Extracting Multipage PDF Table to CSV in Java Using Asynchronous Processing

Template JSON Code

Following is the template JSON code for PDF parsing:

{
  "templateName": "Multipage Table Test",
  "templateVersion": 4,
  "templatePriority": 0,
  "detectionRules": {
    "keywords": [
      "Sample document with multi-page table"
    ]
  },
  "objects": [
    {
      "name": "total",
      "objectType": "field",
      "fieldProperties": {
        "fieldType": "macros",
        "expression": "TOTAL{{Spaces}}({{Number}})",
        "regex": true,
        "dataType": "decimal"
      }
    },
    {
      "name": "table1",
      "objectType": "table",
      "tableProperties": {
        "start": {
          "expression": "Item{{Spaces}}Description{{Spaces}}Price",
          "regex": true
        },
        "end": {
          "expression": "TOTAL{{Spaces}}{{Number}}",
          "regex": true
        },
        "row": {
          "expression": "{{LineStart}}{{Spaces}}(?<itemNo>{{Digits}}){{Spaces}}(?<description>{{SentenceWithSingleSpaces}}){{Spaces}}(?<price>{{Number}}){{Spaces}}(?<qty>{{Digits}}){{Spaces}}(?<extPrice>{{Number}})",
          "regex": true
        },
        "columns": [
          {
            "name": "itemNo",
            "dataType": "integer"
          },
          {
            "name": "description",
            "dataType": "string"
          },
          {
            "name": "price",
            "dataType": "decimal"
          },
          {
            "name": "qty",
            "dataType": "integer"
          },
          {
            "name": "extPrice",
            "dataType": "decimal"
          }
        ],
        "multipage": true
      }
    }
  ]
}

Sample Code in Java

Following is the sample code to parse the PDF file using the above template:

import com.google.gson.JsonObject;
import com.google.gson.JsonParser;
import com.google.gson.JsonPrimitive;
import okhttp3.*;
import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.OutputStream;
import java.nio.charset.StandardCharsets;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;


public class Main {
    // Get your own API Key by registering at https://app.pdf.co
    final static String API_KEY = "************************";


    public static void main(String[] args) throws IOException, InterruptedException {
        // Source PDF file
        final String SourceFileUrl = "https://bytescout-com.s3.amazonaws.com/files/demo-files/cloud-api/document-parser/MultiPageTable.pdf";
        // PDF document password. Leave empty for unprotected documents.
        final String Password = "";
        // Destination JSON file name
        final Path DestinationFile = Paths.get(".\\result.csv");
        final String outputFormat = "CSV";
        // Template text
        String templateText = new String(Files.readAllBytes(Paths.get(".\\MultiPageTable-template1.json")), StandardCharsets.UTF_8);


        // Create HTTP client instance
        OkHttpClient webClient = new OkHttpClient();


        // Parse the uploaded PDF document
        ParseDocumentAsync(webClient, API_KEY, DestinationFile, Password, SourceFileUrl, templateText, outputFormat);
    }


    public static void ParseDocumentAsync(OkHttpClient webClient, String apiKey, Path destinationFile,
                                          String password, String uploadedFileUrl, String templateText, String outputFormat) throws IOException, InterruptedException {


        // Prepare POST request body in JSON format
        JsonObject jsonBody = new JsonObject();
        jsonBody.add("url", new JsonPrimitive(uploadedFileUrl));
        jsonBody.add("template", new JsonPrimitive(templateText));
        jsonBody.add("outputFormat", new JsonPrimitive(outputFormat));
        jsonBody.add("async", new JsonPrimitive(true)); // Enable asynchronous processing


        RequestBody body = RequestBody.create(MediaType.parse("application/json"), jsonBody.toString());


        // Prepare request to `Document Parser` API
        Request request = new Request.Builder()
                .url("https://api.pdf.co/v1/pdf/documentparser")
                .addHeader("x-api-key", apiKey)
                .addHeader("Content-Type", "application/json")
                .post(body)
                .build();


        // Execute request
        Response response = webClient.newCall(request).execute();
        if (response.code() == 200) {
            JsonObject json = new JsonParser().parse(response.body().string()).getAsJsonObject();
            boolean error = json.get("error").getAsBoolean();
            if (!error) {
                String jobId = json.get("jobId").getAsString();
                System.out.println("Job#" + jobId + ": has been created.");


                // Check job status in a loop
                while (true) {
                    String status = CheckJobStatus(webClient, apiKey, jobId);
                    DateTimeFormatter dtf = DateTimeFormatter.ofPattern("MM/dd/yyyy HH:mm:ss");
                    System.out.println("Job#" + jobId + ": " + status + " - " + dtf.format(LocalDateTime.now()));


                    if (status.equalsIgnoreCase("success")) {
                        String resultFileUrl = json.get("url").getAsString();
                        // Download the file
                        downloadFile(webClient, resultFileUrl, destinationFile.toFile());
                        System.out.printf("Generated file saved to \"%s\" file.", destinationFile.toString());
                        break;
                    } else if (status.equalsIgnoreCase("working")) {
                        Thread.sleep(3000); // Pause for a few seconds before retrying
                    } else {
                        System.out.println("Job failed with status: " + status);
                        break;
                    }
                }
            } else {
                System.out.println(json.get("message").getAsString());
            }
        } else {
            System.out.println(response.code() + " " + response.message());
        }
    }


