Re: [racket-users] Case vs. Match

[Robby Findler]

Really this is not the best example either. I guess best to look at
the source, which is in racket/private/case and, according to a
comment in the top of the file, is based on  Clinger's paper, "Rapid
Case Dispatch in Scheme".

Robby

On Fri, Jun 3, 2016 at 6:35 PM, Robby Findler
 wrote:
> I think they need to be bigger, but case does have some smarts. Try
> this one (add a `pretty-write` if you're running in cmd-line racket).
>
> #lang racket
> (syntax->datum
>  (expand
>   #'(case x
>   [(1 2 3 4 5 6 7 8 9 10) 1]
>   [(a b c d e f g h i j k l m n o p q r s t) 2])))
>
>
> On Fri, Jun 3, 2016 at 6:23 PM, Gustavo Massaccesi  wrote:
>>  > [...] I saw in the docs that case specifically mentions it can
>> distribute in log(n) time [...]
>>
>> I think it's strange. I found the reference in
>> https://docs.racket-lang.org/reference/case.html but it's very
>> strange. Is it true???
>>
>> For example, I tried with
>>
>> ;---
>> #lang racket
>> (lambda (x)
>>   (case x
>> [(1 "1" #"1" |1|) 1]
>> [(2 "2" #"2" |2|) 2]))
>> (lambda (x)
>>   (case x
>> [(1 3 5 7) 1]
>> [(2 4 6 8) 2]))
>> ;---
>>
>> but both expand to a bunch of nested if, nothing fancy that may run in
>> O(log(N)).
>> (I don't expect that it has a smarter implementation for bigger N.)
>>
>> Gustavo
>>
>>
>> On Fri, Jun 3, 2016 at 11:40 AM, Gerald Pipes
>>  wrote:
>>> On Friday, June 3, 2016 at 10:30:35 AM UTC-4, Jon Zeppieri wrote:
 There's rarely a good reason. If you were dispatching over a very large 
 set of constants -- fixnums or characters, in particular -- I'd expect 
 better performance from case. That's about it.

 > On Jun 3, 2016, at 10:10 AM, Gerald Pipes  
 > wrote:
 >
 > I have been recently using match a lot and I was wondering what was the 
 > main benefit of using case instead of match?  It seems as though match 
 > has the distribution on a single val-exp except match allows for the 
 > pattern matching and case just uses equal?.
 >
 > --
 > You received this message because you are subscribed to the Google 
 > Groups "Racket Users" group.
 > To unsubscribe from this group and stop receiving emails from it, send 
 > an email to racket-users+unsubscr...@googlegroups.com.
 > For more options, visit https://groups.google.com/d/optout.
>>>
>>> Yes, I saw in the docs that case specifically mentions it can distribute in 
>>> log(n) time but for match there is no time mentioned. I assumed case's 
>>> performance was better because of the performance omission of match. Does 
>>> anyone know the actual performance of match?
>>>
>>> --
>>> You received this message because you are subscribed to the Google Groups 
>>> "Racket Users" group.
>>> To unsubscribe from this group and stop receiving emails from it, send an 
>>> email to racket-users+unsubscr...@googlegroups.com.
>>> For more options, visit https://groups.google.com/d/optout.
>>
>> --
>> You received this message because you are subscribed to the Google Groups 
>> "Racket Users" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to racket-users+unsubscr...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Case vs. Match

[Robby Findler]

I think they need to be bigger, but case does have some smarts. Try
this one (add a `pretty-write` if you're running in cmd-line racket).

