C＃中奇怪的动作循环

Question

注意：我添加了实际的代码段。 滚动到结尾。

// files is created by a OpenFileDialog.
public void Function(String[] files, ...)
{
    for(int i; i<files.Length; i++)
    {
        WriteLine("File " + i + "/" + files.Length + " being processed.");
        //... processing for a long time and printing information to console ...
    }

    //... print results, e.g.: "Results: bla bla"...
}

函数在另一个循环中调用。 我运行了几次代码并认为它​​运行良好，直到我看到它有一次表现得很奇怪。 我提供了上面的函数，其数组长度为6，预期输出如下：

-------------------------
File 0/6 being processed.
...lots of output...
File 1/6 being processed.
...lots of output...
File 2/6 being processed.
...lots of output...
File 3/6 being processed.
...lots of output...
File 4/6 being processed.
...lots of output...
File 5/6 being processed.
...lots of output...

Results: bla bla...
-------------------------

但是，我得到的输出是这样的：

-------------------------
File 0/1 being processed.
...lots of output...

Results: bla bla...

File 0/3 being processed.
...lots of output...
File 1/3 being processed.
...lots of output...
File 2/3 being processed.
...lots of output...

Results: bla bla...

File 0/3 being processed.
...lots of output...
File 1/3 being processed.
...lots of output...
File 2/3 being processed.
...lots of output...

Results: bla bla...

File 0/6 being processed.
...lots of output...
File 1/6 being processed.
...lots of output...
File 2/6 being processed.
...lots of output...
File 3/6 being processed.
...lots of output...
-------------------------

当我看到那个输出时，我在当前循环结束之前退出执行（它运行了很长时间。）

它看起来像函数正常工作（它运行files.Length次并在此之后输出结果。）但是，传递给函数的参数在某种程度上是错误的（该函数有趣地被称为不止一次。通常，它应该只运行在这种情况下一次。我的意思是，脚本文件中的行数决定了调用函数以上的次数，而脚本文件只包含一行。）该参数（文件数组）来自OpenFileDialog，这意味着我跟它无关。 我只是将数组传递给函数。

我仍然试图理解这种奇怪结果的原因。 这只发生过一次，但我仍然需要诊断问题; 因为，我会让程序运行几天。 它应该正常工作。

你对这个废话有什么看法吗？

---

上述功能的实际代码：

public String Start(String[] files, StreamWriter reportWriter)
{
    List<SortedDictionary<int, SortedDictionary<long, int>>>[] allResults
        = new List<SortedDictionary<int,SortedDictionary<long,int>>>[files.Length];
    List<SortedDictionary<int, SortedDictionary<long, int>>> results;
    Simulation_DenemePositionEstimator p;
    Simulation_WimaxStreamReader reader;
    String ret;

    for (int i = 0; i < files.Length; i++)
    {
        System.Console.WriteLine("File " + (i+1) + "/" + files.Length + " being processed.");
        reader = new Simulation_WimaxStreamReader(grids, new StreamReader(files[i]));
        p = new Simulation_DenemePositionEstimator(grids, reader);
        // Using parameters in script file which were saved into
        // different variables when Simulation instance was created.
        results = 
            p.StartInvestigation(maxRssiDiff, maxCinrDiff, maxAvgTxPwrDiff, 
                maxUncontinuity, radiusForNeighbors, expansionFactor, increment,
                n, numberOfIterations, resetCountForPositioning);
        allResults[i] = results;
        reader.Close();
    }

    ret = Statistics(allResults);
    System.Console.WriteLine(ret);
    reportWriter.WriteLine(ret);
    reportWriter.Flush();

    return ret;
}

来电功能代码：

    // read a line from script file.
    while((line = reader.ReadLine()) != null)
    {
        // line starting with # is comment.
        if (line.StartsWith("#") == false)
        {
            // save parameters retrieved from script file into an array.
            values = line.Split(delimiters);
            // new Simulation instance with new parameters
            sim = new Simulation(map, values);
            // Start simulation. scenarioFiles comes from OpenFileDialog.
            report = sim.Start(scenarioFiles, reportWriter);
            //reportWriter.WriteLine(report);
            reportWriter.WriteLine("---------------NEW-PARAMETERS---------------");
            reportWriter.Flush();
        }
    }

脚本文件：

# Horizontal grid count
# Vertical grid count
# maxRssiDiff is the maximum RSSI difference allowed.
# maxCinrDiff is the maximum CINR difference allowed.
# maxAvgTxPwrDiff is the maximum AvgTxPwr difference allowed.
# maxUncontinuity
# radiusForNeighbors
# expansionFactor
# increment
# n -> MT'den gelen kaç değerin ortalaması alınıp yer bulma algoritmasına girdi olarak verilsin?
# Algoritma kaç adımda bir sonuçları dosyaya yazsın?
# Kaç adımdan sonra yer bulma işlemine sıfırdan başlamış gibi devam etsin?
# 
# Örnek:
# 118   90  4   3   4   2   1   1   1   3   10  100
118 90  6   4   6   2   1   1   1   3   250 500
# 200   140 4   3   4   2   1   1   1   3   10  100

Answer 1

似乎有些东西比你期望的更频繁地调用方法。

在方法的第一行放置一个断点，看看它被调用的时间和原因。 这个bug几乎必然存在于调用代码中，而不是方法本身，这意味着我们无法提供更多帮助，而不仅仅是建议断点，记录堆栈跟踪等。