    // Check Job Status
    private static String CheckJobStatus(OkHttpClient webClient, String apiKey, String jobId) throws IOException {
        String url = "https://api.pdf.co/v1/job/check?jobid=" + jobId;


        Request request = new Request.Builder()
                .url(url)
                .addHeader("x-api-key", apiKey)
                .build();


        Response response = webClient.newCall(request).execute();
        if (response.code() == 200) {
            JsonObject json = new JsonParser().parse(response.body().string()).getAsJsonObject();
            return json.get("status").getAsString();
        } else {
            System.out.println(response.code() + " " + response.message());
        }
        return "Failed";
    }


    public static void downloadFile(OkHttpClient webClient, String url, File destinationFile) throws IOException {
        Request request = new Request.Builder()
                .url(url)
                .build();


        Response response = webClient.newCall(request).execute();
        byte[] fileBytes = response.body().bytes();
        OutputStream output = new FileOutputStream(destinationFile);
        output.write(fileBytes);
        output.flush();
        output.close();
        response.close();
    }
}

Sample PDF with Multi-Page Table for PDF to CSV Parsing

Below is the source PDF file for parsing:

Output CSV File

Below is the CSV output of the above code and the source file:

total,tableNames,tables
450.00,table1,"itemNo,description,price,qty,extPrice
1,Item 1,10.00,1,10.00
2,Item 2,10.00,1,10.00
3,Item 3,10.00,1,10.00
4,Item 4,10.00,1,10.00
5,Item 5,10.00,1,10.00
6,Item 6,10.00,1,10.00
7,Item 7,10.00,1,10.00
8,Item 8,10.00,1,10.00
9,Item 9,10.00,1,10.00
10,Item 10,10.00,1,10.00
11,Item 11,10.00,1,10.00
12,Item 12,10.00,1,10.00
13,Item 13,10.00,1,10.00
14,Item 14,10.00,1,10.00
15,Item 15,10.00,1,10.00
16,Item 16,10.00,1,10.00
17,Item 17,10.00,1,10.00
18,Item 18,10.00,1,10.00
19,Item 19,10.00,1,10.00
20,Item 20,10.00,1,10.00
21,Item 21,10.00,1,10.00
22,Item 22,10.00,1,10.00
23,Item 23,10.00,1,10.00
24,Item 24,10.00,1,10.00
25,Item 25,10.00,1,10.00
26,Item 26,10.00,1,10.00
27,Item 27,10.00,1,10.00
28,Item 28,10.00,1,10.00
29,Item 29,10.00,1,10.00
30,Item 30,10.00,1,10.00
31,Item 31,10.00,1,10.00
32,Item 32,10.00,1,10.00
33,Item 33,10.00,1,10.00
34,Item 34,10.00,1,10.00
35,Item 35,10.00,1,10.00
36,Item 36,10.00,1,10.00
37,Item 37,10.00,1,10.00
38,Item 38,10.00,1,10.00
39,Item 39,10.00,1,10.00
40,Item 40,10.00,1,10.00
41,Item 41,10.00,1,10.00
42,Item 42,10.00,1,10.00
43,Item 43,10.00,1,10.00
44,Item 44,10.00,1,10.00
45,Item 45,10.00,1,10.00

Import Required Libraries and Set Up

You can use the IDE of your choice, whether it's VS Code, IntelliJ, or another environment that you're comfortable with.Start by importing necessary packages and libraries to handle the API request. These libraries will manage file I/O, HTTP requests, and JSON parsing. Ensure the file is accessible via a URL or upload it if needed.

Set Up the API Key

Obtain your API key by signing up or logging into your PDF.co account. This key authenticates your requests to the API. Initialize the API key in your code.

Prepare the API Request Payload

Customize the payload with the following:

• Source File URL: The location of your PDF form.
• File Password (if applicable): Add if the PDF is password-protected.
• Template: The template provided here is a JSON file that contains all the information needed to extract data from the document, like the keywords, objects, expressions, and other information. The users can use built-in templates provided by the API or customize them to get the required output.
• Output Format: Set this to "CSV" for the desired output.

Enable Asynchronous Processing

Set the "async" parameter in your request body to true. This ensures the processing happens asynchronously, allowing you to poll for results without waiting for the process to finish.

Send the API Request

Use an HTTP POST request to send the payload to the Document Parser endpoint. Ensure the payload is JSON-formatted with all necessary parameters.

Poll for Job Status

The API will return a jobId and a statusUrl to track the processing status. Use these details to periodically check if the job is complete.

Retrieve and Save the Results

Once the job status is "success", download the resulting CSV file using the resultUrl provided in the response. Save the file locally, for example, as result.csv.