I am confused about the &environment parameter in common lisp. In particular, what is it useful for, and why is it a parameter, rather than a special variable?
EDIT: It would also be nice to see a concrete example of how &environment would be used in code.
Doc
Macro Lambda Lists:
&environment is followed by a single variable that is bound to an
environment representing the lexical environment in which the macro
call is to be interpreted. This environment should be used with
macro-function, get-setf-expansion, compiler-macro-function, and
macroexpand (for example) in computing the expansion of the macro, to
ensure that any lexical bindings or definitions established in the
compilation environment are taken into account.
Explanation
Operators which define macros (local or global) have to define how code is expanded before being evaluation or compiled, and thus may need to expand existing macros, whose expansion may depend on the environment - so the macro definitions need the environment.
Not a Special Variable
Special variables are more dangerous because the user might rebind them and because they are harder to handle correctly in multi-threaded code.
Example
All over the places.lisp:
(defmacro psetf (&whole whole-form
&rest args &environment env)
(labels ((recurse (args)
(multiple-value-bind (temps subforms stores setterform getterform)
(get-setf-expansion (car args) env)
(declare (ignore getterform))
(when (atom (cdr args))
(error-of-type 'source-program-error
:form whole-form
:detail whole-form
(TEXT "~S called with an odd number of arguments: ~S")
'psetf whole-form))
(wrap-let* (mapcar #'list temps subforms)
`(MULTIPLE-VALUE-BIND ,stores ,(second args)
,#(when (cddr args) (list (recurse (cddr args))))
,#(devalue-form setterform))))))
(when args `(,#(recurse args) NIL))))
From iolib/src/new-cl/definitions.lisp:
(defmacro defconstant (name value &optional documentation
&environment env)
(destructuring-bind (name &key (test ''eql))
(alexandria:ensure-list name)
(macroexpand-1
`(alexandria:define-constant ,name ,value
:test ,test
,#(when documentation `(:documentation ,documentation)))
env)))
For example when Lisp wants to macroexpand a form, it needs to know which name refers to what macro definition. This can depend on the globally available macros, locally bound macros via macrolet or what is available during compilation. Thus the environment object makes it possible to macroexpand a form with respect to different environments.
Environment variables provide to the macro author information about the declarations et. al. enforce when the macro was invoked. The macro author can then use that information to customize how he chooses to expand the macro. For example he might add or subtract debugging code based on declarations he finds in the environment. For example he might infer the type of expressions he was given to work with and add or subtract code to get better performance.
You can write a lot of Common Lisp code and never use &environment. There are situations where it becomes necessary, when those are would make good follow up question. Rainer hints at an answer to that.
Related
Is there a way, in common lisp, to receive a user input, say "foo", and defvar a global variable *foo*?
For example (which does NOT work):
(defun global-name (s)
"Takes s and changes it to *s*"
(concatenate 'string "*" s "*"))
(defun add-global-var (var)
"defvars a global variable and adds it to *global-list*"
(let ((var-name (global-name var)))
(defvar var-name var)
(push var-name *global-list*)))
; Used like this:
(add-global-var "myvar")
In this case, the var-name is a string, and will not work with defvar.
Déjà vu... I asked these kinds of questions 20+ years ago ;-)
Your question
Yes, you can do that (but no, you do not want to!)
(defun add-global-var (var-name &optional (package *package*))
(let ((var (intern var-name package)))
(proclaim `(special ,var))
(push var *global-list*)))
Please see
proclaim
intern
*package*
Alternatively, you can use a macro as the other answer suggests - in
fact, symbol creation at macroexpansion time (which is part of
compilation) is a very common thing,
cf. gensym.
Your problem
There is little reason to do this though.
Global variables created at run time were not available at compile time
and are, therefore, pretty useless.
Why do you want to do this?
If you want to map strings to values, you are much better off using an
equal hash table.
If you want to integrate with read,
you should call it while binding
*package*
to your internal temp package and then use
symbol-value
to store and retrieve values.
You will use intern to
map "variable names" to the symbols.
This is most likely a XY problem since it's very unusual to need to make a variable with a name made up in runtime. It's very common in compile time, but not runtime. #coredump has already covered compile time macros if that is what you are after.
Here is how you do it though:
(defun add-global-var (var)
"defvars a global variable and adds it to *global-list*"
(let ((var-name (intern (string-upcase (global-name var)))))
(set var-name var)
(push var-name *global-list*)))
set is deprecated, but I doubt it will ever be removed. Implementations might not be able to run as fast though since this is like messing with internals.
Since the names are not from source you you have no good use for the bidnings. because of this I would rather use a hash:
(defvar *bindings* (make-hash-table :test #'eq))
(defun add-binding (var)
(let ((var-name (intern (string-upcase (global-name var)))))
(setf (gethash var-name *bindings*) var)
*bindings*))
A reason to do this is as a part of your own little interpreter symbol table or something. You don't need a list of them since you can get all the keys from the hash as well as get the bound values.
Yes, with a macro:
(defvar *global-list* nil)
I changed global-name so that it also accepts symbols, to avoid thinking about whether the string should be upcased or not. With a symbol, the case is given by readtable-case (you can use uninterned symbols if you want to avoid polluting packages).
