RED-8481: Use visual layout parsing to detect signatures

integrated service
This commit is contained in:
yhampe 2024-02-12 09:12:59 +01:00
parent 99944c9fef
commit 3664f6800a
10 changed files with 152 additions and 1 deletions

View File

@ -64,6 +64,9 @@ public class MessagingConfiguration {
public static final String CV_ANALYSIS_RESPONSE_QUEUE = "cv_analysis_response_queue";
public static final String CV_ANALYSIS_DLQ = "cv_analysis_dead_letter_queue";
public static final String VISUAL_LAYOUT_PARSING_QUEUE = "visual_layout_parsing_service_queue";
public static final String VISUAL_LAYOUT_RESPONSE_QUEUE = "visual_layout_parsing_service_response_queue";
public static final String VISUAL_LAYOUT_DLQ = "visual_layout_parsing_service_dead_letter_queue";
public static final String ANALYSIS_FLAG_CALCULATION_QUEUE = "analysis_flag_calculation_queue";
public static final String OCR_STATUS_UPDATE_RESPONSE_QUEUE = "ocr_status_update_response_queue";
@ -163,6 +166,28 @@ public class MessagingConfiguration {
return QueueBuilder.durable(CV_ANALYSIS_DLQ).build();
}
@Bean
public Queue visualLayoutParsingRequestQueue() {
return QueueBuilder.durable(VISUAL_LAYOUT_PARSING_QUEUE).withArgument("x-dead-letter-exchange", "").withArgument("x-dead-letter-routing-key", VISUAL_LAYOUT_DLQ).build();
}
@Bean
public Queue visualLayoutParsingResponseQueue() {
return QueueBuilder.durable(VISUAL_LAYOUT_RESPONSE_QUEUE)
.withArgument("x-dead-letter-exchange", "")
.withArgument("x-dead-letter-routing-key", VISUAL_LAYOUT_DLQ)
.build();
}
@Bean
public Queue visualLayoutParsingDLQ() {
return QueueBuilder.durable(VISUAL_LAYOUT_DLQ).build();
}
@Bean
public Queue ocrStatusUpdateResponseQueue() {

View File

@ -0,0 +1,24 @@
package com.iqser.red.service.persistence.management.v1.processor.model;
import com.iqser.red.service.persistence.service.v1.api.shared.model.dossiertemplate.dossier.file.FileType;
import lombok.AllArgsConstructor;
import lombok.Builder;
import lombok.Data;
import lombok.NoArgsConstructor;
@Data
@Builder
@AllArgsConstructor
@NoArgsConstructor
public class VisualLayoutParsingServiceRequest {
public static final String VISUAL_LAYOUT_PARSING_FILE_EXTENSION= FileType.VISUAL_LAYOUT.name() + FileType.VISUAL_LAYOUT.getExtension() + ".gz";
public static final String TARGET_FILE_EXTENSION = "ORIGIN.pdf.gz";
private String dossierId;
private String fileId;
private String targetFileExtension;
private String responseFileExtension;
}

View File

@ -0,0 +1,16 @@
package com.iqser.red.service.persistence.management.v1.processor.model.image;
import lombok.AllArgsConstructor;
import lombok.Builder;
import lombok.Data;
import lombok.NoArgsConstructor;
@Data
@Builder
@AllArgsConstructor
@NoArgsConstructor
public class VisualLayoutParsingServiceResponse {
private String dossierId;
private String fileId;
}

View File

@ -24,6 +24,7 @@ import com.iqser.red.service.persistence.management.v1.processor.model.CvAnalysi
import com.iqser.red.service.persistence.management.v1.processor.model.FileIdentifier;
import com.iqser.red.service.persistence.management.v1.processor.model.NerServiceRequest;
import com.iqser.red.service.persistence.management.v1.processor.model.OCRStatusUpdateResponse;
import com.iqser.red.service.persistence.management.v1.processor.model.VisualLayoutParsingServiceRequest;
import com.iqser.red.service.persistence.management.v1.processor.model.image.ImageServiceRequest;
import com.iqser.red.service.persistence.management.v1.processor.service.layoutparsing.LayoutParsingRequestFactory;
import com.iqser.red.service.persistence.management.v1.processor.service.manualredactions.ManualRedactionProviderService;
@ -176,6 +177,12 @@ public class FileStatusService {
return;
}
if (settings.isVisualLayoutParsingEnabled() && !fileManagementStorageService.objectExists(dossierId, fileId, FileType.VISUAL_LAYOUT)) {
log.info("Add file: {} from dossier {} to Visual Layout Parsing queue", fileId, dossierId);
addToVisualLayoutParsingQueue(dossierId, fileId);
return;
}
if (settings.isFigureDetectionEnabled() && !fileManagementStorageService.objectExists(dossierId, fileId, FileType.FIGURE)) {
log.debug("Add file: {} from dossier {} to Figure Detection queue", fileId, dossierId);
addToFigureDetectionRequestQueue(dossierId, fileId);
@ -247,6 +254,21 @@ public class FileStatusService {
}
}
private void addToVisualLayoutParsingQueue(String dossierId, String fileId) {
fileStatusPersistenceService.updateProcessingStatus(fileId, ProcessingStatus.VISUAL_LAYOUT_PARSING_ANALYZING);
rabbitTemplate.convertAndSend(MessagingConfiguration.VISUAL_LAYOUT_PARSING_QUEUE,
VisualLayoutParsingServiceRequest.builder()
.dossierId(dossierId)
.fileId(fileId)
.targetFileExtension(VisualLayoutParsingServiceRequest.TARGET_FILE_EXTENSION)
.responseFileExtension(VisualLayoutParsingServiceRequest.VISUAL_LAYOUT_PARSING_FILE_EXTENSION)
.build(),
message -> {
message.getMessageProperties().setPriority(1);
log.info("answer from visual layout parsing: "+message);
return message;
});
}
@SneakyThrows
public void addToPreprocessingQueue(String dossierId, String fileId, String filename) {

