<div>I like your suggestion of emitting TRY/CATCH only in the cases where</div><div>it's necessary, i.e. only when trying to escape out of more than one</div><div>level of function nesting, seems like a good way to go. Then you're</div>
<div>only paying for the ugliness when you need it. It's in keeping with</div><div>PS's philosophy of staying close to what one would write by hand.</div><div><br><br><div class="gmail_quote">On Wed, Aug 18, 2010 at 6:12 AM, Red Daly <span dir="ltr"><<a href="mailto:reddaly@gmail.com">reddaly@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">I added RETURN-FROM and BLOCK without too much effort using the<br>
implicit return functionality and try/catch. In my view this is the<br>
most reasonable way to implement this in the general case, since<br>
BLOCK/RETURN-FROM require non-local exit much in the same way that<br>
lisp's TRY/CATCH do.<br>
<br>
The alternative to this approach is to exit from each function in the<br>
call stack via a Javascript `return' statement. Unfortunately, the<br>
call stack can contain many functions code over which the Parenscript<br>
compiler exerts little control, requiring throw as the control<br>
transfer mechanism. Thus, in the general case of unknown code on the<br>
call stack, there is no means to exit without a throw. I do not view<br>
throwing as an ugly solution at all, since try/catch was designed for<br>
non-local exits of all sorts.<br>
<br>
Nonetheless, using try/catch to implement Parenscript features<br>
deserves some attention. Programs will need to ensure that they do<br>
not use try/catch in a way that interferes with the Parenscript<br>
convention. Generally, try/catch blocks should only catch specific<br>
exceptions and re-throw PS's exceptions. I'm happy to also implement<br>
a safe TRY/CATCH wrapper that re-throws Parenscript errors and catches<br>
everything else, too. However, we may want to make an official<br>
interface change to try/catch if any lisp-style non-local exit code<br>
becomes part of the language.<br>
<br>
I present an example of why try/catch is unavoidable inline below:<br>
<div><div></div><div class="h5"><br>
On Fri, Apr 9, 2010 at 3:42 PM, Vladimir Sedach <<a href="mailto:vsedach@gmail.com">vsedach@gmail.com</a>> wrote:<br>
> Makes sense to me. I'll add this to my todo list (which I'll publish<br>
> in an email as soon as I'm done my current work on the PS compiler).<br>
><br>
> Vladimir<br>
><br>
> 2010/4/9 Daniel Gackle <<a href="mailto:danielgackle@gmail.com">danielgackle@gmail.com</a>>:<br>
>> I just pushed a patch (authored by Scott) to implement JS's LABEL and BREAK<br>
>> in PS. (Note that this patch deprecates LABELED-FOR since you can get the<br>
>> same effect by combining LABEL and FOR. Was anybody using LABELED-FOR?)<br>
>> Here's an example:<br>
>> (label scope<br>
>> (foo)<br>
>> (when (bar)<br>
>> (break scope))<br>
>> (blee))<br>
>> =><br>
>> scope: {<br>
>> foo();<br>
>> if (bar()) {<br>
>> break scope;<br>
>> };<br>
>> blee();<br>
>> };<br>
>> I was astonished to discover recently that JS has supported this ability all<br>
>> along in the form of labeled statements and labeled breaks. I'd always<br>
>> assumed that to get explicit returns from an arbitrary scope, you'd have to<br>
>> resort to the ugly hack of muscling TRY/CATCH to do it, thinking that this<br>
>> was the closest JS counterpart.<br>
>> (See <a href="http://news.ycombinator.com/item?id=793092" target="_blank">http://news.ycombinator.com/item?id=793092</a> for a thread in which<br>
>> several people believe this.) But it appears we were all wrong.<br>
>> What's not clear yet is how far this can be taken. Can you use it inside a<br>
>> nested local function to return immediately from the top-level function?<br>
>> That is one thing I've wanted for a long time.<br>
