FAQ
No.
You have to store the two modules in different places and take care
which one you want to use. You could do it with the following shell
script:
<![CDATA[
for L in /usr/local/Minerva /usr/local/Hughes ; do
echo ------------------------------------
echo Configuring for $L
echo ------------------------------------
perl Makefile.PL INSTALLSITELIB=$L/perl INSTALLSITEARCH=$L/perl \
INSTALLSCRIPT=$L/bin
make
make test
make install
done
]]>
From then you can talk to mSQL-1 by running
<![CDATA[
perl -I/usr/local/Minerva/perl yourprogram
]]>
and you talk to mSQL-2 by running
<![CDATA[
perl -I/usr/local/Hughes/perl yourprogram
]]>
Or you insert the library location in your own program with
<![CDATA[
use lib '/usr/local/Minerva/perl';
]]>
After above procedure the two installed pmsql(1) programs are
identical. You could insert such a 'use lib' statement in each of them
to have them talk to the appropriate server. Or you call them as perl
programs the same way as ``yourprogram'' above.
Precanned reply:
<![CDATA[
read the README
read the manpage
read the FAQ that comes with MsqlPerl
read the tutorial in t/msql.t
]]>
subscribe to the msqlperl mailing list by sending mail to
msqlperl-request@franz.ww.tu-berlin.de with "subscribe msqlperl
your.name@your.domain" in the body of the message. Ask the question to
the mailing list by writing to msqlperl@franz.ww.tu-berlin.de. Wait
for the answer. Keep subscribed to the mailing list and answer the
same question as soon as somebody else asks it.
Please, read the manpage. Type 'man Msql' and if that doesn't help,
try 'perldoc Msql'.
Msql.pm IS the manpage in pod format. If you're on UNIX you should be
able to read it with 'man Msql'. On other operating systems try
'perldoc Msql'. Also, the test script, t/msql.t might help you to sort
things out. The most recent addition is the pmsql program which may
also help you to learning by example.
<![CDATA[
From: Yiorgos Adamopoulos <Y.Adamopoulos@noc.ntua.gr>
I successfully compiled MsqlPerl for a HP817 (HP-UX 9.04/s800) using Perl5.002,
msql-1.0.14 and gcc-2.7.2. I added CC_ONLY = -fPIC to site.mm after running
setup.
]]>
The internal fetchhash routine is quite good. If you need speed, you
should hardcode your column names into a hash slice. Everything else
is a terrible waste.
- closure: 8 secs ( 7.47 usr 0.03 sys = 7.50 cpu)
- hashslice via name: 10 secs ( 9.51 usr 0.04 sys = 9.55 cpu)
- internal_fetchhash: 7 secs ( 7.05 usr 0.03 sys = 7.08 cpu)
- brute force fetchhash: 15 secs (13.90 usr 0.05 sys = 13.95 cpu)
- hashslice via constants: 6 secs ( 5.57 usr 0.02 sys = 5.59 cpu)
<![CDATA[
#!/usr/bin/perl
use Msql;
my $db = Msql->Connect("","mod");
$sth = $db->Query("select * from mods");
sub Msql::Statement::brute_fetchhash {
my $self = shift;
my %hash;
@hash{ $self->name } = $self->FetchRow and \%hash;
}
sub Msql::Statement::gimme_a_fetchhash_routine {
my $self = shift;
my @z = $sth->name;
sub { my %hash; @hash{@z} = $self->FetchRow and \%hash;};
}
use Benchmark;
timethese(20, {
'hashslice via constants' => q{
$sth->DataSeek(0);
while (@row{qw/modid userid maillistid chapterid seqdummy statd stats statl stati description changed changedby/} = $sth->FetchRow){
$X = $row{modid};
print "$X\n" ;
}
},
' hashslice via name' => q{
$sth->DataSeek(0);
while (@row{$sth->name} = $sth->FetchRow){
$X = $row{modid};
print "$X\n" ;
}
},
' brute force fetchhash' => q{
$sth->DataSeek(0);
while ($hashref = $sth->brute_fetchhash){
$X = $hashref->{modid};
print "$X\n" ;
}
},
' closure' => q{
$sth->DataSeek(0);
$clos = $sth->gimme_a_fetchhash_routine;
while ($hashref = &$clos){
$X = $hashref->{modid};
print "$X\n" ;
}
},
' internal_fetchhash' => q{
$sth->DataSeek(0);
while (%hash = $sth->fetchhash){
$X = $hash{modid};
print "$X\n" ;
}
},
});
]]>
Hello I am running XXX and would really like to use Msqlperl but I am
getting this error message.
<![CDATA[
Can't load module Msql, dynamic loading not available in this perl.
(You may need to build a new perl executable which either supports
dynamic loading or has the Msql module statically linked into it.)
at /usr/lib/perl5/site_perl/Msql.pm line 56
BEGIN failed--compilation aborted at ./sql.pl line 5.
]]>
There is a manpage dedicated to building and installing modules:
ExtUtils::MakeMaker. The long answer is in there. The short answer is:
run 'make perl' in the MsqlPerl build directory. The third answer is,
get an operating system that supports dynamic loading. With Linux it's
already quite standard to have dynamic loading, and so is with many
other OSs. Check out the newgroups for your OS and consult the FAQs
there.
The icc routine below should do what you want. It's hardly tested and
I'm reluctant to add it to the distribution. But feel free to convince
me that including it is a good thing.
<![CDATA[
#!/usr/bin/perl -w
use Msql;
use strict;
sub Msql::icc {
my($db,$table,@fields)=@_;
my($s,@types)=$db->listfields($table) or return;
if (@fields){
# The user wants to input the fields in a different order
my(@tfields,@ttypes,%tfields);
@tfields = $s->name;
@ttypes = $s->type;
@tfields{@tfields} = @ttypes;
@types = @tfields{@ttypes};
} else {
# They are gonna give the arguments in table's order
@fields = $s->name;
@types = $s->type;
}
# return a subroutine reference.
sub {
my(@arr)=@_;
return join " ",
"insert into $table values (",
join(
",",
map {
defined $arr[$_] ?
$types[$_] == &Msql::REAL_TYPE
||
$types[$_] == &Msql::INT_TYPE ?
$arr[$_]+0 :
Msql->quote($arr[$_]) :
"NULL"
}
0..$#types
),
")\n";
}
}
# Example for usage:
{
my $dbh=Msql->connect("","test");
my $f=$dbh->icc("pwent");
my(@info,$query,$sth);
while (@info = getpwent()) {
print $query =&$f(@info);
$sth = $dbh->query($query) or die Msql->errmsg;
}
}
]]>