- parenscript-devel - mailman3.common-lisp.net

[parenscript-devel] Conflict with % between Parenscript and CSS-Lite
by Andrew Peterson 07 Dec '08

07 Dec '08

Vladimir, There is a conflict between parenscript and css-lite. Css-lite redefines the parenscript operator % (javascript modulo operator) css-lite.lisp: #+parenscript (ps:defpsmacro % (val) `(ps:+ ,val "%")) causes this behavior: ; SLIME 2008-07-05 CL-USER> (lisp-implementation-type) "SBCL" CL-USER> (lisp-implementation-version) "1.0.18" CL-USER> (require :parenscript) NIL CL-USER> (in-package :parenscript) #<PACKAGE "PARENSCRIPT"> PS> (ps (% 3 2)) "3 % 2;" PS> (require :css-lite) NIL PS> (ps (% 3 2)) error while parsing arguments to DESTRUCTURING-BIND: invalid number of elements in (3 2) to satisfy lambda list (CSS-LITE::VAL): exactly 1 expected, but 2 found [Condition of type SB-KERNEL::ARG-COUNT-ERROR] andy

3 3

[parenscript-devel] Problems with css
by Andrei Stebakov 05 Dec '08

05 Dec '08

Hi I started to use parenscript and javascript portion of it works just fine. Now, I wanted to incorporate some css into the code and ran into problems. When I try to evaluate the sample code: (html-stream *standard-output* (html (:html (:head (css (* :border "1px solid black") (div.bl0rg :font-family "serif") (("a:active" "a:hoover") :color "black" :size "200%")))))) I get an error: Execution of a form compiled with errors. Form: ((STYLE TYPE text/css) (PRINC <!-- ) (PRINC (CSS-RULE-TO-STRING (CSS-RULE * BORDER 1px solid black)) (CSS-RULE-TO-STRING (CSS-RULE DIV.BL0RG FONT-FAMILY serif)) (CSS-RULE-TO-STRING (CSS-RULE (a:active a:hoover) COLOR black SIZE 200%))) .. [Condition of type SB-INT:COMPILED-PROGRAM-ERROR] I am using sbcl 1.0.18 on Ubuntu. How can I solve it? Thank you! Andrew

5 6

[parenscript-devel] Generating HTML in parenscript
by Gary King 26 Nov '08

26 Nov '08

I've just started playing with Parenscript and feel confused about HTML generation. Since Parenscript already includes (the unexported) process-html-forms, it seems to be able to (almost) handle much of the logic for simple HTML generation. Consider that > (parenscript::process-html-forms > `(:html > (:head (:title "ParenScript tutorial: 1st example")) > (:body (:h1 "ParenScript tutorial: 1st example") > (:p "Please click the link below." :br > ((:a :href "#" :onclick ,(ps-inline > (alert "Hello World"))) > "Hello World"))))) produces > (+ "<html/><head><title>ParenScript tutorial: 1st example</title></ > head><body><h1>ParenScript tutorial: 1st example</h1><p>Please click > the link below.<br/><a href=\"#\" onclick=\"javascript:alert(\"Hello > World\")\">Hello World</a></p></body>") It's easy to refactor process-html-forms to arrive at the (slightly incorrect) HTML in a string: > "<html/><head><title>ParenScript tutorial: 1st example</title></ > head><body><h1>ParenScript tutorial: 1st example</h1><p>Please click > the link below.<br/><a href=\"#\" onclick=\"javascript:alert(\"Hello > World\")\">Hello World</a></p></body>" and probably easy to futz with *js-inline-string-delimiter* and the like to get the correct quotes around the body of the onclick handler. So... since Parenscript doesn't do this, I have to ask "Why not"?! It would be nice to use Parenscript for simple HTML generation without having to get (Portable) AServe or some other more full-featured HTML generator. Am I missing something? thanks, -- Gary Warren King, metabang.com Cell: (413) 559 8738 Fax: (206) 338-4052 gwkkwg on Skype * garethsan on AIM

2 4

[parenscript-devel] &optional bugfix push request and tracing macro.
by William Halliburton 16 Nov '08

