使用NPOI和线程池快速加载EXCEL数据

Posted 2020-07-02 MessageBox

  private void FilterData()
            {
                  List<Task> tasks = new List<Task>();
                  IWorkbook workbook = Cshap.Common.ExcelUnit.LoadWorkbook(excelfile_);
                  for (int sheetIndex = 0; sheetIndex < workbook.NumberOfSheets; sheetIndex++)
                  {
                        rowCount_ += workbook.GetSheetAt(sheetIndex).PhysicalNumberOfRows;
                  }
                  for (int sheetIndex = 0; sheetIndex < workbook.NumberOfSheets; sheetIndex++)
                  {
                        ISheet sheet = workbook.GetSheetAt(sheetIndex);
                        tasks.Add(Task.Factory.StartNew(FilterSheet, sheet));
                  }
                  Task.WaitAll(tasks.ToArray());
                  mre.WaitOne();
            }

 
            void FilterSheet(object state)
            {
                  ISheet sheet = state as ISheet;
                  for (int rowindex = 0; rowindex < sheet.PhysicalNumberOfRows; rowindex++)
                  {
                        ThreadPool.QueueUserWorkItem(new WaitCallback(FilterRow), sheet.GetRow(rowindex));
                  }
            }

            void FilterRow(object state)
            {
                  IRow row = state as IRow;

                  var rowEntity = new RowEntity
                  {
                        SheetName = row.Sheet.SheetName,
                        RowData = row, //取行的单元格数据时，不可通过RowData.Cells[#]获取，因为当cell为null，列数会不一致
                        IsBlank = false,
                        Cells=new List<ICell>()//必须通过这里获取单元格数据
                  };

                  for (int i = 0; i < row.Sheet.GetRow(0).Cells.Count; i++)//标题列单元格一定不为null
                  {
                        ICell c = row.GetCell(i);
                        rowEntity.Cells.Add(c);//无论是否为null，保证列数一致
                  }

                  if (row.Cells[0].CellType == CellType.String && row.Cells[0].StringCellValue != null
                        && row.Cells[0].StringCellValue.ToUpper() == "123456")
                  {
                        lock (lockObject)
                        {
                              string pattern = @"(\s*\d{1,2}/\d{1,2}/\d{4}\s*)-(\s*\d{1,2}/\d{1,2}/\d{4}\s*)";
                              
                              ICell cellOfG = rowEntity.Cells[6];

                              if (cellOfG!=null && (cellOfG.CellType != CellType.Blank && cellOfG.CellType != CellType.String
                                    || !string.IsNullOrEmpty(cellOfG.StringCellValue) && !Regex.IsMatch(cellOfG.StringCellValue, pattern)))
                              {
                                    rowEntity.Message = "Date Format Exception";
                                    InvalidRows.Add(rowEntity);
                              }
                              else
                              {
                                    ValidRows.Add(rowEntity);
                              }
                        }
                  }

                  Interlocked.Increment(ref increment_);

                  if (increment_ == rowCount_)
                        mre.Set();
            }