/// <summary>
/// Parses the file.
/// </summary>
/// <returns>IEnumerable<dynamic>.</returns>
/// <exception cref="System.IO.FileNotFoundException"></exception>
public IEnumerable <dynamic> ParseFile()
{
int rowIndex = this.RowStart;
// TextFieldParser is in the Microsoft.VisualBasic.FileIO namespace.
using (TextFieldParser parser = new TextFieldParser(_fileName))
{
parser.SetFieldWidths(_fixedWidths.Maybe().FirstOrDefault());
parser.SetDelimiters(_delimiters.Maybe().DefaultIfEmpty(new string[] { ",", "\t" }).First());
parser.TextFieldType = parser.FieldWidths == null ? FieldType.Delimited : FieldType.FixedWidth;
parser.TrimWhiteSpace = TrimWhiteSpace;
parser.HasFieldsEnclosedInQuotes = HasQuotedFields;
parser.CommentTokens = new string[] { "#" };
string[] headers = null;
string[] values = null;
if (FirstRowContainsHeaders)
{
headers = FileIOUtilities.FixupHeaders(parser.ReadFields());
}
for (int i = 0; i < RowsToSkip; i++)
{
parser.ReadLine();
}
while (!parser.EndOfData)
{
values = parser.ReadFields();
if (headers == null)
{
headers = Enumerable.Range(0, values.Length).Select(x => "Col" + x).ToArray <string>();
}
//if all of the values are empty in the row, do not emit it. just continue on
if (values.Any(x => !String.IsNullOrWhiteSpace(x)))
{
dynamic retObj = FileIOUtilities.RowToExpando(headers, values);
retObj.RowId = rowIndex++;
yield return(retObj);
}
}
parser.Close();
}
}
/// <summary>
/// Parses the file.
/// </summary>
/// <returns>IEnumerable<dynamic>.</returns>
/// <exception cref="System.IO.FileNotFoundException"></exception>
public IEnumerable <dynamic> ParseFile()
{
var comparer = StringComparer.InvariantCultureIgnoreCase;
Package spreadsheetPackage = null;
var startColumnIndex = FileIOUtilities.ConvertExcelColumnNameToNumber(this.StartColumnKey);
var endColumnIndex = FileIOUtilities.ConvertExcelColumnNameToNumber(this.EndColumnKey);
try
{
spreadsheetPackage = Package.Open(_fileName, FileMode.Open, FileAccess.Read);
using (var spreadsheetDocument = SpreadsheetDocument.Open(spreadsheetPackage))
{
var workbookPart = spreadsheetDocument.WorkbookPart;
//find the sheet with the matching name
var sheet = spreadsheetDocument.WorkbookPart.Workbook.Descendants <Sheet>().FirstOrDefault(x => comparer.Equals(x.Name, this.SheetName));
if (sheet == null)
{
yield break;
}
//this is used by the reader to load the sheet for processing
var worksheetPart = workbookPart.GetPartById(sheet.Id) as WorksheetPart;
//used to get the rowcount of the sheet
// ReSharper disable once PossibleNullReferenceException
var sheetData = worksheetPart.Worksheet.Elements <SheetData>().First();
//check to ensure that we have any data rows at all, that have cell values
var hasDataRow = sheetData.Descendants <Row>().Any(row =>
row.RowIndex >= this.DataRowStart &&
row.Descendants <Cell>().Any(cell => cell.CellValue != null && !string.IsNullOrWhiteSpace(cell.CellValue.Text))
);
if (!hasDataRow)
{
yield break;
}
//needed to look up text values from cells
var sstpart = workbookPart.GetPartsOfType <SharedStringTablePart>().FirstOrDefault();
SharedStringTable sst = null;
if (sstpart != null)
{
sst = sstpart.SharedStringTable;
}
var cellFormats = workbookPart.WorkbookStylesPart.Stylesheet.CellFormats;
IList <NumberingFormat> numberingFormats = null;
if (workbookPart.WorkbookStylesPart.Stylesheet.NumberingFormats != null)
{
numberingFormats = workbookPart.WorkbookStylesPart.Stylesheet.NumberingFormats.OfType <NumberingFormat>().ToList();
}
//open the reader from the part
var reader = OpenXmlReader.Create(worksheetPart);
// dictionary for the headers, the key will end up being the cell address minus the row number so that the headers and values can match up by key
IDictionary <string, string> headers = new Dictionary <string, string>();
//the values dictionary for each row
Dictionary <string, string> values = null;
while (reader.Read())
{
//read until we find our rows, then loop through them
if (reader.ElementType == typeof(Row))
{
do
{
var row = (Row)reader.LoadCurrentElement();
uint rowIndex = row.RowIndex;
if (!(rowIndex == this.HeaderRow) && rowIndex < this.DataRowStart)
{
continue;
}
//if they have specified a end read row bail out if the rowindex exceeds that end value
if (this.DataRowEnd >= this.DataRowStart && rowIndex > this.DataRowEnd)
{
break;
}
if (row.HasChildren)
{
//loop through all of the cells in the row, building a list of either header keys, or value keys depending on which row it is.
