public static IIndexDocument FromFile(string file, IIndexDataSource parent)
{
string s = file.ToLower();
if (s.EndsWith(".pdf"))
{
// PDF
return(new IndexPDFDocument(file, parent));
}
else
if (s.EndsWith(".txt"))
{
return(new IndexedTextFile(file, parent));
}
else
if ((s.EndsWith(".html")) || (s.EndsWith(".html")))
{
using (FileStream fs = File.OpenRead(file))
{
return(WebDataSource.FromHtml(fs, file, parent.Name));
}
}
return(null);
}
/// Add document from BLOB
public virtual void AddRecord(string name, Stream stream, string fields, ConcurrentQueue <IIndexDocument> queue)
{
bool isText = false;
IIndexDocument doc = null;
AddRecordBase(name, fields);
if ((indexType == IndexType.File) || (indexType != IndexType.Blob))
{
throw new InvalidDataException("Adding record of wrong IndexType");
}
BinaryReader reader = new BinaryReader(stream);
byte[] buff = new byte[4000];
reader.Read(buff, 0, 4000);
String det = Encoding.UTF8.GetString(buff, 0, buff.Length);
stream.Seek(0, SeekOrigin.Begin);
reader.Dispose();
// detect type
if ((buff[0] == '%') && (buff[1] == 'P') && (buff[2] == 'D') && (buff[3] == 'F'))
{
DocumentsDataSource.IndexPDFDocument pdf = new DocumentsDataSource.IndexPDFDocument(name, stream, this);
if (fields != null)
{
pdf.headers = () => { return(fields); }
}
;
doc = pdf;
}
else
if (det.Contains("<html"))
{
IndexPagedTextFile file = WebDataSource.FromHtml(stream, name, Name);
if (fields != null)
{
file.SetHeaders(fields);
}
}
else
{
// detect charset
Ude.CharsetDetector detector = new Ude.CharsetDetector();
detector.Feed(buff, 0, buff.Length);
detector.DataEnd();
if (detector.Charset != null)
{
Encoding enc = Portable.Text.Encoding.GetEncoding(detector.Charset);
using (StreamReader sreader = new StreamReader(stream, enc, false)) {
doc = new IndexPagedTextFile("", sreader.ReadToEnd(), fields != null ? fields : "");
}
}
}
if (doc != null)
{
Enqueue(queue, doc);
}
}