#lang racket
(syntax->datum
 (expand
  #'(case x
  [(1 2 3 4 5 6 7 8 9 10) 1]
  [(a b c d e f g h i j k l m n o p q r s t) 2])))


On Fri, Jun 3, 2016 at 6:23 PM, Gustavo Massaccesi  wrote:
>  > [...] I saw in the docs that case specifically mentions it can
> distribute in log(n) time [...]
>
> I think it's strange. I found the reference in
> https://docs.racket-lang.org/reference/case.html but it's very
> strange. Is it true???
>
> For example, I tried with
>
> ;---
> #lang racket
> (lambda (x)
>   (case x
> [(1 "1" #"1" |1|) 1]
> [(2 "2" #"2" |2|) 2]))
> (lambda (x)
>   (case x
> [(1 3 5 7) 1]
> [(2 4 6 8) 2]))
> ;---
>
> but both expand to a bunch of nested if, nothing fancy that may run in
> O(log(N)).
> (I don't expect that it has a smarter implementation for bigger N.)
>
> Gustavo
>
>
> On Fri, Jun 3, 2016 at 11:40 AM, Gerald Pipes
>  wrote:
>> On Friday, June 3, 2016 at 10:30:35 AM UTC-4, Jon Zeppieri wrote:
>>> There's rarely a good reason. If you were dispatching over a very large set 
>>> of constants -- fixnums or characters, in particular -- I'd expect better 
>>> performance from case. That's about it.
>>>
>>> > On Jun 3, 2016, at 10:10 AM, Gerald Pipes  
>>> > wrote:
>>> >
>>> > I have been recently using match a lot and I was wondering what was the 
>>> > main benefit of using case instead of match?  It seems as though match 
>>> > has the distribution on a single val-exp except match allows for the 
>>> > pattern matching and case just uses equal?.
>>> >
>>> > --
>>> > You received this message because you are subscribed to the Google Groups 
>>> > "Racket Users" group.
>>> > To unsubscribe from this group and stop receiving emails from it, send an 
>>> > email to racket-users+unsubscr...@googlegroups.com.
>>> > For more options, visit https://groups.google.com/d/optout.
>>
>> Yes, I saw in the docs that case specifically mentions it can distribute in 
>> log(n) time but for match there is no time mentioned. I assumed case's 
>> performance was better because of the performance omission of match. Does 
>> anyone know the actual performance of match?
>>
>> --
>> You received this message because you are subscribed to the Google Groups 
>> "Racket Users" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to racket-users+unsubscr...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Case vs. Match

[Pierpaolo Bernardi]

On Sat, Jun 4, 2016 at 1:23 AM, Gustavo Massaccesi  wrote:
>  > [...] I saw in the docs that case specifically mentions it can
> distribute in log(n) time [...]

> but both expand to a bunch of nested if, nothing fancy that may run in
> O(log(N)).
> (I don't expect that it has a smarter implementation for bigger N.)

It indeed has some smarts.  Try expanding the following:

#lang racket

(lambda (x)
  (case x
((1) 1)
((2) 1)
((3) 1)
((4) 1)
((5) 1)
((6) 1)
((7) 1)
((8) 1)
((9) 1)
((10) 1)
((11) 1)
((12) 1)
((13) 1)
((14) 1)
((15) 1)
((16) 1)
((17) 1)
((18) 1)
((19) 1)
((20) 1)
))

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Case vs. Match

[Gustavo Massaccesi]

 > [...] I saw in the docs that case specifically mentions it can
distribute in log(n) time [...]

I think it's strange. I found the reference in
https://docs.racket-lang.org/reference/case.html but it's very
strange. Is it true???

For example, I tried with

;---
#lang racket
(lambda (x)
  (case x
[(1 "1" #"1" |1|) 1]
[(2 "2" #"2" |2|) 2]))
(lambda (x)
  (case x
[(1 3 5 7) 1]
[(2 4 6 8) 2]))
;---

but both expand to a bunch of nested if, nothing fancy that may run in
O(log(N)).
(I don't expect that it has a smarter implementation for bigger N.)

