Compare commits

...
6 Commits
Author SHA1 Message Date
Dominique Eifländer d9a3bbbd30 Merge branch 'RED-5253' into 'main'
RED-5253: Ported last documine changes

See merge request fforesight/layout-parser!7
2023-08-04 09:59:57 +02:00
deiflaender 150aea55c0 RED-5253: Ported last documine changes 2023-08-04 09:55:35 +02:00
Kilian Schuettler 676f0c9d09 cleanup dependency versions 2023-08-01 10:27:14 +02:00
Kilian Schuettler ded00df11e fix build 2023-08-01 09:57:58 +02:00
Kilian Schuettler 286556cbb6 mark non nullable fields in request 2023-08-01 00:50:33 +02:00
Kilian Schuettler d6a74dc9f9 add field id to image data 2023-07-31 16:32:11 +02:00
16 changed files with 77 additions and 48 deletions
-1
View File
@@ -12,7 +12,6 @@
<exclude name="DataflowAnomalyAnalysis"/>
<exclude name="MissingSerialVersionUID"/>
<exclude name="NullAssignment"/>
<exclude name="BeanMembersShouldSerialize"/>
<exclude name="AvoidLiteralsInIfCondition"/>
<exclude name="AvoidDuplicateLiterals"/>
<exclude name="AvoidFieldNameMatchingMethodName"/>
-1
View File
@@ -13,7 +13,6 @@
<exclude name="DataflowAnomalyAnalysis"/>
<exclude name="MissingSerialVersionUID"/>
<exclude name="NullAssignment"/>
<exclude name="BeanMembersShouldSerialize"/>
<exclude name="AvoidLiteralsInIfCondition"/>
<exclude name="AvoidDuplicateLiterals"/>
<exclude name="AvoidFieldNameMatchingMethodName"/>
@@ -35,6 +35,7 @@ public class DocumentStructure {
public static final String TRANSPARENT = "transparent";
public static final String IMAGE_TYPE = "imageType";
public static final String POSITION = "position";
public static final String ID = "id";
}
@@ -4,21 +4,21 @@ import java.util.Map;
import java.util.Optional;
import lombok.Builder;
import lombok.NonNull;
@Builder
public record LayoutParsingRequest(
LayoutParsingType layoutParsingType,
@NonNull LayoutParsingType layoutParsingType,
Map<String, String> identifier,
String originFileStorageId,
@NonNull String originFileStorageId,
Optional<String> tablesFileStorageId,
Optional<String> imagesFileStorageId,
String structureFileStorageId,
@NonNull String structureFileStorageId,
String researchDocumentStorageId,
String textBlockFileStorageId,
String positionBlockFileStorageId,
String pageFileStorageId,
String simplifiedTextStorageId,
String sectionGridStorageId) {
@NonNull String textBlockFileStorageId,
@NonNull String positionBlockFileStorageId,
@NonNull String pageFileStorageId,
@NonNull String simplifiedTextStorageId,
@NonNull String sectionGridStorageId) {
}
@@ -5,6 +5,9 @@ plugins {
description = "layoutparser-service-processor"
val jacksonVersion = "2.15.0-rc2"
val pdfBoxVersion = "3.0.0-alpha2"
dependencies {
implementation(project(":layoutparser-service-internal-api"))
@@ -15,9 +18,9 @@ dependencies {
implementation("com.knecon.fforesight:tenant-commons:0.10.0")
implementation("com.iqser.red.commons:storage-commons:2.1.0")
implementation("org.apache.pdfbox:pdfbox:3.0.0-alpha2")
implementation("org.apache.pdfbox:pdfbox-tools:3.0.0-alpha2")
implementation("com.fasterxml.jackson.module:jackson-module-afterburner:2.15.0-rc2")
implementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:2.15.0-rc2")
implementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
implementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
implementation("com.fasterxml.jackson.module:jackson-module-afterburner:${jacksonVersion}")
implementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")
implementation("org.springframework.boot:spring-boot-starter-web:3.0.6")
}
@@ -77,6 +77,9 @@ public class PdfParsingService {
stripper.setStartPage(pageNumber);
stripper.setEndPage(pageNumber);
stripper.setPdpage(pdPage);
if(layoutParsingType.equals(LayoutParsingType.DOCUMINE)){
stripper.setSortByPosition(true);
}
stripper.getText(pdDocument);
PDRectangle pdr = pdPage.getMediaBox();
@@ -5,6 +5,9 @@ import static java.util.stream.Collectors.toSet;
import java.util.ArrayList;
import java.util.Comparator;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import java.util.stream.Collectors;
import org.springframework.stereotype.Service;
@@ -23,6 +26,8 @@ public class DocuMineBlockificationService {
static final float THRESHOLD = 1f;
Pattern pattern = Pattern.compile("^(\\d{1,2}\\.){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z ()-]{2,50}", Pattern.CASE_INSENSITIVE);
/**
* This method is building blocks by expanding the minX/maxX and minY/maxY value on each word that is not split by the conditions.
