/// <summary>
/// Initialises a new instance of the Algorithms.CandidateWordPriorityCalculator class.
/// </summary>
/// <param name="maximumSourceWordToCandidateWordDistance">The maximum possible distance between any two words in the graph.</param>
/// <param name="sourceWordToCandidateWordDistanceWeight">The weight which should be applied when calculating the overall priority, to the distance from the source word to the candidate word (the 'g(n)' score in the A* algorithm).</param>
/// <param name="numberOfCharactersMatchingDestinationWeight">The weight which should be applied when calculating the overall priority, to the number of matching characters between the candidate word and the destination word.</param>
/// <param name="popularityOfChangeToCharacterWeight">The weight which should be applied when calculating the overall priority, to the frequency that the character changed to in the candidate word is the 'changed from' character in a substitution between adjacent words.</param>
/// <param name="popularityOfCharacterChangeWeight">The weight which should be applied when calculating the overall priority, to the frequency/popularity of the character substitution (i.e. character changed from current word to candidate word).</param>
/// <param name="allWordsTrieRoot">The root of a character trie containing all the words in the graph.</param>
/// <param name="fromCharacterFrequencies">A FrequencyTable containing the number of times each character is the 'from' character in a substitution between adjacent words (i.e. represented by an edge of the graph).</param>
/// <param name="characterSubstitutionFrequencies">A FrequencyTable containing the number of times each pair of characters in a substitution between adjacent words (i.e. represented by an edge of the graph) occurs.</param>
/// <exception cref="System.ArgumentException">Parameter 'maximumSourceWordToCandidateWordDistance' is less than 0.</exception>
/// <exception cref="System.ArgumentException">Parameter 'sourceWordToCandidateWordDistanceWeight' is less than 0.</exception>
/// <exception cref="System.ArgumentException">Parameter 'numberOfCharactersMatchingDestinationWeight' is less than 0.</exception>
/// <exception cref="System.ArgumentException">Parameter 'popularityOfChangeToCharacterWeight' is less than 0.</exception>
/// <exception cref="System.ArgumentException">Parameter 'popularityOfCharacterChangeWeight' is less than 0.</exception>
/// <exception cref="System.ArgumentException">At least one of parameters 'sourceWordToCandidateWordDistanceWeight', 'numberOfCharactersMatchingDestinationWeight', 'popularityOfChangeToCharacterWeight', and 'popularityOfCharacterChangeWeight' must be greater than 0.</exception>
public CandidateWordPriorityCalculator(Int32 maximumSourceWordToCandidateWordDistance, Int32 sourceWordToCandidateWordDistanceWeight, Int32 numberOfCharactersMatchingDestinationWeight, Int32 popularityOfChangeToCharacterWeight, Int32 popularityOfCharacterChangeWeight, Dictionary <Char, TrieNode <Char> > allWordsTrieRoot, FrequencyTable <Char> fromCharacterFrequencies, FrequencyTable <CharacterSubstitution> characterSubstitutionFrequencies)
{
if (maximumSourceWordToCandidateWordDistance < 1)
{
throw new ArgumentException("Parameter 'maximumSourceWordToCandidateWordDistance' must be greater than or equal to 1.", "maximumSourceWordToCandidateWordDistance");
}
if (sourceWordToCandidateWordDistanceWeight < 0)
{
throw new ArgumentException("Parameter 'sourceWordToCandidateWordDistanceWeight' must be greater than or equal to 0.", "sourceWordToCandidateWordDistanceWeight");
}
if (numberOfCharactersMatchingDestinationWeight < 0)
{
throw new ArgumentException("Parameter 'numberOfCharactersMatchingDestinationWeight' must be greater than or equal to 0.", "numberOfCharactersMatchingDestinationWeight");
}
if (popularityOfChangeToCharacterWeight < 0)
{
throw new ArgumentException("Parameter 'popularityOfChangeToCharacterWeight' must be greater than or equal to 0.", "popularityOfChangeToCharacterWeight");
}
if (popularityOfCharacterChangeWeight < 0)
{
throw new ArgumentException("Parameter 'popularityOfCharacterChangeWeight' must be greater than or equal to 0.", "popularityOfCharacterChangeWeight");
}
if (sourceWordToCandidateWordDistanceWeight == 0 && numberOfCharactersMatchingDestinationWeight == 0 && popularityOfChangeToCharacterWeight == 0 && popularityOfCharacterChangeWeight == 0)
