Marmelator/ocr-service
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 64 Wiki Activity

RED-6126: In the OCRService, OCR Text is not applied to Document

Browse Source 
*called PDFNet.initialize and terminate before and after message receive
*updated comments
*renamed some variables
This commit is contained in:
Kilian Schuettler 2023-02-21 10:42:14 +01:00
parent a6d99f5916
commit 430ad45a67
3 changed files with 26 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
ocr-service-v1/ocr-service-server-v1/src/main/java/com/iqser/red/service/ocr/v1/server/initializer/PDFNetInitializer.java
View File

@ -1,14 +1,12 @@
package com.iqser.red.service.ocr.v1.server.initializer;
import javax.annotation.PostConstruct;
import com.pdftron.pdf.PDFNet;
import lombok.RequiredArgsConstructor;
import lombok.SneakyThrows;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.stereotype.Component;
import com.pdftron.pdf.PDFNet;
import lombok.RequiredArgsConstructor;
import lombok.SneakyThrows;
import javax.annotation.PostConstruct;
@Component
@RequiredArgsConstructor
@ -22,7 +20,6 @@ public class PDFNetInitializer {
@SneakyThrows
@PostConstruct
// Do not change back to application runner, if it is application runner it takes messages from the queue before PDFNet is initialized, that leads to UnsatisfiedLinkError.
public void init() {
@ -32,4 +29,5 @@ public class PDFNetInitializer {
}
}

42
ocr-service-v1/ocr-service-server-v1/src/main/java/com/iqser/red/service/ocr/v1/server/service/OCRService.java
View File

@ -1,34 +1,23 @@
package com.iqser.red.service.ocr.v1.server.service;
import static java.lang.String.format;
import java.io.ByteArrayInputStream;
import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.util.Map;
import org.springframework.amqp.rabbit.core.RabbitTemplate;
import org.springframework.stereotype.Service;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.iqser.red.service.ocr.v1.api.model.OCRStatusUpdateResponse;
import com.iqser.red.service.ocr.v1.server.configuration.MessagingConfiguration;
import com.iqser.red.service.ocr.v1.server.settings.OcrServiceSettings;
import com.pdftron.common.PDFNetException;
import com.pdftron.pdf.OCRModule;
import com.pdftron.pdf.OCROptions;
import com.pdftron.pdf.Optimizer;
import com.pdftron.pdf.PDFDoc;
import com.pdftron.pdf.Page;
import com.pdftron.pdf.RectCollection;
import com.pdftron.pdf.*;
import com.pdftron.sdf.SDFDoc;
import io.micrometer.core.annotation.Timed;
import lombok.RequiredArgsConstructor;
import lombok.SneakyThrows;
import lombok.extern.slf4j.Slf4j;
import org.springframework.amqp.rabbit.core.RabbitTemplate;
import org.springframework.stereotype.Service;
import java.io.*;
import java.util.Map;
import static java.lang.String.format;
@Slf4j
@Service
@ -67,17 +56,17 @@ public class OCRService {
try (ByteArrayOutputStream transferOutputStream = new ByteArrayOutputStream()) {
try (InputStream fileStream = fileStorageService.getOriginalFileAsStream(dossierId, fileId)) {
long start = System.currentTimeMillis();
long removalStart = System.currentTimeMillis();
log.debug("Start invisible element removal for file with dossierId {} and fileId {}", dossierId, fileId);
invisibleElementRemovalService.removeInvisibleElements(fileStream, transferOutputStream, false);
long end = System.currentTimeMillis();
log.debug("Invisible element removal successful for file with dossierId {} and fileId {}, took {}s", dossierId, fileId, format("%.1f", (end - start) / 1000.0));
long removalEnd = System.currentTimeMillis();
log.debug("Invisible element removal successful for file with dossierId {} and fileId {}, took {}s", dossierId, fileId, format("%.1f", (removalEnd - removalStart) / 1000.0));
}
try (InputStream transferInputStream = new ByteArrayInputStream(transferOutputStream.toByteArray())) {
long start = System.currentTimeMillis();
long ocrStart = System.currentTimeMillis();
runOcr(transferInputStream, out, fileId);
long end = System.currentTimeMillis();
log.info("ocr successful for file with dossierId {} and fileId {}, took {}s", dossierId, fileId, format("%.1f", (end - start) / 1000.0));
long ocrEnd = System.currentTimeMillis();
log.info("ocr successful for file with dossierId {} and fileId {}, took {}s", dossierId, fileId, format("%.1f", (ocrEnd - ocrStart) / 1000.0));
}
}
}
@ -94,7 +83,7 @@ public class OCRService {
// When a page does not have a TextZone, PDFTron whites out the page. But, PDFTron scans it anyway, resulting in a longer runtime.
// So, we need to remove pages without images.
// Furthermore, creating a new document is *much* faster than reusing the same document and adding/removing pages one by one.
// Therefore, we create a new Document with a single page for every page that contains text.
// Therefore, we create a new Document with a single page for every page that contains an image.
int numProcessedPages = 0;
for (Integer pageId : pageIdToRectCollection.keySet()) {
try {
@ -139,6 +128,7 @@ public class OCRService {
log.error("Processed File with fileId {} could not be saved", fileId);
throw new RuntimeException(e);
}
pdfDoc.close();
}

5
ocr-service-v1/ocr-service-server-v1/src/main/java/com/iqser/red/service/ocr/v1/server/service/OcrMessageReceiver.java
View File

@ -15,6 +15,8 @@ import com.fasterxml.jackson.databind.ObjectMapper;
import com.iqser.red.service.ocr.v1.api.model.DocumentRequest;
import com.iqser.red.service.ocr.v1.server.client.FileStatusProcessingUpdateClient;
import com.iqser.red.service.ocr.v1.server.configuration.MessagingConfiguration;
import com.iqser.red.service.ocr.v1.server.initializer.PDFNetInitializer;
import com.pdftron.pdf.PDFNet;
import feign.FeignException;
import lombok.RequiredArgsConstructor;
@ -28,6 +30,7 @@ public class OcrMessageReceiver {
private final ObjectMapper objectMapper;
private final FileStorageService fileStorageService;
private final FileStatusProcessingUpdateClient fileStatusProcessingUpdateClient;
private final PDFNetInitializer pdfNetInitializer;
private final OCRService ocrService;
@ -36,6 +39,7 @@ public class OcrMessageReceiver {
@RabbitListener(queues = MessagingConfiguration.OCR_QUEUE, concurrency = "1")
public void receiveOcr(String in) throws JsonProcessingException {
pdfNetInitializer.init();
DocumentRequest ocrRequestMessage = objectMapper.readValue(in, DocumentRequest.class);
log.info("Start ocr for file with dossierId {} and fileId {}", ocrRequestMessage.getDossierId(), ocrRequestMessage.getFileId());
@ -58,6 +62,7 @@ public class OcrMessageReceiver {
}
fileStatusProcessingUpdateClient.ocrSuccessful(ocrRequestMessage.getDossierId(), ocrRequestMessage.getFileId());
PDFNet.terminate();
}