@@ -39,11 +44,7 @@ public class DocuMineBlockificationService {
List<TextPositionSequence> chunkWords = new ArrayList<>();
List<AbstractPageBlock> chunkBlockList1 = new ArrayList<>();
float minX = 1000;
float maxX = 0;
float minY = 1000;
float maxY = 0;
float minX = 1000, maxX = 0, minY = 1000, maxY = 0;
TextPositionSequence prev = null;
boolean wasSplitted = false;
@@ -60,7 +61,10 @@ public class DocuMineBlockificationService {
boolean splitByOtherFontAndOtherY = prev != null && prev.getMaxYDirAdj() != word.getMaxYDirAdj() && (word.getFontStyle().contains("bold") && !prev.getFontStyle()
.contains("bold") || prev.getFontStyle().contains("bold") && !word.getFontStyle().contains("bold"));
if (prev != null && (lineSeparation || startFromTop || splitByDir || isSplitByRuling || splitByOtherFontAndOtherY || negativeXGap)) {
Matcher matcher = pattern.matcher(chunkWords.stream().collect(Collectors.joining(" ")).toString());
boolean startsOnSameX = Math.abs(minX - word.getMinXDirAdj()) < 5 && matcher.matches();
if (prev != null && (lineSeparation || startFromTop || splitByDir || isSplitByRuling || splitByOtherFontAndOtherY || negativeXGap || startsOnSameX)) {
Orientation prevOrientation = null;
if (!chunkBlockList1.isEmpty()) {
@@ -231,3 +235,4 @@ public class DocuMineBlockificationService {
}
}
@@ -20,15 +20,14 @@ import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
@Slf4j
@Service
@RequiredArgsConstructor
public class DocuMineClassificationService {
private final BodyTextFrameService bodyTextFrameService;
private static final Pattern pattern = Pattern.compile("^(\\d{1,1}\\.?){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z\\[\\]\\-]{2,50}", Pattern.CASE_INSENSITIVE);
private static final Pattern pattern2 = Pattern.compile(".*\\d{4}$", Pattern.CASE_INSENSITIVE);
private static final Pattern pattern = Pattern.compile("^(\\d{1,2}\\.){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
private static final Pattern pattern2 = Pattern.compile("\\p{L}{3,}", Pattern.CASE_INSENSITIVE);
private static final Pattern pattern3 = Pattern.compile("^(\\d{1,1}\\.){1,3}\\d{1,2}\\.?\\s[a-z]{1,2}\\/[a-z]{1,2}.*");
@@ -18,6 +18,7 @@ public class PropertiesMapper {
properties.put(DocumentStructure.ImageProperties.IMAGE_TYPE, image.getImageType().toString());
properties.put(DocumentStructure.ImageProperties.TRANSPARENT, String.valueOf(image.isTransparent()));
properties.put(DocumentStructure.ImageProperties.POSITION, toString(image.getPosition()));
properties.put(DocumentStructure.ImageProperties.ID, image.getId());
return properties;
}
@@ -12,6 +12,9 @@ plugins {
description = "layoutparser-service-server"
val jacksonVersion = "2.15.0-rc2"
val pdfBoxVersion = "3.0.0-alpha2"
dependencies {
implementation(project(":layoutparser-service-processor"))
implementation(project(":layoutparser-service-internal-api"))
@@ -29,10 +32,10 @@ dependencies {
testImplementation("org.springframework.boot:spring-boot-starter-amqp:3.0.6")
testImplementation("com.iqser.red.service:persistence-service-shared-api-v1:2.36.0")
testImplementation("com.iqser.red.commons:jackson-commons:1.0.0")
testImplementation("com.fasterxml.jackson.module:jackson-module-afterburner:2.15.0-rc2")
testImplementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:2.15.0-rc2")
testImplementation("org.apache.pdfbox:pdfbox:3.0.0-alpha2")
testImplementation("org.apache.pdfbox:pdfbox-tools:3.0.0-alpha2")
testImplementation("com.fasterxml.jackson.module:jackson-module-afterburner:${jacksonVersion}")
testImplementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")
testImplementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
testImplementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
testImplementation("org.springframework.boot:spring-boot-starter-test:3.0.6")
}
@@ -11,6 +11,7 @@ import com.fasterxml.jackson.databind.ObjectMapper;
import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingFinishedEvent;
import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames;
import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingPipeline;
import lombok.RequiredArgsConstructor;
@@ -34,6 +35,9 @@ public class MessageHandler {
LayoutParsingRequest layoutParsingRequest = objectMapper.readValue(message.getBody(), LayoutParsingRequest.class);
if (layoutParsingRequest.layoutParsingType().equals(LayoutParsingType.TAAS) && layoutParsingRequest.researchDocumentStorageId() == null) {
throw new IllegalArgumentException("ResearchDocumentDataStorageId is null!");
}
log.info("Layout parsing request received {}", layoutParsingRequest.identifier());
if (message.getMessageProperties().isRedelivered()) {
throw new AmqpRejectAndDontRequeueException(String.format("Error during last layout parsing of request with identifier: %s, do not retry.",
