[Bioperl-l] Knock on effects of defaulting to zero in Locations

[Ewan Birney]

Jason  ---

impressive work on the Location model stuff ;)


The effect of setting the start/end to 0 when it is fuzzy (the latest
change) is having a knock on effect on the Location.t as when the
location  

(speaking-genbank...)


   join((1..30),(<50..61)) 

is made the sublocations are sorted

     (<50..61),(1..30) 

due to the start position on the fuzzy being 0.


Therefore max_end gives out 30 rather than 61 as tested


Now --- solutions here are:


    (a) client code provided an idiotic splitlocation - change test to
only test "sane" splitlocation constructs

    (b) max_end loops over all sub loctions finding the max of all the
max_ends of its components.
           - could do this at every call to max_end
           - could do this once, cache and have a dirty cache call in
add_sub_Location
           - could do this once at the end of each add_sub_Location and
store




Not sure what you are looking for here. Feel free to weigh in - I'll fix
;)






    

-----------------------------------------------------------------
Ewan Birney. Mobile: +44 (0)7970 151230, Work: +44 1223 494420
<birney@ebi.ac.uk>. 
-----------------------------------------------------------------