Gustavo


On Fri, Jun 3, 2016 at 11:40 AM, Gerald Pipes
 wrote:
> On Friday, June 3, 2016 at 10:30:35 AM UTC-4, Jon Zeppieri wrote:
>> There's rarely a good reason. If you were dispatching over a very large set 
>> of constants -- fixnums or characters, in particular -- I'd expect better 
>> performance from case. That's about it.
>>
>> > On Jun 3, 2016, at 10:10 AM, Gerald Pipes  
>> > wrote:
>> >
>> > I have been recently using match a lot and I was wondering what was the 
>> > main benefit of using case instead of match?  It seems as though match has 
>> > the distribution on a single val-exp except match allows for the pattern 
>> > matching and case just uses equal?.
>> >
>> > --
>> > You received this message because you are subscribed to the Google Groups 
>> > "Racket Users" group.
>> > To unsubscribe from this group and stop receiving emails from it, send an 
>> > email to racket-users+unsubscr...@googlegroups.com.
>> > For more options, visit https://groups.google.com/d/optout.
>
> Yes, I saw in the docs that case specifically mentions it can distribute in 
> log(n) time but for match there is no time mentioned. I assumed case's 
> performance was better because of the performance omission of match. Does 
> anyone know the actual performance of match?
>
> --
> You received this message because you are subscribed to the Google Groups 
> "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Case vs. Match

[Gerald Pipes]

On Friday, June 3, 2016 at 10:30:35 AM UTC-4, Jon Zeppieri wrote:
> There's rarely a good reason. If you were dispatching over a very large set 
> of constants -- fixnums or characters, in particular -- I'd expect better 
> performance from case. That's about it.
> 
> > On Jun 3, 2016, at 10:10 AM, Gerald Pipes  
> > wrote:
> > 
> > I have been recently using match a lot and I was wondering what was the 
> > main benefit of using case instead of match?  It seems as though match has 
> > the distribution on a single val-exp except match allows for the pattern 
> > matching and case just uses equal?.  
> > 
> > -- 
> > You received this message because you are subscribed to the Google Groups 
> > "Racket Users" group.
> > To unsubscribe from this group and stop receiving emails from it, send an 
> > email to racket-users+unsubscr...@googlegroups.com.
> > For more options, visit https://groups.google.com/d/optout.

Yes, I saw in the docs that case specifically mentions it can distribute in 
log(n) time but for match there is no time mentioned. I assumed case's 
performance was better because of the performance omission of match. Does 
anyone know the actual performance of match?

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Case vs. Match

[Jon Zeppieri]

There's rarely a good reason. If you were dispatching over a very large set of 
constants -- fixnums or characters, in particular -- I'd expect better 
performance from case. That's about it.

> On Jun 3, 2016, at 10:10 AM, Gerald Pipes  
> wrote:
> 
> I have been recently using match a lot and I was wondering what was the main 
> benefit of using case instead of match?  It seems as though match has the 
> distribution on a single val-exp except match allows for the pattern matching 
> and case just uses equal?.  
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Case vs. Match

[Matthias Felleisen]

match is far more expressive than case. Think of 

> (match 3 [(? even? x) (+ x 1)][(? odd? x) (- x 1)])

But of course you’re paying for in terms of run-time cost. 

Engineering is a trade-off game. 



> On Jun 3, 2016, at 10:10 AM, Gerald Pipes  
> wrote:
> 
> I have been recently using match a lot and I was wondering what was the main 
> benefit of using case instead of match?  It seems as though match has the 
> distribution on a single val-exp except match allows for the pattern matching 
> and case just uses equal?.  
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[racket-users] Case vs. Match

[Gerald Pipes]

I have been recently using match a lot and I was wondering what was the main 
benefit of using case instead of match?  It seems as though match has the 
distribution on a single val-exp except match allows for the pattern matching 
and case just uses equal?.  

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] auto-defining variables mentioned in a program

[Matthew Butterick]

On Jun 2, 2016, at 11:47 AM, Jack Firth  wrote:

> - Have #%top somehow magically figure out if it should make a new box or 
> reuse an existing one for the identifier it's given.

(call-with-future-miracle ...)

I'll think about it more — but my hunch is this problem is more related to 
parsing the program into an AST, not macro expansion per se. (Moreover, it's 
easy for me to see how certain adjustments to the parser grammar can make the 
variable names easier to find & extract; harder to see how to do the same with 
the expander.)

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.