooks
------------
This is the standard type of hook, it is a variable whose value is a
list of functions. When the hook is evaluated each of the functions
will be called in turn.
The names of hooks of this type will normally end in `-hook'.
These functions are exported by the `rep.system' module.
- Function: add-hook hook function #!optional at-end
This function adds a new function FUNCTION to the list of
functions installed in the (list) hook HOOK (a symbol).
If AT-END is true the new function is added at the end of the
hook's list of functions (and therefore will be called last when
the hook is evaluated), otherwise the new function is added to the
front of the list.
text-mode-hook
=> (#)
(add-hook 'text-mode-hook my-function)
=> (# #)
- Function: remove-hook hook function
This function removes the function FUNCTION from the list of
functions stored in the (list) hook HOOK (a symbol).
_All_ instances of FUNCTION are deleted from the hook.
There are actually three calling conventions for this type of hook,
differing in how many of the functions in the list actually get called.
In this simplest form, _all_ functions are called. In an `and' type
hook, functions are only invoked while all others have returned true.
As soon as the first function in the hook returns false, no others will
be called. Finally, an `or' type hook aborts when a function returns a
true result.
- Function: call-hook hook arg-list #!optional type
Call the hook named by the symbol HOOK, passing all functions the
arguments in the list ARG-LIST. Note that HOOK may also be the
actual list of functions to call.
TYPE defines how the return values of each function in the hook
are treated. If TYPE is false they are ignored, if TYPE is the
symbol `and' the hook aborts after a function returns false, if
TYPE is `or' the hook aborts when a function returns true.
In all cases the value returned by the last-evaluated function is
returned.
back to top
Files
=====
`librep' allows you to manipulate files in the operating system's
filing system; a special type of Lisp object, a "file object", is used
to represent files which have been opened for reading or writing
(through the streams mechanism, *note Streams::).
Names of files are represented by strings, the syntax of file names
is defined by the underlying operating system: `librep' simply treats
it as a string.
Unless otherwise stated, all functions and variables described in the
following sections are exported by the `rep.io.files' module.
* Menu:
* File Names:: Files are named by a string
* File Objects:: Lisp objects representing files
* File Information:: Predicates on files
* Manipulating Files:: Deleting, renaming and copying files
* Manipulating Directories:: Accessing directories
* Manipulating Symlinks:: Handling symbolic links
* File Handlers:: Extending the file name-space
* Remote Files:: Accessing remote files
back to top
File Names
----------
A "file name" is a string identifying an individual file (or
directory) in the filing system (i.e. the disk). The exact syntax of
file names depends on the operating system. There are several functions
for manipulating file names.
- Function: file-name-absolute-p file-name
Returns true when FILE-NAME is not specified relative to the
current directory.
- Function: file-name-directory file-name
This function returns the directory part of the file name string
FILE-NAME. This is the substring of FILE-NAME defining the
directory containing the file.
(file-name-directory "/tmp/foo")
=> "/tmp/"
(file-name-directory "foo")
=> ""
(file-name-directory "foo/bar/")
=> "foo/bar/"
- Function: file-name-nondirectory file-name
Returns the substring of the file name FILE-NAME which is _not_
the directory part.
(file-name-nondirectory "/tmp/foo")
=> "foo"
(file-name-nondirectory "foo")
=> "foo"
(file-name-nondirectory "foo/bar/")
=> ""
- Function: file-name-as-directory file-name
Returns a string through which the item in the file system named by
FILE-NAME can be referred to as a directory.
(file-name-as-directory "./foo")
=> "./foo/"
(file-name-as-directory "./foo/")
=> "./foo/"
- Function: directory-file-name directory-name
Returns a string through which the directory named by
DIRECTORY-NAME can be referred to as a file.
(directory-file-name "./foo/")
=> "./foo"
(directory-file-name "./foo")
=> "./foo"
- Function: expand-file-name file-name #!optional base-dir
Expands FILE-NAME assuming that it specifies a file relative to
BASE-DIR. If BASE-DIR is undefined it is taken as the current
value of the `default-directory' variable. While expanding the
file name, any obvious simplifications will be performed (e.g. on
Unix the removal of "." and ".." where possible).
Note that the returned file name will only be absolute if one of
the following conditions is met:
1. BASE-DIR (or `default-directory') is absolute,
2. FILE-NAME is already absolute.
(expand-file-name "foo" "./bar")
=> "bar/foo"
Note for file handler implementors: when a handler is called for
the `expand-file-name' operation, it will only ever receive one
argument, the already expanded file name. The only action that may
be need to be taken is to simplify the file name (e.g. removing `.'
and `..' entries or whatever).
- Function: canonical-file-name file-name
This function returns the canonical name of the file referred to
by the string FILE-NAME. The canonical name of a file is defined
such that two files can be compared simply by comparing their
canonical names; if the names match, they refer to the same file.
(Note that the opposite isn't always true, if two canonical names
don't match the files could still be the same, for example via
hard links. On most operating systems, symbolic links will be
expanded where possible.
(canonical-file-name "foo")
=> "/home/john/src/librep/man/foo"
- Function: local-file-name file-name
`librep' supports extensible file handling (*note File
Handlers::), so file names may refer to files not residing in the
system's local file structure, and thus which are unavailable to
other programs.
This function returns either the absolute name of the file
FILE-NAME, if it is found in the local system, or false, if the
file does not.
(local-file-name "foo")
=> "/home/john/src/librep/man/foo"
(local-file-name "/john@tango:foo")
=> ()
- Function: make-temp-name
This function returns the name of a file which, when created, may
be used for temporary storage. Each time this function is called a
unique name is computed.
(make-temp-name)
=> "/tmp/00088aaa"
(make-temp-name)
=> "/tmp/00088baa"
- Variable: default-directory
This variable names the current working directory. All relative
file names are interpreted starting from this location in the file
system.
back to top
File Objects
------------
A file object is a Lisp object which represents an open file in the
filing system. Any file object may be used as a stream (either input or
output) to access the contents of the file (*note Streams::).
- Function: filep object
This function returns true when its argument is a file object.
* Menu:
* Creating File Objects:: Opening files
* Destroying File Objects:: Closing files
* Functions on File Objects:: Functions operating on file objects
back to top
Creating File Objects
.....................
- Function: open-file file-name mode
This function opens the file called FILE-NAME (*note File Names::)
and returns the new file object.
The MODE argument is a symbol defining the access modes used to
open the file with, the options are:
`read'
Open an existing file for reading only.
`write'
Open the file for writing only, if the file exists it is
truncated to zero length. Otherwise a new file is created.
`append'
Open the file for appending to, i.e. writing to the end of
the file. If the file doesn't exist it is created.
The three standard I/O streams are also available as file handles.
- Function: stdin-file
Return a file object representing the interpreters standard input.
- Function: stdout-file
Return a file object representing the interpreters standard output.
- Function: stderr-file
Return a file object representing the interpreters standard error.
Attempting to close any of these files will result in the associated
stream being connected to `/dev/null'.
back to top
Destroying File Objects
.......................
The easiest way to close a file is simply to eliminate all
references to it, subsequently the garbage collector will close it for
you. It is better to close files explicitly though since only a limited
number of files may be opened concurrently.
- Function: close-file file-object
This function closes the file pointed to by the file object
FILE-OBJECT.
