[GRASS-dev] Re: [bug #5341] Data corruption problem with v.db.select

[Hamish]

> http://intevation.de/rt/webrt?serial_num=5341

Well, that didn't take long. We now have a nasty 6.2.1 blocker bug.
See the spearfish example below.


HB:
> > v.db.select segfaults if SQL query is malformed:
> > 
> > G63> v.db.select -c fish_sites col=site_id where="CAT = 1"
> > Ai1
> > 
> > G63> v.db.select -c fish_sites col=site_id where="CAT=1"
> > Segmentation fault
MS:
> Why do you think "CAT=1" is malformed and "CAT = 1" is not?
> See eg. http://www.w3schools.com/sql/sql_where.asp. Both notations are
> allowed.
>
> Also GRASS manual mentions both, eg.
> http://grass.itc.it/grass63/manuals/html63_user/sql.html:
> 
> d.vect map where="individual='juvenile'

yes you are correct. (fyi, I put that example in, we were tracking
the movements of sea lion pups)

re-testing for that sea lion map, it is only v.db.select that is broken.
d.vect, v.extract work without the extra whitespace.

the column= where= options are newer than 6.0.2, so can't test that, but
they do break in 6.1.0.


> This bug is similar to my recent
> http://intevation.de/rt/webrt?serial_num=5341,

It is the same bug. Merged.


#spearfish

#includes Brad's updates from 2 days ago

G63> v.db.select -c archsites where="CAT = 1"
1|Signature Rock


G63> v.db.select -c archsites where="CAT=1"
DBMI-DBF driver error:
SQL parser error in statement:
SELECT * FROM archsites WHERE CAT=1)
Error in db_open_select_cursor()
ERROR: Cannot open select cursor


G63> v.db.select -c archsites column=str1 where="CAT = 1"
Signature Rock

G63> v.db.select -c archsites column=str1 where="CAT=1"
No Name

G63> v.db.select archsites where="CAT = 1" col=cat
cat
1

G63> v.db.select archsites where="CAT=1" col=cat
cat
11

==> that is very bad, it reports the attribute of the wrong cat !


Hamish