package Oracle::Schema; # Perl standard modules use strict; use warnings; use Carp; use DBI; use Debug::EchoMessage; use Oracle::DML::Common qw(:db_conn); require 5.003; $Oracle::Schema::VERSION = 0.02; require Exporter; our @ISA = qw(Exporter); our @EXPORT = qw(); our @EXPORT_OK = qw( get_table_definition ); our %EXPORT_TAGS = ( all => [@EXPORT_OK], table => [qw(get_table_definition)], ); our @IMPORT_OK = qw( get_dbh is_object_exist debug echoMSG disp_param ); =head1 NAME Oracle::Schema - Perl class for Oracle Schema Information and Management =head1 SYNOPSIS use Oracle::Schema; my %cfg = ('conn_string'=>'usr/pwd@db'); my $os = Oracle::Schema->new; # or combine the two together my $os = Oracle::Schema->new('cs'=>'usr/pwd@db'); $os->display_objects; =head1 DESCRIPTION This class includes methods to query (find, retrieve, and compare) objects in an Oracle schema and to manage (create, drop, update, merge, and move) Oracle objects. =cut =head2 new (cs=>'usr/pwd@db',tn=>'my_table') Input variables: $cs - Oracle connection string in usr/pwd@db $tn - Oracle table name without schema Variables used or routines called: None How to use: my $obj = new Oracle::Schema; # or my $obj = Oracle::Schema->new; # or my $cs = 'usr/pwd@db'; my $tn = 'my_table'; my $obj = Oracle::Schema->new(cs=>$cs,tn=>$tn); # or my $obj = Oracle::Schema->new('cs',$cs, 'tn',$tn); Return: new empty or initialized Oracle::Schema object. This method constructs a Perl object and capture any parameters if specified. It creates and defaults the following variables: $self->{conn_string} = ""; # or $self->{cs} $self->{table_name} = ""; # or $self->{tn} =cut sub new { my $caller = shift; my $caller_is_obj = ref($caller); my $class = $caller_is_obj || $caller; my $self = bless {}, $class; my %arg = @_; # convert rest of inputs into hash array foreach my $k ( keys %arg ) { if ($caller_is_obj) { $self->{$k} = $caller->{$k}; } else { $self->{$k} = $arg{$k}; } } my $vs = 'conn_string,table_name,cs,tn'; foreach my $k (split /,/, $vs) { $self->{$k} = "" if ! exists $arg{$k}; $self->{$k} = $arg{$k} if exists $arg{$k}; } my $cs1 = $self->{conn_string}; my $tn1 = $self->{table_name}; $self->{cs} = ($cs1)?$cs1:$self->{cs}; $self->{tn} = ($tn1)?$tn1:$self->{tn}; $self->{conn_string} = ($self->{cs})?$self->{cs}:$cs1; $self->{table_name} = ($self->{tn})?$self->{tn}:$tn1; return $self; } =head1 METHODS The following are the common methods, routines, and functions defined in this class. =head2 Exported Tag: All The I<:all> tag includes all the methods or sub-rountines defined in this class. use Oracle::Schema qw(:all); It includes the following sub-routines: =head2 Table Methods The I<:table> tag includes sub-rountines for creating, checking and manipulating tables. use Oracle::DML::Common qw(:table); It includes the following sub-routines: =head3 get_table_definition($dbh,$tn,$cns,$otp) Input variables: $dbh - database handler, required. $tn - table/object name, required. schema.table_name is allowed. $cns - column names separated by comma. Default is null, i.e., to get all the columns. If specified, only get definition for those specified. $otp - output array type: AR|ARRAY - returns ($cns,$df1,$cmt) AH1|ARRAY_HASH1 - returns ($cns,$df2,$cmt) HH|HASH - returns ($cns,$df3,$cmt) AH2|ARRAY_HASH2 - returns ($cns,$df4,$cmt) Variables used or routines called: echoMSG - display messages. How to use: ($cns,$df1,$cmt) = $self->getTableDef($dbh,$table_name,'','array'); ($cns,$df2,$cmt) = $self->getTableDef($dbh,$table_name,'','ah1'); ($cns,$df3,$cmt) = $self->getTableDef($dbh,$table_name,'','hash'); ($cns,$df4,$cmt) = $self->getTableDef($dbh,$table_name,'','ah2'); Return: $cns - a list of column names separated by comma. $df1 - column definiton array ref in [$seq][$cnn]. where $seq is column sequence number, $cnn is array index number corresponding to column names: 0 - cname, 1 - coltype, 2 - width, 3 - scale, 4 - precision, 5 - nulls, 6 - colno, 7 - character_set_name. $df2 - column definiton array ref in [$seq]{$itm}. where $seq is column number (colno) and $itm are: col - column name seq - column sequence number typ - column data type wid - column width max - max width min - min width dec - number of decimals req - requirement: null or not null dft - date format dsp - description or comments $df3 - {$cn}{$itm} when $otp = 'HASH' where $cn is column name in lower case and $itm are the same as the above $df4 - [$seq]{$itm} when $otp = 'AH2' where $seq is the column number, and $itm are: cname - column name (col) coltype - column data type (typ) width - column width (wid) scale - column scale (dec) precision - column precision (wid for N) nulls - null or not null (req) colno - column sequence number (seq) character_set_name - character set name $cmt - {$cn}: contains comments for each column =cut sub get_table_definition { my $self = shift; my($dbh, $tn, $cns, $otp) = @_; # Input variables: # $dbh - database handler # $tn - table name # $cns - column names # # 0. check inputs croak "ERR: could not find database handler.\n" if !$dbh; croak "ERR: no table or object name is specified.\n" if !$tn; $tn = uc($tn); $self->echoMSG(" - reading table $tn definition...", 1); $otp = 'ARRAY' if (! defined($otp)); $otp = uc $otp; if ($cns) { $cns =~ s/,\s*/','/g; $cns = "'$cns'"; } # # 1. retrieve column definitions my($q,$msg); if (index($tn,'.')>0) { # it is in schema.table format my ($sch,$tab) = ($tn =~ /([-\w]+)\.([-\w]+)/); $q = " SELECT column_name,data_type,data_length,"; $q .= "data_scale,data_precision,\n "; $q .= "nullable,column_id,character_set_name\n"; $msg = "$q"; $q .= " FROM dba_tab_columns\n"; $msg .= " FROM dba_tab_columns\n"; $q .= " WHERE owner = '$sch' AND table_name = '$tab'\n"; $msg .= " WHERE owner = '$sch' AND table_name = '$tab'\n"; } else { $q = " SELECT cname,coltype,width,scale,precision,nulls,"; $q .= "colno,character_set_name\n"; $msg = "$q"; $q .= " FROM col\n WHERE tname = '$tn'"; $msg .= " FROM col\n WHERE tname = '$tn'\n"; } if ($cns) { $q .= " AND cname in (" . uc($cns) . ")\n"; $msg .= " AND cname in (" . uc($cns) . ")\n"; } if (index($tn,'.')>0) { # it is in schema.table format $q .= "\n ORDER BY table_name,column_id"; $msg .= " ORDER BY table_name, column_id\n"; } else { $q .= "\n ORDER BY tname, colno"; $msg .= " ORDER BY tname, colno\n"; } $self->echoMSG(" $msg", 2); my $sth=$dbh->prepare($q) || croak "ERR: Stmt - $dbh->errstr"; $sth->execute() || croak "ERR: Stmt - $dbh->errstr"; my $arf = $sth->fetchall_arrayref; # = output $df1 # # 2. construct column name list my $r = ${$arf}[0][0]; for my $i (1..$#{$arf}) { $r .= ",${$arf}[$i][0]"; } $msg = $r; $msg =~ s/,/, /g; $self->echoMSG(" $msg", 5); # # 3. get column comments $q = "SELECT column_name, comments\n FROM user_col_comments"; $q .= "\n WHERE table_name = '$tn'"; $msg = "SELECT column_name, comments\nFROM user_col_comments"; $msg .= "\nWHERE table_name = '$tn'

