我维护了一个 NPOI 的扩展(WeihanLi.Npoi),主要用来导入导出 Excel 数据,最近有网友提出了导入 Excel 的时候解析图片的需求,因而就有了本文的探索git
xls
和 xlsx
的 API 稍有不一样,详细能够直接参考如下代码,实现代码以下:github
public static Dictionary<CellPosition, IPictureData> GetPicturesAndPosition(this ISheet sheet) { var dictionary = new Dictionary<CellPosition, IPictureData>(); if (sheet.Workbook is HSSFWorkbook) { foreach (var shape in ((HSSFPatriarch)sheet.DrawingPatriarch).Children) { if (shape is HSSFPicture picture) { var position = new CellPosition(picture.ClientAnchor.Row1, picture.ClientAnchor.Col1); dictionary[position] = picture.PictureData; } } } else if (sheet.Workbook is XSSFWorkbook) { foreach (var shape in ((XSSFDrawing)sheet.DrawingPatriarch).GetShapes()) { if (shape is XSSFPicture picture) { var position = new CellPosition(picture.ClientAnchor.Row1, picture.ClientAnchor.Col1); dictionary[position] = picture.PictureData; } } } return dictionary; }
CellPosition
是一个自定义的结构体,表示当前单元格的位置,源码以下:c#
public readonly struct CellPosition : IEquatable<CellPosition> { public CellPosition(int row, int col) { Row = row; Column = col; } public int Row { get; } public int Column { get; } public bool Equals(CellPosition other) { return Row == other.Row && Column == other.Column; } public override bool Equals(object? obj) => obj is CellPosition other && Equals(other); public override int GetHashCode() => $"{Row}_{Column}".GetHashCode(); }
根据上面的代码,咱们就能够获取到获取到全部的图片以及图片的所在位置,这样根据单元格位置去找图片信息的时候就会很方便了数组
实现代码以下:async
public static bool TryAddPicture(this ISheet sheet, int row, int col, byte[] pictureBytes, PictureType pictureType = PictureType.PNG) { if (sheet is null) { throw new ArgumentNullException(nameof(sheet)); } try { var pictureIndex = sheet.Workbook.AddPicture(pictureBytes, pictureType); var clientAnchor = sheet.Workbook.GetCreationHelper().CreateClientAnchor(); clientAnchor.Row1 = row; clientAnchor.Col1 = col; var picture = (sheet.DrawingPatriarch ?? sheet.CreateDrawingPatriarch()) .CreatePicture(clientAnchor, pictureIndex); picture.Resize(); return true; } catch (Exception e) { Debug.WriteLine(e); } return false; }
经过上面的代码咱们就能够在指定的单元格设置图片,目前没有支持单元格合并操做,有须要本身进行修改ide
WeihanLi.Npoi
WeihanLi.Npoi
在 1.15.0
版本中增长了图片导入导出的支持,使用示例能够参考下面的单元测试:单元测试
[Theory] [ExcelFormatData] public async Task ImageImportExportTest(ExcelFormat excelFormat) { using var httpClient = new HttpClient(); var imageBytes = await httpClient.GetByteArrayAsync("https://weihanli.xyz/assets/avator.jpg"); var list = Enumerable.Range(1, 5) .Select(x => new ImageTest() { Id = x, Image = imageBytes }) .ToList(); var excelBytes = list.ToExcelBytes(excelFormat); var importResult = ExcelHelper.ToEntityList<ImageTest>(excelBytes, excelFormat); Assert.NotNull(importResult); Assert.Equal(list.Count, importResult.Count); for (var i = 0; i < list.Count; i++) { Assert.NotNull(importResult[i]); var result = importResult[i]!; Assert.Equal(list[i].Id, result.Id); Assert.NotNull(result.Image); Assert.True(list[i].Image.SequenceEqual(result.Image)); } } private class ImageTest { public int Id { get; set; } public byte[] Image { get; set; } = null!; }
导入时会自动将 byte[]
类型的属性尝试获取对应的单元格位置的图片,若是在对应的位置找到了图片就可以读取到图片的字节数组信息映射到 model 里的字节数组属性测试
除此以外,还支持直接导入的时候将图片信息 Map 到 IPictureData
属性上,可是以为仍是字节数组更通用一些,若是对此感兴趣能够参考项目源码以及单元测试this
感谢 @ZeguangZhang94 童鞋提出的需求和帮忙测试~excel
若是你也有相似的读取指定单元格的图片或者在指定单元格插入图片的需求,能够试一下上面的方法,但愿对你有帮助,能够直接引用个人类库或者直接拷贝源码到本身的项目里使用