排序列表 <string> 通过自定义规则

Question

I am currently writing a program that reads directories, writes all of the .pdf files' paths in the directory into a List, and merges the files into a single .pdf file

public void ProcessFiles(string path)
{
    try
    {
        List<string> pathsForFiles = new List<string>();
        String storeNum = "";
        //Every subdirectory in the directory
        foreach (string d in Directory.GetDirectories(path))
        {
            storeNum = (Path.GetFileName(d));

            //Every item in the folder
            foreach (string filePaths in Directory.GetFiles(d))
            {
                pathsForFiles.Add(filePaths);
            }

            pathsForFiles.Sort(new PdfFileComparer());

            foreach (string f in Directory.GetFiles(d))
            {
                Console.WriteLine(f);
            }

            string dirForOutput = @"C:\Users\EMC\Desktop";
            MergeFiles(pathsForFiles, storeNum, dirForOutput);
            pathsForFiles.Clear();

            Console.WriteLine(" - - - - - - - - - - - - - - - - ");
        }
        Console.ReadKey();
    }
    catch (Exception excpt)
    {
        Console.WriteLine(excpt.Message);
    }
} // end ProcessFiles()

public void MergeFiles(List<string> f, string storeNum, string outputDirectory)
{
    PdfVision v = new PdfVision();
    string[] files = new string[f.Count];
    int i = 0;
    foreach (string s in f)
    {
        files[i] = s;
        i++;
    }
    string outName = outputDirectory + "\\" + storeNum + " Original As-Builts.pdf";
    int ret = v.MergePDFFileArrayToPDFFile(files, @outName);
    //0 - merged successfully
    //1 - error, can't merge PDF documents
    //2 - error, can't create output file, probably it used by another application
    //3 - merging failed
    //4 - merged successfully, but some files were not merged
    Console.WriteLine("Merge Status - " + ret);
}

The program runs as expected, however the .pdf files are not in order. I cannot rename the actual files, however I need to come up with a way to sort them. My guidelines for the order of the .pdf's are as follows -

1. Cover Page
2. S - Pages
3. E - Pages
4. L - Pages
5. C - Pages
6. D - Pages
7. Anything else (H, R, etc)

I have come up with a class to write my comparing, however it only slightly works the way I want it to. Here is the class, currently it assures that the COVER page is always first, however I don't know where to continue.

class PdfFileComparer : IComparer<string>
{
    public int Compare(string x, string y)
    {
        string file1 = Path.GetFileNameWithoutExtension(x);
        string file2 = Path.GetFileNameWithoutExtension(y);
        // Get last part of filenames (after the last '_')
        string lastBit1 = file1.Substring(file1.LastIndexOf('_') + 1);
        string lastBit2 = file2.Substring(file2.LastIndexOf('_') + 1);

        // Ensure COVER file always first
        if (string.Compare(lastBit1, "COVER", StringComparison.InvariantCultureIgnoreCase) == 0)
        {
            return -1;
        }
        if (string.Compare(lastBit2, "COVER", StringComparison.InvariantCultureIgnoreCase) == 0)
        {
            return 1;
        }

        // Else just sort by the last part alphabetically (ignoring case)
        return String.Compare(lastBit1, lastBit2, StringComparison.CurrentCultureIgnoreCase);
    }
}

My question is how do I conform this class to fit the above criteria for sorting the files. The files' format looks somewhat like this, where the ending indicates the page type.

C:\\Users\\Desktop\\PDFs\\T-0023\\T-0023_H1.pdf

C:\\Users\\Desktop\\PDFs\\T-0023\\T-0023_c3_S01.pdf

C:\\Users\\Desktop\\PDFs\\T-0023\\T-0023_L01.pdf

C:\\Users\\Desktop\\PDFs\\T-0023\\T-0023_H3A.pdf

C:\\Users\\Desktop\\PDFs\\T-0023\\T-0023_H3B.pdf

Answer 1

I'd create a collection of the types you want in order:

var types = "SELCD";

int file1Index = types.IndexOf(Char.ToUpper(lastBit1[0]));
int file2Index = types.IndexOf(Char.ToUpper(lastBit2[0]));

if (file1Index == -1 && file2Index != -1)
   return 1;

if (file1Index > file2Index)
    return 1;
if (file1Index < file2Index)
   return -1;

// Else do your alphabetic sort

EDIT Updated code to account for file2's character being in the collection, but not file1's.

Answer 2

Here's another similar way to do it, where you keep a list of the items you want to order by, and then for each file you determine if the "last bit" starts with any of the items in the list by getting the IndexOf for the FirstOrDefault item in the list where the item StartsWith that string.

The logic goes:

• If the indexes of both items are the same, then return the comparison of their 'lastBit' strings, and if those are the same use the whole file name.

• If the first item is not contained in the list then return 1 (first item is greater than second)

• If the second item is not in the list then return -1 (first item is less than the second)

• Finally, return the comparison of their indexes in the list

Here's the code:

class PdfFileComparer : IComparer<string>
{
    public int Compare(string x, string y)
    {
        var first = Path.GetFileNameWithoutExtension(x);
        var second = Path.GetFileNameWithoutExtension(y);

        // Get last part of filenames (after the last '_')
        var lastBit1 = first.Substring(first.LastIndexOf('_') + 1).ToUpper();
        var lastBit2 = second.Substring(second.LastIndexOf('_') + 1).ToUpper();

        var orders = new List<string>
        {
            "COVER", "S", "E", "L", "C", "D"
        };

        var firstIndex = orders.IndexOf(
            orders.FirstOrDefault(sortTerm => lastBit1.StartsWith(sortTerm)));
        var secondIndex = orders.IndexOf(
            orders.FirstOrDefault(sortTerm => lastBit2.StartsWith(sortTerm)));

        if (firstIndex == secondIndex)
        {
            var result = lastBit1.CompareTo(lastBit2);
            if (result == 0) result = x.CompareTo(y);
            return result;
        }

        if (firstIndex < 0)
        {
            return 1;
        }
        if (secondIndex < 0)
        {
            return -1;
        }

        return firstIndex.CompareTo(secondIndex);
    }
}

To test it, I just created a dummy list of file names based on the ones you provided, and added a few more in an unsorted way:

private static void Main()
{
    var fileNames = new List<string>
    {
        @"C:\Users\Desktop\PDFs\T-0023\T-0023_H1.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_c3_S01.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_L01.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_E3A.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_H3B.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_c3_COVER.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_D3A.pdf",
        @"C:\Users\Desktop\PDFs\T - 0023\T - 0023_C3B.pdf",
    };

    fileNames.Sort(new PdfFileComparer());

    fileNames.ForEach(Console.WriteLine);

    GetKeyFromUser("\nPress any key to exit...");
}

Output