source: branches/2886_SymRegGrammarEnumeration/HeuristicLab.Algorithms.DataAnalysis.SymRegGrammarEnumeration/Analysis/SentenceLogger.cs @ 16017

using System;
using System.Diagnostics;
using System.IO;
using System.IO.Compression;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;

namespace HeuristicLab.Algorithms.DataAnalysis.SymRegGrammarEnumeration {
  [Item("SentenceLogger", "")]
  [StorableClass]
  public class SentenceLogger : Item, IGrammarEnumerationAnalyzer {
    private readonly string workingDir = Environment.GetFolderPath(Environment.SpecialFolder.DesktopDirectory);
    private readonly string columnDelimiter = ";";
    private readonly string header = "hash;length;infix";

    private string distinctSentencesFileName;
    private string allSentencesFileName;
    private string shortestDistinctSentencesFileName;

    private TextWriterTraceListener distinctSentencesFileTrace;
    private TextWriterTraceListener allSentencesFileTrace;

    public SentenceLogger() { }

    [StorableConstructor]
    protected SentenceLogger(bool deserializing) : base(deserializing) { }

    protected SentenceLogger(SentenceLogger original, Cloner cloner) : base(original, cloner) { }

    public override IDeepCloneable Clone(Cloner cloner) {
      return new SentenceLogger(this, cloner);
    }

    public void Register(GrammarEnumerationAlgorithm algorithm) {
      algorithm.Stopped += GrammarEnumerationAlgorithmOnStopped;
      algorithm.ExceptionOccurred += GrammarEnumerationAlgorithmOnStopped;

      algorithm.SentenceGenerated += SentenceGenerated;
      algorithm.DistinctSentenceGenerated += DistinctSentenceGenerated;
    }

    public void Deregister(GrammarEnumerationAlgorithm algorithm) {
      algorithm.Stopped -= GrammarEnumerationAlgorithmOnStopped;
      algorithm.ExceptionOccurred -= GrammarEnumerationAlgorithmOnStopped;

      algorithm.SentenceGenerated -= SentenceGenerated;
      algorithm.DistinctSentenceGenerated -= DistinctSentenceGenerated;
    }

    private void GrammarEnumerationAlgorithmOnStopped(object sender, EventArgs eventArgs) {
      distinctSentencesFileTrace.Close();
      allSentencesFileTrace.Close();

      // Remove duplicates afterwards using bash commands from the git bash
      // string bashExecutable = Environment.GetFolderPath(Environment.SpecialFolder.LocalApplicationData) + @"\Programs\Git\git-bash.exe";
      // 
      // string commandCreate = $"echo \"{header}\" > {shortestDistinctSentencesFileName.Replace("\\", "/")}";
      // string commandFill = $"sort -s {distinctSentencesFileName.Replace("\\", "/")} | uniq -w 10 >> {shortestDistinctSentencesFileName.Replace("\\", "/")}";
      // 
      // ProcessStartInfo startInfo = new ProcessStartInfo {
      //   WindowStyle = ProcessWindowStyle.Hidden,
      //   UseShellExecute = false,
      //   CreateNoWindow = true,
      //   FileName = bashExecutable,
      //   RedirectStandardError = true,
      //   RedirectStandardOutput = true,
      //   Arguments = $"-c '{commandCreate};{commandFill}'"
      // };
      // Process.Start(startInfo);
    }

    private void Init(object sender) {
      string datePostfix = $"_{DateTime.Now:yyyy-MM-dd_HH-mm}_TreeSize-{((GrammarEnumerationAlgorithm)sender).MaxComplexity}.csv.gz";
      distinctSentencesFileName = workingDir + @"\distinctSentences" + datePostfix;
      allSentencesFileName = workingDir + @"\allSentences" + datePostfix;
      shortestDistinctSentencesFileName = workingDir + @"\shortestDistinctSentences" + datePostfix;

      distinctSentencesFileTrace = new TextWriterTraceListener(new GZipStream(new FileStream(distinctSentencesFileName, FileMode.Create), CompressionMode.Compress));
      allSentencesFileTrace = new TextWriterTraceListener(new GZipStream(new FileStream(allSentencesFileName, FileMode.Create), CompressionMode.Compress));
      ((StreamWriter)distinctSentencesFileTrace.Writer).AutoFlush = true;
      ((StreamWriter)allSentencesFileTrace.Writer).AutoFlush = true;

      allSentencesFileTrace.WriteLine(header);
    }

    private void DistinctSentenceGenerated(object sender, PhraseAddedEventArgs phraseAddedEventArgs) {
      if (distinctSentencesFileName == null) Init(sender);
      distinctSentencesFileTrace.WriteLine(ToCsvLine(
        ((uint)phraseAddedEventArgs.NewHash).ToString("D10"),
        phraseAddedEventArgs.NewPhrase.Count().ToString("D3"),
        //phraseAddedEventArgs.NewPhrase.ToString(),
        ((GrammarEnumerationAlgorithm)sender).Grammar.ToInfixString(phraseAddedEventArgs.NewPhrase)));
    }


    private void SentenceGenerated(object sender, PhraseAddedEventArgs phraseAddedEventArgs) {
      if (allSentencesFileTrace == null) Init(sender);
      allSentencesFileTrace.WriteLine(ToCsvLine(
        ((uint)phraseAddedEventArgs.NewHash).ToString("D10"),
        phraseAddedEventArgs.NewPhrase.Count().ToString("D3"),
        //phraseAddedEventArgs.NewPhrase.ToString(),
        ((GrammarEnumerationAlgorithm)sender).Grammar.ToInfixString(phraseAddedEventArgs.NewPhrase)));
    }

    private string ToCsvLine(params string[] cols) {
      return string.Join(columnDelimiter, cols);
    }
  }
}