/// <summary>
/// Guesses the not a delimited file.
/// </summary>
/// <param name="setting"><see cref="ICsvFile" /> with the information</param>
/// <returns><c>true</c> if this is most likely not a delimited file</returns>
public static bool GuessNotADelimitedFile(ICsvFile setting)
{
Contract.Requires(setting != null);
using (var improvedStream = ImprovedStream.OpenRead(setting))
using (var streamReader = new StreamReader(improvedStream.Stream, setting.GetEncoding(), setting.ByteOrderMark))
{
for (int i = 0; i < setting.SkipRows; i++)
{
streamReader.ReadLine();
}
// If the file doe not have a good delimiter
// has empty lines
var dc = GetDelimiterCounter(streamReader, '\0', 300);
// Have a proper delimiter
for (var sep = 0; sep < dc.Separators.Length; sep++)
{
if (dc.SeparatorRows[sep] >= dc.LastRow * 9 / 10)
{
Log.Info("Not a delimited file");
return(false);
}
}
}
Log.Info("Delimited file");
return(true);
}
/// <summary>
/// Determines the start row in the file
/// </summary>
/// <param name="setting"><see cref="ICsvFile" /> with the information</param>
/// <returns>
/// The number of rows to skip
/// </returns>
public static int GuessStartRow(ICsvFile setting)
{
Contract.Requires(setting != null);
using (var improvedStream = ImprovedStream.OpenRead(setting))
using (var streamReader = new StreamReader(improvedStream.Stream, setting.GetEncoding(), setting.ByteOrderMark))
{
return(GuessStartRow(streamReader, setting.FileFormat.FieldDelimiterChar,
setting.FileFormat.FieldQualifierChar));
}
}
/// <summary>
/// Try to guess the new line sequence
/// </summary>
/// <param name="setting"><see cref="ICsvFile" /> with the information</param>
/// <returns>The NewLine Combination used</returns>
public static string GuessNewline(ICsvFile setting)
{
Contract.Requires(setting != null);
using (var improvedStream = ImprovedStream.OpenRead(setting))
using (var streamReader = new StreamReader(improvedStream.Stream, setting.GetEncoding(), setting.ByteOrderMark))
{
for (int i = 0; i < setting.SkipRows; i++)
{
streamReader.ReadLine();
}
return(GuessNewline(streamReader, setting.FileFormat.FieldQualifierChar));
}
}
/// <summary>
/// Guesses the delimiter for a files. Done with a rather simple csv parsing, and trying to find
/// the delimiter that has the least variance in the read rows, if that is not possible the
/// delimiter with the highest number of occurrences.
/// </summary>
/// <param name="setting">The CSVFile fileSetting</param>
/// <returns>
/// A character with the assumed delimiter for the file
/// </returns>
/// <remarks>
/// No Error will not be thrown.
/// </remarks>
public static string GuessDelimiter(ICsvFile setting)
{
Contract.Requires(setting != null);
Contract.Ensures(Contract.Result <string>() != null);
using (var improvedStream = ImprovedStream.OpenRead(setting))
using (var streamReader = new StreamReader(improvedStream.Stream, setting.GetEncoding(), setting.ByteOrderMark))
{
for (int i = 0; i < setting.SkipRows; i++)
{
streamReader.ReadLine();
}
return(GuessDelimiter(streamReader, setting.FileFormat.EscapeCharacterChar));
}
}
private void UpdateView()
{
m_DisplayedAt = ScrollBarVertical.Value;
if (string.IsNullOrEmpty(m_CsvFile.FileName))
{
return;
}
try
{
using (var procDisp = new ProcessDisplayTime(System.Threading.CancellationToken.None))
using (var istream = ImprovedStream.OpenRead(m_CsvFile))
using (var sr = new StreamReader(istream.Stream, m_CsvFile.GetEncoding(), m_CsvFile.ByteOrderMark))
{
// Some stream do not support seek...
