public async Task <ActionResult <DocumentVM> > CreateDocumentContents([FromRoute] Guid id, [FromBody] DocumentContentVM documentContentVM)
{
// Validation
if (!ModelState.IsValid)
{
return(BadRequest(ModelState));
}
// Mapping
DocumentContent documentContent = this.mapper.Map <DocumentContentVM, DocumentContent>(documentContentVM);
Document document = await this.bll.CreateDocumentContentAsync(id, documentContent);
// Mapping
return(Ok(this.mapper.Map <Document, DocumentVM>(document)));
}
private async Task HandleMessage(string content)
{
List <DocumentContentVM> documentContentVMs = new List <DocumentContentVM>();
this.logger.LogInformation($"Received content: {content}");
this.logger.LogInformation("Parsing content to Document");
Models.Document document = JsonConvert.DeserializeObject <Models.Document>(content);
this.logger.LogInformation($"Content parsed to Document model: {document}");
this.logger.LogInformation($"Document mime type: {document.MimeType}");
switch (document.MimeType)
{
// Text
case "text/plain":
this.logger.LogInformation($"Start text detection of document with ID: {document.Id}");
// Retrieve file
GetObjectResponse response = await this.awsS3Service.DownloadFile(document.Path);
this.logger.LogInformation("File downloaded from AWS S3 storage");
this.logger.LogInformation("Reading contents of file");
// Read contents
using (StreamReader streamReader = new StreamReader(response.ResponseStream))
{
while (streamReader.Peek() != -1)
{
string line = streamReader.ReadLine();
// Check contents of line
if (string.IsNullOrEmpty(line))
{
continue;
}
// Found a line
// Checking for words in line
string[] words = line.Split(" ");
if (words.Length > 1)
{
// If multiple words in the line, add line as Content
DocumentContentVM newLineDocumentContentVM = new DocumentContentVM()
{
DocumentId = document.Id,
Text = line,
TextType = DocumentContentTextType.LINE
};
documentContentVMs.Add(newLineDocumentContentVM);
// Add the words seperatly
foreach (string word in words)
{
if (string.IsNullOrEmpty(word))
{
continue;
}
DocumentContentVM newWordDocumentContentVM = new DocumentContentVM()
{
DocumentId = document.Id,
Text = word,
TextType = DocumentContentTextType.WORD
};
documentContentVMs.Add(newWordDocumentContentVM);
}
}
else
{
// If only 1 word in the line, add word as Content
DocumentContentVM newWordDocumentContentVM = new DocumentContentVM()
{
DocumentId = document.Id,
Text = line,
TextType = DocumentContentTextType.WORD
};
documentContentVMs.Add(newWordDocumentContentVM);
}
}
;
}
break;
// PDF, JPG, PNG
case "application/pdf":
case "image/jpeg":
case "image/png":
this.logger.LogInformation($"Amazon Textract - Start text detection of document with ID: {document.Id}");
string jobId = await this.awsTextractService.StartDocumentTextDetection(
this.configuration.GetSection("AWS")
.GetSection("S3")
.GetSection("Bucket")
.GetValue <string>("Name"),
document.Path
);
this.logger.LogInformation($"Amazon Textract - Created a new job with ID: {jobId}");
this.logger.LogInformation("Amazon Textract - Waiting for job completion");
await this.awsTextractService.WaitForJobCompletion(jobId);
this.logger.LogInformation($"Amazon Textract - Job with ID: {jobId} is completed");
this.logger.LogInformation($"Amazon Textract - Retrieving results for job with ID: {jobId}");
List <GetDocumentTextDetectionResponse> textDetectionResponses = await this.awsTextractService.GetJobResult(jobId);
this.logger.LogInformation("Amazon Textract - Saving results to JSON text file");
string jsonResponse = JsonConvert.SerializeObject(textDetectionResponses, Formatting.Indented);
File.WriteAllText($"Results/DocumentParser/{DateTime.Now.ToString("yyyyMMdd_HHmmss")}_{document.Id}.textracted.json", jsonResponse);
this.logger.LogInformation("Amazon Textract - Converting results to document contents");
foreach (GetDocumentTextDetectionResponse textDetectionResponse in textDetectionResponses)
{
foreach (Block block in textDetectionResponse.Blocks)
{
if (!string.IsNullOrEmpty(block.Text))
{
DocumentContentVM newDocumentContentVM = new DocumentContentVM()
{
DocumentId = document.Id,
Text = block.Text,
Confidence = block.Confidence
};
if (block.BlockType.Equals("WORD"))
{
newDocumentContentVM.TextType = DocumentContentTextType.WORD;
}
if (block.BlockType.Equals("LINE"))
{
newDocumentContentVM.TextType = DocumentContentTextType.LINE;
}
documentContentVMs.Add(newDocumentContentVM);
}
}
}
break;
}
if (documentContentVMs.Any())
{
this.logger.LogInformation($"Text detected in document with ID: {document.Id}");
this.logger.LogInformation($"Saving results in database through API");
HttpResponseMessage httpResponseMessage = await this.apiService.DocumentsCreateDocumentContents(document.Id, documentContentVMs);
if (httpResponseMessage.IsSuccessStatusCode)
{
this.logger.LogInformation($"Saved in database");
}
}
else
{
this.logger.LogInformation($"No text detected in document with ID: {document.Id}");
}
this.logger.LogInformation($"Document with ID: {document.Id} parsed");
}