values = new Dictionary <string, string>();
foreach (var cell in row.Descendants <Cell>())
{
var cellKey = FileIOUtilities.GetColumnKey(cell.CellReference.Value);
if (startColumnIndex != -1 || endColumnIndex != -1)
{
var cellIndex = FileIOUtilities.ConvertExcelColumnNameToNumber(cellKey);
if (startColumnIndex >= 0 && cellIndex < startColumnIndex)
{
continue;
}
if (endColumnIndex >= 0 && cellIndex > endColumnIndex)
{
break;
}
}
var value = String.Empty;
if (cell.DataType != null && cell.DataType == CellValues.SharedString && sst != null)
{
//read the text value out of the shared string table.
value = sst.ChildElements[int.Parse(cell.CellValue.Text)].InnerText;
}
else if (cell.CellValue != null && !String.IsNullOrWhiteSpace(cell.CellValue.Text))
{
if (cell.StyleIndex != null) //style index?
{
//frakking excel dates. wth. determing if a cell is formatted as a date is a huge pita
var cellFormat = (CellFormat)cellFormats.ElementAt((int)cell.StyleIndex.Value);
NumberingFormat numberingFormat = null;
if (numberingFormats != null && cellFormat.NumberFormatId != null && cellFormat.NumberFormatId.HasValue)
{
numberingFormat =
numberingFormats.FirstOrDefault(fmt => fmt.NumberFormatId.Value == cellFormat.NumberFormatId.Value);
}
if ((cell.DataType != null && cell.DataType == CellValues.Date) //just in case
||
(cellFormat.NumberFormatId != null &&
(cellFormat.NumberFormatId >= 14 && cellFormat.NumberFormatId <= 22)) //built in date formats
||
(numberingFormat != null &&
!numberingFormat.FormatCode.Value.Contains("[") && //so we dont match [Red] in numbering formats.... /sigh
Regex.IsMatch(numberingFormat.FormatCode, "d|h|m|s|y", RegexOptions.IgnoreCase))
//custom date formats, would an isdate be too hard msft?
) // Dates
{
value = Convert.ToString(DateTime.FromOADate(double.Parse(cell.CellValue.Text)),
CultureInfo.InvariantCulture);
}
else
{
value = cell.CellValue.Text;
}
}
else
{
value = cell.CellValue.Text;
}
}
if (rowIndex >= this.DataRowStart)
{
values.Add(cellKey, (value ?? "").Trim());
}
else if (rowIndex == this.HeaderRow)
{
headers.Add(cellKey, value);
}
}
}
//we have accumulated either our headers or values for this row, so now we need to handle them
if (rowIndex >= this.DataRowStart)
{
//sometimes excel reports the last row as higher than it actually has values, and we end up with an empty row.