(defun global-name (name)
(check-type name (or string symbol))
(intern
(concatenate 'string "*" (string name) "*")))
I named the macro defvar*:
(defmacro defvar* (name)
`(push
(defvar ,(global-name name) ',name)
*global-list*))
Tests:
CL-USER> (defvar* #:foo)
(*FOO*)
CL-USER> (defvar* #:bar)
(*BAR* *FOO*)
Note:
You can also add an optional package argument like in #sds's answer, that's better.
Consider this question. Here the basic problem is the code:
(progv '(op arg) '(1+ 1)
(eval '(op arg)))
The problem here is that progv binds the value to the variable as variable's symbol-value not symbol-function. But, that's obvious because we didn't explicitly suggest which values are functions.
The Plan
So, in order to solve this problem, I thought of manually dynamically binding the variables, to their values based on the type of values. If the values are fboundp then they should be bound to the symbol-function of the variable. A restriction, is that match-if can't be a macro. It has to be a function, because it is called by a funcall.
Macro : functioner:
(defmacro functioner (var val)
`(if (and (symbolp ',val)
(fboundp ',val))
(setf (symbol-function ',var) #',val)
(setf ,var ,val)))
Function: match-if:
(defun match-if (pattern input bindings)
(eval `(and (let ,(mapcar #'(lambda (x) (list (car x))) bindings)
(declare (special ,# (mapcar #'car bindings)))
(loop for i in ',bindings
do (eval `(functioner ,(first i) ,(rest i))))
(eval (second (first ,pattern))))
(pat-match (rest ,pattern) ,input ,bindings))))
Here, the let part declares all the variables lexically (supposedly). Then declare declares them special. Then functioner binds the variables and their values aptly. Then the code in the pattern is evaluated. If the code part is true, then only the pattern-matcher function pat-match is invoked.
The Problem
The problem is that in the function, all it's arguments are evaluated. Thus bindings in the let and declare parts will be replaced by something like :
((v1 . val1)(v2 . val2)(v3 . val3))
not
'((v1 . val1)(v2 . val2)(v3 . val3))
So, it's treated as code, not a list.
So, I'm stuck here. And macros won't help me on this one.
Any help appreciated.
Not the answer you are looking for, but PROGV is a special operator; it is granted the ability to modify the dynamic bindings of variables at runtime; AFAIK, you can't simply hack it to operate on "dynamic function bindings".
The point of progv is to use list of symbols and values that are evaluated, meaning that you can generate symbols at runtime and bind them dynamically to the corresponding values.
You might be able to find a solution with eval but note that if you macroexpand into (eval ...), then you loose the surrounding lexical context, which is generally not what you want ("eval" operates on the null lexical environment). I speculate that you could also have a custom code walker which works on top-level forms but reorganizes them, when it finds your special operator, to bring the context back in, producing something like (eval '(let (...) ...)).
I'm reviving an old LISP program from the early 1980s.
(It's the Nelson-Oppen simplifier, an early proof system.
This version was part of the Ford Pascal-F Verifier,
and was running in Franz LISP in 1982.) Here's
the entire program:
https://github.com/John-Nagle/pasv/tree/master/src/CPC4
I'm converting the code to run under clisp on Linux,
and need some advice. Most of the problems are with
macros.
HUNKSHELL
Hunkshell was a 1970s Stanford SAIL hack to support records with named fields in LISP. I think I've converted this OK; it seems to work.
https://github.com/John-Nagle/pasv/blob/master/src/CPC4/hunkshell.l
The original macro generated more macros as record update functions.
I'm generating defuns. Is there any reason to generate macros?
By the way, look at what I wrote for "CONCAT". Is there a better way
to do that?
DEFMAC
More old SAIL macros, to make macro definition easier before defmacro became part of the language.
https://github.com/John-Nagle/pasv/blob/master/src/CPC4/defmac.l
I've been struggling with "defunobj". Here's my CL version, partly converted:
; This macro works just like defun, except that both the value and the
; function-binding of the symbol being defined are set to the function
; being defined. Therefore, after (defunobj f ...), (f ...) calls the
; defined function, whereas f evaluates to the function itself.
;
(defmacro defunobj (fname args &rest b)
`(progn
(defun ,fname ,args ,b)
;;;;(declare (special ,fname)) ;;;; ***declare not allowed here
(setq ,fname (getd ',fname))))
If I made that declare a proclaim, would that work right?
And what replaces getd to get a function pointer?
SPECIAL
There are lots of (DECLARE (SPECIAL FOO)) declarations at top
level in this code. That's not allowed in CL. Is it
appropriate to use (PROCLAIM (SPECIAL FOO)) instead?
Concat
Essentially correct, but indentation is broken (like everywhere else - I suggest Emacs to fix it).
Also, you don't need the values there.
defunobj
I suggest defparameter instead
of setq. Generally speaking, before setting a variable (with, e.g., setq) one should establish it (with, e.g., let or defvar).
fdefinition is what
you are looking for instead of getd.
