使用CsvHelper读取分块CSV，仅提取第一块数据的技术问题

我的CSV文件内容如下：

Metals:,E10
Al,0.1906
Ca,0.1132
Co,0.01951
Cu,0.5824
Cu,0.02383
Fe,0.03828
K,0.09577
Li,0.03024
Mg,0.007145
Na,0.1833
Ni,0.3236
Pb,0.0005787
Ti,0.4931
Tl,0.001887
Zn,0.07644

GLot,id,Slot,Scribe,Diameter,MPD,SResistivity,SThickness,TTV,LTV,Warp,Bow,S_U_A,Ep,Epi_L,Epi_Layer,Epi_Layer_2,EThick,E2thick,E2Dope,E2DopeT,E2DopeMax,E2DopeMin
31075046-001,XFB-LE00674.CP10023+001-12,1,22C1285,149.98,0,0.0217,334.71,1.91,1.03,5.35,-0.91,99.590582,1.0,1.0E18,9.8,1.12,9.9,9.6,9926193600000000,4.5574,10834500800000000,9551876800000000

该文件包含两个格式不同的数据块：第一块是金属元素与对应数值的两列结构，第二块是多表头多数据列的结构。

我编写的.NET CsvHelper代码如下：

namespace CsvHelperTest
{
    class CsvHelperTester
    {
        static void Main(string[] args)
        {
            var csvConfig = new CsvConfiguration(CultureInfo.InvariantCulture)
            {
                HasHeaderRecord = false,
                HeaderValidated = null,
                IgnoreBlankLines = true,
                MissingFieldFound = null,
                AllowComments = true,
                Comment = ';',
                Delimiter = ",",
                TrimOptions = TrimOptions.Trim, 
                PrepareHeaderForMatch = header => Regex.Replace(header.Header, ",", "\n"),
            };

            using (var streamReader = new StreamReader("C:\\Users\\eyoung\\Desktop\\parse test files\\XFB-1C2002A_62152_CoA.csv"))
            {
                using (var csvReader = new CsvReader(streamReader, csvConfig))
                {
                    for (var i = 0; i < 1; i++)
                    {
                        csvReader.Read();
                    }

                    var records = csvReader.GetRecords<EpiDataNames>().ToList();

                    var table = records[0];

                    records.RemoveAt(0);

                    var columns = records;

                    using (var writer = new CsvWriter(Console.Out, CultureInfo.InvariantCulture))
                    {
                        //writer.WriteField(records[0].Type);
                        //writer.NextRecord();

                        //records.RemoveAt(0);
                        //foreach (var item in records.Select(r => r.Type))
                        //{
                        //    writer.WriteField(item);
                        //}
                        //writer.NextRecord();
                        //foreach (var item in records.Select(r => r.Value))
                        //{
                        //    writer.WriteField(item);
                        //}
                        //writer.NextRecord(); 
                    }
                }
            }
        }

        public class EpiDataNames
        {
            [Index(0)]
            public string Type { get; set; }
            [Index(1)]
            public string Value { get; set; }
        }
    }
}

当前代码能将第一块数据转换为Type和Value两列，但无法只读取第一块数据。我尝试用以下代码跳过行：

for (var i = 0; i < 1; i++)
{
    csvReader.Read(); //this skips the first line of data
}

for (var i = 0; i > 18; i++)
{
    csvReader.Read(); //I thought this would skip the last lines of data, but it doesn't.
}

但这种方式无效，第二块的表头还被错误解析为如下结构：

我无法预先修改原CSV文件，求一个仅读取第一块数据的解决方案。

要只读取第一块数据，核心思路是逐行读取直到遇到空行（两个数据块的分隔符）或非两列结构的行，停止后续读取。具体实现如下：

namespace CsvHelperTest
{
    class CsvHelperTester
    {
        static void Main(string[] args)
        {
            var csvConfig = new CsvConfiguration(CultureInfo.InvariantCulture)
            {
                HasHeaderRecord = false,
                HeaderValidated = null,
                IgnoreBlankLines = false, // 关闭空行自动跳过，才能检测到数据块分隔空行
                MissingFieldFound = null,
                AllowComments = true,
                Comment = ';',
                Delimiter = ",",
                TrimOptions = TrimOptions.Trim,
            };

            var metalRecords = new List<EpiDataNames>();

            using (var streamReader = new StreamReader("C:\\Users\\eyoung\\Desktop\\parse test files\\XFB-1C2002A_62152_CoA.csv"))
            using (var csvReader = new CsvReader(streamReader, csvConfig))
            {
                // 跳过第一行"Metals:,E10"，如果需要保留该行可删除此句
                csvReader.Read();

                while (csvReader.Read())
                {
                    // 遇到空行，说明到达第一块末尾
                    if (string.IsNullOrWhiteSpace(csvReader.Context.Record[0]))
                    {
                        break;
                    }

                    // 仅读取两列结构的行，避免解析第二块的多列数据
                    if (csvReader.Context.Record.Length == 2)
                    {
                        var record = csvReader.GetRecord<EpiDataNames>();
                        metalRecords.Add(record);
                    }
                    else
                    {
                        // 遇到非两列行，终止读取
                        break;
                    }
                }
            }

            // 输出读取结果
            using (var writer = new CsvWriter(Console.Out, CultureInfo.InvariantCulture))
            {
                writer.WriteHeader<EpiDataNames>();
                writer.NextRecord();
                foreach (var record in metalRecords)
                {
                    writer.WriteRecord(record);
                    writer.NextRecord();
                }
            }
        }

        public class EpiDataNames
        {
            [Index(0)]
            public string Type { get; set; }
            [Index(1)]
            public string Value { get; set; }
        }
    }
}

关键修改说明 #

1. 关闭空行自动跳过：将IgnoreBlankLines设为false，这样才能检测到分隔两个数据块的空行，以此作为第一块结束的标志。
2. 逐行循环读取：不再使用GetRecords（会读取整个文件），而是逐行判断，遇到空行或非两列行就终止读取，精准控制读取范围。
3. 增加列数校验：即使空行被意外忽略，遇到第二块的多列行时也能立即停止，避免错误解析多列数据为两列结构。

内容的提问来源于stack exchange，提问作者Kevin Jablowski