Subsequently, any stream accesses FILE-OBJECT are illegal and will
signal an error.
back to top
Functions on File Objects
.........................
- Function: seek-file file #!optional offset where
When called as `(seek-file FILE)', returns the distance in bytes
from the start of the file that the next character would be read
from.
When called as `(seek-file FILE OFFSET [WHERE])' alters the
position from which the next byte will be read. WHERE can be one
of the values:
`()'
OFFSET bytes after the current position.
`start'
OFFSET bytes after the beginning of the file.
`end'
OFFSET bytes before the end of the file.
Note that not all files may be seekable; if `(seek-file FILE)'
returns false, indicating that the current position is unknown,
any attempts to set the current position will also fail.
- Function: flush-file file
This function flushes any buffered output to the file object FILE
to disk.
- Function: file-binding file
Returns the name of the file which the file object FILE is
currently bound to. Returns false if the file is currently unbound
(i.e. `close-file' was called on it).
The next three functions are used when non-local files are being
accessed. See *Note File Handlers:: for more details.
- Function: file-bound-stream file
If the file object FILE doesn't refer to a file in the local
filing system, return the stream that it is bound to (allowing the
file's contents to be accessed), this may or may not be another
file object.
- Function: file-handler-data file
Return the file-handler-specific data associated with the file
object FILE.
- Function: set-file-handler-data file data
Set the handler-specific data of file object FILE to DATA. This
should only be done by the handler owning the file.
It's also possible to register a callback function to be invoked when
input is available on a file,
- Function: set-input-handler local-file function
Arrange for FUNCTION to be called whenever pending input is
available on LOCAL-FILE, a file object bound to a file in the
local file space.
Note that this makes LOCAL-FILE subsequently do non-blocking input.
This function is normally only useful when LOCAL-FILE represents a
pipe or socket.
back to top
File Information
----------------
A number of functions exist which when given the name of a file
return one of the attributes relating to that file.
- Function: file-exists-p file-name
Returns true when a file FILE-NAME exists.
- Function: file-regular-p file-name
Returns true when the file FILE-NAME is a `normal' file. This means
that it isn't a directory, device, symbolic link or whatever.
- Function: file-directory-p file-name
Returns true when the file FILE-NAME is a directory.
- Function: file-symlink-p file-name
Returns true when the file FILE-NAME is a symbolic link.
- Function: file-readable-p file-name
Returns true when the file FILE-NAME is readable.
- Function: file-writable-p file-name
Returns true when the file FILE-NAME is writable.
- Function: file-owner-p file-name
Returns true when the ownership of the file FILE-NAME is the same
as that of any files written by the editor.
- Function: file-size file-name
Returns the number of bytes stored in the file named FILE-NAME.
- Function: file-nlinks file-name
Returns the number of hard links pointing to the file FILE-NAME. If
FILE-NAME has only one name the number will be one.
- Function: file-modes file-name
This function returns the access permissions of the file FILE-NAME.
This will be an integer whose format is undefined; it differs from
operating system to operating system.
- Function: file-modes-as-string file-name
Returns a ten-character string describing the attibutes of the file
called FILE-NAME
(file-modes-as-string ".")
=> "drwxr-sr-x"
- Function: set-file-modes file-name modes
This function sets the access permissions of the file FILE-NAME to
the integer MODES (as returned by the `file-modes' function).
- Function: file-modtime file-name
Returns the system time at the last modification to the file
FILE-NAME, this will be in the usual timestamp format, *Note
Timestamps::.
- Function: file-newer-than-file-p file-name1 file-name2
This function returns true if the file FILE-NAME1 was modified
more recently than the file FILE-NAME2 was.
back to top
Manipulating Files
------------------
- Command: delete-file file-name
This function deletes the file called FILE-NAME. When called
interactively FILE-NAME is prompted for.
- Command: rename-file file-name new-name
This function attempts to change the name of the file NEW-NAME to
NEW-NAME.
This won't work from one file system to another or if a file called
NEW-NAME already exists, in these cases an error is signalled.
This prompts for its arguments when called interactively.
- Command: copy-file file-name destination-name
Creates a new copy of the file FILE-NAME with the name
DESTINATION-NAME.
The access modes of the new file will be the same as those of the
original file.
The arguments are prompted for when this function is called
interactively.
back to top
Manipulating Directories
------------------------
- Command: make-directory directory-name
Create a new directory called DIRECTORY-NAME.
- Command: delete-directory directory-name
Delete the directory called DIRECTORY-NAME. This only succeeds if
the directory in question is empty, otherwise an error is
signalled.
- Function: directory-files directory-name
Returns a list of the names of all items in the directory whose
name is DIRECTORY-NAME. The names in the list will be relative to
the directory DIRECTORY-NAME.
(directory-files "/tmp/foo"
=> ("bar" "subdir" "xyz" "." "..")
back to top
Manipulating Symbolic Links
---------------------------
- Function: make-symlink name contents
Create a symbolic link called NAME, containing the string CONTENTS.
- Function: read-symlink name
Return the string that is the contents of the symbolic link called
NAME. Signals an error if no such link exists.
back to top
File Handlers
-------------
As noted earlier, `librep' supports virtual files; that is it allows
file names to be accessed that don't reside on the local filing system,
or aren't normally valid as file names. This is achieved through the
use of "file handlers", Lisp functions that have signalled that they
should be used to redirect all accesses to files whose names match a
particular regular expression (*note Regular Expressions::).
For example, there is a convention under Unix that a user's home
directory can be accessed via the file name `~', even though there is
no such support from the operating system itself. So a file handler can
be (and has been) written that recognises all file names starting with
a tilde, translating them to the actual file in the file system.
- Variable: file-handler-alist
This variable contains a list of file handler declarations, each
one of the form `(REGEXP . FUNCTION)'. Whenever a file operation
is performed on a file whose name matches REGEXP, FUNCTION is
invoked to perform the action. The function is called as
`(FUNCTION OPERATION ARGS...)', where OPERATION and ARGS are from
the original call.
For example if the `file-handler-alist' contains the entry `("^~"
. tilde-file-handler)', then all file operations on files starting
with a tilde are redirected to the `tilde-file-handler' function.
Thus if a form `(file-exists-p "~/foo")' is executed, it would
result in a call to `tilde-file-handler' as `(tilde-file-handler
'file-exists-p "~/foo")'.
The list of operations that may be redirected to a file handler is:
`file-name-absolute-p', `expand-file-name', `local-file-name',
`canonical-file-name', `file-name-nondirectory', `file-name-directory',
`file-name-as-directory', `directory-file-name', `open-file',
`close-file', `flush-file', `seek-file', `write-buffer-contents',
`read-file-contents', `insert-file-contents', `delete-file',
`rename-file', `copy-file', `copy-file-to-local-fs',
`copy-file-from-local-fs', `make-directory', `delete-directory',
`file-exists-p', `file-regular-p', `file-readable-p', `file-writable-p',
`file-directory-p', `file-symlink-p', `file-owner-p', `file-nlinks',
`file-size', `file-modes', `file-modes-as-string', `set-file-modes',
`file-modtime', `directory-files', `make-symlink', `read-symlink'.
There are several undefined functions in this list. The
`write-buffer-contents', `read-file-contents', and
`insert-file-contents' pertain to the Jade text editor. The other two
are defined as follows.
- Operation: copy-file-to-local-fs file-name local-name
Called when copying files between file handlers, this operation
should copy the file matching the handler FILE-NAME, to the file
on the local file system LOCAL-NAME.
