public static void DoPorter() //todo: не помешает добавить чистку от лишних пробелов, а то лагает чутка
{
var porter = new Porter();
foreach (var fileName in initialFileNames)
{
var textToProcess = File.ReadAllText(pathForPorterFiles + fileName);
var textAfterProcess = "";
var startIndex = 0;
var count = 0;
for (var index = 0; index < textToProcess.Length; index++)
{
var symbol = textToProcess[index];
if (symbol == ' ' || symbol == '!' || symbol == '?' || symbol == '.' || symbol == ',' ||
symbol == '(' || symbol == ')' || symbol == ':' || symbol == ';' || symbol == '_')
{
var endIndex = index;
var word = "";
var temp = string.Copy(textToProcess);
var len = endIndex - startIndex;
if (len > 0)
{
count++;
word += temp.Substring(startIndex, len);
word = word.ToLower();
word = !porter.particles.Contains(word) ? porter.Stemm(word) : "";
textAfterProcess += word + " ";
startIndex = endIndex + 1;
if (!string.IsNullOrEmpty(word))
{
if (invertIndexes.Keys.Contains(word))
{
invertIndexes[word].Add(fileName);
}
else
{
var linksForWord = new HashSet <string>
{
fileName
}; //todo: нужна ссылка или имя фйала?
invertIndexes.Add(word, linksForWord);
}
}
}
}
}
countsWordsInDocuments.Add(fileName, count);
File.WriteAllText(pathForPorterFiles + fileName, textAfterProcess);
WriteInverIndexesInFile();
}
}
public void Search(string request)
{
var porter = new Porter();
var textToProcess = request;
var textAfterProcess = "";
var startIndex = 0;
var count = 0;
List <string> words = new List <string>();
for (var index = 0; index < textToProcess.Length; index++)
{
var symbol = textToProcess[index];
if (symbol == ' ' || symbol == '!' || symbol == '?' || symbol == '.' || symbol == ',' ||
symbol == '(' || symbol == ')' || symbol == ':' || symbol == ';' || symbol == '_')
{
var endIndex = index;
var word = "";
var temp = string.Copy(textToProcess);
var len = endIndex - startIndex;
if (len > 0)
{
count++;
word += temp.Substring(startIndex, len);
word = word.ToLower();
word = !porter.particles.Contains(word) ? porter.Stemm(word) : "";
textAfterProcess += word + " ";
startIndex = endIndex + 1;
if (!string.IsNullOrEmpty(word))
{
words.Add(word);
}
}
}
}
Dictionary <string, int> results = new Dictionary <string, int>();
DirectoryInfo dir = new DirectoryInfo(pathForPorterFiles);
foreach (var item in dir.GetFiles())
{
int countWords = 0;
var text = File.ReadAllText(pathForPorterFiles + item.Name);
foreach (var word in words)
{
if (text.Contains(word))
{
countWords++;
}
}
results.Add(item.Name, countWords);
}
Console.WriteLine("Булевый поиск: для запроса \"" + request + "\" составлен список совпадений: ");
foreach (var result in results)
{
Console.WriteLine("в документе \"" + result.Key + "\" совпало слов из запроса: " +
result.Value);
}
}
public void Search(string request)
{
var porter = new Porter();
var textToProcess = request;
var textAfterProcess = "";
var startIndex = 0;
var count = 0;
List <string> words = new List <string>();
for (var index = 0; index < textToProcess.Length; index++)
{
var symbol = textToProcess[index];
if (symbol == ' ' || symbol == '!' || symbol == '?' || symbol == '.' || symbol == ',' ||
symbol == '(' || symbol == ')' || symbol == ':' || symbol == ';' || symbol == '_')
{
var endIndex = index;
var word = "";
var temp = string.Copy(textToProcess);
var len = endIndex - startIndex;
if (len > 0)
{
count++;
word += temp.Substring(startIndex, len);
word = word.ToLower();
word = !porter.particles.Contains(word) ? porter.Stemm(word) : "";
textAfterProcess += word + " ";
startIndex = endIndex + 1;
if (!string.IsNullOrEmpty(word))
{
words.Add(word);
}
}
}
}
var file = File.ReadLines("C:\\Users\\Elina\\RiderProjects\\IS\\Spider\\Spider\\outputs\\TFIDF\\TFIDF.txt");
Dictionary <string, double> terms = new Dictionary <string, double>();
int i = 0;
string w = "";
bool readNext = false;
foreach (var line in file)
{
if (i == 0 || readNext && line.Length > 1)
{
w = line.Substring(0, line.Length - 1);
readNext = false;
}
string num = "";
if (line.StartsWith("IDF:"))
{
readNext = true;
num = line.Substring(5);
var idf = double.Parse(num);
if (!terms.ContainsKey(w))
{
terms.Add(w, idf);
}
else
{
terms.Add(w + " ", idf);
}
}
i++;
}
List <double> vectorRequest = new List <double>();
foreach (var term in terms)
{
if (words.Contains(term.Key))
{
vectorRequest.Add(term.Value);
}
else
{
vectorRequest.Add(0);
}
}
int j = 0;
bool readN = false;
var vectorsFile = File.ReadLines("C:\\Users\\Elina\\RiderProjects\\IS\\Spider\\Spider\\outputs\\vectorSearch\\vectorSearch.txt");
Dictionary <string, List <double> > vectorsInDocuments = new Dictionary <string, List <double> >();
List <double> tempVector = new List <double>();
string tempLine = "";
foreach (var line in vectorsFile)
{
if (line.Contains("vector:"))
{
readN = true;
continue;
}
if (readN)
{
readN = false;
if (line.EndsWith(j.ToString()))
{
var num = double.Parse(line.Substring(0, line.IndexOf("-") - 1));
tempVector.Add(num);
}
}
if (line.Contains(".txt"))
{
vectorsInDocuments.Add(tempLine, tempVector);
tempVector = new List <double>();
tempLine = line.Substring(0, line.Length - 5);
j++;
}
}
Dictionary <string, double> results = new Dictionary <string, double>();
foreach (var vector in vectorsInDocuments)
{
var result = 0d;
var len1 = 0d;
var len2 = 0d;
for (int k = 0; k < vector.Value.Count; k++)
{
result += vector.Value[k] * vectorRequest[k];
len1 += vector.Value[k] * vector.Value[k];
len2 += vectorRequest[k];
}
result /= Math.Sqrt(len1);
result /= Math.Sqrt(len2);
results.Add(vector.Key, result);
}
Console.WriteLine("для запроса \"" + request + "\" составлен список совпадений: ");
foreach (var result in results)
{
Console.WriteLine("вероятность найдти информацию по запросу в документе \"" + result.Key + "\" равна: " + result.Value * 100);
}
}