public void Batch()
{
List<Document> list = new List<Document>();
for (int i = 0; i < 5; i++)
{
Document task = new Document();
task.Id = System.Guid.NewGuid().ToString();
task.Text = _message;
list.Add(task);
}
int res = _session.QueueBatchOfDocuments(list, _configId);
Assert.IsTrue((res == 200 || res == 202), "RESULT: " + res);
System.Threading.Thread.Sleep(5000);
IList<DocAnalyticData> data = _session.GetProcessedDocuments(_configId);
Assert.IsNotNull(data);
}
public static List<TweetModel> AddScore(List<TweetModel> tweetList)
{
// Creates JSON serializer instance
ISerializer serializer = new JsonSerializer();
// Initializes new session with the serializer object and the keys.
using (Session session = Session.CreateSession(consumerKey, consumerSecret, serializer))
{
// Error callback handler. This event will occur in case of server-side error
session.Error += new Session.ErrorHandler(delegate(object sender, ResponseErrorEventArgs ea)
{
Console.WriteLine(string.Format("{0}: {1}", (int)ea.Status, ea.Message));
});
//Obtaining subscription object to get user limits applicable on server side
Subscription subscription = session.GetSubscription();
Dictionary<string, Semantria.Com.TaskStatus> docsTracker = new Dictionary<string, Semantria.Com.TaskStatus>();
List<Document> outgoingBatch = new List<Document>(subscription.BasicSettings.BatchLimit);
foreach (var item in tweetList)
{
if (docsTracker.ContainsKey(item.TweetIdStr))
continue;
docsTracker.Add(item.TweetIdStr, Semantria.Com.TaskStatus.QUEUED);
Document doc = new Document()
{
Id = item.TweetIdStr,
Text = item.Text
};
outgoingBatch.Add(doc);
if (outgoingBatch.Count == subscription.BasicSettings.BatchLimit)
{
break;
}
}
if (outgoingBatch.Count > 0)
{
// Queues batch of documents for processing on Semantria service
if (session.QueueBatchOfDocuments(outgoingBatch) != -1)
{
Console.WriteLine(string.Format("{0} documents queued successfully.", outgoingBatch.Count));
}
}
List<DocAnalyticData> results = new List<DocAnalyticData>();
int count = 0;
while (docsTracker.Any(item => item.Value == Semantria.Com.TaskStatus.QUEUED) && count <= 200)
{
count++;
System.Threading.Thread.Sleep(500);
// Requests processed results from Semantria service
Console.WriteLine("Retrieving your processed results...");
IList<DocAnalyticData> incomingBatch = session.GetProcessedDocuments();
foreach (DocAnalyticData item in incomingBatch)
{
if (docsTracker.ContainsKey(item.Id))
{
docsTracker[item.Id] = item.Status;
results.Add(item);
}
}
}
foreach (DocAnalyticData data in results)
{
var currentTweet = tweetList.First(res=>res.TweetIdStr == data.Id);
float score = 0;
// Printing of document entities
if (data.Phrases != null && data.Phrases.Count > 0)
{
foreach (var entity in data.Phrases)
{
score += entity.SentimentScore;
}
score = score / data.Phrases.Count();
}
currentTweet.Score = score;
}
}
return tweetList;
}
static void Main(string[] args)
{
// Use correct Semantria API credentias here
string consumerKey = "";
string consumerSecret = "";
// A dictionary that keeps IDs of sent documents and their statuses. It's required to make sure that we get correct documents from the API.
Dictionary<string, TaskStatus> docsTracker = new Dictionary<string, TaskStatus>(4);
List<string> initialTexts = new List<string>();
Console.WriteLine("Semantria Detailed processing mode demo.");
string path = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "source.txt");
if (!File.Exists(path))
{
Console.WriteLine("Source file isn't available.");
return;
}
//Reads collection from the source file
Console.WriteLine("Reading dataset from file...");
using (StreamReader stream = new StreamReader(path))
{
while (!stream.EndOfStream)
{
string line = stream.ReadLine();
if (string.IsNullOrEmpty(line) || line.Length < 3)
continue;
initialTexts.Add(line);
}
}
// Creates JSON serializer instance
ISerializer serializer = new JsonSerializer();
// Initializes new session with the serializer object and the keys.
