public static string getDocumentContent(string path)
{
var extractor = new TikaOnDotNet.TextExtraction.TextExtractor();
var something = extractor.Extract(path);
return(extractor.Extract(path).Text);
}
public string ExtractFromDocument(string filePath)
{
try
{
var extractor = new TikaOnDotNet.TextExtraction.TextExtractor();
var text = extractor.Extract(filePath);
return(text.Text);
}
catch (Exception) { return(null); }
}
public static void AddDocument(string path)
{
var extractor = new TikaOnDotNet.TextExtraction.TextExtractor();
var result = extractor.Extract(path);
string description = result.Metadata?.FirstOrDefault(x => x.Key.ToLower() == "description").Value; //pobiera wartości z metadanych
string author = result.Metadata?.FirstOrDefault(x => x.Key.ToLower() == "author").Value;
string title = result.Metadata?.FirstOrDefault(x => x.Key.ToLower() == "title").Value;
var document = new Models.DocumentModel()
{
FileName = Path.GetFileName(path),
Author = (author == null) ? "": author,
Description = (description == null) ? "" : description,
Title = (title == null) ? "" : title
};
using (var db = new DocumentDbContext())
{
db.Documents.Add(document);
db.SaveChanges();
}
Services.LuceneSearch.AddUpdateLuceneIndex(document);
}
public static async Task <HttpResponseMessage> Run([HttpTrigger(AuthorizationLevel.Anonymous, "get", "post", Route = null)] HttpRequestMessage req, TraceWriter log)
{
string message = "";
List <string> keyPhrases = new List <string>();
string description = "";
// Get request body
dynamic data = await req.Content.ReadAsAsync <object>();
string id = data.ID;
log.Info("Document id: " + id);
message += "Document id: " + id;
OfficeDevPnP.Core.AuthenticationManager authManager = new OfficeDevPnP.Core.AuthenticationManager();
try
{
// Connects to SharePoint online site
using (var ctx = authManager.GetSharePointOnlineAuthenticatedContextTenant(siteUrl, userName, password))
{
// List Name input
// Retrieves list object using title
List list = ctx.Site.RootWeb.GetListByTitle(listName);
if (list != null)
{
// Returns required result
ListItem li = list.GetItemById(id);
ctx.Load(li);
ctx.Load(li.File);
ctx.ExecuteQuery();
ctx.ExecuteQuery();
// We CAN extract text out of most documents with the library, but for this demo I'm limiting our options to these 2 that I know to be working :)
if (li.File.Name.IndexOf(".pdf") >= 0 || li.File.Name.IndexOf(".doc") >= 0)
{
li.File.OpenBinaryStream();
ctx.Load(li.File);
ctx.ExecuteQuery();
log.Info("It was a valid file! Continuing into handling...");
try
{
log.Info("Got a file! Name: " + li.File.Name);
var fileRef = li.File.ServerRelativeUrl;
var fileInfo = Microsoft.SharePoint.Client.File.OpenBinaryDirect(ctx, fileRef);
fileInfo = Microsoft.SharePoint.Client.File.OpenBinaryDirect(ctx, fileRef);
using (var ms = new MemoryStream())
{
log.Info("Extracting text..");
fileInfo.Stream.CopyTo(ms);
byte[] fileContents = ms.ToArray();
var extractor = new TikaOnDotNet.TextExtraction.TextExtractor();
var extractionResult = extractor.Extract(fileContents);
string text = extractionResult.Text;
List <MultiLanguageInput> analyzable = FormatAnalyzableText(ref text);
log.Info("Formed altogether " + analyzable.Count + " sentences to analyze!");
int snippetEnd = 500 < text.Length ? 500 : text.Length;
log.Info("Extracted text! First few rows here.. \r\n " + text.Substring(0, snippetEnd));
RunTextAnalysis(ref keyPhrases, analyzable, log);
}
log.Info("Found " + keyPhrases.Count + " key phrases! First 20 are here: ");
foreach (var kp in keyPhrases.Take(20))
{
log.Info(kp);
}
try
{
log.Info("Saving to SharePoint..");
TextInfo ti = new CultureInfo("en-US", false).TextInfo;
li["Title"] = ti.ToTitleCase(li.File.Name);
// then write the most important keyphrases back
foreach (var s in keyPhrases.Take(keywordCount))
{
description += s + "\r\n";
}
li.Update();
try
{
ctx.Load(list.Fields);
ctx.ExecuteQuery();
log.Info("Updating Managed Metadata...");
var fieldnames = new string[] { "Keywords" };
var field = list.GetFields(fieldnames).First();
// setting managed metadata
log.Info("Updating keywords to taxonomy! Taking: " + keywordCount);
UpdateManagedMetadata(keyPhrases.Take(keywordCount).ToArray(), log, ctx, li, field, wantedGuid);
ctx.ExecuteQuery();
}
catch (Exception ex)
{
log.Error(ex.Message);
}
}
catch (Exception ex)
{
log.Error(ex.Message);
}
}
catch (Exception ex)
{
log.Error(ex.Message);
return(req.CreateResponse(HttpStatusCode.InternalServerError, ex.Message));
}
}
else
{
return(req.CreateResponse(HttpStatusCode.OK, "File was not pdf or doc"));
}
return(req.CreateResponse(HttpStatusCode.OK, list.Id));
}
else
{
log.Info("List is not available on the site");
return(req.CreateResponse(HttpStatusCode.NotFound, "List is not available on the site"));
}
}
}
catch (Exception ex)
{
log.Info("Error Message: " + ex.Message);
message += "Error Message: " + ex.Message;
}
log.Info("");
var returnable = JsonConvert.SerializeObject(keyPhrases);
return(keyPhrases.Count <= 0
? req.CreateResponse(HttpStatusCode.BadRequest, "Couldn't analyze file. Please verify the POST payload!")
: req.CreateResponse(HttpStatusCode.OK, returnable));
}