- Operation: copy-file-from-local-fs local-name file-name
Called when copying files between file handlers, this operation
should copy the local file FILE-NAME to the file matching the
handler FILE-NAME.
To prevent infinite recursion, while a particular operation is being
processed by a file handler, that operation may not be passed back to
the same handler.
To allow file handlers to handle the `open-file' operation, it is
possible to create file handles from arbitrary streams.
- Function: make-file-from-stream file-name stream handler
Return a new file object that refers to the logical file called
FILE-NAME, that is not in the local filing system. All access to
the file object will be directed through the stream object STREAM,
and the file handler function HANDLER.
An alternative method of opening remote files is to use a temporary
file in the local file system with either one (`read' or `write'
modes), or two (`append' mode) synchronisations with the remote system.
This is the method used by the FTP remote file backend (see the next
section). It has the advantage of simplifying the `seek-file' operation.
back to top
Remote files
------------
Since one of the intentions for file handlers is to allow remote
files to be accessed, a common method of providing new methods of doing
this has been implemented, in the `remote.jl' Lisp library.
Accessing a file name matching the regular expression:
^/(([a-zA-Z0-9._-]+)@)?([a-zA-Z0-9._-]+):
for example `/john@host.com:file' refers to a file called `file' owned
by the user `john', on the system `host.com'.
If no username is specified explicitly, two variables are used to
select the user:
- Variable: remote-host-user-alist
An alist mapping host regexps to the default user name to use for
remote file connections to that host.
- Variable: remote-default-user
User name to use for remote file connections when otherwise
unspecified. By default the current user name on the local system.
Two variables control how individual hosts are matched to methods of
accessing files.
- Variable: remote-auto-backend-alist
An alist of `(HOST-REGEXP . BACKEND-TYPE)' mapping host names to
methods of accessing remote files.
- Variable: remote-default-backend
A symbol defining the method to use for otherwise unselected hosts.
A method of accessing files, or a "backend" is a symbol whose
`remote-backend' property names a function to call when files need to
be accessed. For example the `ftp' backend is initialised as:
(put 'ftp 'remote-backend remote-ftp-handler)
The backend function is called as `(FUNCTION SPLIT-NAME OPERATION
ARGS)'. The SPLIT-NAME is a three-element list, `(USER-OR-NIL HOST
FILE)' defining the file to be accessed. The other options are as usual.
Further details can be found in the `remote.jl', `remote-ftp.jl' and
`remote-rcp.jl' Lisp source files.
The `ftp' backend is currently the most well-developed, several
functions and variables may be used to customise its behaviour.
- Function: remote-ftp-add-passwd user host passwd
Add the string PASSWD as the password for the FTP session
connecting to USER@HOST.
- Variable: remote-ftp-show-messages
When true (the default), messages are displayed as FTP commands are
executed.
- Variable: remote-ftp-display-progress
When true (the default) display progress of FTP transfers.
- Variable: remote-ftp-anon-users
A regular expression matching the user names for "anonymous" FTP
sessions.
- Variable: remote-ftp-anon-passwd
The string to send as the passwd of an anonymous FTP session. By
default the current uses email address.
There is a problem with the `ftp' backend however; due to
limitations in the FTP protocol, not all `librep' file operations are
supported, with the most obvious exception being the `make-symlink'
function.
When this is a problem it may be possible to use rep's custom file
transfer protocol. If it is possible to use `rsh' to connect to the
remote host in question, then the `rep' backend may be used.
The `rep-remote' program distributed with `librep' must exist on the
remote host, this is executed via `rsh' and provides a protocol for
executing all of `librep''s file operations on that host. See the
`lisp/remote-rep.jl' file in the distribution for more details.
back to top
Processes
=========
When running on a Unix-style operating system `librep' allows you to
launch and control an arbitrary number of subprocesses. These
subprocesses can run either synchronously or asynchronously in respect
to the Lisp system; data can be sent to the `stdin' channel and any
output from the process is automatically written to a specified Lisp
output stream.
Unless otherwise stated, all functions and variables described in the
following sections are exported by the `rep.io.processes' module.
* Menu:
* Process Objects:: Lisp objects associated with subprocesses
* Asynchronous Processes:: Subprocesses running in parallel
* Synchronous Processes:: Subprocesses which run to completion
* Process I/O:: Input and output with subprocesses
* Process States:: Suspending subprocesses
* Signalling Processes:: Sending signals to subprocesses
* Process Information:: Information stored in a process object
back to top
Process Objects
---------------
A "process object" is a type of Lisp object used to provide a link
between a `physical' process running in the operating system and the
Lisp system. Each process object consists of a number of values
(references to other Lisp objects); these values are used when the
object is used to run a subprocess.
Process objects which aren't currently being used to run a subprocess
store the exit value of the last subprocess which was run on that
object.
- Function: processp object
This function returns true when its argument is a process object.
The programmer-accessible components of a process object are,
"Output stream"
A normal Lisp output stream (*note Output Streams::), all data
which the subprocess outputs to its `stdout' channel is copied to
this output stream. *Note Process I/O::.
"Error stream"
A normal Lisp output stream (*note Output Streams::), all data
which the subprocess outputs to its `stderr' channel is copied to
this output stream. Unless explicitly specified error output goes
to the `stdout' stream. *Note Process I/O::.
"State change function"
A Lisp function, called each time the state of the subprocess
being run on the object changes. *Note Process States::.
"Program name"
The name of the program (a string) to execute when the subprocess
is created.
"Program arguments"
A list of strings defining the arguments which the program executed
is given.
"Directory"
When a subprocess is started its current working directory is set
to the directory named by this component of its process object.
"Connection type"
Asynchronous subprocesses (*note Asynchronous Processes::) use this
component to decide how to connect to the I/O channels of the
subprocess. Current options include pseudo-terminals and pipes.
- Function: make-process #!optional output-stream state-function
directory program args
This functions creates and returns a new process object. _No
subprocess will be started._
The optional arguments are used to define the values of the
components of the new process object, any undefined components
will be set to default or null values.
For each component of a process object two functions exist; one to
read the component's value in a specific process object, the other to
set the component's value.
- Function: process-prog process
Returns the value of the program name component of the process
object PROCESS.
- Function: set-process-prog process prog-name
Sets the value of the program name component of the process object
PROCESS to the string PROG-NAME, then returns PROG-NAME.
- Function: process-args process
Returns the value of the program arguments component of the
process object PROCESS.
- Function: set-process-args process arg-list
Sets the value of the program arguments component of the process
object PROCESS to the list ARG-LIST, then returns ARG-LIST.
- Function: process-dir process
Returns the value of the directory component of the process object
PROCESS.
- Function: set-process-dir process directory
Sets the value of the directory component of the process object
PROCESS to the string DIRECTORY, then returns DIRECTORY.
- Variable: process-environment
This is a list of environment variable definitions, as well as
being used by the `setenv' and `getenv' functions (*note
Environment Variables::), it also provides the environment of all
started subprocesses.
(car process-environment)
=> "LOGNAME=john"
- Function: active-processes
Returns a list containing all active (i.e. running or stopped)
process objects.
back to top
Asynchronous Processes
----------------------
An "asynchronous process" is one that runs in parallel with Lisp
evaluation, basically this means that once the subprocess has been
started (by the `start-process' function) `librep' will carry on as
normal.