using (Session session = Session.CreateSession(consumerKey, consumerSecret, serializer))
{
// Error callback handler. This event will occur in case of server-side error
session.Error += new Session.ErrorHandler(delegate(object sender, ResponseErrorEventArgs ea)
{
Console.WriteLine(string.Format("{0}: {1}", (int)ea.Status, ea.Message));
});
//Obtaining subscription object to get user limits applicable on server side
Subscription subscription = session.GetSubscription();
List<Document> outgoingBatch = new List<Document>(subscription.BasicSettings.IncomingBatchLimit);
IEnumerator<string> iterrator = initialTexts.GetEnumerator();
while (iterrator.MoveNext())
{
string docId = Guid.NewGuid().ToString();
Document doc = new Document()
{
Id = docId,
Text = iterrator.Current
};
outgoingBatch.Add(doc);
docsTracker.Add(docId, TaskStatus.QUEUED);
if (outgoingBatch.Count == subscription.BasicSettings.IncomingBatchLimit)
{
// Queues batch of documents for processing on Semantria service
if (session.QueueBatchOfDocuments(outgoingBatch) != -1)
{
Console.WriteLine(string.Format("{0} documents queued successfully.", outgoingBatch.Count));
outgoingBatch.Clear();
}
}
}
if (outgoingBatch.Count > 0)
{
// Queues batch of documents for processing on Semantria service
if (session.QueueBatchOfDocuments(outgoingBatch) != -1)
{
Console.WriteLine(string.Format("{0} documents queued successfully.", outgoingBatch.Count));
}
}
Console.WriteLine();
// As Semantria isn't real-time solution you need to wait some time before getting of the processed results
// In real application here can be implemented two separate jobs, one for queuing of source data another one for retreiving
// Wait ten seconds while Semantria process queued document
List<DocAnalyticData> results = new List<DocAnalyticData>();
while (docsTracker.Any(item => item.Value == TaskStatus.QUEUED))
{
System.Threading.Thread.Sleep(500);
// Requests processed results from Semantria service
Console.WriteLine("Retrieving your processed results...");
IList<DocAnalyticData> incomingBatch = session.GetProcessedDocuments();
foreach (DocAnalyticData item in incomingBatch)
{
if (docsTracker.ContainsKey(item.Id))
{
docsTracker[item.Id] = item.Status;
results.Add(item);
}
}
}
Console.WriteLine();
foreach (DocAnalyticData data in results)
{
// Printing of document sentiment score
Console.WriteLine(string.Format("Document {0}. Sentiment score: {1}", data.Id, data.SentimentScore));
// Printing of intentions
if (data.Topics != null && data.Topics.Count > 0)
{
Console.WriteLine("Document topics:");
foreach (Topic topic in data.Topics)
Console.WriteLine(string.Format("\t{0} (type: {1}) (strength: {2})", topic.Title, topic.Type, topic.SentimentScore));
}
// Printing of intentions
if (data.AutoCategories != null && data.AutoCategories.Count > 0)
{
Console.WriteLine("Document categories:");
foreach (Topic category in data.AutoCategories)
Console.WriteLine(string.Format("\t{0} (strength: {1})", category.Title, category.StrengthScore));
}
// Printing of document themes
if (data.Themes != null && data.Themes.Count > 0)
{
Console.WriteLine("Document themes:");
foreach (DocTheme theme in data.Themes)
Console.WriteLine(string.Format("\t{0} (sentiment: {1})", theme.Title, theme.SentimentScore));
}
// Printing of document entities
if (data.Entities != null && data.Entities.Count > 0)
{
Console.WriteLine("Entities:");
foreach (DocEntity entity in data.Entities)
Console.WriteLine(string.Format("\t{0} : {1} (sentiment: {2})", entity.Title, entity.EntityType, entity.SentimentScore));
}
Console.WriteLine();
}
}
Console.ReadKey(false);
}
public void Document()
{
Document task = new Document();
task.Id = _docId;
task.Text = _message;
int res = _session.QueueDocument(task, _configId);
Assert.IsTrue((res == 200 || res == 202), "RESULT: " + res);
DocAnalyticData data = _session.GetDocument(_docId, _configId);
Assert.IsNotNull(data);
int r = _session.CancelDocument(_docId, _configId);
Assert.IsTrue((r == 200 || r == 202 || r == 404), "RESULT: " + res);
}
private IList<DocAnalyticData> _getSemanticDataForTweets(List<ITweet> tweets)
{
ISerializer serializer = new JsonSerializer();
IList<DocAnalyticData> results = new List<DocAnalyticData>();
// Initializes new session with the serializer weobject and the keys.
