/// <summary>
/// Writes a single probability hash entry as a line
/// in the data file
/// </summary>
/// <param name="data">Writing state structure</param>
private static void WriteLine(CorpusWriteData data)
{
StringBuilder sb = new StringBuilder();
// String each key word together, separating them
// with an item delimiter constant
for (int i = 0; i < data.Key.Length; i++)
{
sb.Append(data.Key[i]);
if (i < data.Key.Length - 1)
sb.Append(ItemDelimiter);
}
// Add the sub-topic to the key portion of the line
sb.Append(TopicDelimiter)
.Append(data.Key.Topic)
.Append(ListDelimiter);
// String the values together along with their probabilities.
for (int i = 0; i < data.ProbabilityHash[data.Key].Count; i++)
{
var word = data.ProbabilityHash[data.Key][i];
sb.Append(word.Key)
.Append(ProbabilityDelimiter)
.Append(word.Value);
if (i < data.ProbabilityHash[data.Key].Count - 1)
sb.Append(ItemDelimiter);
}
data.Writer.WriteLine(sb.ToString());
}
/// <summary>
/// Serializes a probability hash and writes it to disk.
/// </summary>
/// <param name="topic"></param>
/// <param name="probabilityHash"></param>
public static void Write(string topic, ProbabilityHash probabilityHash)
{
// Mangle the topic name to save the file with no spaces
// in the filename.
topic = topic.ToLower().Replace(" ", "-");
// This was a rushed design decision. I actually change the working directory to
// get to the corpus data. A lock needs to be held because the client can be multithreaded,
// and a race condition could cause the directory to be entered twice.
lock (DataDirectories.DirectoryControl)
{
// Create the corpus directory if it doesn't exist already.
DataDirectories.ForceEnterDirectory(DirectoryName);
// Create the filename with string substitution
string filename = String.Format(FilenamePattern, topic, probabilityHash.Order);
// I can't think of a reason why the filename would exist
// already, but just in case, delete it.
if (File.Exists(filename))
File.Delete(filename);
using (var writer = new StreamWriter(filename, false))
{
// Use a struct to simplify argument passing
var data = new CorpusWriteData()
{
Writer = writer,
ProbabilityHash = probabilityHash
};
// Write each probability hash entry as a
// line in the data file
foreach (var key in probabilityHash.Keys)
{
data.Key = key;
WriteLine(data);
}
}
DataDirectories.LeaveDirectory();
}
}