The event loop checks for output from asynchronous processes, any
found is copied to the process' output stream, and calls the the
process' state change function when necessary (*note Process States::).
Alternatively the `accept-process-output' function can be called to
explicitly allow output to be processed.
When using asynchronous processes you have a choice as to the Unix
mechanism used to connect the `stdin', `stdout' and `stderr' streams of
the subprocess to `librep''s process.
The two options currently available are pipes or pseudo-terminals; in
general pseudo-terminals should only be used to provide a direct
interface between the user and a process (i.e. the `*shell*' buffer)
since they allow job control to work properly. At other times pipes
will be more efficient and are used by default. However, there are
cases where the buffering characteristics of pipes mean that ptys must
be used.
- Function: start-process #!optional process program #!rest args
This function starts an asynchronous subprocess running on the
process object PROCESS. If PROCESS is undefined a new process
object is created (by calling the function `make-process' with all
arguments undefined).
The function always returns the process object which the
subprocess has been started on. If for some reason the subprocess
can't be created an error of type `process-error' is signalled.
The optional argument PROGRAM is a string defining the name of the
program to execute, it will be searched for in all the directories
in the `PATH' environment variable. The ARGS are strings to pass
to the subprocess as its arguments.
When defined, the optional arguments overrule the values of the
related components of the process object.
The following example runs the `ls' program asynchronously, its
output is sent to the `standard-output' stream.
(let
((process (make-process standard-output)))
(start-process process "ls" "-s"))
Note that when `librep' exits it kills all of its asynchronous
subprocesses which are still running without warning.
- Function: process-connection-type process
Returns the value of the connection type component of the process
object PROCESS. See the documentation of the
`set-process-connection-type' function for the values this may
take.
- Function: set-process-connection-type process symbol
Sets the value of the connection type component of the process
object PROCESS to SYMBOL, then returns SYMBOL.
SYMBOL should be one of the following symbols,
`pty'
Use pseudo-terminals to connect to subprocesses running
asynchronously on this process object.
`pipe'
Use standard Unix pipes to connect, this is the default value
of this component.
`socketpair'
Uses a connected pair of sockets.
Note that currently only the `pipe' connection type allows the
normal and error output streams of the process to be separated.
back to top
Synchronous Processes
---------------------
When a "synchronous process" is started `librep' waits for it to
terminate before continuing; they are usually used when a Lisp program
must invoke an external program as part of its function, i.e. the
auto-compression feature runs the compression program `gzip'
synchronously when it needs to compress a buffer.
Unlike asynchronous processes their is no choice between pipes and
pseudo-terminals for connecting to a subprocess. Instead, it is
possible to link the `stdin' channel of a synchronous process to a
named file.
- Function: call-process #!optional process input-file-name program
#!rest args
This function starts a process running on the process object
PROCESS. If PROCESS is undefined a new process object is created
by calling the `make' function.
If defined, the string INPUT-FILE-NAME names the file to connect
to the standard input of the subprocess, otherwise the subprocess'
input comes from the null device (`/dev/null' on UNIX).
The optional arguments PROGRAM and ARGS define the name of the
program to invoke and any arguments to pass to it. The program will
be searched for in all directories listed in the
`process-environment' variable.
If any of the optional parameters are unspecified they should have
been set in the PROCESS-OBJECT prior to calling this function.
After successfully creating the new subprocess, this function
simply copies any output from the process to the output stream
defined by the output stream component of the process object. When
the subprocess exits its exit-value is returned (an integer). Note
that the exit-value is the value returned by the
`process-exit-value' function, see *Note Process Information::.
If, for some reason, the new subprocess can't be created an error
of type `process-error' is signalled.
The following function definition is taken from the `gzip.jl' file,
it shows how the `call-process' function can be used to uncompress a
file into a buffer (for Jade).
;; Uncompress FILE-NAME into the current buffer
(defun gzip-uncompress (file-name)
(let
((proc (make-process (current-buffer))))
(message (concat "Uncompressing `" file-name "'") t)
;; gunzip can do .Z files as well
(unless (zerop (call-process proc nil "gunzip" "-c" file-name))
(signal 'file-error (list "Can't gunzip file" file-name)))))
The user is able to interrupt synchronous subprocesses (for example
if they seem to have got wedged somehow). Each time a user-interrupt is
received by `librep' (i.e. the `INT' signal), a stronger signal is sent
to the subprocess. First an interrupt signal, then a termination
signal, before finally a non-ignoreable quit signal is sent.
back to top
Process I/O
-----------
It is only possible for lisp programs to explicitly send input data
to _asynchronous_ processes (by the time it's possible to call a
function to send data to a synchronous process, the process will
already have terminated!). Simply use the process object which an
asynchronous process is running on as a normal Lisp input stream, any
strings or characters written to the stream will immediately be copied
to the `stdin' channel of the subprocess.
With synchronous processes, the only control over input data possible
is by giving the `call-process' function the name of a file containing
the subprocess' input data.
Output data from subprocesses is handled the same way by both
asynchronous and synchronous processes: it is simply copied to the
stream defined by the output stream component of the subprocess'
process object.
- Function: process-output-stream process
Returns the value of the output stream component of the process
object PROCESS.
- Function: set-process-output-stream process stream
Sets the value of the output stream component of the process object
PROCESS to the stream STREAM, then returns STREAM.
By default the `stdout' and `stderr' streams are combined, use the
`set-process-error-stream' function to separate them. (Note that this
currently only works with `pipe' connection types.)
- Function: process-error-stream process
Returns the value of the error stream component of the process
object PROCESS.
- Function: set-process-error-stream process stream
Sets the value of the error stream component of the process object
PROCESS to the stream STREAM, then returns STREAM.
Output from asynchronous subprocesses (this includes changes of state
as well as stream output) is only propagated at well-defined times.
Either when in the read stage of the read-eval-print, or input, loop,
or when the `accept-process-output' or `sit-for' functions are called.
- Function: accept-process-output #!optional seconds milliseconds
Wait SECONDS plus MILLISECONDS for output from any asynchronous
subprocesses. If any arrives, process it, then return false.
Otherwise return true. If either of the arguments is undefined,
they count as zero in the addition.
- Function: sit-for #!optional seconds milliseconds
Wait for input to arrive and be processed. No more than SECONDS
seconds plus MILLISECONDS milliseconds will be waited. If at the
end of this time no input has arrived, return true. Otherwise
return false if input was found.
Note that this function is only distinct to
`accept-process-output' when `librep' is embedded in another
application, or an extension has been loaded that provides an event
loop (such as the `gtk' binding). In this case other input forms,
such as user input, for example, can preempt the timeout.
This function is exported by the `rep.system' module.
*Note Streams::.
back to top
Process States
--------------
Each process object has a "state" associated with it; this depends on
the status of the subprocess currently running on the process object (or
not as the case may be).
The possible states are,
"running"
This state means that the subprocess using this process object is
currently running, i.e. it hasn't been stopped.
"stopped"
Means that the subprocess has been temporarily suspended from
running.
"unused"
This means that the process object is free to have a new
subprocess created on it.
Predicates exist which test whether a given process object is in one
of these states.
- Function: process-running-p process-object
Returns true when PROCESS-OBJECT is in the running state.
- Function: process-stopped-p process-object
Returns true when PROCESS-OBJECT is in the stopped state.
