我不想在任何时候阅读整个文件，我知道这个问题有答案，我想要 o阅读第一行或最后一行。
我知道我的代码锁定了它正在读取的文件有两个原因：1）当我使用此代码运行我的小应用程序时，写入文件的应用程序会间歇性地崩溃，但是当我没有运行此代码时，它永远不会崩溃！ 2）有一些文章会告诉你File.ReadLines会锁定文件。
有一些类似的问题，但答案似乎涉及读取整个文件，这对大文件来说很慢，因此不是我想要做的。我要求大部分时间只阅读最后一行也是我所读到的独特之处。
我知道如何读取第一行（标题行）和最后一行（最新行）。我不希望在我的代码中的任何一点读取所有行，因为这个文件会变得很大并且读取整个文件会变慢。
我知道
line = File.ReadLines（fullFilename）.First（）。Replace（“\”“，”“）;
...与...相同
FileStream fs = new FileStream（@fullFilename，FileMode.Open，FileAccess.Read，FileShare.Read）;
我的问题是，如何重复读取文件的第一行和最后一行，这些行可能是由另一个应用程序写入而不以任何方式锁定它。我无法控制写入文件的应用程序。它是一个可以随时附加的数据日志。我这样听的原因是这个日志可以连续几天附加。我希望在我自己的c＃程序中看到此日志中的最新数据，而无需等待日志完成写入。
我的代码用于调用阅读/收听功能......
//Start Listening to the "data log"
private void btnDeconstructCSVFile_Click(object sender, EventArgs e)
{
MySandbox.CopyCSVDataFromLogFile copyCSVDataFromLogFile = new MySandbox.CopyCSVDataFromLogFile();
copyCSVDataFromLogFile.checkForLogData();
}
我的班级听力。目前它只是将数据添加到2个泛型列表中......
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using MySandbox.Classes;
using System.IO;
namespace MySandbox
{
public class CopyCSVDataFromLogFile
{
static private List<LogRowData> listMSDataRows = new List<LogRowData>();
static String fullFilename = string.Empty;
static LogRowData previousLineLogRowList = new LogRowData();
static LogRowData logRowList = new LogRowData();
static LogRowData logHeaderRowList = new LogRowData();
static Boolean checking = false;
public void checkForLogData()
{
//Initialise
string[] logHeaderArray = new string[] { };
string[] badDataRowsArray = new string[] { };
//Get the latest full filename (file with new data)
//Assumption: only 1 file is written to at a time in this directory.
String directory = "C:\\TestDir\\";
string pattern = "*.csv";
var dirInfo = new DirectoryInfo(directory);
var file = (from f in dirInfo.GetFiles(pattern) orderby f.LastWriteTime descending select f).First();
fullFilename = directory + file.ToString(); //This is the full filepath and name of the latest file in the directory!
if (logHeaderArray.Length == 0)
{
//Populate the Header Row
logHeaderRowList = getRow(fullFilename, true);
}
LogRowData tempLogRowList = new LogRowData();
if (!checking)
{
//Read the latest data in an asynchronous loop
callDataProcess();
}
}
private async void callDataProcess()
{
checking = true; //Begin checking
await checkForNewDataAndSaveIfFound();
}
private static Task checkForNewDataAndSaveIfFound()
{
return Task.Run(() => //Call the async "Task"
{
while (checking) //Loop (asynchronously)
{
LogRowData tempLogRowList = new LogRowData();
if (logHeaderRowList.ValueList.Count == 0)
{
//Populate the Header row
logHeaderRowList = getRow(fullFilename, true);
}
else
{
//Populate Data row
tempLogRowList = getRow(fullFilename, false);
if ((!Enumerable.SequenceEqual(tempLogRowList.ValueList, previousLineLogRowList.ValueList)) &&
(!Enumerable.SequenceEqual(tempLogRowList.ValueList, logHeaderRowList.ValueList)))
{
logRowList = getRow(fullFilename, false);
listMSDataRows.Add(logRowList);
previousLineLogRowList = logRowList;
}
}
//System.Threading.Thread.Sleep(10); //Wait for next row.
