public IEnumerable <TagData> GetTags(IEnumerable <TaggableContent> taggableContents)
{
try
{
var credentials = new AzureCognitiveServicesTextAnalyticsClientCredentials(Key);
var client = new TextAnalyticsClient(credentials)
{
Endpoint = Endpoint
};
//All of the Text Analytics API endpoints accept raw text data.The current limit is 5,120 characters for each document;
//Limit Value
//Maximum size of a single document 5,120 characters as measured by StringInfo.LengthInTextElements.
//Maximum size of entire request 1 MB
//Maximum number of documents in a request 1,000 documents
var documents = new List <MultiLanguageInput>();
int id = 0;
foreach (StringContent taggableContent in taggableContents)
{
var text = (taggableContent != null) ? taggableContent.Content : null;
if (string.IsNullOrEmpty(text))
{
continue;
}
if (text.Length > 5100)
{
text = text.Substring(0, 5100);
}
//just skip it if description is 500 characters or less, not enough content to process
//if (text.Length > 500)
documents.Add(new MultiLanguageInput()
{
Id = id.ToString(), Text = text
});
id++;
}
if (!documents.Any())
{
return(new List <TagData>());
}
var inputDocuments = new MultiLanguageBatchInput(documents);
var results = client.EntitiesBatch(inputDocuments);
if (results.Errors.Count > 0)
{
var messageBus = MessageBus;
if (messageBus != null)
{
foreach (var error in results.Errors)
{
messageBus.SendMessage(new Message()
{
Body = string.Format("Azure Cogntive Services Text Analytics Error: {0} Message: {1}", error.Id, error.Message),
Level = MessageLevel.Error
});
}
}
}
var tagData = new List <TagData>();
var entityNames = new List <string>();
foreach (var document in results.Documents)
{
foreach (var entity in document.Entities)
{
if (!Exclude.Contains(entity.Name.ToLower()) &&
(entity.Type == "Organization" || entity.Type == "Location" ||
(entity.Type == "Other" && IncludeOther.Contains(entity.Name.ToLower())) ||
(entity.Type == "Quantity" && IncludeQuantity.Contains(entity.Name))))
{
var name = NormalizeEntityName(entity.Name);
if (entity.SubType != null)
{
entityNames.Add(string.Format("{0} - {1} - {2}", entity.Type, entity.SubType, name));
}
else
{
entityNames.Add(string.Format("{0} - {1}", entity.Type, name));
}
}
}
}
return(entityNames.Distinct().Select(x => new TagData()
{
TagName = x
}));
}
catch (Exception ex)
{
var messageBus = MessageBus;
if (messageBus != null)
{
messageBus.SendMessage(new Message()
{
Body = string.Format("Azure Cogntive Services Text Analytics Error: {0}", ex.Message),
Level = MessageLevel.Error
});
}
}
return(new List <TagData>());
}