- Function: process-in-use-p process-object
Returns true when PROCESS-OBJECT is _not_ in the unused state.
The following two functions are used to stop and then subsequently
continue a process running.
- Function: stop-process process-object #!optional whole-group
This function suspends execution of the subprocess running on the
process object PROCESS-OBJECT.
If WHOLE-GROUP is true all subprocesses in the process group of
PROCESS-OBJECT are stopped.
- Function: continue-process process-object #!optional whole-group
Use this function to continue a subprocess executing after it has
been stopped (by the `stop-process' function).
If WHOLE-GROUP is true all subprocesses in the process group of
PROCESS-OBJECT are continued.
The state change function component of a process object defines a
function which will be called each time the state of the process object
changes. If your program needs to be informed when an asynchronous
process terminates this function is the way to do it.
- Function: process-function process
Returns the value of the state change function component of the
process object PROCESS.
- Function: set-process-function process function
Sets the value of the state change function component of the
process object PROCESS to the function FUNCTION, then returns
FUNCTION.
back to top
Signalling Processes
--------------------
- Function: interrupt-process process-object #!optional whole-group
Sends the `SIGINT' signal to PROCESS-OBJECT.
- Function: kill-process process-object #!optional whole-group
Sends the `SIGKILL' signal to the PROCESS-OBJECT.
Note that the functions `stop-process' and `continue-process' also
send signals to the subprocess.
- Function: signal-process process signal #!optional whole-group
Send the signal SIGNAL to the process PROCESS; if WHOLE-GROUP is
true the signal is also sent to all processes in the process group
of PROCESS.
PROCESS may be either a Lisp process object, or an integer
defining the pid of the process to signal (not necessarily started
by `librep').
SIGNAL may either be an integer defining the actual signal number,
or a symbol naming the signal. All names are as usual but with the
preceding `SIG' removed, for example the `SIGINT' signal would be
sent by using the symbol `INT'. If a named signal doesn't exist on
the current operating system, an error is raised.
Returns true if the signal was sent successfully.
As with the UNIX `kill' system call, `signal-process' may also be
used to test whether a process with a particular pid is currently
active, by using a signal with value zero.
back to top
Process Information
-------------------
- Function: process-id process-object
This function returns the operating-system identifier associated
with the subprocess currently running on the process object
PROCESS-OBJECT.
- Function: process-exit-value process-object
Returns the integer representing the return code of the last
subprocess to be run on PROCESS-OBJECT.
If no subprocess has been run on PROCESS-OBJECT, PROCESS-OBJECT is
currently in the running state or the last subprocess exited
abnormally (i.e. from a terminal signal) false is returned.
- Function: process-exit-status process-object
This function returns the integer that was the exit status of the
last subprocess which was run on the process object PROCESS-OBJECT.
Note that the exit status is _not_ the value given to the `exit'
function in a C program, use the `process-exit-value' to access
this value.
If no process has been run on PROCESS-OBJECT, or the process is
currently in the running state false is returned.
back to top
Regular Expressions
===================
Regular expressions (or "regexps") are a powerful method of matching
patterns in strings. `librep' uses the `regexp(3)' implementation by
Henry Spencer, with some modifications that I have made. It comes with
this banner:
Copyright (c) 1986 by University of Toronto.
Written by Henry Spencer. Not derived from licensed software.
Permission is granted to anyone to use this software for any
purpose on any computer system, and to redistribute it freely,
subject to the following restrictions:
1. The author is not responsible for the consequences of use of
this software, no matter how awful, even if they arise from
defects in it.
2. The origin of this software must not be misrepresented, either
by explicit claim or by omission.
3. Altered versions must be plainly marked as such, and must not
be misrepresented as being the original software.
* Menu:
* Regexp Syntax:: How to write regular expressions
* Regexp Functions:: How to use them
back to top
Regular Expression Syntax
-------------------------
The syntax of a regular expression is as follows (this is adapted
from the manual page):
A regular expression is zero or more "branches", separated by `|'.
It matches anything that matches one of the branches.
A branch is zero or more "pieces", concatenated. It matches a match
for the first, followed by a match for the second, etc.
A piece is an "atom" possibly followed by `*', `+', or `?'. An atom
followed by `*' matches a sequence of 0 or more matches of the atom. An
atom followed by `+' matches a sequence of 1 or more matches of the
atom. An atom followed by `?' matches a match of the atom, or the null
string.
An atom is a regular expression in parentheses (matching a match for
the regular expression), a "range" (see below), `.' (matching any
single character), `^' (matching the null string at the beginning of
the input string), `$' (matching the null string at the end of the
input string), one of the strings `\s', `\S', `\w', `\W', `\d', `\D',
`\b', `\B', or a `\' followed by a single character (matching that
character), or a single character with no other significance (matching
that character).
A "range" is a sequence of characters enclosed in `[]'. It normally
matches any single character from the sequence. If the sequence begins
with `^', it matches any single character _not_ from the rest of the
sequence. If two characters in the sequence are separated by `-', this
is shorthand for the full list of ASCII characters between them (e.g.
`[0-9]' matches any decimal digit). To include a literal `]' in the
sequence, make it the first character (following a possible `^'). To
include a literal `-', make it the first or last character.
Also, any of the `*', `+' or `?' operators can be suffixed by a `?'
character (i.e. `*?', `+?', `??'). The meaning of the operator remains
the same but it becomes "non-greedy". This means that it will match the
_smallest_ number of characters satisfying the regular expression,
instead of the default behaviour which is to match the _largest_.
The backslash-introduced atoms have the following meanings:
`\s'
Match any whitespace character.
`\S'
Match any non-whitespace character.
`\w'
Match any alphanumeric or underscore character.
`\W'
Match any non-(alphanumeric or underscore) character.
`\d'
Match any numeric character.
`\D'
Match any non-numeric character.
`\b'
Match the null string between two adjacent `\w' and `\W'
characters (in any order).
`\B'
Match the null string that is not between two adjacent `\w' and
`\W' characters.
Some example legal regular expressions could be:
`ab*a+b'
Matches an `a' followed by zero or more `b' characters, followed by
one or more `a' characters, followed by a `b'. For example,
`aaab', `abbbab', etc...
`(one|two)_three'
Matches `one_three' or `two_three'.
`^cmd_[0-9]+'
`^cmd_\d+'
Matches `cmd_' followed by one or more digits, it must start at the
beginning of the line.
back to top
Regexp Functions
----------------
These functions are exported by the `rep.regexp' module.
- Function: quote-regexp string
Return a version of STRING, such that when used as a regexp, it
will match the original contents of STRING verbatim, and nothing
else. This involves quoting regexp meta-characters.
(quote-regexp "abc")
=> "abc"
(quote-regexp "a+c")
=> "a\\+c"
- Function: string-match regexp string #!optional start ignore-case
Returns true if the string STRING matches the regular expression
REGEXP. The string matches if executing the regexp at _any_
position in the string succeeds.
When defined, START is the index of the first character to start
matching at (counting from zero). When IGNORE-CASE is true the
case of matched strings are ignored. Note that character classes
are still case-significant.
(string-match "ab+c" "abbbc")
=> t
(string-match "ab+c" "xxxabbbcyyy")
=> t
- Function: string-looking-at regexp string #!optional start
ignore-case
Similar to `string-match', but only returns true if STRING matches
REGEXP starting at the character at index START in the string (or
the first character if START is undefined).
