C# csv 加载到数据表缺失的列中

Question

我已将 csv 加载到数据表中，但某些列在非空白时被读取为空白。我最初提出这个问题时只有标题有问题，但我现在在我的数据行中也看到了这个问题，所以我需要重新询问......我的数据集有什么问题以及为什么某些列被读取为空白？

目前，此设置将读取第 1-7 列的数据（我不需要 8-10）。为除第 4 列之外的所有列正确填充了数据。奇怪的是，我测试了两个文件，它们的结构都相似，但其中一个在第 4 列中有值，另一个没有。完整的代码被设置为循环遍历许多文件，检查数据的开始和结束，然后加载到 sql server。

CSV 示例：

By OrgID/Location

As of:   December 6, 2017 at 10:13 AM 

Date Range: summaryYM           2017M08 to 2017M08

"orgid=13778 medType=' '"
"col1","col2","col3","col4","col5","col6","col7","col8","col9","col10"
13778,140242,"2A","2017M08",0,0.058,78,".",".",
13778,140242,"2B","2017M08",0,0.014,19,".",".",
13778,140242,"2C","2017M08",0,0.083,133,".",".",
13778,140242,"2ICU","2017M08",0,0.099,114,".",".",
13778,140242,"3 ICU","2017M08",0,0.076,88,".",".",

代码

//open connection to csv
string connStrCsv = string.Format(@"Provider=Microsoft.Jet.OleDb.4.0; Data Source={0};Extended Properties=""Text;HDR=NO;FMT=Delimited"""
      , Path.GetDirectoryName(file));
OleDbConnection connCsv = new OleDbConnection(connStrCsv);
connCsv.Open();

//store csv data in datatable
string readCsv = "select * from [" + Path.GetFileName(file) + "]";
OleDbDataAdapter adapter = new OleDbDataAdapter(readCsv, connCsv);
DataSet ds = new DataSet();
adapter.Fill(ds, "sheet1");
DataTable table = ds.Tables["sheet1"];
connCsv.Close();

//find header to define start of data
int start = 0;
StreamReader headerSearch = null;
int incr = 0;
headerSearch = new StreamReader(file);
while (!headerSearch.EndOfStream)
{
  incr++;
  string line = headerSearch.ReadLine();
  if (line.Contains("\"col1\",\"col2\",\"col3\",\"col4\",\"col5\",\"col6\""))
  {
      start = incr;
  }
}
headerSearch.Close();

//load each row of excel into SQL server until first empty row
string sqlConnStr = "Data Source=mysource;Initial Catalog=mydatabase;Trusted_Connection=Yes;Integrated Security=SSPI;";
SqlConnection connSql = new SqlConnection(sqlConnStr);
connSql.Open();

int end = start;
while (table.Rows[end][0].ToString().Length != 0)
{
  string sql = string.Format
  (@"
  delete from schema.table
  where ss_col1      =  {0}    
      and ss_col2    = '{1}' 
      and ss_col3    = '{2}' 
      and ss_col4    = '{3}';

  insert into schema.table
  values ({4}        
      ,'{5}'         
      ,'{6}'         
      ,'{7}'         
      , {8}          
      ,'{9}'         
      ,'{10}'        
      ,getdate()
      ,user_name()
      ,getdate()
      ,user_name());"
      //delete statement variables
      , table.Rows[end][0].ToString()
      , table.Rows[end][2].ToString()
      , table.Rows[end][3].ToString()
      , infTypes[i]
      //insert statement variables
      , table.Rows[end][0].ToString()
      , table.Rows[end][2].ToString()
      , table.Rows[end][3].ToString()
      , infTypes[i]
      , table.Rows[end][4]
      , table.Rows[end][5].ToString()
      , table.Rows[end][6]
  );

  SqlCommand execSql = new SqlCommand(sql, connSql);
  execSql.ExecuteNonQuery();

  end++;
}
connSql.Close();

Answer 1

您必须将其加载到数据表中吗？ 如果 "header1","header2","header3","header4","header5","header6" 是唯一的，那么在找到之前读取 csv 文件会不会更容易？

例子……

StreamReader Reader = null;
string FilePath = "Your File Path";
try
{
  Reader = new StreamReader(FilePath);
  while(Reader.Peek() > 0)
  {
    string line = Reader.ReadLine();
    bool HeaderFound = false;
    if(line == "What ever your headers are")
    {
      HeaderFound = true;
    }
    if(HeaderFound)
    {
      //Here is all your data you were looking for.
     //Do whatever you need to do with it now.
    }
  }
} catch(exception e)
{/*Deal with the issues*/}
finally
{
  if(Reader != null)
  {
    Reader.Close();
    Reader.Dispose();
  }
}