Re: datacopy syntax
Frediano Ziglio <[email protected]>
| Newsgroups | gmane.comp.db.tds.freetds |
|---|---|
| Message-ID | <CAHt6W4dEkCm_c_-3jG-Y3YmjLi4frcB_BH5ky+Nvv5PTska04Q@mail.gmail.com> |
2011/8/16 <[email protected]>: > I would like to let our "datacopy" utility use Kerberos, but the parameter > syntax makes it awkward. I propose to change the syntax. I'd like feedback. > > Also, datacopy connects to two different servers. I don't know of a way to > use two Kerberos accounts, should that be needed. Suggestions? > > The current parameters are: > > [-S server/username/password/database/table_or_view] > [-D server/username/password/database/table] > > You see the problem immediately: how to omit the username & password? > Use strsep instead of strtok and pass empty strings like -S my_server///my_db/my_table > I propose to change the syntax to emulate a URL: > > -S [username[:passwd]@][dbname.][schema.]tablename > > This would be more similar to our other utilities, and would permit everything > except the tablename to be omitted. It would, however break existing scripts. > why not using another parameter like --source to avoid clash? > I don't want to do what someone will inevitably suggest: allow both. If > you want both and want to implement both, OK, we can discuss it. Otherwise, > the complexity would be hard to document and harder to understand. > > Would this change negatively affect your organization? Would a little Perl > script transforming the above be of any use? Do you think the proposal > is a step in the right direction? > > --jkl > Frediano _______________________________________________ FreeTDS mailing list [email protected] http://lists.ibiblio.org/mailman/listinfo/freetds