Pre-PEP: __hashable__

[Duncan Booth]

Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote in 
news:mailman.1039581793.4260.python-list at python.org:

> Not all tuples are hashable.  Actually, that clears up one final 
> point of discussion I was having with Tim Delaney.  Tuples are NOT 
> replacements for his immutablelists, since his immutablelists ARE 
> meant to always be hashable (I assume, or else they cannot be 
> created...)
> 

This means that the __hashable__ method has to be recursive, and quite 
probably has to take steps to avoid infinite recursion as with deepcopy:

aList = [ {'a': 1} ]

Ok, aList.__hashable__ has to call hashable on each of its elements.

bList = [ 1, 2, 3 ]
blist[1] = blist

At this point bList.__hashable__ had better be maintaining some sort of 
state dictionary to avoid the infinite recursion. Perhaps in this simple 
case it would be alright just to give up and throw an exception when the 
recursion limit is reached, but I think it wouldn't be too hard to produce 
a data structure that explodes its memory usage unreasonably unless you do 
the same sort of checking as for deepcopy. e.g.

cList = [ somelargishdictionary ] * 2000
cList[2000] = cList

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?