问题
Hi I am trying to convert oracle jdbc resultset to csv file. Below is the code used. Issue occures when there is value like below in the field. It deforms the output csv and all this come in separate line rather than in one field.
Value in Field comes in csv as
[<333message:Runtime error in script' ProcessItem: ' Type: 'ITEM'" 1:0).Internal Script error: java.lang.NullPointerException
Script (line 1):
setHours = 0 ;
if(ts.instanceId == null)
" 3 : ts.instanceId = 0 ;"
Step >]
int ncols = result.getMetaData().getColumnCount();
System.out.println("ColumnCout"+ncols);
FileOutputStream fos=new FileOutputStream(new File("C:\\test.csv"),false);
Writer out = new OutputStreamWriter(new BufferedOutputStream(fos),"UTF_8");
for (int j=1; j<(ncols+1); j++) {
out.append(result.getMetaData().getColumnName (j));
if (j<ncols) out.append(","); else out.append("\r\n");
}
int m =1;
while (result.next()) {
for (int k=1; k<(ncols+1); k++) {
out.append(result.getString(k));
if (k<ncols) out.append(","); else out.append("\r\n");
}
//System.out.println("No of rows"+m);
m++;
}
回答1:
Get the value for the column that could have new lines as
String multiLine = null;
if (k == <col_index> && (mutiLine = rs.getString(k)) != null)
out.append(multiLine.replaceAll("\\n", ""));
else
out.append(result.getString(k));
You could filter all the columns as well but then would incur some performance hit.
回答2:
Are you using "java.sql.ResultSet" class? If yes, see the library in this link http://opencsv.sourceforge.net/
See an example:
CSVWriter csvWriter = new CSVWriter(new FileWriter("yourfile.csv"), '\t');
java.sql.ResultSet myResultSet = .... ;
csvWriter.writeAll(myResultSet, includeHeaders);
来源:https://stackoverflow.com/questions/18128802/convert-resultset-to-csv-file-using-java