我正在加载很多xml文档，其中一些返回错误，例如“十六进制值0x12，是无效字符”，并且有不同的字符。如何去除它们？
我在这里做了一个小研究。
这是 ASCII 表。有128个符号 ![可查询][1] [1]：http://i.stack.imgur.com/7guQC.gif
这里有一些小测试代码，它添加 ASCII 表中的每个符号并尝试将其作为 XML 文档加载。
static public void RegexTry()
{
StreamReader stream = new StreamReader(@"test.xml");
string xmlfile = stream.ReadToEnd();
stream.Close();
string text = "";
for (int i = 0; i < 128; i++ )
{
char t = (char) i;
text = xmlfile.Replace('П', t);
XmlDocument xml = new XmlDocument();
try
{
xml.LoadXml(text);
}
catch (Exception ex)
{
Console.WriteLine("Char("+i.ToString() +"): " + t + " => error! " + ex.Message);
continue;
}
Console.WriteLine("Char(" + i.ToString() + "): " + t + " => fine!");
}
Console.ReadKey();
}
结果返回：
Char(0): => error! '.', hexadecimal value 0x00, is an invalid character. Line 5, position 7.
Char(1): => error! '', hexadecimal value 0x01, is an invalid character. Line 5, position 7.
Char(2): => error! '', hexadecimal value 0x02, is an invalid character. Line 5, position 7.
Char(3): => error! '', hexadecimal value 0x03, is an invalid character. Line 5, position 7.
Char(4): => error! '', hexadecimal value 0x04, is an invalid character. Line 5, position 7.
Char(5): => error! '', hexadecimal value 0x05, is an invalid character. Line 5, position 7.
Char(6): => error! '', hexadecimal value 0x06, is an invalid character. Line 5, position 7.
Char(7): => error! '', hexadecimal value 0x07, is an invalid character. Line 5, position 7.
Char(8): => error! '', hexadecimal value 0x08, is an invalid character. Line 5, position 7.
Char(9): => fine!
Char(10):
=> fine!
Char(11): => error! '', hexadecimal value 0x0B, is an invalid character. Line 5, position 7.
Char(12): => error! '', hexadecimal value 0x0C, is an invalid character. Line 5, position 7.
Char(13):
=> fine!
Char(14): => error! '', hexadecimal value 0x0E, is an invalid character. Line 5, position 7.
Char(15): => error! '', hexadecimal value 0x0F, is an invalid character. Line 5, position 7.
Char(16): => error! '', hexadecimal value 0x10, is an invalid character. Line 5, position 7.
Char(17): => error! '', hexadecimal value 0x11, is an invalid character. Line 5, position 7.
Char(18): => error! '', hexadecimal value 0x12, is an invalid character. Line 5, position 7.
Char(19): => error! '', hexadecimal value 0x13, is an invalid character. Line 5, position 7.
Char(20): => error! '', hexadecimal value 0x14, is an invalid character. Line 5, position 7.
Char(21): => error! '', hexadecimal value 0x15, is an invalid character. Line 5, position 7.
Char(22): => error! '', hexadecimal value 0x16, is an invalid character. Line 5, position 7.
Char(23): => error! '', hexadecimal value 0x17, is an invalid character. Line 5, position 7.
Char(24): => error! '', hexadecimal value 0x18, is an invalid character. Line 5, position 7.
Char(25): => error! '', hexadecimal value 0x19, is an invalid character. Line 5, position 7.
Char(26): => error! '', hexadecimal value 0x1A, is an invalid character. Line 5, position 7.
Char(27): => error! '', hexadecimal value 0x1B, is an invalid character. Line 5, position 7.
Char(28): => error! '', hexadecimal value 0x1C, is an invalid character. Line 5, position 7.
Char(29): => error! '', hexadecimal value 0x1D, is an invalid character. Line 5, position 7.
Char(30): => error! '', hexadecimal value 0x1E, is an invalid character. Line 5, position 7.
Char(31): => error! '', hexadecimal value 0x1F, is an invalid character. Line 5, position 7.
Char(32): => fine!
Char(33): ! => fine!
Char(34): " => fine!
Char(35): # => fine!
Char(36): $ => fine!
Char(37): % => fine!
Char(38): => error! An error occurred while parsing EntityName. Line 5, position 8.
Char(39): ' => fine!
Char(40): ( => fine!
Char(41): ) => fine!
Char(42): * => fine!
Char(43): + => fine!
Char(44): , => fine!
Char(45): - => fine!
Char(46): . => fine!
Char(47): / => fine!
Char(48): 0 => fine!
Char(49): 1 => fine!
Char(50): 2 => fine!
Char(51): 3 => fine!
Char(52): 4 => fine!
Char(53): 5 => fine!
