public static void UploadDocuments(SearchIndexClient indexClient, string fileId, string fileName, string ocrText, KeyPhraseResult keyPhraseResult)
{
List<IndexAction> indexOperations = new List<IndexAction>();
var doc = new Document();
doc.Add("fileId", fileId);
doc.Add("fileName", fileName);
doc.Add("ocrText", ocrText);
doc.Add("keyPhrases", keyPhraseResult.KeyPhrases.ToList());
indexOperations.Add(IndexAction.Upload(doc));
try
{
indexClient.Documents.Index(new IndexBatch(indexOperations));
}
catch (IndexBatchException e)
{
// Sometimes when your Search service is under load, indexing will fail for some of the documents in
// the batch. Depending on your application, you can take compensating actions like delaying and
// retrying. For this simple demo, we just log the failed document keys and continue.
Console.WriteLine(
"Failed to index some of the documents: {0}",
String.Join(", ", e.IndexingResults.Where(r => !r.Succeeded).Select(r => r.Key)));
}
}
static void Main(string[] args)
{
var searchPath = "pdf";
var outPath = "image";
// Note, this will create a new Azure Search Index for the OCR text
Console.WriteLine("Creating Azure Search index...");
AzureSearch.CreateIndex(serviceClient, indexName);
// Creating an image directory
if (Directory.Exists(outPath) == false)
{
Directory.CreateDirectory(outPath);
}
foreach (var filename in Directory.GetFiles(searchPath, "*.pdf", SearchOption.TopDirectoryOnly))
{
Console.WriteLine("Extracting images from {0} \r\n", System.IO.Path.GetFileName(filename));
var images = PdfImageExtractor.ExtractImages(filename);
Console.WriteLine("{0} images found.", images.Count);
Console.WriteLine();
var directory = System.IO.Path.GetDirectoryName(filename);
foreach (var name in images.Keys)
{
if (name.LastIndexOf(".") + 1 != name.Length)
{
images[name].Save(Path.Combine(outPath, name));
}
}
string ocrText = string.Empty;
Console.WriteLine("Extracting text from image... \r\n");
foreach (var imagefilename in Directory.GetFiles(outPath))
{
OcrResults ocr = vision.RecognizeText(imagefilename);
ocrText += vision.GetRetrieveText(ocr);
File.Delete(imagefilename);
}
Console.WriteLine("Extracting key phrases from processed text... \r\n");
KeyPhraseResult keyPhraseResult = TextExtraction.ProcessText(ocrText);
// Take the resulting orcText and upload to a new Azure Search Index
// It is highly recommended that you upload documents in batches rather
// individually like is done here
if (ocrText.Length > 0)
{
Console.WriteLine("Uploading extracted text to Azure Search...\r\n");
string fileNameOnly = System.IO.Path.GetFileName(filename);
string fileId = System.Convert.ToBase64String(System.Text.Encoding.UTF8.GetBytes(fileNameOnly));
AzureSearch.UploadDocuments(indexClient, fileId, fileNameOnly, ocrText, keyPhraseResult);
}
}
// Execute a test search
Console.WriteLine("Execute Search...");
AzureSearch.SearchDocuments(indexClient, "Azure Search");
Console.WriteLine("All done. Press any key to continue.");
Console.ReadLine();
}
public static void UploadDocuments(SearchIndexClient indexClient, string fileId, string fileName, string ocrText, KeyPhraseResult keyPhraseResult)
{
List <IndexAction> indexOperations = new List <IndexAction>();
var doc = new Document();
doc.Add("fileId", fileId);
doc.Add("fileName", fileName);
doc.Add("ocrText", ocrText);
doc.Add("keyPhrases", keyPhraseResult.KeyPhrases.ToList());
indexOperations.Add(IndexAction.Upload(doc));
try
{
indexClient.Documents.Index(new IndexBatch(indexOperations));
}
catch (IndexBatchException e)
{
// Sometimes when your Search service is under load, indexing will fail for some of the documents in
// the batch. Depending on your application, you can take compensating actions like delaying and
// retrying. For this simple demo, we just log the failed document keys and continue.
