Hi,

2016/7/25 Mon 3:47:51 UTC+9 Ken Takata wrote:
> Hi,
> 
> 2016/7/22 Fri 22:29:33 UTC+9 Ken Takata wrote:
> > Hi ZyX,
> > 
> > 2016/7/22 Fri 20:15:33 UTC+9 ZyX wrote:
> > > >> I was wrong regarding the consequences and why you should alter the
> > > >> GC: it is not memory leak because cycle is not GC’d. It is *crash*
> > > >> because Vim does not know that dictionary is referenced:
> > > >>
> > > >> ```VimL
> > > >> function F()
> > > >>     let d = {}
> > > >>     return {-> d}
> > > >> endfunction
> > > >> let L = F()
> > > >> call garbagecollect(1)
> > > >> call feedkeys(":echo L()\n", 'n')
> > > >> ```
> > > >>
> > > >> Save to `test.vim` and run as `vim -u NONE -i NONE -N -S test.vim`.
> > > >> When I use it in Vim with optimizations it crashes immediately:
> > 
> > Yes, I also noticed this. I tested with the following script:
> > 
> >   function! Test_lambda_closure()
> >     function! s:foo()
> >       let x = [0]
> >       return {-> [execute("let x[0] += 1"), x[0]][-1]}
> >     endfunction
> >   
> >     let l:F = s:foo()
> >     call test_garbagecollect_now()
> >     call assert_equal(1, l:F())
> >     call assert_equal(2, l:F())
> >     call assert_equal(3, l:F())
> >     call assert_equal(4, l:F())
> >   endfunction
> > 
> > I have updated the patch:
> > https://bitbucket.org/k_takata/vim-ktakata-mq/src/52c8d4fd0af2dd8bd2c79204dbbedd9ded874439/lambda-update.patch?at=default
> > 
> > Now it deals with GC. Also add some tests, reduce memory when local 
> > variables
> > or arguments are not used.
> 
> I have slightly updated the tests:
> https://bitbucket.org/k_takata/vim-ktakata-mq/src/5f8f4f212a09b1fa58a939f34559c7b3c88bb616/lambda-update.patch?at=default
> 
> It seems work well. BTW, I'm thinking the implementation again.
> I thought that capture by value is easier to implement, but it is harder to
> apply to normal functions inside a function. Capture by reference seems easier
> to apply to normal functions.
> 
> I read the comment for mattn's implementation by Bram again:
> 
> > In the implementation it seems the dictionary storing the function-local
> > variables is kept for a very long time. This relies on the garbage
> > collector. It's better to use reference counting to be able to free the
> > dictionary as soon as it's unused.
> > 
> > Also, the lambda always keeps the function-local variable dict, even
> > when it's not actually used. That makes lambdas a expensive.
> > It would be better to explicitly state the lambda is using its context.
> > Then we can also do that with ":function", so that we are not forced to
> > use a lambda if we want a closure.
> 
> Checking if a lambda is a closure is now available with my patch.
> So I tried to implement reference counting based on mattn's patch.
> Unfortunately it doesn't work well yet. I need help for this.
> https://bitbucket.org/k_takata/vim-ktakata-mq/src/5f8f4f212a09b1fa58a939f34559c7b3c88bb616/lambda-capture_by_reference-temp.patch?at=default
> (This patch should be applied after the above patch (lambda-update.patch).)
> Test_circular_reference() in test_lambda.vim doesn't work well.
> 
> Which is better, capture by value or by reference?

I have updated the patches.

Capture by value:
https://bitbucket.org/k_takata/vim-ktakata-mq/src/05cb2721d17d4b1c52b603037e3ec65e523b472d/lambda-update.patch?at=default

Capture by reference:
https://bitbucket.org/k_takata/vim-ktakata-mq/src/05cb2721d17d4b1c52b603037e3ec65e523b472d/lambda-capture_by_reference.patch?at=default
(Apply on top of lambda-update.patch)

Now both seem to work almost good.
And I wrote another patch to support closure with normal functions:
https://bitbucket.org/k_takata/vim-ktakata-mq/src/cf0cce51b390335d5ed5a5dffa933e10f16d3aab/closure.patch?at=default
(Apply on top of lambda-capture_by_reference.patch)

The :function command supports [closure] argument now:

  function! Foo()
    let x = 0
    function! Bar() closure
      let x += 1
      return x
    endfunction
    return function('Bar')
  endfunction
  
  let F = Foo()
  echo F() " 1
  echo F() " 2
  echo F() " 3
  echo F() " 4

(Documents and tests are not updated yet.)

Regards,
Ken Takata

-- 
-- 
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php

--- 
You received this message because you are subscribed to the Google Groups 
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Raspunde prin e-mail lui