Char(54): 6 => fine!
Char(55): 7 => fine!
Char(56): 8 => fine!
Char(57): 9 => fine!
Char(58): : => fine!
Char(59): ; => fine!
Char(60): => error! The '<' character, hexadecimal value 0x3C, cannot be included in a name. Line 5, position 13.
Char(61): = => fine!
Char(62): > => fine!
Char(63): ? => fine!
Char(64): @ => fine!
Char(65): A => fine!
Char(66): B => fine!
Char(67): C => fine!
Char(68): D => fine!
Char(69): E => fine!
Char(70): F => fine!
Char(71): G => fine!
Char(72): H => fine!
Char(73): I => fine!
Char(74): J => fine!
Char(75): K => fine!
Char(76): L => fine!
Char(77): M => fine!
Char(78): N => fine!
Char(79): O => fine!
Char(80): P => fine!
Char(81): Q => fine!
Char(82): R => fine!
Char(83): S => fine!
Char(84): T => fine!
Char(85): U => fine!
Char(86): V => fine!
Char(87): W => fine!
Char(88): X => fine!
Char(89): Y => fine!
Char(90): Z => fine!
Char(91): [ => fine!
Char(92): \ => fine!
Char(93): ] => fine!
Char(94): ^ => fine!
Char(95): _ => fine!
Char(96): ` => fine!
Char(97): a => fine!
Char(98): b => fine!
Char(99): c => fine!
Char(100): d => fine!
Char(101): e => fine!
Char(102): f => fine!
Char(103): g => fine!
Char(104): h => fine!
Char(105): i => fine!
Char(106): j => fine!
Char(107): k => fine!
Char(108): l => fine!
Char(109): m => fine!
Char(110): n => fine!
Char(111): o => fine!
Char(112): p => fine!
Char(113): q => fine!
Char(114): r => fine!
Char(115): s => fine!
Char(116): t => fine!
Char(117): u => fine!
Char(118): v => fine!
Char(119): w => fine!
Char(120): x => fine!
Char(121): y => fine!
Char(122): z => fine!
Char(123): { => fine!
Char(124): | => fine!
Char(125): } => fine!
Char(126): ~ => fine!
Char(127): => fine!
你可以看到有很多符号不能出现在XML代码中。要替换它们，我们可以使用
Regex.Replacestatic string ReplaceHexadecimalSymbols(string txt)
{
string r = "[\x00-\x08\x0B\x0C\x0E-\x1F\x26]";
return Regex.Replace(txt, r,"",RegexOptions.Compiled);
}
PS。抱歉，如果大家都知道的话。
XML 规范定义了如下的有效字符：
Char ::= #x9 | #xA | #xD | [#x20-#xD7FF] | [#xE000-#xFFFD] | [#x10000-#x10FFFF]
如您所见，
#x12您问如何删除它们，但我认为这不是您应该问的问题。他们根本不应该出现。您应该拒绝任何格式错误的此类文件。简单地删除无效字符就可以抑制真正的问题。
如果您正在创建有问题的文档，那么您需要修复生成该文档的代码，以便它生成有效的 XML。
我认为x26“&”是一个有效的字符，它可以通过XML反序列化。
所以要替换非法字符，我们应该使用：
// Replace illegal character in XML documents with blank
// See here for reference http://www.w3.org/TR/xml/#charsets
var regex = "[\x00-\x08\x0B\x0C\x0E-\x1F]";
xml = Regex.Replace(xml, regex, String.Empty, RegexOptions.Compiled);
正则表达式解决方案即使在 100MB XML 文档上也运行得相当快。
以下表达式字符串即可完成此工作。
"[\x00-\x08\x0B\x0C\x0E-\x1F]"
这本质上是这个问题的一个特例。我建议您使用那里的答案之一。
只需使用 jhon 提供的上述修复来更新这些函数，以及您必须更新的地方，检查代码中的这些函数。我已经测试过了，它对你有用。
private static void WriteDataTableToExcelWorksheet(DataTable dt, WorksheetPart worksheetPart)
{
var worksheet = worksheetPart.Worksheet;
var sheetData = worksheet.GetFirstChild<SheetData>();
string cellValue = "";
// Create a Header Row in our Excel file, containing one header for each Column of data in our DataTable.