if (istream.Stream.CanSeek)
{
istream.Stream.Seek(m_DisplayedAt, SeekOrigin.Begin);
if (m_DisplayedAt != 0)
{
// find the line start
var read = sr.Read();
while (read != 13 && read != 10 && !sr.EndOfStream)
{
read = sr.Read();
}
var next = sr.Peek();
if (read == 13 && next == 10 || read == 10 && next == 13)
{
sr.Read();
}
}
}
else
{
ScrollBarVertical.Enabled = false;
}
var buffer = new char[32000];
var len = sr.Read(buffer, 0, buffer.Length);
CSVTextBox.Text = new string(buffer, 0, len);
}
}
catch (Exception exc)
{
CSVTextBox.Text = exc.ExceptionMessages();
}
}
/// <summary>
/// Resets the position and buffer to the first line, excluding headers, use ResetPositionToStart if you want to go to
/// first data line
/// </summary>
private void ResetPositionToStart()
{
if (m_ImprovedStream == null)
{
m_ImprovedStream = ImprovedStream.OpenRead(m_CsvFile);
}
m_ImprovedStream.ResetToStart(delegate(Stream str)
{
// in case we can not seek need to reopen the stream reader
if (!str.CanSeek || m_TextReader == null)
{
if (m_TextReader != null)
{
m_TextReader.Dispose();
}
m_TextReader = new StreamReader(str, m_CsvFile.GetEncoding(), m_CsvFile.ByteOrderMark);
}
else
{
// only need to discard the buffer
m_TextReader.DiscardBufferedData();
}
});
m_CsvFile.CurrentEncoding = m_TextReader.CurrentEncoding;
m_BufferPos = 0;
m_BufferFilled = 0;
// End Line should be at 1, later on as the line is read the start line s set to this value
EndLineNumber = 1;
RecordNumber = 0;
m_EndOfLine = false;
EndOfFile = false;
// Skip the given number of lines
// <= so we do skip the right number
while (EndLineNumber <= m_CsvFile.SkipRows && !EndOfFile && !CancellationToken.IsCancellationRequested)
{
ReadToEOL();
}
}
/// <summary>
/// Guesses the code page ID of a file
/// </summary>
/// <param name="setting">The CSVFile fileSetting</param>
/// <remarks>
/// No Error will be thrown, the CodePage and the BOM will bet set
/// </remarks>
public static void GuessCodePage(ICsvFile setting)
{
Contract.Requires(setting != null);
// Read 256 kBytes
var buff = new byte[262144];
int length;
using (var fileStream = ImprovedStream.OpenRead(setting))
{
length = fileStream.Stream.Read(buff, 0, buff.Length);
}
if (length >= 2)
{
var byBom = EncodingHelper.GetCodePageByByteOrderMark(buff);
if (byBom != 0)
{
setting.ByteOrderMark = true;
setting.CodePageId = byBom;
return;
}
}
setting.ByteOrderMark = false;
var detected = EncodingHelper.GuessCodePageNoBom(buff, length);
// ASCII will be reported as UTF-8, UTF8 includes ASCII as subset
if (detected == 20127)
{
detected = 65001;
}
Log.Info("Detected Code Page: " + EncodingHelper.GetEncodingName(detected, true, setting.ByteOrderMark));
setting.CodePageId = detected;
}
/// <summary>
/// Does check if quoting was actually used in the file
/// </summary>
/// <param name="setting">The setting.</param>
/// <param name="token">The token.</param>
/// <returns>
/// <c>true</c> if [has used qualifier] [the specified setting]; otherwise, <c>false</c>.
/// </returns>
public static bool HasUsedQualifier(ICsvFile setting, CancellationToken token)
{
Contract.Requires(setting != null);
// if we do not have a quote defined it does not matter
if (string.IsNullOrEmpty(setting.FileFormat.FieldQualifier) || token.IsCancellationRequested)
{
return(false);
}
using (var improvedStream = ImprovedStream.OpenRead(setting))
using (var streamReader = new StreamReader(improvedStream.Stream, setting.GetEncoding(), setting.ByteOrderMark))
{
for (int i = 0; i < setting.SkipRows; i++)
{
streamReader.ReadLine();
}
var buff = new char[262144];
var isStartOfColumn = true;
while (!streamReader.EndOfStream)
{
var read = streamReader.ReadBlock(buff, 0, 262143);
// Look for Delimiter [Whitespace] Qualifier or StartofLine [Whitespace] Qualifier
for (var current = 0; current < read; current++)
{
if (token.IsCancellationRequested)
{
return(false);
}
var c = buff[current];
if (c == '\r' || c == '\n' || c == setting.FileFormat.FieldDelimiterChar)
{
isStartOfColumn = true;
continue;
}
// if we are not at the start of a column we can get the next char
if (!isStartOfColumn)
{
continue;
}
// If we are at the start of a column and this is a ", we can stop, this is a real qualifier
if (c == setting.FileFormat.FieldQualifierChar)
{
return(true);
}
// Any non whitespace will reset isStartOfColumn
if (c <= '\x00ff')
{
isStartOfColumn = c == ' ' || c == '\t';
}
else
{
isStartOfColumn = CharUnicodeInfo.GetUnicodeCategory(c) == UnicodeCategory.SpaceSeparator;
}
}
}
}
return(false);
}