Console.WriteLine(
"Failed to index some of the documents: {0}",
String.Join(", ", e.IndexingResults.Where(r => !r.Succeeded).Select(r => r.Key)));
}
}
public static KeyPhraseResult ProcessText(string inputText)
{
string accountKey = ConfigurationManager.AppSettings["textExtractionAccountKey"];
KeyPhraseResult keyPhraseResult = new KeyPhraseResult();
using (var httpClient = new HttpClient())
{
string inputTextEncoded = HttpUtility.UrlEncode(inputText);
httpClient.BaseAddress = new Uri(ServiceBaseUri);
string creds = "AccountKey:" + accountKey;
string authorizationHeader = "Basic " + Convert.ToBase64String(Encoding.ASCII.GetBytes(creds));
httpClient.DefaultRequestHeaders.Add("Authorization", authorizationHeader);
httpClient.DefaultRequestHeaders.Accept.Add(new MediaTypeWithQualityHeaderValue("application/json"));
// get key phrases
string keyPhrasesRequest = "data.ashx/amla/text-analytics/v1/GetKeyPhrases?Text=" + inputTextEncoded;
Task <HttpResponseMessage> responseTask = httpClient.GetAsync(keyPhrasesRequest);
responseTask.Wait();
HttpResponseMessage response = responseTask.Result;
Task <string> contentTask = response.Content.ReadAsStringAsync();
contentTask.Wait();
string content = contentTask.Result;
if (!response.IsSuccessStatusCode)
{
throw new Exception("Call to get key phrases failed with HTTP status code: " +
response.StatusCode + " and contents: " + content);
}
keyPhraseResult = JsonConvert.DeserializeObject <KeyPhraseResult>(content);
Console.WriteLine("Key phrases: {0} \r\n", string.Join(",", keyPhraseResult.KeyPhrases));
// Uncomment the following if you want to retrieve additional details on this text
//// get sentiment
//string sentimentRequest = "data.ashx/amla/text-analytics/v1/GetSentiment?Text=" + inputTextEncoded;
//responseTask = httpClient.GetAsync(sentimentRequest);
//responseTask.Wait();
//response = responseTask.Result;
//contentTask = response.Content.ReadAsStringAsync();
//contentTask.Wait();
//content = contentTask.Result;
//if (!response.IsSuccessStatusCode)
//{
// throw new Exception("Call to get sentiment failed with HTTP status code: " +
// response.StatusCode + " and contents: " + content);
//}
//SentimentResult sentimentResult = JsonConvert.DeserializeObject<SentimentResult>(content);
//Console.WriteLine("Sentiment score: " + sentimentResult.Score);
//// get the language in text
//string languageRequest = "data.ashx/amla/text-analytics/v1/GetLanguage?Text=" + inputTextEncoded;
//responseTask = httpClient.GetAsync(languageRequest);
//responseTask.Wait();
//response = responseTask.Result;
//contentTask = response.Content.ReadAsStringAsync();
//contentTask.Wait();
//content = contentTask.Result;
//if (!response.IsSuccessStatusCode)
//{
// throw new Exception("Call to get language failed with HTTP status code: " +
// response.StatusCode + " and contents: " + content);
//}
//LanguageResult languageResult = JsonConvert.DeserializeObject<LanguageResult>(content);
//Console.WriteLine("Detected Languages: " + string.Join(",", languageResult.DetectedLanguages.Select(language => language.Name).ToArray()));
}
return(keyPhraseResult);
}
public static void UploadDocuments(SearchIndexClient indexClient, string fileId, string fileName, string ocrText, KeyPhraseResult keyPhraseResult)
{
List <IndexAction> indexOperations = new List <IndexAction>();
var doc = new Document();
doc.Add("fileId", fileId);