>> In the ideal case, LABEL/BREAK could be used as a base for implementing a<br>
>> proper BLOCK and RETURN-FROM in PS, something which we'd long believed to be<br>
>> impossible. One challenge is that in CL, RETURN-FROM can take a value, which<br>
>> becomes the value of BLOCK. In other words, BLOCK in CL is an expression<br>
>> while LABEL in JS is not. It seems, though, that most of this challenge has<br>
>> already been conquered with the development of implicit return in PS. The<br>
>> only thing you'd need to add is detecting when BLOCK is being used as an<br>
>> expression, declaring a gensymed variable and assigning whatever is<br>
>> happening inside BLOCK to that variable (much like implicit return already<br>
>> does with e.g. CASE), then put the variable in the expression position that<br>
>> BLOCK was in. It seems like this ought to work. It would also make things<br>
>> like this possible in PS:<br>
>> (1+ (case foo<br>
>> (:eleven 11)<br>
>> (:twelve 12)))<br>
>> Vladimir (and everybody), is the above clear? What do you think of it?<br>
<br>
</div></div>As stated above, I think try/catch is the way to go. There is no<br>
other way to exit a stack of functions in the general case otherwise.<br>
<br>
For example, I can write a Javascript function that calls its argument<br>
infinity times and never returns.<br>
<br>
function mapForever(fn) {<br>
while(true) fn();<br>
}<br>
<br>
Now consider some parenscript:<br>
<br>
(block non-local<br>
(map-forever<br>
(lambda ()<br>
(return-from non-local "we got out!"))))<br>
<br>
To extricate itself from map-forever, there is no alternative but JS's<br>
throw statement.<br>
<br>
Even if we had the ability to alter every function in the system, it<br>
would be necessary to inspect nearly every function call's return<br>
values to properly unwind the stack to the appropriate BLOCK.<br>
<br>
Having said all that, there are cases when try/catch is not necessary<br>
for BLOCK/RETURN-FROM, as you have described. BLOCK should emit code<br>
according to the contexts in which RETURN-FROM appears. If there is a<br>
RETURN-FROM inside the same function, BLOCK can use a label for a<br>
local exit. If RETURN-FROM appears inside a lambda, try/catch is<br>
necessary (except in cases where you want to optimize this away by<br>
inspecting how that lambda gets passed around). If there are no<br>
return-froms, just emit a PROGN.<br>
<br>
My solution does not do the local optimization, but it does refrain<br>
from putting try/catches around code with no return-froms.<br>
<font color="#888888"><br>
<br>
<br>
Red<br>
</font><div><div></div><div class="h5"><br>
>> Daniel<br>
>> _______________________________________________<br>
>> parenscript-devel mailing list<br>
>> <a href="mailto:parenscript-devel@common-lisp.net">parenscript-devel@common-lisp.net</a><br>
>> <a href="http://common-lisp.net/cgi-bin/mailman/listinfo/parenscript-devel" target="_blank">http://common-lisp.net/cgi-bin/mailman/listinfo/parenscript-devel</a><br>
>><br>
>><br>
><br>
> _______________________________________________<br>
> parenscript-devel mailing list<br>
> <a href="mailto:parenscript-devel@common-lisp.net">parenscript-devel@common-lisp.net</a><br>
> <a href="http://common-lisp.net/cgi-bin/mailman/listinfo/parenscript-devel" target="_blank">http://common-lisp.net/cgi-bin/mailman/listinfo/parenscript-devel</a><br>
><br>
<br>
_______________________________________________<br>
parenscript-devel mailing list<br>
<a href="mailto:parenscript-devel@common-lisp.net">parenscript-devel@common-lisp.net</a><br>
<a href="http://common-lisp.net/cgi-bin/mailman/listinfo/parenscript-devel" target="_blank">http://common-lisp.net/cgi-bin/mailman/listinfo/parenscript-devel</a><br>
</div></div></blockquote></div><br></div>