Here is another approach, by using the file record separator ($/) to actually a "record" in your data file. There are a few approaches to it, one approach requires to change the source data file to use two newlines between records; this is because $/ can't take a regex as input.

When $/ value is an empty string it will match \n\n+ or a paragraph.

Also, if you use a hash to generate a table to translate the description to the field number you can simply the script for maintaning and when adding/removing fields as well as generating a template in case your need to feed default values.

Disclosure: I don't have a SQL server to test the code, that's why I commented the code, I focused the attention on how the data is handled.

#!/usr/bin/perl

use warnings;
use strict;

my %fields = (
    'Type'              => 0,
    'No.of students'    => 1,
    'No.of classes'     => 2,
    'No.of teachers'    => 3,
    'Hostel facility'   => 4,
);


my $dbh;
#$dbh->do('DROP TABLE test1'); # uncomment after first run
#my $sql = 'CREATE TABLE test1 (
#             type     char(5),
#             students int, 
#             class    int, 
#             teachers int, 
#             hostel   char(10) )';
             
#$dbh->do($sql);
my $sqli = 'INSERT INTO test1 VALUES (?,?,?,?,?)';
#my $sth = $dbh->prepare($sqli);

my @record = ();

local $/ = "";
foreach(<DATA>) {
    chomp;
    # Fix the type field
    if(!s/^\++School\(B\):Students\++/Type:BOYS/) {
        s/^\++School\(G\):Students\++/Type:GIRLS/;
    }
    my @arr = split /[\r\n]+/;
    foreach(@arr) {
        my ($key, $data) = split /\s*:\s*/;
        $record[$fields{$key}] = $data;
    }
    print "Inserting \"@record\"\n";
#    $sth->execute(@record[0..4]);
}

__DATA__
+++++++++++School(B):Students+++++++++++
No.of students : 120
No.of classes:2
No.of teachers:5
Hostel facility:available

+++++++++++School(G):Students+++++++++++
No.of students:300
No.of classes:3
No.of teachers:10
[download]