I also don't think you are using backquote right:
(defmacro defunobj (fname &body body)
`(progn
(defun ,fname ,#body)
(defparameter ,fname (fdefinition ',fname))))
Special
I think defvar and defparameter
are better than proclaim
special.
PS. Are you aware of the CR.se site?
I'm new to Lisp and I was reading about an text-generator example in ANSI Common Lisp, Chapter 8. I follow the example and defined a function "see" in the scope of a LET variable "prec",
(let ((prec '|.|))
(defun see (symb)
(let ((pair (assoc symb (gethash prev *words*))))
(if (null pair)
(push (cons symb 1) (gethash prev *words*))
(incf (cdr pair))))
(setf prev symb)))
and saved it into a lisp file.
Then when I returned to REPL and tried to invoke see after loading the compiled version of the file, an error occurred:
The variable PREV is unbound.
[Condition of type UNBOUND-VARIABLE]
How do I invoke see properly? And what's a lexical closure for? I'm all confused.
Thanks for helping.
Looks like you've typed prec instead of prev in the enclosing let form.
Lexical closures are functions that 'close over' a part of the lexical environment (hence the name). There are many good introductions to closures in lisp that I will not attempt to repeat here but, essentially, let is the most common way to manipulate the lexical environment; in this case, you want to add the binding for prev, which will then be available to code within the body of the form. Your function see will 'close over' this binding, and so each call to see has access to it, even though when you make these calls, you will no longer be in the lexical environment established by the let form. You could say the function definition takes the binding with it, in a sense.
As you appear to have mis-typed the name of the prev, your function is trying to refer to a binding that has not been established at that point in the code.
I'm writing some methods to emit HTML for various elements. Each method has the same output, but doesn't necessarily need the same input.
The method for echoing a game-board needs to take a player as well (because each player only sees their own pieces)
(defmethod echo ((board game-board) (p player)) ... )
Echoing a board space doesn't require changing per player (that dispatch is actually done in the game-board method, which later calls echo on a space). Ideally, I'd be able to do
(defmethod echo ((space board-space)) ... )
(defmethod echo ((space empty-space)) ... )
It's also conceivable that I later run into an object that will need to know more than just the player in order to display itself properly. Since there are already methods specializing on the same generic, though, that would give the error
The generic function #<STANDARD-GENERIC-FUNCTION ECHO (4)> takes 2 required arguments;
It seems less than ideal to go back and name these methods echo-space, echo-board and so on.
Is there a canonical way of varying other arguments based on the specialized object? Should I do something like
(defgeneric echo (thing &key player ...) ...)
or
(defgeneric echo (thing &rest other-args) ...)
? More generally, can anyone point me to a decent tutorial on defgeneric specifically? (I've read the relevant PCL chapters and some CLOS tutorials, but they don't cover the situation I'm asking about here).
Generally speaking if interfaces of two functions are too different it indicates that they are not actually a specializations of the same operation and should not have the same name. If you only want to specialize on optional/key arguments the way to achieve that is to use a normal function which calls a generic function and provides it with default values for missing arguments to specialize on.
Keene's book is, I believe, the most comprehensive guide to CLOS. Unfortunately it seems to be available only in book form.
It's easier whenever the methods take the same parameters but only differ on the data type. But in cases where the generic function and methods all use the same name (of course) yet have lambda lists that vary significantly, I personally tend to use &key parameters to make my intentions explicit, rather than using &optional parameters. I think it helps with readability later.
Try something like this (pretending we already have classes class-a and class-b):
(defgeneric foo (object &key &allow-other-keys)
(:documentation "The generic function. Here is where the docstring is defined."))
(defmethod foo ((object class-a) &key &allow-other-keys)
(print "Code goes here. We dispatched based on type Class A."))
(defmethod foo ((object class-b) &key (x 1) (y 2) &allow-other-keys)
(print "Code goes here. We dispatched based on type Class B. We have args X and Y."))
Since "method combination" is involved, in order for the dispatching logic to "flow" through its possible choices of methods to use, we need to think of the method definitions somewhat like a chain where incompatible lambda lists (i.e. parameter lists) will break that chain. That's why there's the &key and &allow-other-keys in the method definitions that don't specifically need them. Putting them in DEFGENERIC and the method definitions allows us to have the method definition where we dispatch based on class-b.
DISCLAIMER: I'm a Common Lisp newbie myself, so take this with a grain of salt!
What about restructuring the objects/classes so that each object you want to echo has all required properties in its own (and inherited) slots?
That way you don't have to pass them as arguments when you call echo on the object, since what you need is already stored in the object.
Stepped in the same problem today and found a nice workaround, but it will work only if you have a fixed number of class-specified args, placed at the beginning.
The example for the single class-specific arg:
(defgeneric foo (obj &rest args))
(defmacro my-defmethod (name args code)
(let ((rst (gensym))
`(defmethod ,name (,(car args) &rest ,rst)
(apply (lambda (,(caar args) ,(cdr args)) ,code) (cons ,(caar args) ,rst)))))
(my-defmethod foo ((x class1) y z &optional a b c)
...)
(my-defmethod foo ((x class2) &key a b)
...)