//
// We'll also create an array, showing which type each column of data is (Text or Numeric), so when we come to write the actual
// cells of data, we'll know if to write Text values or Numeric cell values.
int numberOfColumns = dt.Columns.Count;
bool[] IsNumericColumn = new bool[numberOfColumns];
string[] excelColumnNames = new string[numberOfColumns];
for (int n = 0; n < numberOfColumns; n++)
excelColumnNames[n] = GetExcelColumnName(n);
//
// Create the Header row in our Excel Worksheet
//
uint rowIndex = 1;
var headerRow = new Row { RowIndex = rowIndex }; // add a row at the top of spreadsheet
sheetData.Append(headerRow);
for (int colInx = 0; colInx < numberOfColumns; colInx++)
{
DataColumn col = dt.Columns[colInx];
AppendTextCell(excelColumnNames[colInx] + "1", col.ColumnName, headerRow);
IsNumericColumn[colInx] = (col.DataType.FullName == "System.Decimal") || (col.DataType.FullName == "System.Int32");
}
//
// Now, step through each row of data in our DataTable...
//
double cellNumericValue = 0;
foreach (DataRow dr in dt.Rows)
{
// ...create a new row, and append a set of this row's data to it.
++rowIndex;
var newExcelRow = new Row { RowIndex = rowIndex }; // add a row at the top of spreadsheet
sheetData.Append(newExcelRow);
for (int colInx = 0; colInx < numberOfColumns; colInx++)
{
cellValue = dr.ItemArray[colInx].ToString();
// Create cell with data
if (IsNumericColumn[colInx])
{
// For numeric cells, make sure our input data IS a number, then write it out to the Excel file.
// If this numeric value is NULL, then don't write anything to the Excel file.
cellNumericValue = 0;
if (double.TryParse(cellValue, out cellNumericValue))
{
cellValue = ReplaceHexadecimalSymbols(cellNumericValue.ToString());
AppendNumericCell(excelColumnNames[colInx] + rowIndex.ToString(), cellValue, newExcelRow);
}
}
else
{
// For text cells, just write the input data straight out to the Excel file.
AppendTextCell(excelColumnNames[colInx] + rowIndex.ToString(), cellValue, newExcelRow);
}
}
}
}
static string ReplaceHexadecimalSymbols(string txt)
{
string r = "[\x00-\x08\x0B\x0C\x0E-\x1F\x26]";
return Regex.Replace(txt, r, "", RegexOptions.Compiled);
}
private static void AppendTextCell(string cellReference, string cellStringValue, Row excelRow)
{
// Add a new Excel Cell to our Row
Cell cell = new Cell() { CellReference = cellReference, DataType = CellValues.String };
CellValue cellValue = new CellValue();
cellValue.Text = ReplaceHexadecimalSymbols(cellStringValue);
cell.Append(cellValue);
excelRow.Append(cell);
}
private static void AppendNumericCell(string cellReference, string cellStringValue, Row excelRow)
{
// Add a new Excel Cell to our Row
Cell cell = new Cell() { CellReference = cellReference };
CellValue cellValue = new CellValue();
cellValue.Text = ReplaceHexadecimalSymbols(cellStringValue);
cell.Append(cellValue);
excelRow.Append(cell);
}
谢谢，如果您需要进一步帮助，请告诉我。
对我有用的是检查特殊字符。失败的特殊字符的字符代码类似于“&”，也用于“<'
”public static string CleanInvalidEscapedXmlCharacters(string s)
{
if(s == null ) return null;
StringBuilder sbOutput = new StringBuilder();
char ch;
//keeps track of which character the previous character was.
bool hitAmp = false;
bool hitPound = false;
bool hitX = false;
string escapedHold = "";
for( int i = 0; i < s.Length; i++ )
{
ch = s[i];
//check this first so that the x gets ignored.
if(hitX)
{
//found the end of the escaped portion
if(ch == ';')
{
ch = (char) Int32.Parse(escapedHold, NumberStyles.AllowHexSpecifier);
escapedHold = "";
hitX = false;
hitPound = false;
}
else
{
//found another digit in the escaped portion
escapedHold += ch;
continue;
}
}
if(hitPound)
{
if(ch == 'x')
{
//found &#x
hitX = true;
continue;
}
else
{
//found &# but no x
//reset hits and output &# and current character.
hitAmp = false;
hitPound = false;
sbOutput.Append('&');
sbOutput.Append('#');
sbOutput.Append(ch);
continue;
}
}
if(ch == '&')
{
//found an initial &
hitAmp = true;
continue;
}
if (hitAmp)
{
if (ch == '#')
{
//found &#
hitPound = true;
hitAmp = false;
continue;
}
else
{
//found & but no # so this is something like <
//reset hits and output the & and current character
hitAmp = false;
hitPound = false;
sbOutput.Append('&');
sbOutput.Append(ch);
continue;
}
}
if(!hitAmp && !hitPound && !hitX)
{
if ((ch >= 0x0020 && ch <= 0xD7FF) ||
(ch >= 0xE000 && ch <= 0xFFFD) ||
ch == 0x0009 || ch == 0x000A || ch == 0x000D)
{
sbOutput.Append(ch);
}
}
}
return sbOutput.ToString();
}