/// <summary>
/// Parses the named indexes from the header record.
/// </summary>
private void ParseHeaderRecord()
{
// We assume all columns contain headers by default
_columnCount = _sheet.Columns.Count;
// First make sure we have a header record
if (IsEmptyRecord())
{
throw new ExcelReaderException("No header record was found.");
}
// Process each column in the header row
for (var i = 0; i < _columnCount; i++)
{
// Get the header name
var name = _sheet[_row, i].Value as string;
if (string.IsNullOrWhiteSpace(name))
{
// Header is null or empty, so we are done. This can happen if the file has more total columns
// in it than header rows, which can happen if some white space ends up in a right column
// or there are extra rows below the records. Also check that the next column is
// also empty and throw and exception if not.
if (i + 1 < _columnCount && !string.IsNullOrWhiteSpace(_sheet[_row, i + 1].Value as string))
{
throw new Exception($"Blank column header found at column {ColumnHelper.ColumnIndexToColumnLetter(i + 1)}");
}
_columnCount = i;
break;
}
// Now store the named index for later for this header column
if (!_configuration.IsHeaderCaseSensitive)
{
name = name.ToLower();
}
if (_namedIndexes.ContainsKey(name))
{
_namedIndexes[name].Add(i);
}
else
{
_namedIndexes[name] = new List <int> {
i
};
}
}
// Move to the next row
_row++;
}
/// <summary>
/// Gets all the records in the Excel file and converts each to dictionary of strings to strings.
/// </summary>
/// <returns>An enumeration of dictionaries.</returns>
public IEnumerable <Dictionary <string, string> > GetRecordsAsDictionary()
{
// We assume all columns contain headers by default
_columnCount = _sheet.Columns.Count;
// First make sure we have a header record
ReadRow();
if (IsEmptyRecord())
{
throw new ExcelReaderException("No header record was found.");
}
// Process each column in the header row
var headers = new List <string>();
for (var i = 0; i < _columnCount; i++)
{
// Get the header name
var name = _sheet[_row, i].Value as string;
if (string.IsNullOrWhiteSpace(name))
{
// Header is null or empty, so we are done. This can happen if the file has more total columns
// in it than header rows, which can happen if some white space ends up in a right column
// or there are extra rows below the records. Also check that the next column is
// also empty and throw and exception if not.
if (i + 1 < _columnCount && !string.IsNullOrWhiteSpace(_sheet[_row, i + 1].Value as string))
{
throw new Exception($"Blank column header found at column {ColumnHelper.ColumnIndexToColumnLetter(i + 1)}");
}
_columnCount = i;
break;
}
// Now store the named index for later for this header column
if (!_configuration.IsHeaderCaseSensitive)
{
name = name.ToLower();
}
headers.Add(name.Trim());
}
// Move to the next row
_row++;
// Read each record one at a time and yield it
while (!IsEmptyRecord())
{
var record = new Dictionary <string, string>();
for (var i = 0; i < _columnCount; i++)
{
try {
var cell = _sheet[_row, i];
record.Add(headers[i], FormatValueAsString(cell, _row, i, true));
} catch (Exception ex) {
// Build the details about the error so it can be logged
var details = new ExcelReadErrorDetails {
Row = _row + 1,
Column = i + 1,
FieldName = headers[i],
FieldValue = _sheet[_row, i].Value,
};
// Use the inner exception if we have one so the message is more clear
ex = ex.InnerException ?? ex;
// Add the details to the exception
ExceptionHelper.AddExceptionDataMessage(ex, null, details);
// If we are ignoring errors, optionally call the callback and continue
if (_configuration.IgnoreReadingExceptions)
{
_configuration.ReadingExceptionCallback?.Invoke(ex, details);
_row++;
continue;
}
throw;
}
}
_row++;
yield return(record);
}
}
