Unable to correctly extract tables from pdf document using pdf extract api

[Anil B Dugar]

Hello Everyone

Use Case : I am using the PDF Extract API service to extract the tables within the pdf

Tech Stack : .Net Nuget version is Adobe.PDFServicesSDK :  3.0.0

Problem : In a given table, if all the cells for a particular column are empty, they get merged with the next column, (Both the pdf file and output file is attached)

Expected output : CSV Files

Here is the sample code

Adobe.PDFServicesSDK.ExecutionContext executionContext = Adobe.PDFServicesSDK.ExecutionContext.Create(credentials);
                
ExtractPDFOperation extractPdfOperation = ExtractPDFOperation.CreateNew();
FileRef sourceFileRef = FileRef.CreateFromStream(pdfFileStream, "application/pdf");
                    extractPdfOperation.SetInputFile(sourceFileRef);

                    // Build ExtractPDF options and set them into the operation.
                    ExtractPDFOptions extractPdfOptions = ExtractPDFOptions.ExtractPDFOptionsBuilder()
                          .AddElementsToExtract(new List<ExtractElementType>(new[] { ExtractElementType.TABLES }))
                          .AddTableStructureFormat(TableStructureType.CSV)
                          .Build();

                    extractPdfOperation.SetOptions(extractPdfOptions);

                    // Lock & Execute the operation.
                    FileRef resultZipFile = extractPdfOperation.Execute(executionContext);

Error CSV : expected are 6 columns but only 5 are being shown in the csv

Pdf File being parsed :

Pls help

thanks

AD

[Anil B Dugar]

Try parsing the attached file and you will be able to reproduce the issue,

Also note i tried parsing using Amazon Textextract Service and it works !!