I attach argument-variables that is passed to BaseAdapter.select() (by using pydev).
Do you mean as such? I'm working for sahana-eden project. It will be used for the earthquake rescue of Japan. http://eden.sahanafoundation.org/wiki PostgreSQL errors are caused by some querys which succeeds on other DBs. I want you to cooperate in the solution. Thanks 2011/4/4 Massimo Di Pierro <[email protected]>: > Can you show us the web2py dal expression that caused the invalid > query? Thanks! > > On Apr 3, 8:19 am, fuga <[email protected]> wrote: >> web2py generate SQL that PostgreSQL doesn't accept. >> >> For instance, >> >> SELECT DISTINCT a.id, a.uid, a.name FROM a, b WHERE ((((a.deleted <> >> 'T') AND (a.id > 0)) AND (((b.deleted <> 'T') AND (b.id > 0)) AND >> (b.id = 1))) AND (b.uid = a.uid)) ORDER BY b.id, a.id LIMIT 1 OFFSET >> 0; >> >> It fails because select list is not contain "b.id". >> >> I change gluon/dal.py(Version 1.94.5) as follows. >> It is work well. >> But I am not sure that it is a good practice. >> >> === modified file 'gluon/dal.py' Version 1.94.5 >> @@ -1039,2 +1039,6 @@ >> -------------------------------------------------------------- >> if not orderby and tablenames: >> - sql_o += ['%s.%s'%(t,x) for t in tablenames for x in >> ((hasattr(self.db[t],'_primarykey') and self.db[t]._primarykey) or >> [self.db[t]._id.name])] >> + orderby_fields = ['%s.%s'%(t,x) for t in tablenames for x in >> ((hasattr(self.db[t],'_primarykey') and self.db[t]._primarykey) or >> [self.db[t]._id.name])] >> + sql_o += ' ORDER BY %s' % ', '.join(orderby_fields) >> + for orderby_field in orderby_fields: >> + if orderby_field not in self._colnames: >> + sql_f += ', %s' % orderby_field >> >> Your opinion will be much appreciated, >> Cheers,
variables
Description: Binary data
