Obwohl [EMAIL PROTECTED] schon am 27.12.01 schrieb:
> Hi,
>
> I get on your nerves, I know... But I need the opinion of experts:
> I am using Linux, so there is no SQL-Studio. My database constists of more
> than 80 tables. I want to create the database in a batch-like mode. What is
> the best way to create tables? Using dbmcli with sql_execute is a little bit
> long-winded (you need to insert '/' and no comments allowed for the
> fields..). Is it a better choice to use PERL? Are there samples?
>
> Thanks in advance for comments
>
> Thomas
Well, you can also write a batch of sql statements and have them
executed by the replication manager. That could look like this:
/* --------------------------------------------------------
/ A line beginning with a slash is a remark
/
/ and you can even emulate the remark identifiers of
/ most popular sql processors. But you must have empty
/ lines between sql statements.
/*/
drop table AdditionalInfo
create table AdditionalInfo (
ADDINFO_ID varchar(40),
EVENT_ID varchar(40),
parameter varchar(30),
my_value varchar(30),
unit varchar(30),
description varchar(100),
primary key (ADDINFO_ID)
)
create index ix1_AdditionalInfo on AdditionalInfo (EVENT_ID)
grant select on table AdditionalInfo to plr_user
grant select, insert, update, delete on table AdditionalInfo to plr_worker
/* -------------------------------------------------------- */
Execute the sql batch file with the following command in a DOS box:
repmcli -d <db> -u <username>,<password> -b <sqlbatchfilename>
For the example above, the -E (number of allowed errors) option
might be useful, to avoid stopping execution when the table of the
drop table statement does not exist.
Rüdiger
______________________________________________________________________________
Darf es ein bisschen mehr sein? Mehr Speicher, mehr Mail, mehr Erlebnis,
mehr Prämie, mehr WEB.DE. Der WEB.DE Club - http://club.web.de
