Trivial bug or bad documentation of Tor (from homolog.lib)

[Oleksandr]

Tor's documentation says:

USAGE:   Tor(v,M,N[,any]);  v int resp. intvec, M,N modules
COMPUTE: a presentation of Tor_k(M',N'), for k=v[1],v[2],...
...
RETURN:  - if v is of type int: module Tor, a presentation of Tor_k(M',N');@*
         - if v is of type intvec: a list of Tor_k(M',N') (k=v[1],v[2],...);@*
         - in case of a third argument of any type: 

BUT: if 'v' contains negative last entries it may ONLY print:

// Tor_i=0 for i<0!

and return [1] without doing anything for other v[i]-s, since sort is wrong for such intvec-s:

> sort(3..-3)[1];
0,1,2,3,-1,-2,-3

Please consider using sortIntvec from dmodapp.lib

Besides, the line

RETURN:  - if v is of type int: module Tor, a presentation of Tor_k(M',N');@*

looks a bit strange (v<->k).

[Oleksandr]

In fact using sort on an integer vector is an overkill just for computing the maximal entry... (the sorted intvec v1 is never used afterwards)

[seelisch]

sort should do what it is supposed to; check kernel implementation vs. LIB implementation (general.lib), check usage in LIBs

[Oleksandr]

with the commit (​http://www.singular.uni-kl.de:8002/trac/changeset/13800) I have fixed the bugs I found in homolog.lib.

The buggy sort goes to another ticket...