使用存在运算符避免 DBI 重复主键

Question

我想将 FASTA 文件中的信息插入 MySQL 数据库的表中。我使用Ensembl_id 列作为主键。

我的一些Ensembl_id 不是唯一的，所以我尝试使用exists 运算符来解决这个问题。但是表中只插入了 5 行，其中只有 1 行具有重复的 Ensembl_id 值。

#!/usr/bin/perl -w

#usage script.pl <username> <password> <database_name> <mouse_genes> <mouse_transcripts>

use DBI;
use Data::Dumper;

my $user              = shift @ARGV or die $!;
my $password          = shift @ARGV or die $!;
my $database          = shift @ARGV or die $!;
my $mouse_genes       = shift @ARGV or die $!;
my $mouse_transcripts = shift @ARGV or die $!;

my $dbh = DBI->connect( "dbi:mysql:$database:localhost", "$user", "$password",
    { RaiseError => 1 } );
my %gene;

$/ = "\n>";

open( FILE, "gzip -d -c /data.dash/class2016/student/Mus_musculus.GRCm38.cdna.all.fa.gz |" )
        or die $!;

LOOP:
while ( <FILE> ) {

    my $line = $_;
    chomp $line;

    if ( $line =~ /[a-z]/ ) {

        my @array = split( "\t", $line );

        if ( m/gene:(\w+\d+\.\w+)/ ) {

            my $Ensembl_id = $1;

            if ( !exists $gene{$Ensembl_id} ) {
                $gene{$Ensembl_id} = 1;
            }
            else {
                next;
            }

            if ( m/gene_biotype:(\w+)/ ) {

                my $gene_biotype = $1;

                if ( m/gene_symbol:(\w+\D\d+)/ ) {

                    my $gene_symbol = $1;

                    if ( m/description:(\w+\s+\w+\s+\w+\s+)/ ) {

                        my $gene_description = $1;

                        if ( m/MGI:(\d+)/ ) {

                            my $MGI_accession = $1;
                            my $sth           = $dbh->prepare(
                                qq{insert into $mouse_genes (Ensembl_id,gene_biotype,gene_symbol,gene_description,MGI_accession) values ("$Ensembl_id","$gene_biotype","$gene_symbol","$gene_description","$MGI_accession")}
                            );
                            $sth->execute();
                            $sth->finish();

                            next LOOP;
                        }
                    }
                }
            }
        }
    }
}

close FILE;

$dbh->disconnect();

如果主键$Ensembl_id 重复，我如何使用exists 运算符移动到文件的下一行？

Answer 1

我以为我看到了一个与此非常相似的问题，但我找不到它

解决方案是忘记散列并使用IGNORE 关键字来避免引发错误。 MySQL documentation 是这样说的：

如果您使用 IGNORE 关键字，则在执行 INSERT 语句时发生的错误将被忽略。例如，如果没有 IGNORE，复制表中现有 UNIQUE 索引或 PRIMARY KEY 值的行会导致重复键错误并且语句被中止。使用 IGNORE，该行将被丢弃并且不会发生错误。忽略的错误可能会生成警告，尽管重复键错误不会。

您还应该在 SQL 语句中使用占位符，所以它应该看起来像这样

请注意，END_SQL 必须在其前后不带和空格出现。您可能希望在程序顶部定义 SQL 语句以避免破坏缩进

my $sth = $dbh->prepare(<<END_SQL);
INSERT IGNORE INTO $mouse_genes (
    Ensembl_id,
    gene_biotype,
    gene_symbol,
    gene_description,
    MGI_accession
)
VALUES ( ?, ?, ?, ?, ? )
END_SQL

$sth->execute($Ensembl_id, $gene_biotype, $gene_symbol, $gene_description, $MGI_accession);

---

更新

您的程序可以进行大量整理以使其更易于阅读。下面是我的写法

#!/usr/bin/perl

use strict;
use warnings 'all';

# usage script.pl <username> <password> <database_name> <mouse_genes> <mouse_transcripts>

use DBI;

my $user              = shift @ARGV or die $!;
my $password          = shift @ARGV or die $!;
my $database          = shift @ARGV or die $!;
my $mouse_genes       = shift @ARGV or die $!;
my $mouse_transcripts = shift @ARGV or die $!; # Not used at present

my $dbh = DBI->connect( "dbi:mysql:$database:localhost", $user, $password,
        { RaiseError => 1, PrintError => 0 } );

my $sth = $dbh->prepare( <<END_SQL );
INSERT IGNORE INTO $mouse_genes (
    Ensembl_id,
    gene_biotype,
    gene_symbol,
    gene_description,
    MGI_accession
)
VALUES ( ?, ?, ?, ?, ? )
END_SQL

my $cmd = 'gzip -d -c /data.dash/class2016/student/Mus_musculus.GRCm38.cdna.all.fa.gz';

open my $cmd_fh, '-|', $cmd or die $!;

$/ = "\n>";

while ( <$cmd_fh> ) {

    next unless my ( $ensembl_id )       = /gene:(\w+\d+\.\w+)/;
    next unless my ( $gene_biotype )     = /gene_biotype:(\w+)/;
    next unless my ( $gene_symbol )      = /gene_symbol:(\w+\D\d+)/;
    next unless my ( $gene_description ) = /description:(\w+\s+\w+\s+\w+)\s/;
    next unless my ( $mgi_accession )    = /MGI:(\d+)/;

    $sth->execute( $ensembl_id, $gene_biotype, $gene_symbol, $gene_description, $mgi_accession );
}

$dbh->disconnect;

Answer 2

我想出了如何使用哈希来克服重复的键：

#!/usr/bin/perl -w

#this script inserts sequences from Mus_musculus.GRCm38.cdna.all.fa.gz into mouse_genes table
#usage lab5_2.pl <username> <password> <database_name> <mouse_genes> <mouse_transcripts>

use DBI;

use Data::Dumper;

my $user = shift @ARGV or die $!;
my $password = shift @ARGV or die $!;
my $database = shift @ARGV or die $!;
my $mouse_genes = shift @ARGV or die $!;

my $dbh = DBI->connect("dbi:mysql:$database:localhost",
                       "$user",
                       "$password",
                       {RaiseError => 1}
                       );
my %gene;

$/ = "\n>";

open (FILE, "gzip -d -c /data.dash/class2016/student/Mus_musculus.GRCm38.cdna.all.fa.gz |") or die $!;

 LOOP: while (<FILE>) {
     if (m/gene:(\w+\d+\.\d+)/) {
         my $Ensembl_id = $1;
         if ( !exists $gene{$Ensembl_id} ) {
             $gene{$Ensembl_id} = 1;
             if (m/gene_biotype:(\w+)/) {
                 my $gene_biotype = $1;
                 my $gene_symbol;
                 if (m/gene_symbol:(\w+\D\d+)/) {
                     $gene_symbol = $1;
                 }
                 if (! defined $gene_symbol) {
                     $gene_symbol = "NULL";
                 }
                 if (m/description:([^\[]*)/) {
                     my $gene_description = $1;
                     if (m/MGI:(\d+)/) {
                         my $MGI_accession = $1;
                         $sth = $dbh->prepare (qq{insert into mouse_genes (Ensembl_id, gene_biotype, gene_symbol, gene_description, MGI_accession) values ("$Ensembl_id","$gene_biotype","$gene_symbol","$gene_description","$MGI_accession")});
                         $sth->execute();
                         $sth->finish();
                         next LOOP;
                     }
                 }
             }
close FILE;
$dbh->disconnect ();