[Parseperl-discuss] PPI::Cache and PPI version

[Chris D. <ch...@cl...>]

I think that the PPI version number should be part of the  
PPI::Document MD5 identifier.  On my machine, I have a mixture of PPI  
SVN trunk and PPI v1.118 from CPAN.  I test against both when I work  
on Perl::Critic, and use the PPI::Cache to speed things up.  However,  
due to the recent addition of new PPI::Token::Number subclasses, I  
get errors when the cached version is from trunk and I'm running with  
the older CPAN code.

For example:
    Cannot restore overloading on HASH(0x1a6f574) (package  
PPI::Token::Number::Float) (even after a "require  
PPI::Token::Number::Float;") at blib/lib/Storable.pm (autosplit into  
blib/lib/auto/Storable/_retrieve.al) line 331, at /Users/chris/perl/ 
lib/perl5/site_perl/PPI/Cache.pm line 238

If the $PPI::VERSION were mixed into the cache ID, then this would  
not happen (except to developers working against SVN).  This would  
lead to wasted disk space (new copies in cache for each PPI version),  
but that's a minor inconvenience.

Sound good?  It will be very easy to implement.

Alternatively, the cache could store the version number of the last  
PPI that accessed it and clear itself if the version numbers don't  
match.  However, if any part of the cache were not writable this  
would cause odd failures which would not occur in the above  
implementation.

Chris

-- 
Chris Dolan, Software Developer, Clotho Advanced Media Inc.
608-294-7900, fax 294-7025, 1435 E Main St, Madison WI 53703
vCard: http://www.chrisdolan.net/ChrisDolan.vcf

Clotho Advanced Media, Inc. - Creators of MediaLandscape Software  
(http://www.media-landscape.com/) and partners in the revolutionary  
Croquet project (http://www.opencroquet.org/)