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

[parenscript-devel] Log macro
by Daniel Gackle 01 Apr '08

01 Apr '08

I wrote a simple ps macro to work like CL's log. It seems like a candidate for inclusion in PS itself. Daniel (defpsmacro log (number &optional base) (cond ((null base) `(*math.log ,number)) ((= base 10) `(* (log ,number) *math.*log10e*)) (t `(/ (log ,number) (log ,base)))))

2 2

[parenscript-devel] Merging some improvements, additions, and cleanup
by Travis Cross 31 Mar '08

31 Mar '08

Greetings, I've made a number of improvements to the ParenScript code base. What's the best way for us to merge these changes into the darcs repo on common-lisp.net? Most of the updates fall into five basic broad categories: * Rewriting PS special forms to better match their CL equivalents. For example, DO now does parallel assignment, supports a return value, and can properly initialize variables that don't increment. * Improving coverage of CL by adding such forms as SETQ and PSETF. * Improving access to JS capabilities, such as labeled for loops. * Documentation and test case updates. * Miscellaneous code cleanups, whitespaceification, and global standardizations. The changes are fully backward compatible, as long as you weren't relying on the 'broken' sequential assignment that DO and LET were doing. ;) DO* and LET* are now available for that purpose, of course. Looks like in total I added somewhere around 350 lines and removed 210 lines from the code base, excluding documentation and test cases. To the documentation I added 368 lines and removed 227. The changes are reasonably tested and have been in use on my production systems for a couple of months. I've included the darcs changelog below my signature. Since I use ParenScript rather heavily, it's very likely that I'll be making more substantial changes on an ongoing basis. I'd be happy to help maintain the project. Let me know. Cheers, -- Travis P.S. My branch of parenscript can currently be found here: http://darcs.tcross.org/parenscript/ $ darcs chan --reverse [snip...] Tue Feb 26 22:59:53 UTC 2008 Travis Cross <tc@tra...> * Corrected spelling error. Tue Feb 26 23:09:02 UTC 2008 Travis Cross <tc@tra...> * Whitespace standardization in src/package.lisp. Sun Mar 30 01:46:18 UTC 2008 Travis Cross <tc@tra...> * Added support for breaking and continuing to arbitrary labels. Sun Mar 30 02:05:25 UTC 2008 Travis Cross <tc@tra...> * Greatly simplified implementation of lexical-let*. Sun Mar 30 02:08:29 UTC 2008 Travis Cross <tc@tra...> * Added lexical-let. Sun Mar 30 02:14:00 UTC 2008 Travis Cross <tc@tra...> * Added simple-let* and aliased let* to it. Sun Mar 30 02:22:18 UTC 2008 Travis Cross <tc@tra...> * Added simple-let, which does parallel binding without a lambda. Sun Mar 30 02:23:16 UTC 2008 Travis Cross <tc@tra...> * Aliased let to simple-let. Sun Mar 30 02:25:53 UTC 2008 Travis Cross <tc@tra...> * Added psetf. Sun Mar 30 02:27:55 UTC 2008 Travis Cross <tc@tra...> * Added setq and psetq. Sun Mar 30 02:33:34 UTC 2008 Travis Cross <tc@tra...> * Updated let1-dynamic to clean up its temporary stack variable after use. Sun Mar 30 19:35:12 UTC 2008 Travis Cross <tc@tra...> * Improved, extended, and refactored iteration special forms. * Added support for DO*. * DO now does parallel assignment. * DO/DO* now correctly support init-forms such as: (do (a (b) (c 1) (d 1 (1+ d))) ...) * DO/DO*/DOTIMES/DOLIST/DOEACH now support return values via a lambda transform. * DOEACH now supports implicit key/value destructuring, e.g.: (doeach ((key val) arr res) ...) * Added labeled-for to facilitate creating labeled for loops. * Restructured low level printer operators to better match the underlying JS. Sun Mar 30 19:57:12 UTC 2008 Travis Cross <tc@tra...> * Whitespaceification: removed indentation tabs throughout project. ...and replaced each tab with 8 spaces. Tab indentation had previously been used inconsistently and was dominated by space-based indentation. In the future, disable indent-tabs-mode in emacs! (setq indent-tabs-mode nil) Sun Mar 30 20:11:34 UTC 2008 Travis Cross <tc@tra...> * Some minor comment and whitespace cleanup. Sun Mar 30 22:28:56 UTC 2008 Travis Cross <tc@tra...> * Updated the documentation and test cases. Sun Mar 30 22:35:50 UTC 2008 Travis Cross <tc@tra...> * Cleaned up and rearranged the package exports a bit. Sun Mar 30 22:38:31 UTC 2008 Travis Cross <tc@tra...> * Removed deprecated exports that are no longer implemented. Sun Mar 30 22:51:16 UTC 2008 Travis Cross <tc@tra...> * Added a mechanism for expanding parenscript forms in a clean environment. Sun Mar 30 22:53:08 UTC 2008 Travis Cross <tc@tra...> * Added myself to the list of contributors. Sun Mar 30 23:39:49 UTC 2008 Travis Cross <tc@tra...> * Cleaned up and standardized indentation of JS code in the reference documentation.