doc.Add("fileName", fileName);
doc.Add("ocrText", ocrText);
doc.Add("keyPhrases", keyPhraseResult.KeyPhrases.ToList());
indexOperations.Add(IndexAction.Upload(doc));
try
{
indexClient.Documents.Index(new IndexBatch(indexOperations));
}
catch (IndexBatchException e)
{
Console.WriteLine(
"Failed to index some of the documents: {0}",
String.Join(", ", e.IndexingResults.Where(r => !r.Succeeded).Select(r => r.Key)));
}
}
public static KeyPhraseResult ProcessText(string inputText)
{
string accountKey = ConfigurationManager.AppSettings["textExtractionAccountKey"];
KeyPhraseResult keyPhraseResult = new KeyPhraseResult();
using (var httpClient = new HttpClient())
{
string inputTextEncoded = HttpUtility.UrlEncode(inputText);
httpClient.BaseAddress = new Uri(ServiceBaseUri);
string creds = "AccountKey:" + accountKey;
string authorizationHeader = "Basic " + Convert.ToBase64String(Encoding.ASCII.GetBytes(creds));
httpClient.DefaultRequestHeaders.Add("Authorization", authorizationHeader);
httpClient.DefaultRequestHeaders.Accept.Add(new MediaTypeWithQualityHeaderValue("application/json"));
// get key phrases
string keyPhrasesRequest = "data.ashx/amla/text-analytics/v1/GetKeyPhrases?Text=" + inputTextEncoded;
Task<HttpResponseMessage> responseTask = httpClient.GetAsync(keyPhrasesRequest);
responseTask.Wait();
HttpResponseMessage response = responseTask.Result;
Task<string> contentTask = response.Content.ReadAsStringAsync();
contentTask.Wait();
string content = contentTask.Result;
if (!response.IsSuccessStatusCode)
{
throw new Exception("Call to get key phrases failed with HTTP status code: " +
response.StatusCode + " and contents: " + content);
}
keyPhraseResult = JsonConvert.DeserializeObject<KeyPhraseResult>(content);
Console.WriteLine("Key phrases: {0} \r\n", string.Join(",", keyPhraseResult.KeyPhrases));
// Uncomment the following if you want to retrieve additional details on this text
//// get sentiment
//string sentimentRequest = "data.ashx/amla/text-analytics/v1/GetSentiment?Text=" + inputTextEncoded;
//responseTask = httpClient.GetAsync(sentimentRequest);
//responseTask.Wait();
//response = responseTask.Result;
//contentTask = response.Content.ReadAsStringAsync();
//contentTask.Wait();
//content = contentTask.Result;
//if (!response.IsSuccessStatusCode)
//{
// throw new Exception("Call to get sentiment failed with HTTP status code: " +
// response.StatusCode + " and contents: " + content);
//}
//SentimentResult sentimentResult = JsonConvert.DeserializeObject<SentimentResult>(content);
//Console.WriteLine("Sentiment score: " + sentimentResult.Score);
//// get the language in text
//string languageRequest = "data.ashx/amla/text-analytics/v1/GetLanguage?Text=" + inputTextEncoded;
//responseTask = httpClient.GetAsync(languageRequest);
//responseTask.Wait();
//response = responseTask.Result;
//contentTask = response.Content.ReadAsStringAsync();
//contentTask.Wait();
//content = contentTask.Result;
//if (!response.IsSuccessStatusCode)
//{
// throw new Exception("Call to get language failed with HTTP status code: " +
// response.StatusCode + " and contents: " + content);
//}
//LanguageResult languageResult = JsonConvert.DeserializeObject<LanguageResult>(content);
//Console.WriteLine("Detected Languages: " + string.Join(",", languageResult.DetectedLanguages.Select(language => language.Name).ToArray()));
}
return keyPhraseResult;
}