using (Session session = Session.CreateSession(ConsumerKey, ConsumerSecret, serializer))
{
// Error callback handler. This event will be rised in case of server-side error
session.Error += new Session.ErrorHandler(delegate(object sender, ResponseErrorEventArgs ea)
{
Console.WriteLine(string.Format("{0}: {1}",
(int)ea.Status, ea.Message));
});
foreach (ITweet tweet in tweets)
{
string docId = Guid.NewGuid().ToString();
// Creates a sample document which need to be processed on Semantria
Document doc = new Document()
{
// Unique document ID
Id = string.Format("{0}", docId),
// Source text which need to be processed
Text = tweet.Text,
Tag = tweet.IdStr
};
// Queues document for processing on Semantria service
if (session.QueueDocument(doc) != -1)
Console.WriteLine(string.Format("\"{0}\" document queued successfully.", docId));
}
// Console.WriteLine();
// As Semantria isn't real-time solution you need to wait some time before getting of the processed results
// In real application here can be implemented two separate jobs, one for queuing of source data another one for retreiving
// Wait ten seconds while Semantria process queued document
int resCount = tweets.Count;
while (resCount > 0)
{
System.Threading.Thread.Sleep(2000);
// Requests processed results from Semantria service
Console.WriteLine("Retrieving your processed results...");
((List<DocAnalyticData>)results).AddRange(session.GetProcessedDocuments());
resCount -= results.Count;
}
}
return results;
}
static void Main(string[] args)
{
// Use correct Semantria API credentias here
string consumerKey = string.Empty;
string consumerSecret = string.Empty;
// null - send every single document separately
// false - send uniqueJobIdCount batches
// true - send all documents in single batch
bool? dataSendingMode = true;
int uniqueJobIdCount = 4;
Dictionary<string, int> jobIds = new Dictionary<string, int>(uniqueJobIdCount);
Dictionary<string, List<Document>> documents = new Dictionary<string, List<Document>>();
Console.WriteLine("Semantria JobId feature demo.");
string path = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "source.txt");
if (!File.Exists(path))
{
Console.WriteLine("Source file isn't available.");
return;
}
//Generates N unique jobId values
for (int index = 0; index < uniqueJobIdCount; index++)
{
string id = Guid.NewGuid().ToString();
jobIds.Add(id, 0);
documents.Add(id, new List<Document>());
}
//Reads documents from the source file
Console.WriteLine("Reading documents from file...");
Console.WriteLine();
using (StreamReader stream = new StreamReader(path))
{
Random rnd = new Random();
while (!stream.EndOfStream)
{
string line = stream.ReadLine();
if (string.IsNullOrEmpty(line) || line.Length < 3)
continue;
string jobId = jobIds.ElementAt(rnd.Next(uniqueJobIdCount)).Key;
jobIds[jobId]++;
Document doc = new Document()
{
Id = Guid.NewGuid().ToString(),
Text = line,
JobId = jobId
};
documents[jobId].Add(doc);
}
}
// Creates JSON serializer instance
ISerializer serializer = new JsonSerializer();
// Initializes new session with the serializer object and the keys.
using (Session session = Session.CreateSession(consumerKey, consumerSecret, serializer))
{
// Error callback handler. This event will occur in case of server-side error
session.Error += new Session.ErrorHandler(delegate(object sender, ResponseErrorEventArgs ea)
{
if ((int)ea.Status < 500)
{
Console.WriteLine(string.Format("{0}: {1}", (int)ea.Status, ea.Message));
}
else
{
Console.WriteLine(string.Format("{0}: {1}", (int)ea.Status, "Unhandled server error happened."));
}
});
if (dataSendingMode == null)
{
foreach (KeyValuePair<string, List<Document>> pair in documents)
{
foreach (Document doc in pair.Value)
{
// Queues document for processing on Semantria service
session.QueueDocument(doc);
}
Console.WriteLine("{0} documents queued for {1} job ID", pair.Value.Count, pair.Key);
}
}
else if (!dataSendingMode.Value)
{
foreach (KeyValuePair<string, List<Document>> pair in documents)
{
// Queues batch of documents for processing on Semantria service
if (session.QueueBatchOfDocuments(pair.Value) != -1)
{
Console.WriteLine("{0} documents queued for {1} job ID", pair.Value.Count, pair.Key);
}
}
}
else
{
List<Document> aBatch = new List<Document>();
foreach (KeyValuePair<string, List<Document>> pair in documents)
{
aBatch.AddRange(pair.Value);
}
if (session.QueueBatchOfDocuments(aBatch) != -1)
{
Console.WriteLine("{0} documents queued in single batch", aBatch.Count);
}
}
Console.WriteLine();
Dictionary<string, int> aCopy = new Dictionary<string, int>(jobIds);
foreach (KeyValuePair<string, int> pair in aCopy)
{
int count = 0;
while (jobIds[pair.Key] > 0)
{
// Waits half of second while Semantria process queued document
System.Threading.Thread.Sleep(500);
