Delphi ADO Query | 易学教程

问题

Is there any faster way to iterate through an ADO Dataset than

while (not ADOQuery1.Eof) do
    begin
      /* Do something */
      ADOQuery1.Next;
    end;

I need to scan a dataset of around 9000 items and only extract records matching a pre-defined set of branch numbers.

回答1:

It is much faster to use ADORecordset for such tasks:

 while not ADOQuery1.Recordset.EOF do
  begin
    ADOQuery1.Recordset.MoveNext;
    // get value
    SomeVar := ADOQuery1.Recordset.Fields['FieldName'].Value;  
  end;

回答2:

Be sure that you use DisableControls/EnableControls if it's not necesary for not spend time updating visible controls associated at DataSet.

try
  ADOQuery1.DisableControls;
  while (not ADOQuery1.Eof) do
    begin
      /* Do something */
      ADOQuery1.Next;
    end;
finally
  ADOQuery1.EnableControls;
end;

Regards.

回答3:

@Pieter, two options

1) you can modify your sql sentence before to execute, adding the where condition wich match with the pre-defined set of branch numbers.

2) using the Filter property of TAdoQuery.

AdoQuery1.close;
AdoQuery1.filter := 'your condition goes here';
AdoQuery1.filtered := true;
AdoQuery1.Open;

回答4:

Additional performance gains can be made by avoiding any string comparisons until as late as possible (when everything else matches). If you have a large number of duplicate strings in your database, then consider placing your strings in a separate table, linked back to the first table by an integer.

回答5:

You may want to change the query to include a SQL where clause, something like

Select whatever fields From whatevertable
where branchnumber in (
    select branchnumber from whatevertable where branchid=xxz
)

I would also highly suggest looking at forward-only, read-only cursors to give the biggest speed increase.

回答6:

Delphi ADO stuff (TADOQuery or TADOTable) is not bad, it is awful (checked with Delphi XE2 / 2007). Was exporting data from Transbase tables (ODBC driver) to MySQL through sql files and Navicat. For table with near million records it takes many hours through ADO (8 million records table was at 10% done after 2 days), several minutes by using TQuery (but can crash due to BDE bugs with big tables, BDE was not updated last 15 years), several minutes through pure ODBC or Navicat. My advise: use anything instead ADO (at least until seriously reworked by developers).

来源：https://stackoverflow.com/questions/2312444/delphi-ado-query

标签

Delphi

dataset

ado