public void Index(string contentPath, List <string> foldersToExclude)
{
if (foldersToExclude == null)
{
throw new ArgumentException("Must not have null collection of folders", "foldersToExclude");
}
var indexDirectory = new SimpleFSDirectory(new DirectoryInfo(_configuration.IndexPath));
Log(string.Format("Begining to index {0}. Index location: {1}", contentPath, indexDirectory.Directory.FullName));
var stopWatch = new Stopwatch();
stopWatch.Start();
var analyzer = AnalyzerBuilder.CreateAnalyzer();
using (var writer = new IndexWriter(indexDirectory, analyzer, true, IndexWriter.MaxFieldLength.UNLIMITED))
{
IndexDirectory(writer, new DirectoryInfo(contentPath), foldersToExclude);
}
stopWatch.Stop();
Log(string.Format("Indexed {0:N0} files in {1:00}:{2:00}.{3:00}", _fileCount, stopWatch.Elapsed.Hours, stopWatch.Elapsed.Minutes, stopWatch.Elapsed.Seconds));
}
public ReadOnlyHitCollection Search(string searchExpression, Model.Filter filter = null, string sortField = null, int page = 1, int hitsPerPage = 10)
{
if (string.IsNullOrEmpty(searchExpression))
{
throw new SearchException("Must have searchExpression");
}
string defaultFieldName = Fields.Content;
var analyzer = AnalyzerBuilder.CreateAnalyzer();
Query query;
try
{
query = new QueryParser(Version.LUCENE_30, defaultFieldName, analyzer).Parse(searchExpression.ToLower());
}
catch (ParseException ex)
{
throw new SearchException(string.Format("Sorry, '{0}' isn't something we can search for so far.", searchExpression), ex);
}
var indexDirectory = new SimpleFSDirectory(new DirectoryInfo(_configuration.IndexPath));
List <Hit> onePageOfHits;
int totalHits;
using (var reader = IndexReader.Open(indexDirectory, true))
{
// Get one page of hits
var hits = new List <Hit>();
var searcher = new IndexSearcher(reader);
var termsFilter = filter != null && !string.IsNullOrEmpty(filter.Field)
? new FieldCacheTermsFilter(filter.Field, filter.Terms.ToArray())
: null;
var sort = !string.IsNullOrEmpty(sortField)
? new Sort(new SortField(sortField, SortField.STRING))
: Sort.RELEVANCE;
ScoreDoc[] scoreDocs = searcher.Search(query, termsFilter, MaxNumberOfHits, sort).ScoreDocs;
totalHits = scoreDocs.Length;
foreach (var scoreDoc in scoreDocs)
{
int docId = scoreDoc.Doc;
string filePath = searcher.Doc(docId).Get(Fields.Path);
string language = searcher.Doc(docId).Get(Fields.Language);
var hit = new Hit(docId, _configuration.ContentRootPath, filePath, scoreDoc.Score, language);
hits.Add(hit);
}
onePageOfHits = hits.GetPage(page, hitsPerPage).ToList();
// Get offsets and higlights on the page we are going to return
foreach (var hit in onePageOfHits)
{
var primitiveQuery = query.Rewrite(reader);
var terms = new HashSet <Term>();
primitiveQuery.ExtractTerms(terms);
string searchField = string.Empty;
if (terms.Count == 0)
{
// There can be all kinds of queires
var prefixQuery = query as PrefixQuery;
if (prefixQuery != null)
{
searchField = prefixQuery.Prefix.Field;
primitiveQuery = prefixQuery;
}
}
else
{
// TODO: There can be multiple term fields, like code: and method:
searchField = terms.First().Field;
}
var termFreqVector = reader.GetTermFreqVector(hit.DocId, Fields.Content);
var termPositionVector = termFreqVector as TermPositionVector;
if (termFreqVector == null || termPositionVector == null)
{
throw new ArgumentException("Must have term frequencies and positions vectors");
}
// No offsets for prefix and other non-term based queries
const int maxOffsetNumber = 10;
foreach (var term in terms)
{
int termIndex = termFreqVector.IndexOf(term.Text); // Meaning get me this term, not text yet.
if (termIndex != -1)
{
foreach (var offset in termPositionVector.GetOffsets(termIndex))
{
if (hit.Offsets.Count < maxOffsetNumber)
{
hit.Offsets.Add(new Offset
{
StartOffset = offset.StartOffset,
EndOffset = offset.EndOffset
});
}
}
}
}
// Highlighter from contrib package
var tokenStream = TokenSources.GetTokenStream(termPositionVector);
var scorer = new QueryScorer(primitiveQuery, searchField);
var fragmenter = new SimpleSpanFragmenter(scorer);
var formatter = new SimpleHTMLFormatter("<kbd>", "</kbd>");
var highlighter = new Highlighter(formatter, scorer)
{
TextFragmenter = fragmenter
};
string text;
using (var sr = new StreamReader(hit.FilePath))
{
text = sr.ReadToEnd();
}
string bestFragment = highlighter.GetBestFragment(tokenStream, text);
if (!string.IsNullOrEmpty(bestFragment))
{
hit.BestFragment = EscapeHtmlMarkup(bestFragment);
}
}
}
return(new ReadOnlyHitCollection(onePageOfHits, totalHits));
}