// Requests processed results from Semantria service
List<DocAnalyticData> results = new List<DocAnalyticData>(pair.Value);
results.AddRange(session.GetProcessedDocumentsByJobId(pair.Key));
jobIds[pair.Key] -= results.Count;
count += results.Count;
}
Console.WriteLine("{0} documents received for {1} Job ID.", count, pair.Key);
}
}
Console.WriteLine();
Console.WriteLine("Done!");
Console.ReadKey(false);
}
static void Main(string[] args)
{
// Use correct Semantria API credentias here
string consumerKey = "";
string consumerSecret = "";
// Initial texts for processing
List<string> initialTexts = new List<string>() {
@"Lisa - there's 2 Skinny cow coupons available $5 skinny cow ice cream coupons on special k boxes and Printable FPC from facebook - a teeny tiny cup of ice cream. I printed off 2 (1 from my account and 1 from dh's). I couldn't find them instore and i'm not going to walmart before the 19th. Oh well sounds like i'm not missing much ...lol",
@"In Lake Louise - a guided walk for the family with Great Divide Nature Tours rent a canoe on Lake Louise or Moraine Lake go for a hike to the Lake Agnes Tea House. In between Lake Louise and Banff - visit Marble Canyon or Johnson Canyon or both for family friendly short walks. In Banff a picnic at Johnson Lake rent a boat at Lake Minnewanka hike up Tunnel Mountain walk to the Bow Falls and the Fairmont Banff Springs Hotel visit the Banff Park Museum. The ""must-do"" in Banff is a visit to the Banff Gondola and some time spent on Banff Avenue - think candy shops and ice cream.",
@"On this day in 1786 - In New York City commercial ice cream was manufactured for the first time."
};
System.Console.WriteLine("Semantria service demo.");
// Creates JSON serializer instance
ISerializer serializer = new XmlSerializer();
// Initializes new session with the serializer object and the keys.
using (Session session = Session.CreateSession(consumerKey, consumerSecret, serializer))
{
// Error callback handler. This event will be rised in case of server-side error
session.Error += new Session.ErrorHandler(delegate(object sender, ResponseEventArgs ea)
{
System.Console.WriteLine(string.Format("{0}: {1}", (int)ea.Status, ea.Message));
});
foreach (string text in initialTexts)
{
string docId = Guid.NewGuid().ToString();
// Creates a sample document which need to be processed on Semantria
Document doc = new Document()
{
// Unique document ID
Id = string.Format("{0}", docId),
// Source text which need to be processed
Text = text
};
// Queues document for processing on Semantria service
if (session.QueueDocument(doc) != -1)
System.Console.WriteLine(string.Format("\"{0}\" document queued successfully.", docId));
}
System.Console.WriteLine();
// As Semantria isn't real-time solution you need to wait some time before getting of the processed results
// In real application here can be implemented two separate jobs, one for queuing of source data another one for retreiving
// Wait ten seconds while Semantria process queued document
int resCount = initialTexts.Count;
IList<DocAnalyticData> results = new List<DocAnalyticData>();
while (resCount > 0)
{
System.Threading.Thread.Sleep(2000);
// Requests processed results from Semantria service
System.Console.WriteLine("Retrieving your processed results...");
((List<DocAnalyticData>)results).AddRange(session.GetProcessedDocuments());
resCount -= results.Count;
}
foreach (DocAnalyticData data in results)
{
// Printing of document sentiment score
System.Console.WriteLine(string.Format("Document {0}. Sentiment score: {1}", data.Id, data.SentimentScore));
// Printing of document themes
if (data.Themes != null && data.Themes.Count > 0)
{
System.Console.WriteLine("Document themes:");
foreach (DocTheme theme in data.Themes)
System.Console.WriteLine(string.Format("\t{0} (sentiment: {1})", theme.Title, theme.SentimentScore));
}
// Printing of document entities
if (data.Entities != null && data.Entities.Count > 0)
{
System.Console.WriteLine("Entities:");
foreach (DocEntity entity in data.Entities)
System.Console.WriteLine(string.Format("\t{0} : {1} (sentiment: {2})", entity.Title, entity.EntityType, entity.SentimentScore));
}
System.Console.WriteLine();
}
}
System.Console.ReadKey(false);
}
public void CreateDocument()
{
Document task = new Document();
task.Id = _id;
task.Text = _message;
int res = _session.QueueDocument(task);
Assert.IsTrue((res == 200 || res == 202), "RESULT: " + res);
}
public void CreateBatch()
{
List<Document> list = new List<Document>();
for (int i = 0; i < 5; i++)
{
Document task = new Document();
task.Id = System.Guid.NewGuid().ToString();
task.Text = _message;
list.Add(task);
}
int res = _session.QueueBatchOfDocuments(list);
Assert.IsTrue((res == 200 || res == 202), "RESULT: " + res);
}
