/// <summary>
/// Parse an document
/// </summary>
/// <param name="document">Document to parse</param>
/// <returns>Chunk of text</returns>
public virtual IEnumerable<DocumentChunk> Parse(DocumentIdentity document)
{
List<DocumentChunk> result = new List<DocumentChunk>();
Contract.Assert(document != null);
Contract.Result<IEnumerable<DocumentChunk>>();
Contract.Ensures(result != null, "Empty collection can be returned but not null reference");
using (StreamReader reader = File.OpenText(document.FilePath))
{
DocumentChunk chunk = new DocumentChunk();
StringBuilder text = new StringBuilder();
while (!reader.EndOfStream)
{
string line = reader.ReadLine();
bool isEmptyParagraph = line.Length == 0;
if (isEmptyParagraph && text.Length > 0)
{
chunk.Metadata = "Content";
chunk.Text = text.ToString();
result.Add(chunk);
text.Clear();
chunk = new DocumentChunk();
}
else
{
text.Append(line);
}
}
// Post-Loop action : add remaining chunk
if (text.Length > 0)
{
chunk.Text = text.ToString();
chunk.Metadata = "Content";
result.Add(chunk);
}
}
return result;
}
/// <summary>
/// Parse an PDF document and extract the text content
/// </summary>
/// <param name="document">Document to analyze</param>
/// <returns>One text chunk per page</returns>
public IEnumerable<DocumentChunk> Parse(DocumentIdentity document)
{
IList<DocumentChunk> result = new List<DocumentChunk>();
Contract.Assert(document != null);
Contract.Result<IEnumerable<DocumentChunk>>();
Contract.Ensures(result != null, "Empty collection can be returned but not null reference");
using (PdfReader reader = new PdfReader(File.Open(document.FilePath, FileMode.Open)))
{
for (int i = 1; i <= reader.NumberOfPages; i++)
{
DocumentChunk chunk = new DocumentChunk();
chunk.Text = PdfTextExtractor.GetTextFromPage(reader, i);
chunk.Metadata = "Content";
if (!String.IsNullOrEmpty(chunk.Text))
{
result.Add(chunk);
}
}
}
return result;
}