"; $self->echoMSG(" $msg", 5); my $s2=$dbh->prepare($q) || croak "ERR: Stmt - $dbh->errstr"; $s2->execute() || croak "ERR: Stmt - $dbh->errstr"; my $brf = $s2->fetchall_arrayref; my (%cmt, $j, $k, $cn); for my $i (0..$#{$brf}) { $j = lc(${$brf}[$i][0]); # column name $cmt{$j} = ${$brf}[$i][1]; # comments } # # 4. construct output $df2($def) and $df3($df2) my $def = bless [], ref($self)||$self; # = output $df2 my $df2 = bless {}, ref($self)||$self; # = output $df3 for my $i (0..$#{$arf}) { $j = ${$arf}[$i][6]-1; # column seq number ${$def}[$j]{seq} = $j; # column seq number $cn = lc(${$arf}[$i][0]); # column name ${$def}[$j]{col} = uc($cn); # column name ${$def}[$j]{typ} = ${$arf}[$i][1]; # column type if (${$arf}[$i][4]) { # precision > 0 # it is NUMBER data type ${$def}[$j]{wid} = ${$arf}[$i][4]; # column width ${$def}[$j]{dec} = ${$arf}[$i][3]; # number decimal } else { # CHAR or VARCHAR2 ${$def}[$j]{wid} = ${$arf}[$i][2]; # column width ${$def}[$j]{dec} = "" # number decimal } ${$def}[$j]{max} = ${$def}[$j]{wid}; if (${$def}[$j]{typ} =~ /date/i) { # typ is DATE ${$def}[$j]{max} = 17; # set width to 17 ${$def}[$j]{wid} = 17; # set width to 17 ${$def}[$j]{dft} = 'YYYYMMDD.HH24MISS'; } else { ${$def}[$j]{dft} = ''; # set date format to null } if (${$arf}[$i][5] =~ /^(not null|N)/i) { ${$def}[$j]{req} = 'NOT NULL'; } else { ${$def}[$j]{req} = ''; } if (exists $cmt{$cn}) { ${$def}[$j]{dsp} = $cmt{$cn}; } else { ${$def}[$j]{dsp} = ''; } ${$def}[$j]{min} = 0; ${$df2}{$cn}{seq} = $j; ${$df2}{$cn}{col} = ${$def}[$j]{col}; ${$df2}{$cn}{typ} = ${$def}[$j]{typ}; ${$df2}{$cn}{dft} = ${$def}[$j]{dft}; ${$df2}{$cn}{wid} = ${$def}[$j]{wid}; ${$df2}{$cn}{dec} = ${$def}[$j]{dec}; ${$df2}{$cn}{max} = ${$def}[$j]{max}; ${$df2}{$cn}{min} = ${$def}[$j]{min}; ${$df2}{$cn}{req} = ${$def}[$j]{req}; ${$df2}{$cn}{dsp} = ${$def}[$j]{dsp}; } # # 5. construct output array $df4 my $df4 = bless [],ref($self)||$self; # = output $df4 for my $i (0..$#{$arf}) { $j = lc(${$arf}[$i][0]); # column name push @$df4, {cname=>$j, coltype=>${$arf}[$i][1], width=>${$arf}[$i][2], scale=>${$arf}[$i][3], precision=>${$arf}[$i][4], nulls=>${$arf}[$i][5], colno=>${$arf}[$i][6], character_set_name=>${$arf}[$i][7]}; } # # 6. output based on output type if ($otp =~ /^(AR|ARRAY)$/i) { return ($r, $arf, \%cmt); # output ($cns,$df1,$cmt) } elsif ($otp =~ /^(AH1|ARRAY_HASH1)$/i) { return ($r, $def, \%cmt); # output ($cns,$df2,$cmt) } elsif ($otp =~ /^(HH|HASH)$/i) { return ($r, $df2, \%cmt); # output ($cns,$df3,$cmt) } else { return ($r, $df4, \%cmt); # output ($cns,$df4,$cmt); } } 1; =head1 HISTORY =over 4 =item * Version 0.01 This version is to set the framework and move the get_table_definition from Oracle:;DML::Common. =item * Version 0.02 Added table tag for export. =cut =head1 SEE ALSO (some of docs that I check often) Data::Describe, Oracle::Loader, CGI::Getopt, File::Xcopy, perltoot(1), perlobj(1), perlbot(1), perlsub(1), perldata(1), perlsub(1), perlmod(1), perlmodlib(1), perlref(1), perlreftut(1). =head1 AUTHOR Copyright (c) 2005 Hanming Tu. All rights reserved. This package is free software and is provided "as is" without express or implied warranty. It may be used, redistributed and/or modified under the terms of the Perl Artistic License (see http://www.perl.com/perl/misc/Artistic.html) =cut