(string-looking-at "ab+c" "abbbc" 0)
=> t
(string-looking-at "ab+c" "xxxabbbcyyy" 0)
=> ()
(string-looking-at "ab+c" "xxxabbbcyyy" 3)
=> t
- Function: match-start #!optional n
Returns the position at which the N'th parenthesised expression
started in the last successful regexp match. If N is false or zero
the position of the start of the whole match is returned instead.
When matching strings, all positions are integers, with the first
character in the string represented by zero. However, extensions
that allow regexps to be matched against other textual inputs may
return different position values.
(string-match "x*(foo|bar)y" "xxxbary")
=> t
(match-start 1)
=> 3
- Function: match-end #!optional n
Similar to `match-start', but returns the position of the
character following the matched item.
(string-match "x*(foo|bar)y" "xxxbary")
=> t
(match-end 1)
=> 6
A common use of regular expressions is to match a string, then
replace certain portions of the string with other text.
- Function: expand-last-match template
Expand the TEMPLATE substituting the parenthesised expressions
from the most recent successfully matched regular expression.
TEMPLATE may contain the following substitution-inducing escape
sequences:
`\0'
`\&'
Substitute the whole string matched by the last regexp
`\N'
Substitute the N'th parenthensised expression, where 1 <= N
<= 9.
`\\'
Substitute a single backslash character.
(string-match "x*(foo|bar)y" "xxxbary")
=> t
(expand-last-match "test-\\1-ing")
=> "test-bar-ing"
Note that double backslashes are required due to the read syntax of
strings (*note Strings::).
- Function: string-replace regexp template string
Returns the string created by replacing all matches of REGEXP in
STRING with the result of expanding TEMPLATE using the
`expand-last-match' function.
(string-replace "-" "_" "foo-bar-baz")
=> "foo_bar_baz"
(string-replace "&(optional|rest)" "#!\\1" "(a &optional b &rest c)")
=> "(a #!optional b #!rest c)"
back to top
Time and Date
=============
This section describes how time and date values are handled in
`librep'.
* Menu:
* Timestamps:: Internal representation of time
* Formatting Dates:: Creating strings from timestamps
* Parsing Dates:: Reading textual dates
back to top
Timestamps
----------
As in UNIX, `librep' measures time as the number of seconds since
January 1st, 1970 (known as the "epoch"). For historical reasons rep
stores timestamps as a pair of integers, using a cons cell.
The first integer records the number of whole days since the epoch,
the second records the number of seconds since the start of the day (in
universal time).
These function are exported by the `rep.system' module:
- Function: current-time
Return the number of seconds since the epoch, in a cons-cell.
(current-time)
=> (10744 . 61063)
- Function: fix-time timestamp
Ensure that the two parts of TIMESTAMP (a pair or integers) are
consistent, simply that the number of seconds is less than the
number of seconds in a whole day. If not, the timestamp is
adjusted to meet this constraint.
- Function: time-later-p timestamp-1 timestamp-2
Returns true if TIMESTAMP-1 is later than TIMESTAMP-2.
On the plus side, this scheme won't wrap around as quickly as UNIX's
`time_t' will ;-)
The `rep.util.time' module also provides some functions for
manipulating timestamps:
- Function: time->seconds timestamp
Convert TIMESTAMP to an integer, the number of seconds since the
epoch that it represents.
- Function: seconds->time seconds
Convert from an integer SECONDS to a timestamp object.
- Function: time- timestamp-1 timestamp-2
Return the number of seconds difference between TIMESTAMP-1 and
TIMESTAMP-2.
- Constant: seconds-per-day
The number of seconds in a 24-hour day.
back to top
Formatting Dates
----------------
Given a timestamp value it is possible to format it as a string, in
many different formats.
- Function: current-time-string #!optional timestamp format
Return a string defining TIMESTAMP according to the string FORMAT.
If TIMESTAMP is undefined, the current time is used.
The FORMAT string may include any of the formatting characters
from the C library's `strftime(3)' function. If undefined a
standard, fixed-width, format is used:
(current-time-string)
=> "Wed Jun 2 18:07:53 1999"
Some of the possible formatting substitutions include (this is
copied from the GNU libc manual, *note (libc)Formatting Date and
Time::):
`%a'
The abbreviated weekday name according to the current locale.
`%A'
The full weekday name according to the current locale.
`%b'
The abbreviated month name according to the current locale.
`%B'
The full month name according to the current locale.
`%c'
The preferred date and time representation for the current
locale.
`%d'
The day of the month as a decimal number (range `01' to `31').
`%H'
The hour as a decimal number, using a 24-hour clock (range
`00' to `23').
`%I'
The hour as a decimal number, using a 12-hour clock (range
`01' to `12').
`%j'
The day of the year as a decimal number (range `001' to
`366').
`%m'
The month as a decimal number (range `01' to `12').
`%M'
The minute as a decimal number.
`%p'
Either `am' or `pm', according to the given time value; or the
corresponding strings for the current locale.
`%S'
The second as a decimal number.
`%U'
The week number of the current year as a decimal number,
starting with the first Sunday as the first day of the first
week.
`%W'
The week number of the current year as a decimal number,
starting with the first Monday as the first day of the first
week.
`%w'
The day of the week as a decimal number, Sunday being `0'.
`%x'
The preferred date representation for the current locale, but
without the time.
`%X'
The preferred time representation for the current locale, but
with no date.
`%y'
The year as a decimal number, but without a century (range
`00' to `99').
`%Y'
The year as a decimal number, including the century.
`%Z'
The time zone or name or abbreviation (empty if the time zone
can't be determined).
`%%'
A literal `%' character.
(current-time-string nil "%Y-%m-%d")
=> "1999-06-02"
back to top
Parsing Dates
-------------
The `date' Lisp library provides rudimentary support for parsing
date and time descriptions to their individual components, and to
timestamps. Evaluate the form `(require 'date)' to load this library.
- Function: parse-date string #!optional start
Returns a vector encoding the date described by STRING. If START
is defined, it specifies the index of the character in the string
to start parsing from.
Each element of the vector contains a separate component of the
overall point in time described by the string. The indices of
these elements are defined by the following constants:
`date-vec-day-abbrev'
The abbreviated name of the day of the week.
`date-vec-day'
The numeric day of the month, counting from one.
`date-vec-month-abbrev'
The abbreviated name of the month.
`date-vec-month'
The numeric month of the year, counting from January equals
one.
`date-vec-year'
The numeric year.
`date-vec-hour'
The numeric hour of the day.
`date-vec-minute'
The numeric minute of the hour.
`date-vec-second'
The numeric second of the minute.
`date-vec-timezone'
If true, a string defining the timezone.
`date-vec-epoch-time'
The timestamp (*note Timestamps::), including the effects of
the timezone, if given.
(current-time-string)
=> "Wed Jun 2 18:37:17 1999"
(parse-date (current-time-string))
=> ["Wed" 2 "Jun" 6 1999 18 37 17 0 (10744 . 67037)]
(parse-date "1999-06-02")
=> ["Tue" 2 "Jun" 6 1999 0 0 0 0 (10744 . 0)]
(parse-date "June 6, 1999")
=> ["" 0 "Jun" 6 1999 0 0 0 0 (10742 . 0)]
(aref (parse-date "June 6, 1999") date-vec-epoch-time)
=> (10742 . 0)
XXX provide more information on accepted formats, outputs for
incomplete descriptions, etc...
back to top
Internationalisation
====================
`librep' has support for internationalisation (or i18n) of text
messages, using the GNU `gettext' implementation (*note Overview:
(gettext)Top.), a run-time library managing the mapping between text
strings in the programmer's native language and in the language of the
end user.
