如何从DataTable1.Row [i] [j]读取字符串值到DataTable2的数组或多行中

Question

I am reading a csv file into a dataTable1 . One of the column has multiple values for each row. I want to store those multiple values in an array.

my csv file is something like this

my dataTable1 has ID and Admin values. 
    dataTable1.Rows[i][1] = 
    [{"@odata.type":"#Microsoft.Azure","Claims","DisplayName1","Department"}]

i want to loop through each Row in dataTable1 and split this string using "," delimiter and write to dataTable2. Something like

dataTable2.Rows[i][1] = "....Microsft..."
dataTable2.Rows[i][2] = "Claims"
dataTable2.Rows[i][3] = "DisplayName1"

. .

Answer 1

I couldn't really make sense of the example code you posted because c# arrays don't work like that but you either want to spread your csv column into multiple datatable columns:

DataTable dt = new DataTable();
dt.Columns.Add("something");
dt.Columns.Add("name");
dt.Columns.Add("displayname");
dt.Columns.Add("location");
dt.Columns.Add("ID");

string csv = @"someth1,{jon}{jonno}{newyork}{1}
Someth2,{jane}{janey}{paris}{2}";

foreach(string line in csv.Split()){
  string bits = line.Split(',');
  List<string> itm = new List<string>();
  itm.Add(bits[0]);
  itm.AddRange(bits[1].Split("{}".ToCharArray(), StringSplitOptions.RemoveEmptyEntries);
  dt.Rows.Add(itm.ToArray());
}

or you want to store it as an array in a single datatable columns:

DataTable dt = new DataTable();
dt.Columns.Add("something");
dt.Columns.Add("info", typeof(string[]));

string csv = @"someth1,{jon}{jonno}{newyork}{1}
someth2,{jane}{janey}{paris}{2}";

foreach(string line in csv.Split()){
  string bits = line.Split(',');
  string info = bits[1].Split("{}".ToCharArray(), StringSplitOptions.RemoveEmptyEntries);
  dt.Rows.Add(new object[] { bits[0], info });
}

This is not an exercise/example of model CSV parsing. I'm assuming you have a good library doing that already and the bits array hence represents the output from that library

Key here is how to deal with your curlybracket delimited list; there are plenty of options and rhino by no means the only one. It's relatively unsophisticated, but you haven't posted a massive example to know what kind of variety you're dealing with. It may need some upgrades if eg curly brackets can be escaped. You might also have some joy in telling your CSV parser that ,{} are all tokens that delimit fields

---

Edit, in response to your question revision:

DataTable dt1 = new DataTable();
dt1.Columns.Add("ID");
dt1.Columns.Add("CSV");
dt1.Rows.Add(new object[] { "1", "[{\"@odata.type\":\"#Microsoft.Azure\",\"Claims\",\"DisplayName1\",\"Department\"}]" } );

DataTable dt2 = new DataTable();
dt2.Columns.Add("ID");
dt2.Columns.Add("Provider");
dt2.Columns.Add("Department");
dt2.Columns.Add("Displayname");
dt2.Columns.Add("OtherDept");



foreach(DataRow ro in dt1.Rows){
  string csv = ro[1].ToString();
  csv = csv.Trim('[',']','{','}','"'); //remove leading and trailing chars that mess up our splitting
  string bits = csv.Split(new []{"},{"}); //split on a string delimiter of "},{". note: new[]{"abc","def"} is c# shorthand to declare an array of strings containing [0]="abc" and [1]="def"

  // our dt2 has 5 columns
  dt2.Rows.Add(new object[] {
       ro["ID"],
       bits[1],
       bits[2],
       bits[3],
       bits[4]
     }
  );

}