// *** PUBLIC ***
// CONSTRUCTOR
/// <summary>
/// Initializes DawgCreator, the only way of initializing
/// </summary>
/// <param name="lexiconName">Name of the lexicon</param>
/// <param name="numNodes">Number of nodes (after compression typically)</param>
/// <param name="numWords">Number of full words in the Dawg</param>
public DawgCreator(string lexiconName, uint numNodes, uint numWords, uint numReversePartWords)
{
// initialization
CreateHeader(lexiconName, numNodes, numWords, numReversePartWords);
nodes = new DawgNode[numNodes];
numAddedNodes = 0;
}
// ADD NODE
/// <summary>
/// Adds the given node to DAWG. Please note the nodes need to be added sequentially!
/// </summary>
/// <param name="dawgNode">Node to be added</param>
public void AddNode(DawgNode dawgNode)
{
// validations
Debug.Assert(numAddedNodes < header.NumNodes);
nodes[numAddedNodes++] = dawgNode;
}
private const uint MINIMUM_NUMBER_OF_NODES = 3; // Root, forward and reverse
// *** PUBLIC ***
// CONSTRUCTOR
public Dawg(string fileName)
{
// open the file
using (BinaryReader reader = new BinaryReader(File.Open(fileName, FileMode.Open)))
{
// read the header
header.VersionNumber = reader.ReadUInt32();
header.NumNodes = reader.ReadUInt32();
header.LexiconName = reader.ReadString();
header.LexiconDate = reader.ReadString();
header.NumWords = reader.ReadUInt32();
header.NumReversePartWords = reader.ReadUInt32();
// validations
if (header.VersionNumber != 1U)
{
string exceptionMessage = string.Format(
"Unxpected Dawg header version: {0}. Only Version 1 is supported!",
header.VersionNumber);
throw new InvalidDataException(exceptionMessage);
}
if (header.NumNodes < Dawg.MINIMUM_NUMBER_OF_NODES)
{
string exceptionMessage = string.Format(
"Number of nodes specified in Dawg header: {0} is less than the minimum number of nodes: {1}!",
header.NumNodes, Dawg.MINIMUM_NUMBER_OF_NODES);
throw new InvalidDataException(exceptionMessage);
}
// read the nodes
nodes = new DawgNode[header.NumNodes];
for (uint idx = 0; idx < header.NumNodes; idx++)
{
nodes[idx].ChildNodeIdx = reader.ReadUInt32();
nodes[idx].Letter = reader.ReadByte();
nodes[idx].IsTerminal = reader.ReadBoolean();
nodes[idx].IsLastChild = reader.ReadBoolean();
}
}
// validate the counts
uint numCountedWords = CountNumWords();
if (header.NumWords != numCountedWords)
{
string exceptionMessage = string.Format(
"Number of words specified in Dawg header: {0} is less than the actual number of words: {1}!",
header.NumWords, numCountedWords);
throw new InvalidDataException(exceptionMessage);
}
uint numCountedReversePartWords = CountNumReversePartWords();
if (header.NumReversePartWords != numCountedReversePartWords)
{
string exceptionMessage = string.Format(
"Number of words specified in Dawg header: {0} is less than the actual number of words: {1}!",
header.NumReversePartWords, numCountedReversePartWords);
throw new InvalidDataException(exceptionMessage);
}
}