Three functions are provided to access the message catalogues
maintained by GNU `gettext'. Import the `rep.i18n.gettext' module to
load them.
- Function: _ string
Attempt to find a native language equivalent of STRING. If no
equivalent is found the original string is returned.
Note that this function is always defined, even if the `gettext'
module hasn't been required. In this case it always returns the
original string.
- Function: bindtextdomain domain directory
Tell `gettext' that message catalogues for message domain DOMAIN
(a string) can be found under the directory called DIRECTORY.
- Function: textdomain domain
Note that any strings that are to be translated in the future
(until the next call to `textdomain') are in the domain called
DOMAIN (a string).
The usual method of constructing message catalogue templates (`.pot'
files) is to run `xgettext' on the C source files of the program (that
have been annotated for i18n). librep provides the `rep-xgettext'
program to perform the same task for files of Lisp code.
back to top
System Information
==================
These definitions are all exported by the `rep.system' module.
- Variable: operating-system
A symbol naming the current operating system. The only current
option is `unix'.
- Function: system-name
This function returns a string naming the host that the
interpreter is running on. When possible this be a fully-qualified
name (i.e. including the domain)
- Variable: rep-build-id
A string describing the environment under which `librep' was
built. This will always have the format `DATE by USER@HOST, for
ARCH.'.
rep-build-id
=> "Mon May 17 1999 by john@tizer.dcs.warwick.ac.uk, for sparc-sun-solaris2.6."
- Variable: rep-version
A string describing the current release of `librep'.
rep-version
=> "1.0"
back to top
User Information
================
These functions are exported by the `rep.system' module.
- Function: user-login-name
This function returns a string containing the login name of the
user.
(user-login-name)
=> "john"
- Function: user-full-name #!optional real-name
This function returns a string containing the `real' name of the
user; the format of the string will depend on the host system.
If REAL-NAME is a string, it defines the name that will be
returned by subsequent calls to this function.
(user-full-name)
=> "John Harper"
- Function: user-home-directory #!optional user
This function returns the home directory of the user whose login
name is USER, or the current user if USER is undefined. The
returned string will be as returned by `file-name-as-directory'
(i.e. terminated by a `/' character under UNIX)
(user-home-directory)
=> "/home/john/"
back to top
Environment Variables
=====================
These functions are exported by the `rep.system' module.
- Function: getenv variable-name
This function returns the value (a string) of the environment
variable called VARIABLE-NAME. If the specified variable doesn't
exist false is returned.
(getenv "OSTYPE")
=> "Linux"
- Function: setenv variable-name new-value
This function sets the value of the environment variable called
VARIABLE-NAME to NEW-VALUE. NEW-VALUE can either be a string
containing the new contents of the variable or false, in which
case the environment variable is deleted.
- Function: unsetenv variable-name
Deletes any variable in `process-environment' named VARIABLE-NAME.
See also *Note Process Objects:: for the description of the
`process-environment' variable.
back to top
String Functions
================
- Function: translate-string string map
Applies the MAP to each character in the STRING. MAP is also
string, each character represents the translation for an ASCII
character of that characters position in the string. If the string
is less than 256 chars long any undefined characters will remain
unchanged.
For example, if STRING contains the character `A', with ASCII code
65, then it would be replaced by the 65th character in the string
MAP.
Note that the STRING really is modified, no copy is made
- Variable: upcase-table
A `translate-string' compatible translation map to convert
lowercase characters to uppercase characters.
- Variable: downcase-table
A map to convert uppercase characters to lowercase.
- Variable: flatten-table
A translation map to convert newline characters to spaces.
(translate-string "Foo" upcase-table)
=> "FOO"
(translate-string "Foo" downcase-table)
=> "foo"
- Function: complete-string template list #!optional ignore-case
Return a string whose beginning matches the string TEMPLATE, and
is unique in the set of all strings in LIST which also match
TEMPLATE. If IGNORE-CASE is true, all matching ignores case of
characters.
(complete-string "foo" '("bar" "foobar" "forbarf" "foobat"))
=> "fooba"
- Function: string-head-eq string-1 string-2
Returns t if STRING-2 matches the beginning of STRING-1.
(string-head-eq "foobar" "foo")
=> t
(string-head-eq "foo" "foobar")
=> ()
- Function: string-upper-case-p string
Return true if STRING contains no lower case characters.
- Function: string-lower-case-p string
Return true if STRING contains no upper case characters.
- Function: string-capitalized-p string
Return true if the first character of STRING is upper case.
- Function: string-upcase string
Return a new string, an upper case copy of STRING.
- Function: string-downcase string
Return a new string, a lower case copy of STRING.
- Function: capitalize-string string
Return a new string, a copy of STRING with the first character in
upper case.
- Function: mapconcat function sequence separator
Call FUNCTION for each member of SEQUENCE, concatenating the
results. Between each pair of results, insert SEPARATOR. Return
the resulting string.
back to top
Sleeping
========
- Function: sleep-for seconds #!optional milliseconds
Pause for a SECONDS (plus the optional MILLISECONDS component)
long period of time. Input becoming available will _not_ break the
sleep (*note Process I/O::).
This function is exported by the `rep.system' module.
back to top
Beeping
=======
Use this function to attract the user's attention.
- Function: beep
Ring a bell somewhere.
back to top
Messages
========
The `message' function will show the user a small message (typically
no more than a single column of text). In graphical applications it
_won't_ bring up a separate window, only displaying the text in a
status bar or something similar. In a console-based environment, the
message will be printed to the `stderr' stream, followed by a line
break.
- Function: message #!optional display-now
Displays a one-line message, the string MESSAGE. If DISPLAY-NOW,
every effort will be made to display the message as soon as
possible, possibly before the next scheduled screen update (if
applicable).
This function is exported by the `rep.system' module.
back to top
Command Line Options
====================
As noted earlier any unused command line arguments are made available
to scripts through the `command-line-args' variable (*note
Invocation::).
- Variable: command-line-args
The list of unused command line arguments, in their original order.
The `get-command-line-option' function may be used to scan this list
of arguments. The `rep.system' module exports this function.
- Function: get-command-line-option option #!optional requires-arg
Returns t if OPTION was specified on the command line (OPTION is
typically a phrase beginning with `--').
If REQUIRES-ARG is true, the option requires a parameter, the
value of which is returned. If a parameter isn't supplied an error
is signalled.
(setq command-line-args '("--foo" "bar"))
=> ("--foo" "bar")
(get-command-line-option "--foo" t)
=> "bar"
command-line-args
=> ()
(setq command-line-args '("--foo=bar"))
=> ("--foo=bar")
(get-command-line-option "--foo" t)
=> "bar"
command-line-args
=> ()
back to top
Executing Shell Commands
========================
The subprocess handling of `librep' provides a comprehensive
interface to starting and controlling external processes (*note
Processes::). However it can be overkill when all that is required is
to invoke a shell command, with its I/O going to the same places as the
interpreter's.