//skip the row if this happens. otherwise we can output a weird object value
if (values.Any(x => !String.IsNullOrWhiteSpace(x.Value)))
{
dynamic retObj = FileIOUtilities.RowToExpando(headers, values, Convert.ToInt32(rowIndex));
retObj.RowId = Convert.ToInt32(rowIndex);
//stream the data row back to the caller
yield return(retObj);
}
}
else if (rowIndex == this.HeaderRow)
{
//remove all characters that are not allowed in .net property names
headers = FileIOUtilities.FixupHeaders(headers);
//string headersString = "\t[" + String.Join("] varchar(500),\r\n\t[", headers.Values) + "] varchar(500)";
//Debug.WriteLine(headersString);
}
} while (reader.ReadNextSibling());
//rows are all done, break out of the loop
break;
}
}
}
}
finally
{
spreadsheetPackage.Close();
}
}
public IDictionary <string, string> GetHeaders(bool removeInvalidChars = false)
{
var comparer = StringComparer.InvariantCultureIgnoreCase;
// dictionary for the headers, the key will end up being the cell address minus the row number so that the headers and values can match up by key
IDictionary <string, string> headers = new Dictionary <string, string>(comparer);
Package spreadsheetPackage = null;
var startColumnIndex = FileIOUtilities.ConvertExcelColumnNameToNumber(this.StartColumnKey);
var endColumnIndex = FileIOUtilities.ConvertExcelColumnNameToNumber(this.EndColumnKey);
try
{
spreadsheetPackage = Package.Open(_fileName, FileMode.Open, FileAccess.Read);
using (var spreadsheetDocument = SpreadsheetDocument.Open(spreadsheetPackage))
{
var workbookPart = spreadsheetDocument.WorkbookPart;
//find the sheet with the matching name
var sheet = spreadsheetDocument.WorkbookPart.Workbook.Descendants <Sheet>().FirstOrDefault(x => comparer.Equals(x.Name, this.SheetName));
if (sheet == null)
{
return(headers);
}
//this is used by the reader to load the sheet for processing
var worksheetPart = workbookPart.GetPartById(sheet.Id) as WorksheetPart;
//needed to look up text values from cells
var sstpart = workbookPart.GetPartsOfType <SharedStringTablePart>().FirstOrDefault();
SharedStringTable sst = null;
if (sstpart != null)
{
sst = sstpart.SharedStringTable;
}
//open the reader from the part
var reader = OpenXmlReader.Create(worksheetPart);
uint rowIndex = 0;
while (reader.Read())
{
//read until we find our rows, then loop through them
if (reader.ElementType != typeof(Row))
{
continue;
}
if (rowIndex > this.HeaderRow)
{
break;
}
do
{
var row = (Row)reader.LoadCurrentElement();
rowIndex = row.RowIndex;
if (rowIndex > this.HeaderRow)
{
break;
}
if (rowIndex != this.HeaderRow)
{
continue;
}
if (row.HasChildren)
{
foreach (var cell in row.Descendants <Cell>())
{
var cellKey = FileIOUtilities.GetColumnKey(cell.CellReference.Value);
if (startColumnIndex != -1 || endColumnIndex != -1)
{
int cellIndex = FileIOUtilities.ConvertExcelColumnNameToNumber(cellKey);
if (startColumnIndex >= 0 && cellIndex < startColumnIndex)
{
continue;
}
if (endColumnIndex >= 0 && cellIndex > endColumnIndex)
{
break;
}
}
string value;
if (cell.CellValue != null)
{
if (cell.DataType != null && cell.DataType == CellValues.SharedString && sst != null)
{
//read the text value out of the shared string table.
value = sst.ChildElements[int.Parse(cell.CellValue.Text)].InnerText;
}
else
{
value = cell.CellValue.Text;
}
if (rowIndex == this.HeaderRow)
{
headers.Add(cellKey, value);
}
}
}
}
if (rowIndex != this.HeaderRow)
{
continue;
}
if (removeInvalidChars)
{
//remove all characters that are not allowed in .net property names
headers = FileIOUtilities.FixupHeaders(headers);
}
} while (reader.ReadNextSibling());
//rows are all done, break out of the loop
break;
}
}
}
finally
{
if (spreadsheetPackage != null)
{
spreadsheetPackage.Close();
}
}
return(headers);
}