16 Nov '08

Hi guys. Parenscript rocks. I feel like I'm really rollerskating now.. There was a bug reported on *Tue Sep 30 21:28:05 UTC 2008 by* Daniel Gackle about the failure of &optional arguements. I just got bit by the bug, can we please get that one-liner pushed into the repository. Muchos gracais. On a happier note here is a macro I wrote to enable tracing-ala-cl. Works with firebug. You'll need to (defvar *trace-level*). I don't do indentation but that would be an easy addition. (defpsmacro console (&rest rest) `(console.log ,@rest)) (defpsmacro defun-trace (name args &rest body) (let* ((sname (ps::symbol-to-js name)) (tname (ps-gensym name)) (arg-names (loop for arg in args unless (eq arg '&optional) collect (if (consp arg) (car arg) arg))) (argpairs (loop for arg in arg-names nconc (list (ps::symbol-to-js arg) arg)))) `(progn (defun ,name ,arg-names (console *trace-level* ,sname ":" ,@argpairs) (incf *trace-level*) (let* ((rtn (,tname ,@arg-names))) (decf *trace-level*) (console *trace-level* ,sname "returned" rtn) (return rtn))) (defun ,tname ,args ,@body)))) Thanks much, Will

2 1

[parenscript-devel] wrong operator priority?
by Alex Mizrahi 16 Nov '08

16 Nov '08

this piece of parenscript: (defun my-element-by-id (cache id) (return (or (slot-value cache id) (setf (slot-value cache id) (document.get-element-by-id id))))) gets transformed to: function myElementById(cache, id) { return cache[id] || cache[id] = document.getElementById(id); }; which i believe is wrong -- browsers interpret it (according to JS operator priority rules, i guess) as return (cache[id] || cache[id]) = document.getElementById(id); and complain about wrong assignment to non-lvalue. while with correctly set parentheses it works fine: function myElementById(cache, id) { return cache[id] || (cache[id] = document.getElementById(id)); }; so, i guess, parenscript should detect this operator priority problem and insert parentheses correctly?

2 1

[parenscript-devel] A class of bugs in PS: expressions not parenthesized when necessary
by Daniel Gackle 16 Nov '08

16 Nov '08

The following bug in PS broke my code today: (ps (aref (if (and x (> (length x) 0)) (aref x 0) y) z)) => "x && x.length > 0 ? x[0] : y[z];" Obviously, this should be: (x && x.length > 0 ? x[0] : y)[z] You can see a similar bug here: (ps (aref (or (slot-value x 'y) (slot-value a 'b)) z)) => "x.y || a.b[z];" and here again: (ps (- (if x y z))) => "-x ? y : z;" Clearly, there's a problem with not parenthesizing things. Since the last patch or two I've sent have been ignored, I won't send a new one here, but rather ask: is PS being actively maintained or not? This project is too valuable to let fade. Dan

3 3

[parenscript-devel] Bug: &optional arguments stay undefined when they shouldn't
by Daniel Gackle 15 Nov '08

15 Nov '08

PS' implementation of &optional function arguments breaks because of JS's weird short-circuiting. For example, (ps (defun blah (&optional (x 0)) (return x))) => "function blah(x) { x = undefined === x && 0 || x; return x; };" With this implementation, blah() is undefined when it should return 0. A patch is below. Daniel hunk ./src/special-forms.lisp 268 - `(setf ,place (or (and (=== undefined ,place) ,value) - ,place))) + `(setf ,place (if (=== ,place undefined) ,value ,place)))

2 1

[parenscript-devel] insert verbatim JS code
by Alex Mizrahi 15 Nov '08

15 Nov '08

is there a trick to insert verbatim JS code directly into parenscript source? there used to be one -- via quote -- in older version, undocumented, but it seems new version just blocks such attempts. while this might look like a hack, in some cases it is very important feature

2 1

[parenscript-devel] Bug: array referencing doesn't work on conditional expression
by Daniel Gackle 14 Nov '08

14 Nov '08

(ps (aref (if a b c) 0)) => "a ? b : c[0];" instead of the correct "(a ? b : c)[0]" Daniel

2 2

[parenscript-devel] Bug in PS operator precedence
by Daniel Gackle 15 Oct '08

15 Oct '08

I got bitten by this bug: (ps (or x (if (= x 0) "zero" "empty"))) => "x || x == 0 ? 'zero' : 'empty'" This is wrong, because it evaluates to 'zero' for any value of x, instead of just when x is 0. The correct expansion is: x || (x == 0 ? 'zero' : 'empty') It seems that Parenscript simply has JS operator precedence wrong for the "if" (question-mark) operator. The following web page (which claims to be reporting what the ECMA spec says) makes it clear that the "if" operator's precedence is between that of "or" and the assignment operators. http://www.codehouse.com/javascript/precedence/ So a simple fix is: hunk ./src/printer.lisp 110 - (in js-expression-if) + (in) hunk ./src/printer.lisp 118 + (js-expression-if) Daniel

2 1