View File

@ -39,12 +39,16 @@ public class LayoutParsingRequestFactory {
Optional<String> optionalTableFileId = fileManagementStorageService.objectExists(dossierId, fileId, FileType.TABLES) //
? Optional.of(StorageIdUtils.getStorageId(dossierId, fileId, FileType.TABLES)) : Optional.empty();
Optional<String> optionalVisualLayoutParsingFileId = fileManagementStorageService.objectExists(dossierId, fileId, FileType.VISUAL_LAYOUT) //
? Optional.of(StorageIdUtils.getStorageId(dossierId, fileId, FileType.VISUAL_LAYOUT)) : Optional.empty();
return LayoutParsingRequest.builder()
.layoutParsingType(type)
.identifier(layoutParsingRequestIdentifierService.buildIdentifier(dossierId, fileId, priority))
.originFileStorageId(StorageIdUtils.getStorageId(dossierId, fileId, FileType.ORIGIN))
.imagesFileStorageId(optionalImageFileId)
.tablesFileStorageId(optionalTableFileId)
.visualLayoutParsingFileId(optionalVisualLayoutParsingFileId)
.pageFileStorageId(StorageIdUtils.getStorageId(dossierId, fileId, FileType.DOCUMENT_PAGES))
.structureFileStorageId(StorageIdUtils.getStorageId(dossierId, fileId, FileType.DOCUMENT_STRUCTURE))
.textBlockFileStorageId(StorageIdUtils.getStorageId(dossierId, fileId, FileType.DOCUMENT_TEXT))

View File

@ -0,0 +1,54 @@
package com.iqser.red.service.persistence.management.v1.processor.service.queue;
import java.time.OffsetDateTime;
import java.time.temporal.ChronoUnit;
import org.springframework.amqp.core.Message;
import org.springframework.amqp.rabbit.annotation.RabbitListener;
import org.springframework.stereotype.Service;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.iqser.red.service.persistence.management.v1.processor.configuration.MessagingConfiguration;
import com.iqser.red.service.persistence.management.v1.processor.model.image.VisualLayoutParsingServiceResponse;
import com.iqser.red.service.persistence.management.v1.processor.service.FileStatusProcessingUpdateService;
import com.iqser.red.service.persistence.management.v1.processor.service.FileStatusService;
import com.iqser.red.service.persistence.service.v1.api.shared.model.dossiertemplate.dossier.file.FileErrorInfo;
import lombok.RequiredArgsConstructor;
import lombok.SneakyThrows;
import lombok.extern.slf4j.Slf4j;
@Slf4j
@Service
@RequiredArgsConstructor
public class VisualLayoutParsingMessageReceiver {
private final ObjectMapper objectMapper;
private final FileStatusService fileStatusService;
private final FileStatusProcessingUpdateService fileStatusProcessingUpdateService;
@SneakyThrows
@RabbitListener(queues = MessagingConfiguration.VISUAL_LAYOUT_RESPONSE_QUEUE)
public void receive(VisualLayoutParsingServiceResponse response) {
fileStatusService.setStatusAnalyse(response.getDossierId(), response.getFileId(), false);
log.info("Received message in {} for dossierId {} and fileId {}", MessagingConfiguration.VISUAL_LAYOUT_RESPONSE_QUEUE, response.getDossierId(), response.getFileId());
}
@SneakyThrows
@RabbitListener(queues = MessagingConfiguration.VISUAL_LAYOUT_DLQ)
public void handleDLQMessage(Message failedMessage) {
var response = objectMapper.readValue(failedMessage.getBody(), VisualLayoutParsingServiceResponse.class);
log.warn("Received message from {} for dossierId {} and fileId {}", MessagingConfiguration.VISUAL_LAYOUT_DLQ, response.getDossierId(), response.getFileId());
fileStatusProcessingUpdateService.analysisFailed(response.getDossierId(),
response.getFileId(),
new FileErrorInfo("table extractor failed", MessagingConfiguration.VISUAL_LAYOUT_DLQ, "table-extractor", OffsetDateTime.now().truncatedTo(ChronoUnit.MILLIS)));
}
}

View File

@ -24,6 +24,7 @@ public class FileManagementServiceSettings {
private boolean imageServiceEnabled = true;
private boolean nerServiceEnabled = true;
private boolean visualLayoutParsingEnabled = false;
private boolean storeImageFile = true;

View File

@ -30,6 +30,7 @@ cors.enabled: true
persistence-service:
imageServiceEnabled: false
visualLayoutParsingEnabled: true
nerServiceEnabled: false
storeImageFile: false
applicationName: RedactManager

View File

@ -16,6 +16,8 @@ public enum FileType {
TEXT_HIGHLIGHTS(".json"),
FIGURE(".json"),
TABLES(".json"),
VISUAL_LAYOUT(".json"),
COMPONENTS(".json"),
// document is split into 4 files, all should be overridden/deleted at the same time
DOCUMENT_TEXT(".json"),

View File

@ -18,5 +18,7 @@ public enum ProcessingStatus {
PRE_PROCESSING,
PRE_PROCESSED,
FIGURE_DETECTION_ANALYZING,
TABLE_PARSING_ANALYZING
TABLE_PARSING_ANALYZING,
VISUAL_LAYOUT_PARSING_ANALYZING
}