/// <see cref="IIntegerFileInfoCollector.GetIntegerFileInfo(string, int)"/>
public IntegerFileInfo GetIntegerFileInfo(string filePath, int chunkSize)
{
Debug.Assert(filePath != null);
Debug.Assert(chunkSize > 0);
IntegerFileInfo fileInfo = new IntegerFileInfo()
{
NumOfChunks = 0,
NumOfIntegers = 0
};
//Open a text reader to the file
using (StreamReader fileReader = fileIO.CreateFileStreamReader(filePath))
{
while (!fileReader.EndOfStream)
{
//Read each line to count the number of lines in the file
fileReader.ReadLine();
fileInfo.NumOfIntegers++;
}
}
//Now that we have the number of integers, calculate the number of chunks
fileInfo.NumOfChunks = (int)Math.Ceiling(Decimal.Divide(fileInfo.NumOfIntegers, chunkSize));
return(fileInfo);
}
/// <summary>
/// Sorts the integers found in an input file and writes the sorted integers to an output file
/// </summary>
/// <param name="options">The command line options</param>
private static List <string> SortIntegers(CommandLineOptions options)
{
IFileIO fileIO = serviceProvider.GetService <IFileIO>();
IIntegerFileInfoCollector integerFileInfoCollector = serviceProvider.GetService <IIntegerFileInfoCollector>();
//Find the output directory where the output file will be written
string outputDirectory = fileIO.GetDirectoryFromFilePath(options.OutputFilePath);
//Indicate to the user that we are retrieving information regarding the input file. This can take a long
//time for very large files.
DisplayInputFileInfoInProgress();
//Collect information on the input file
IntegerFileInfo inputFileInfo = integerFileInfoCollector.GetIntegerFileInfo(options.InputFile, (int)options.ChunkSize);
//Store the intermediate files so that they can be deleted later on
List <string> intermediateFilePaths = new List <string>();
List <string> chunkFiles = CreateChunkFiles(options, inputFileInfo, outputDirectory);
//Translate the file names to file paths
var chunkFilePaths = chunkFiles.Select(file => Path.Combine(outputDirectory, file)).ToList();
intermediateFilePaths.AddRange(chunkFilePaths);
List <string> mergeIntermediateFilePaths = MergeChunkFiles(chunkFiles, options, inputFileInfo);
intermediateFilePaths.AddRange(mergeIntermediateFilePaths);
return(intermediateFilePaths);
}
/// <summary>
/// Creates the chunk files
/// </summary>
/// <param name="options">The command line options</param>
/// <param name="inputFileInfo">Information regarding the input file</param>
/// <param name="outputDirectory">The directory where the chunk files will be written</param>
/// <returns>The chunk files that were created</returns>
private static List <string> CreateChunkFiles(CommandLineOptions options, IntegerFileInfo inputFileInfo,
string outputDirectory)
{
DisplayInputFileInfo(inputFileInfo);
//Create the progress bar for the chunk file creation
var chunkProgressBar = CreateProgressBar(inputFileInfo.NumOfChunks, ChunkProgressMessage);
//Create the progress callback that will be passed to the chunk file creation code
Action <int> updateChunkProgress = (int chunkNum) => UpdateProgressBar(chunkProgressBar, chunkNum);
//Create the sorted chunk files
string chunkFileTemplate = string.Format(GenFileTemplate, "1", "{0}");
List <string> chunkFiles = CreateSortedChunkFiles(options.InputFile, (int)options.ChunkSize,
outputDirectory, chunkFileTemplate, updateChunkProgress);
if (chunkFiles.Count > 0)
{
OnProgressBarCompleted();
}
DisplayChunkFileCreationResults(chunkFiles.Count);
return(chunkFiles);
}
/// <summary>
/// Displays the integer file information for the input file
/// </summary>
/// <param name="inputFileInfo">The input file information</param>
private static void DisplayInputFileInfo(IntegerFileInfo inputFileInfo)
{
string integerOutput = "Number of Integers: {0}";
string chunkOutput = "Number of Chunks: {0}";
Console.WriteLine(string.Format(integerOutput, inputFileInfo.NumOfIntegers));
Console.WriteLine(string.Format(chunkOutput, inputFileInfo.NumOfChunks));
}
/// <summary>
/// Merges a set of chunk files into a single output file
/// </summary>
/// <param name="chunkFiles">The chunk files to be merged</param>
/// <param name="options">The command line options</param>
/// <param name="inputFileInfo">Information regarding the input file</param>
/// <returns>The intermediate files that were created while merging</returns>
private static List <string> MergeChunkFiles(List <string> chunkFiles, CommandLineOptions options,
IntegerFileInfo inputFileInfo)
{
List <string> intermediateFilePaths = new List <string>();
IFileIO fileIO = serviceProvider.GetService <IFileIO>();
IChunkFileMerger chunkFileMerger = serviceProvider.GetService <IChunkFileMerger>();
//Find the output directory where the output file will be written
string outputDirectory = fileIO.GetDirectoryFromFilePath(options.OutputFilePath);
//Convert the chunk file names into chunk file paths
List <string> chunkFilePaths = chunkFiles
.Select(chunkFileName => Path.Combine(outputDirectory, chunkFileName))
.ToList();
//Keep a dictionary of progress bars where the key is the merge generation number, and
//the value is the progress bar for that merge generation
Dictionary <int, ProgressBar> mergeProgressBars = new Dictionary <int, ProgressBar>();
Action <int, int> updateMergeProgress = (int gen, int integersProcessed) =>
{
if (!mergeProgressBars.ContainsKey(gen))
{
//Create a progress bar for the current generation
mergeProgressBars[gen] = CreateProgressBar(inputFileInfo.NumOfIntegers,
string.Format(MergeProgressMessage, gen - 1));
}
bool stepsComplete = UpdateProgressBar(mergeProgressBars[gen], integersProcessed);
if (stepsComplete)
{
OnProgressBarCompleted();
//Display the results of the merge operation. We can calculate the number of output
//files that will result from each merge generation
int mergedFilesCount = CalculateNumberOfGenerationOutputFiles(chunkFiles.Count, MergeCount,
gen - FirstMergeGeneration + 1);
DisplayGenMergeResults(gen, mergedFilesCount);
}
};
List <string> mergeFiles = chunkFileMerger.MergeChunkFilesIntoSingleFile(chunkFilePaths, MergeCount, GenFileTemplate,
Path.GetFileName(options.OutputFilePath), outputDirectory, FirstMergeGeneration, !options.KeepIntermediate,
updateMergeProgress);
intermediateFilePaths.AddRange(mergeFiles);
return(intermediateFilePaths);
}