}
});
}
private static LogRowData getRow(string fullFilename, bool isHeader)
{
string line;
string[] logDataArray = new string[] { };
LogRowData logRowListResult = new LogRowData();
try
{
if (isHeader)
{
//Asign first (header) row data.
//Works but seems to block writting to the file!!!!!!!!!!!!!!!!!!!!!!!!!!!
line = File.ReadLines(fullFilename).First().Replace("\"", "");
}
else
{
//Assign data as last row (default behaviour).
line = File.ReadLines(fullFilename).Last().Replace("\"", "");
}
logDataArray = line.Split(',');
//Copy Array to Generics List and remove last value if it's empty.
for (int i = 0; i < logDataArray.Length; i++)
{
if (i < logDataArray.Length)
{
if (i < logDataArray.Length - 1)
{
//Value is not at the end, from observation, these always have a value (even if it's zero) and so we'll store the value.
logRowListResult.ValueList.Add(logDataArray[i]);
}
else
{
//This is the last value
if (logDataArray[i].Replace("\"", "").Trim().Length > 0)
{
//In this case, the last value is not empty, store it as normal.
logRowListResult.ValueList.Add(logDataArray[i]);
}
else { /*The last value is empty, e.g. "123,456,"; the final comma denotes another field but this field is empty so we will ignore it now. */ }
}
}
}
}
catch (Exception ex)
{
if (ex.Message == "Sequence contains no elements")
{ /*Empty file, no problem. The code will safely loop and then will pick up the header when it appears.*/ }
else
{
//TODO: catch this error properly
Int32 problemID = 10; //Unknown ERROR.
}
}
return logRowListResult;
}
}
}
答案 0 :(得分：0)
我在其他问题的组合中找到了答案。一个答案解释了如何从文件的末尾读取，我对其进行了调整，使其只读取文件末尾的1行。另一个解释了如何在不锁定的情况下读取整个文件（我不想读取整个文件但是非锁定部分很有用）。所以现在你可以读取文件的最后一行（如果它包含行尾字符）而不锁定它。对于行结束的另一端，只需用你的行尾字符字节替换我的10和13 ......
将以下方法添加到公共类CopyCSVDataFromLogFile
private static string Reverse(string str)
{
char[] arr = new char[str.Length];
for (int i = 0; i < str.Length; i++)
arr[i] = str[str.Length - 1 - i];
return new string(arr);
}
并替换此行...
line = File.ReadLines(fullFilename).Last().Replace("\"", "");
使用此代码块...
Int32 endOfLineCharacterCount = 0;
Int32 previousCharByte = 0;
Int32 currentCharByte = 0;
//Read the file, from the end, for 1 line, allowing other programmes to access it for read and write!
using (FileStream reader = new FileStream(fullFilename, FileMode.Open, FileAccess.Read, FileShare.ReadWrite, 0x1000, FileOptions.SequentialScan))
{
int i = 0;
StringBuilder lineBuffer = new StringBuilder();
int byteRead;
while ((-i < reader.Length) /*Belt and braces: if there were no end of line characters, reading beyond the file would give a catastrophic error here (to be avoided thus).*/
&& (endOfLineCharacterCount < 2)/*Exit Condition*/)
{
reader.Seek(--i, SeekOrigin.End);
byteRead = reader.ReadByte();
currentCharByte = byteRead;
//Exit condition: the first 2 characters we read (reading backwards remember) were end of line ().
//So when we read the second end of line, we have read 1 whole line (the last line in the file)
//and we must exit now.
if (currentCharByte == 13 && previousCharByte == 10)
{
endOfLineCharacterCount++;
}
if (byteRead == 10 && lineBuffer.Length > 0)
{
line += Reverse(lineBuffer.ToString());
lineBuffer.Remove(0, lineBuffer.Length);
}
lineBuffer.Append((char)byteRead);
previousCharByte = byteRead;
}
reader.Close();
}