Perl to Ruby

[Arich Chanachai]

Also looking for a Perl to Ruby conversion utility.

Much appreciated.

- Arich

 

[James Edward Gray II]

Also looking for a Perl to Ruby conversion utility.

You have a lot of questions about language cross-over. What is it you
are looking for exactly? Perhaps we can provide better information if
we understand what you're looking to do.

James Edward Gray II

 

[Ara.T.Howard]

Also looking for a Perl to Ruby conversion utility.

Much appreciated.

- Arich

i work for cheap.

;-)

-a
--
===============================================================================
| EMAIL :: Ara [dot] T [dot] Howard [at] noaa [dot] gov
| PHONE :: 303.497.6469
| A flower falls, even though we love it;
| and a weed grows, even though we do not love it.
| --Dogen
===============================================================================

 

[Arich Chanachai]

You have a lot of questions about language cross-over. What is it you
are looking for exactly? Perhaps we can provide better information if
we understand what you're looking to do.

James Edward Gray II

I need speed comparable to that of the .NET framework with a dynamic
language such Ruby or Python. So far I see .NET/Mono and Java as the
only JIT frameworks, at least worth looking at. I need to maintain the
libraries of Python or Ruby in this cross-over however, and I do not
know if such implementations have been created. I am aware PyCs is in
development for .NET/Mono and Jython for Java. I am looking for a
similar implementation of Ruby. It seems that JRuby does not compile to
Java bytecode and Jython does not match pure Java code in speed. I have
read that alot of Jython code is still interpreted and is thus "10x
slower" than a Java implementation.

Someone suggested that JRuby could be layed onto Groovy to take
advantage of Groovy's bytecode compilation. Anyone here know anything
about the feasibility of this or whether this would achieve the speed I
desire? I couldn't find anywhere on the Groovy website what the speed
comparability was between it and pure Java, but perhaps I am blind.

Thanks all.

 

[Arich Chanachai]

You have a lot of questions about language cross-over. What is it you
are looking for exactly? Perhaps we can provide better information if
we understand what you're looking to do.

James Edward Gray II

Also, I have large libraries of Perl, Python, Ruby, Tcl, and C/C++ code.
Which ever language + JIT implementation I choose, I will need to port
these libraries. So SWIG can be used a tool for wrapping up C/C++
libraries? If so, then perhaps all I need is to convert my modules and
libraries into C/C++. Then I can wrap in SWIG and interface using my
choice of language + JIT compiler.

I remember reading somewhere on the Python site about a compiler in the
workers, anyone know about that? Guess I'll pop over to their
newsgroup. How about Ruby to C/C++? I may have found one in Japanese
but who knows.

Thanks all.

 

[Gavin Sinclair]

I need speed comparable to that of the .NET framework with a dynamic
language such Ruby or Python. So far I see .NET/Mono and Java as the
only JIT frameworks, at least worth looking at. I need to maintain the
libraries of Python or Ruby in this cross-over however, and I do not
know if such implementations have been created. I am aware PyCs is in
development for .NET/Mono and Jython for Java. I am looking for a
similar implementation of Ruby. It seems that JRuby does not compile to
Java bytecode and Jython does not match pure Java code in speed. I
have read that alot of Jython code is still interpreted and is thus
"10x slower" than a Java implementation.

Someone suggested that JRuby could be layed onto Groovy to take
advantage of Groovy's bytecode compilation. Anyone here know anything
about the feasibility of this or whether this would achieve the speed I
desire? I couldn't find anywhere on the Groovy website what the speed
comparability was between it and pure Java, but perhaps I am blind.

Thanks all.

Sorry I can't help with the specifics, but my general advice is this: all
the projects you mention are immature, with the possible exception of
JPython. So I wouldn't accept any general statements about their
performance -- the figures would already be out of date. I don't even
recall any discussion on this list about their performance over the last
couple of years.

Summary: you'll have to test them out yourself, which shouldn't be too
hard, since Jython, JRuby, and Groovy are all demonstrated in those IBM
articles.

The most mature of those projects is Jython, I suspect, so that's likely
to be your best bet, given your requirements.

Cheers,
Gavin

 

[Martin DeMello]

Someone suggested that JRuby could be layed onto Groovy to take
advantage of Groovy's bytecode compilation. Anyone here know anything
about the feasibility of this or whether this would achieve the speed I
desire? I couldn't find anywhere on the Groovy website what the speed
comparability was between it and pure Java, but perhaps I am blind.

If you're specifically looking for something to run atop the JVM, I'd
say Groovy was a better bet than JRuby. Also, don't miss Scala
(http://scala.epfl.ch/index.html) which does compile to bytecode, and
which has a .NET compiler planned too.

martin

 

[Robert Klemme]

Sorry I can't help with the specifics, but my general advice is this: all
the projects you mention are immature, with the possible exception of
JPython.

Although I didn't test it completely, my impression is that Groovy is
quite mature, too.

So I wouldn't accept any general statements about their
performance -- the figures would already be out of date. I don't even
recall any discussion on this list about their performance over the last
couple of years.

Yes, performance for the application case at hand must be measured.
General comparisons don't help much here.

Summary: you'll have to test them out yourself, which shouldn't be too
hard, since Jython, JRuby, and Groovy are all demonstrated in those IBM
articles.
Yes!

The most mature of those projects is Jython, I suspect, so that's likely
to be your best bet, given your requirements.

As I said, I didn't do extensive testing of Groovy and I don't know
Jython - but Groovy seems to be quite mature (as well).

Kind regards

robert

 

[gabriele renzi]

Arich Chanachai ha scritto:

I need speed comparable to that of the .NET framework with a dynamic
language such Ruby or Python.

I my opinion, the right thing to do is to go the way that usually other
people go: write it in full ruby (or python). Then benchmark it and
rewrite the cpu intensive code in C (or c++ or ocaml or whatever).

 

[James Edward Gray II]

I need speed comparable to that of the .NET framework with a dynamic
language such Ruby or Python.

I agree with the poster who suggested Ruby with a little C mixed in for
speed, but...

You've run benchmarks I assume, by your above statement. How for
behind .NET is Ruby for you? Perhaps we could help optimize, if it's
not a great gap we must cross.

What's the application?

James Edward Gray II