加速perlDBIfetchrow_hashref-SpeedingupperlDBIfetchrow_hashref

8  

You can use the fetchall_arrayref function which accepts a 'maxrows' argument:

您可以使用接受'maxrows'参数的fetchall_arrayref函数：

while (my $data = $dbc->fetchall_arrayref(undef, 10000)) {
  for my $row( @{$data} ) {
    $report->process_record($row);
  }
}

You could also look at the RowCacheSize property which attempts to control how many records are returned in a fetch from your driver.

您还可以查看RowCacheSize属性，该属性尝试控制从驱动程序获取中返回的记录数。

4  

Which bit is slow? Is it the call to execute, fetchrow_hashref or process_record? It seems unlikely to me that fetchrow_hashref is the problem. It's far more likely to be the execution of the query or the black-box of process_record.

哪个位慢？是执行调用，fetchrow_hashref还是process_record？我认为fetchrow_hashref不是问题所在。它更有可能是执行查询或process_record的黑盒子。

But this all guesswork. It's impossible to really help here. I recommend you get some real data about the performance of the code by using Devel::NYTProf.

但这一切都在猜测。这里真的不可能有所帮助。我建议您使用Devel :: NYTProf获取有关代码性能的一些真实数据。

3  

The fastest way to fetch rows as hashes using the DBI is to use bind_columns() like this:

使用DBI将行作为哈希获取的最快方法是使用bind_columns（），如下所示：

  $sth->execute;
  my %row;
  $sth->bind_columns( \( @row{ @{$sth->{NAME_lc} } } ));
  while ($sth->fetch) {
      print "$row{region}: $row{sales}\n";
  }

That's only appropriate if you're happy for each row to reuse the same hash.

如果您对每一行重用相同的哈希感到满意，那么这是唯一合适的。

Beyond that, I agree with davorg, avoid guesswork: measure first.

除此之外，我同意davorg，避免猜测：先测量。

For much more information on using the DBI, including performance, see my tutorial slides (from 2007, but still relevant).

有关使用DBI的更多信息（包括性能），请参阅我的教程幻灯片（从2007年开始，但仍然相关）。