Re: [PATCH] lilypond-book performance with new hash subdirs

[John Mandereau]

Han-Wen Nienhuys wrote:
> 2008/4/14, John Mandereau <address@hidden>:

> >  in current lilypond-book, it seems that using sets for reading existing
> >  output file names is a significant speed bottleneck.

> >   def split_output_files(directory):

> please convert back to a set before returning.
> 
> Subsequent functions do a lot of
> 
>  if file in files
> 
> lists get very slow when there are 10000 elements in them.

Oh yes, in split_output_files, using lists and converting to a set when
returning gives best performance as far as I tested.  I pushed that to
Git.

Cheers,
John