4. Running an Example Database
After completing the system installation we can now finally run a model application. Depending on the version of msql installed and the perl database interface used, we have to modify the sample programs in a few points.
First however, the file index.html residing in
/home/httpd/html/ must be modified to allow calling a
sample database application. We can place our database (which we call
database.cgi or inventur.cgi here despite its archive name
perl.lst.ck) in /home/httpd/html/test/.
We add one line (of course, depending on your installation choices)
similar to the following to index.html:
<LI>Test the <A HREF="test/database.cgi">Database, DBI:DBD-mSQL style!</A> <LI>Test the <A HREF="test/inventur.cgi">Database, MsqlPerl style!</A>
Usually you should only pick one of these two choices but if you have both types of database interface installed you can leave both lines here as they are. You can then compare performance, etc.
4.1 Adapting the sample script for MsqlPerl
Our sample script
has to be told
to use the MsqlPerl
interface. The modification takes place in several locations. First,
near the beginning of the file, we change the use clause:
# # use DBI; # Generisches Datenbank-Interface use Msql;
Then, near line 27, the MsqlPerl syntax does not require the mentioning of a specific driver:
# $dbh = DBI->connect($host, $database, '', $driver) || $dbh = Msql->connect($host, $database) ||
Then, from line 33 onward throughout the whole script, we have to
change all instances of do against query:
# $dbh->do("SELECT * FROM hw") || db_init($dbh);
$dbh->query("SELECT * FROM hw") || db_init($dbh);
Finally, in MsqlPerl speak, line 207 can be commented out:
# $sth->execute || msg("SQL Error:", $sth->errstr);
In addition, it may become necessary to swap all errstr calls
like the one in the preceding code fragment against errmsg.
This is also version dependent.
After these modifications, the script should run smoothly.
4.2 Adapting the sample script for msql-2
The SQL syntax was redefined during the development of mslq-2. The
original script will fail to execute the table initialization
statements in lines 45 -- 58. The primary key modifier is no
longer supported by msql-2, and should simply be skipped:
$dbh->do(<<EOT) || die $dbh->errstr; # Neue Personen-Tabelle
create table person (
# We do not need the 'primary key' modifier anymore in msql-2!
# pn int primary key, # Personalnummer
pn int, # Personalnummer
name char(80), # Nachname, Vorname
raum int # Raumnummer
)
EOT
$dbh->do(<<EOT) || die $dbh->errstr; # Neue Hardware-Tabelle
create table hw (
# We do not need the 'primary key' modifier anymore in msql-2!
# asset int primary key, # Inventurnummer
asset int, # Inventurnummer
name char(80), # Bezeichnung
person int # Besitzer
)
EOT
Unfortunately, this specific script will then accept new entries
with identical personnel numbers; the msql-1 modifier primary
key intends to prevent exactly this behaviour. The msql-2
documentation shows how to use the CREATE INDEX clause to
create unique entries.
Next Previous Contents