public void Index_Then_Search()
{
var search = new DocumentIndex();
var docs = TestData.TestCorpus().Select(t => XDocument.Parse(t)).ToList().AsQueryable();
search.IndexDocuments(docs, d => d.Root.Attribute("id").Value);
var matches = search.SearchInternal("love time");
Assert.That(matches.All(m => m.Value > 0));
Assert.That(matches.First().Key, Is.EqualTo("3"));
}
public void Index_Then_Export_ReturnsExpectedXml()
{
var search = new DocumentIndex();
var docs = TestData.TestCorpus().Select(t => XDocument.Parse(t)).ToList().AsQueryable();
search.IndexDocuments(docs, d => d.Root.Attribute("id").Value);
var xml = search.ExportAsXml();
Assert.That(xml.Root.Name.LocalName, Is.EqualTo("index"));
Assert.That(xml.Root.Attribute("doc-count").Value, Is.EqualTo(docs.Count().ToString()));
Assert.That(xml.Root.Elements().All(e => e.Name.LocalName == "term"));
xml.WriteTo(XmlWriter.Create(Console.Out, new XmlWriterSettings() { Indent = true }));
}
public void Export_Then_Import_RestoresState()
{
var index1 = new DocumentIndex();
var docs = TestData.TestCorpus().Select(t => XDocument.Parse(t)).ToList().AsQueryable();
index1.IndexDocuments(docs, d => d.Root.Attribute("id").Value);
var xml = index1.ExportAsXml();
var index2 = new DocumentIndex(index1.Tokeniser);
index2.ImportXml(xml);
var xml2 = index2.ExportAsXml();
Assert.That(xml.ToString(), Is.EqualTo(xml2.ToString()));
}
public void CreateVectorExtractor()
{
var index = new DocumentIndex();
var docs = TestData.TestCorpus().Select(t => XDocument.Parse(t)).ToList().AsQueryable();
int id = 0;
var blocks = docs.SelectMany(d => new Corpus(index.Tokeniser.Tokenise(d.Root.Value)).Blocks.ToList()).ToList();
var tdocs = blocks
.Select(b => new TokenisedTextDocument((id++).ToString(), b))
.ToList();
index.IndexDocuments(tdocs);
var ve = index.CreateVectorExtractor();
var vect = ve.ExtractColumnVector(index.Tokeniser.Tokenise("love time fortune"));
Console.WriteLine(vect);
}
public void CreateVectorExtractor_LargeCorpus()
{
var index = new DocumentIndex();
using (var corpusStream = GetResource("shakespeare.txt"))
{
var corpus = new Corpus(corpusStream.Tokenise());
int id = 0;
index.IndexDocuments(corpus.Blocks.Select(b => new TokenisedTextDocument((id++).ToString(), b)));
}
var ve = index.CreateVectorExtractor(1024);
var vect = ve.ExtractColumnVector(index.Tokeniser.Tokenise("love time fortune"));
Console.WriteLine(vect);
}