2 2

[parenscript-devel] how to represent "bandInfos[1].highlight = true; "?
by Kevin Layer 06 Mar '08

06 Mar '08

(setf (.highlight (aref band-infos 1)) t) almost does it, but adds ()'s: bandInfos[1].highlight() = true; I've tried every permutation I can think of... thanks. Kevin

2 1

[parenscript-devel] automate javascript -> parenscript conversion
by Lui Fungsin 15 Feb '08

15 Feb '08

Hi, I found that manually converting javascript (sample code) to parenscript is quite annoying. Does anyone know if writing a parser/converter (in lisp, perhaps) to automate this task is a major undertaking or is relatively easy? (maybe using some of the existing lisp libraries) The reason to convert existing javascript to parenscript is that it's easier to write/maintain afterwards (you have macro, emacs's paredit mode, etc). Also, it's a lot easier to integrate in-line with cl-who or other similar libraries whereas with existing javascript you need to link them with the script tag (otherwise you have to escape all the quotes, backslashes etc to include it as a really longggggg string). >From my experience of converting existing javascript code to parenscript, the oo method call conversion is most troublesome. i.e. x.getobj().slot1.slot2.method1() /* hard to read backward if you're used to js */ => (.method (aref (aref (.getobj x) "slot1") "slot2")) /* easier to read but required a temp variable */ or (let ((obj (.getobj x))) (.method1 obj.slot1.slot2)) thanks

2 2

[parenscript-devel] Add #\= to *special-chars*?
by Daniel Gackle 01 Feb '08

01 Feb '08

Can we add #\= to *special-chars* so that instead of this: (ps blah=) => "blah=;" we get something like this: (ps blah=) => "blahequals;" ? Dan

2 1

[parenscript-devel] Bug in setf
by Daniel Gackle 01 Feb '08

01 Feb '08

There's a bug when using setf to decrement a variable: (ps (setf x (- x 1 2))) => "x -= 1 - 2;" Note that this works correctly: (ps (setf x (- 1 x 2))) => "x = 1 - x - 2;" Dan

2 1

[parenscript-devel] New Parenscript release.
by Vladimir Sedach 04 Nov '07

04 Nov '07

Available for your programming pleasure is a new Parenscript release. It fixes a critical bug with define-symbol-macro, and adds several handy macro utilities and a list of Parenscript utility functions (just mapcar for now) that you can include in your Parenscript programs. Happy hacking, Vladimir

1 0

[parenscript-devel] "Halloween" Parenscript release.
by Vladimir Sedach 31 Oct '07

31 Oct '07

A new Parenscript release is now up on common-lisp.net. It mostly addresses bug fixes, although some more control over the way JavaScript is printed is exposed as well. Happy hacking, Vladimir

1 0

[parenscript-devel] js closures in ps?
by Kevin Layer 13 Oct '07

13 Oct '07

Greetings. I'm new to parenscript. I'm currently playing with converting a lisp program that generates js "the hard way". I have some code that looks like this (excerpted from a format string): function addPlace(lat, lon, label, icon) { var point = new GLatLng(lat, lon); var marker = new GMarker(point, { icon : icon }); GEvent.addListener(marker,\"click\", function() { map.openInfoWindowHtml(point, label); }); map.addOverlay(marker); } I can find nothing in the ps docs that tell me how to deal with the function() { ... }. Did I just miss it? Kevin

2 2

[parenscript-devel] Bug: nested symbol-macrolet
by Daniel Gackle 11 Oct '07

11 Oct '07

This works in CL: (symbol-macrolet ((x 123) (y (* 2 x))) y) => 246 ... but produces an error in PS. Seems wrong? Dan

2 1