- Function: system command
Execute the shell command COMMAND synchronously, returning its
exit status. An error will be signalled if the shell process could
not be started.
The `stdin', `stdout' and `stderr' streams of the shell are left
as in the interpreter process.
The subprocesses environment is copied from the current value of
the `process-environment' variable.
Note that the exit status is _not_ the same as the return code of
the command. It depends on the operating system, but under UNIX the
return code can be found through right-shifting the exit status by
eight bits. Low non-zero values represent that the process was killed
by a signal.
It is possible to interrupt a running shell process in the same way
as with a normal synchronous process (*note Synchronous Processes::).
Interrupt the interpreter, it will send progressively harder-to-ignore
signals to the child each interrupt, until it is eventually terminated.
back to top
Asynchronous Timers
===================
The `rep.io.timers' module (*note Modules::) allows a Lisp program
to create multiple asynchronous timers, each of which is primed to call
a specified function after a specified period of time. These functions
only work when the Lisp event loop is being used (i.e. at least one
`recursive-edit' is currently in progress).
- Function: make-timer function #!optional seconds milliseconds
Create and return a new timer object. It will be set to call the
Lisp function FUNCTION after SECONDS seconds plus MILLISECONDS
milliseconds. FUNCTION will be called with a single argument, the
timer object that has just fired.
If both SECONDS and MILLISECONDS are undefined, or zero, the timer
will be created but won't call FUNCTION.
After the time interval has passed, and FUNCTION has been called,
the timer _will not_ be restarted. Use the `set-timer' function to
reset it.
- Function: delete-timer timer
Prevent the timer object TIMER from calling the Lisp function
associated with it. Use the `set-timer' function to reset it.
- Function: set-timer timer #!optional seconds milliseconds
Reset the timer object TIMER. If either/both of SECONDS and
MILLISECONDS are defined the interval of the timer will be set to
the specified time period. If neither are defined then the current
interval of the timer is preserved.
back to top
Debugging
=========
When you have written a Lisp program you will have to debug it
(unless all your programs work first time?). There are two main classes
of errors; syntax errors and semantic errors.
Syntax errors occur when the text you've typed out to represent your
program is not a valid representation of a Lisp object (since a program
is simply an ordered set of Lisp objects). When you try to load your
program the Lisp reader will find the syntax error and tell you about,
unfortunately though it probably won't be able to tell you exactly
where the error is.
The most common source of syntax errors is too few or too many
parentheses; the Jade or Emacs `Ctrl-Meta-f' and `Ctrl-Meta-b' commands
can be used to show the structure of the program as the Lisp reader
sees it.
Semantic errors are what we normally call bugs--errors in logic, the
program is syntactically correct but doesn't do what you want it to.
For these types of errors librep provides hooks to allow interactive
debugging. The debugger supplied with librep uses these hooks to
implement a simple command line debugger; programs using librep as an
extension language may provide their own debugger interface.
There are several ways to enter the Lisp debugger; functions can be
marked so that they cause the debugger to be entered when they are
called, breakpoints can be written in functions or it can be called
explicitly with a form to step through.
- Command: trace symbol
This command marks the symbol SYMBOL so that each time its value
is dereferenced the debugger is entered when the next form is
evaluated. This can be used to set breakpoints on functions (or
variables).
When called interactively SYMBOL is prompted for.
- Command: untrace symbol
The opposite of `trace'--unmarks the symbol.
- Function: break
This function causes the debugger to be entered immediately. By
putting the form `(break)' at suitable points in your program
simple breakpoints can be created.
- Command: step form
This function invokes the debugger to step through the form FORM.
When called interactively FORM is prompted for.
- Function: backtrace #!optional stream
Prints a description of the current Lisp function call stack to
STREAM (or `standard-output' if STREAM is undefined).
(backtrace (current-buffer))
-| # ((current-buffer)) nil
-| # ((backtrace (current-buffer))) t
=> t
Each line represents a stack frame, the first item is the called
function, the second is the list of arguments applied to it. The
third item is true if the list of arguments as displayed has
already been evaluated.
Whenever the Lisp debugger is entered the form waiting to be
evaluated is printed, preceded by the current depth of execution in
angular brackets. At this point the special debugger commands available
are,
`step'
`s'
Step into the current form; this means that in a list form the
debugger is used to evaluated each argument in turn.
`next'
`n'
Continue evaluating forms normally until the next form at the
current level is entered, then re-enter the debugger.
`continue'
`c'
Continue execution normally. Note that this command is the one to
use when an error has been trapped.
`return FORM'
`r FORM'
Evaluate FORM then return this value as the result of the current
form.
`print FORM'
`p FORM'
Evaluate FORM, then print its value.
`form'
`f'
Print the form being debugged.
`backtrace'
`b'
Print a backtrace of the current Lisp call stack.
Entering a null string repeats the previous `next', `step', or
`continue' command.
After the form has been evaluated (i.e. after you've typed one of the
commands above) the value of the form is printed in the buffer,
prefixed by the string `=> '.
Note that it is also possible to make certain types of errors invoke
the debugger immediately they are signalled, see *Note Errors::. Also
note that the debugger is unable to step through compiled Lisp code.
back to top
Tips
====
This section of the manual gives advice about programming in
`librep'.
For advice on getting the most out of the compiler, see *Note
Compilation Tips::.
* Menu:
* Comment Styles:: Different types of comments
back to top
Comment Styles
--------------
As already described, single-line comments in Lisp are introduced by
a semi-colon (`;') character. By convention a different number of
semi-colons is used to introduce different types of comments,
`;'
A comment referring to the line of Lisp code that it occurs on,
comments of this type are usually indented to the same depth, on
the right of the Lisp code. When editing in Jade's Lisp mode the
command `Meta-;' can be used to insert a comment of this type.
For example,
(defconst op-call #x08) ;call (stk[n] stk[n-1] ... stk[0])
; pops n values, replacing the
; function with the result.
(defconst op-push #x10) ;pushes constant # n
`;;'
Comments starting with two semi-colons are written on a line of
their own and indented to the same depth as the next line of Lisp
code. They describe the following lines of code.
For example,
(let
((fname (concat file-name ?c)))
;; Be sure to remove any partially written dst-file.
(when (file-exists-p fname)
(delete-file fname)))
Comments of this type are also placed before a function definition
to describe the function. This saves wasting memory with a
documentation string in a module's internal functions.
For example,
;; Compile a form which occurred at the `top-level' into a
;; byte code form.
;; defuns, defmacros, defvars, etc... are treated specially.
;; require forms are evaluated before being output uncompiled;
;; this is so any macros are brought in before they're used.
(defun comp-compile-top-form (form)
...
`;;;'
This type of comment always starts in the first column of the
line, they are used to make general comments about a program and
don't refer to any function or piece of code in particular.
For example,
;;; Notes:
;;; Instruction Encoding
;;; ====================
;;; Instructions which get an argument (with opcodes of zero up to
...
`;;;;'
Each program should have a comment of this type as its first line,
the body of the comment is the name of the file, two dashes and a
brief description of what the program does. They always start in
the first column.
For example,
;;;; compiler.jl -- Simple compiler for Lisp files/forms
If you adhere to these standards the indentation functions provide by
the Lisp mode will indent your comments to the correct depth.
back to top
librep : The language
table of contents
Copying
Introduction
Invocation
The language
The REPL
librep Internals
Reporting bugs
News
Function index
Variable index
Concept index