{
throw new ArgumentException("At least one of parameters 'sourceWordToCandidateWordDistanceWeight', 'numberOfCharactersMatchingDestinationWeight', 'popularityOfChangeToCharacterWeight', and 'popularityOfCharacterChangeWeight' must be greater than 0.");
}
// Initialize priority functions and weights
priorityFunctions = new List <Func <String, String, String, Int32, Double> >();
priorityFunctionWeights = new List <Int32>();
functionWeightsTotal = 0;
priorityFunctionWeights.Add(sourceWordToCandidateWordDistanceWeight);
priorityFunctions.Add(CalculateSourceWordToCandidateWordDistancePriority);
functionWeightsTotal += Convert.ToInt64(sourceWordToCandidateWordDistanceWeight);
priorityFunctionWeights.Add(numberOfCharactersMatchingDestinationWeight);
priorityFunctions.Add(CalculateNumberOfCharactersMatchingDestinationPriority);
functionWeightsTotal += Convert.ToInt64(numberOfCharactersMatchingDestinationWeight);
priorityFunctionWeights.Add(popularityOfChangeToCharacterWeight);
priorityFunctions.Add(CalculatePopularityOfChangeToCharacterPriority);
functionWeightsTotal += Convert.ToInt64(popularityOfChangeToCharacterWeight);
priorityFunctionWeights.Add(popularityOfCharacterChangeWeight);
priorityFunctions.Add(CalculatePopularityOfCharacterChangePriority);
functionWeightsTotal += Convert.ToInt64(popularityOfCharacterChangeWeight);
this.maximumSourceWordToCandidateWordDistance = maximumSourceWordToCandidateWordDistance;
this.allWordsTrieRoot = allWordsTrieRoot;
this.fromCharacterFrequencies = fromCharacterFrequencies;
this.characterSubstitutionFrequencies = characterSubstitutionFrequencies;
wordUtilities = new WordUtilities();
PopulateMaximumFrequencyMembers();
}
/// <summary>
/// Reads a set of words from a stream, and adds the words and statistics pertaining to the words to a set of data structures.
/// </summary>
/// <param name="reader">The stream reader to used to read the set of words (allows specifying a mocked reader for unit testing).</param>
/// <param name="trieBuilder">The trie builder to use to add the words to the trie (allows specifying a mocked builder for unit testing).</param>
/// <param name="wordFilterFunction">A Func to filter whether or not the specified word should be added to the trie. Accepts the word as a parameter, and returns a bookean indicating whether that word should be added to the trie.</param>
/// <param name="allWordsTrieRoot">The root of a character trie to populate with the words.</param>
/// <param name="allWords">A HashSet to populate with the words.</param>
/// <param name="fromCharacterFrequencies">A FrequencyTable to populate with the number of times each character is the 'from' character in a substitution.</param>
/// <param name="characterSubstitutionFrequencies">A FrequencyTable to populate with the number of times each pair of characters in a substitution occur.</param>
public void PopulateAdjacentWordDataStructures(IStreamReader reader, ICharacterTrieBuilder trieBuilder, Func <String, Boolean> wordFilterFunction, Dictionary <Char, TrieNode <Char> > allWordsTrieRoot, HashSet <String> allWords, FrequencyTable <Char> fromCharacterFrequencies, FrequencyTable <CharacterSubstitution> characterSubstitutionFrequencies)
{
// Read all words and add them to the HashSet and trie
using (reader)
{
while (reader.EndOfStream == false)
{
String currentWord = reader.ReadLine();
if (wordFilterFunction.Invoke(currentWord) == true)
{
if (allWords.Contains(currentWord) == false)
{
allWords.Add(currentWord);
trieBuilder.AddWord(allWordsTrieRoot, currentWord, true);
}
}
}
}
// Populate the frequency tables
CharacterTrieUtilities trieUtilities = new CharacterTrieUtilities();
WordUtilities wordUtilities = new WordUtilities();
foreach (String currentWord in allWords)
{
foreach (String adjacentWord in trieUtilities.FindAdjacentWords(allWordsTrieRoot, currentWord))
{
// Find the character which was substitued between the word and the adjacent word
Tuple <Char, Char> differingCharacters = wordUtilities.FindDifferingCharacters(currentWord, adjacentWord);
Char fromCharacter = differingCharacters.Item1, toCharacter = differingCharacters.Item2;
// Increment the data structures
fromCharacterFrequencies.Increment(fromCharacter);
characterSubstitutionFrequencies.Increment(new CharacterSubstitution(fromCharacter, toCharacter));
}
}
}
