public async Task <List <string> > GetPDFasHtml(Uri uri)
{
PdfDocument document = PdfReader.Open(await _azureService.DownloadToStream(uri));
var pages = new List <string>();
try
{
for (int i = 0; i < document.PageCount; i++)
{
StringWriter stringWriter = new StringWriter();
var pageText = PdfTextExtractor.GetText(document.Pages[i]);
pageText = pageText.Replace("\r\n", "<br/>");
pageText = pageText.Replace(" ", "<p></p>");
using (HtmlTextWriter writer = new HtmlTextWriter(stringWriter))
{
string classValue = "pdfHtml";
writer.AddAttribute(HtmlTextWriterAttribute.Class, classValue);
writer.RenderBeginTag(HtmlTextWriterTag.Div); // Begin #1
writer.Write(pageText);
writer.RenderEndTag(); // End #1
}
pages.Add(stringWriter.ToString());
}
// Return the result.
return(pages);
}
catch (Exception e)
{
_logger.Error(e.Message);
return(pages);
}
}