@@ -2,9 +2,9 @@ package com.knecon.fforesight.service.layoutparser.server;
import org.junit.jupiter.api.Test;
import com.knecon.fforesight.service.layoutparser.server.utils.BaseTest;
import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
class ApplicationTests extends BaseTest {
class ApplicationTests extends AbstractTest {
@Test
void contextLoads() {
@@ -27,12 +27,12 @@ import com.knecon.fforesight.service.layoutparser.processor.python_api.model.tab
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
import com.knecon.fforesight.service.layoutparser.processor.services.mapper.TaasDocumentDataMapper;
import com.knecon.fforesight.service.layoutparser.server.utils.BaseTest;
import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
import com.knecon.fforesight.service.layoutparser.server.utils.visualizations.PdfDraw;
import lombok.SneakyThrows;
public class BdrJsonBuildTest extends BaseTest {
public class BdrJsonBuildTest extends AbstractTest {
@Autowired
private ObjectMapper objectMapper;
@@ -15,11 +15,11 @@ import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsi
import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingPipeline;
import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
import com.knecon.fforesight.service.layoutparser.server.utils.BaseTest;
import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
import lombok.SneakyThrows;
public class BuildDocumentGraphTest extends BaseTest {
public class BuildDocumentGraphTest extends AbstractTest {
@Autowired
protected LayoutParsingPipeline layoutParsingPipeline;
@@ -33,11 +33,11 @@ import com.knecon.fforesight.service.layoutparser.processor.model.table.TablePag
import com.knecon.fforesight.service.layoutparser.processor.services.PdfParsingService;
import com.knecon.fforesight.service.layoutparser.processor.services.SectionsBuilderService;
import com.knecon.fforesight.service.layoutparser.processor.services.classification.RedactManagerClassificationService;
import com.knecon.fforesight.service.layoutparser.server.utils.BaseTest;
import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
import lombok.SneakyThrows;
public class PdfSegmentationServiceTest extends BaseTest {
public class PdfSegmentationServiceTest extends AbstractTest {
@Autowired
private PdfParsingService pdfParsingService;
@@ -20,6 +20,7 @@ import org.springframework.test.context.junit.jupiter.SpringExtension;
import com.iqser.red.storage.commons.service.StorageService;
import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingStorageService;
import com.knecon.fforesight.service.layoutparser.server.Application;
import com.knecon.fforesight.tenantcommons.TenantContext;
@@ -29,8 +30,8 @@ import lombok.SneakyThrows;
@ExtendWith(SpringExtension.class)
@SpringBootTest(classes = Application.class, webEnvironment = SpringBootTest.WebEnvironment.RANDOM_PORT)
@Import(BaseTest.TestConfiguration.class)
public class BaseTest {
@Import(AbstractTest.TestConfiguration.class)
public abstract class AbstractTest {
@Autowired
protected LayoutParsingStorageService layoutParsingStorageService;
@@ -52,18 +53,23 @@ public class BaseTest {
protected final static String POSITION_FILE_ID = "positions";
protected final static String PAGES_FILE_ID = "pages";
protected final static String TENANT_ID = "tenant";
protected final static String SECTION_GRID_ID = "section";
protected final static String SIMPLIFIED_ID = "simplified";
protected LayoutParsingRequest buildStandardLayoutParsingRequest() {
return LayoutParsingRequest.builder()
.imagesFileStorageId(Optional.of(IMAGE_FILE_ID))
.layoutParsingType(LayoutParsingType.REDACT_MANAGER)
.originFileStorageId(ORIGIN_FILE_ID)
.tablesFileStorageId(Optional.of(TABLE_FILE_ID))
.pageFileStorageId(PAGES_FILE_ID)
.positionBlockFileStorageId(POSITION_FILE_ID)
.imagesFileStorageId(Optional.of(IMAGE_FILE_ID))
.structureFileStorageId(STRUCTURE_FILE_ID)
.textBlockFileStorageId(TEXT_FILE_ID)
.positionBlockFileStorageId(POSITION_FILE_ID)
.pageFileStorageId(PAGES_FILE_ID)
.simplifiedTextStorageId(SIMPLIFIED_ID)
.sectionGridStorageId(SECTION_GRID_ID)
.build();
}
@@ -94,13 +100,16 @@ public class BaseTest {
storageService.storeObject(TenantContext.getTenantId(), ORIGIN_FILE_ID, fileInputStream);
return LayoutParsingRequest.builder()
.imagesFileStorageId(Optional.empty())
.layoutParsingType(LayoutParsingType.REDACT_MANAGER)
.originFileStorageId(ORIGIN_FILE_ID)
.tablesFileStorageId(Optional.empty())
.pageFileStorageId(PAGES_FILE_ID)
.positionBlockFileStorageId(POSITION_FILE_ID)
.tablesFileStorageId(Optional.of(TABLE_FILE_ID))
.imagesFileStorageId(Optional.of(IMAGE_FILE_ID))
.structureFileStorageId(STRUCTURE_FILE_ID)
.textBlockFileStorageId(TEXT_FILE_ID)
.positionBlockFileStorageId(POSITION_FILE_ID)
.pageFileStorageId(PAGES_FILE_ID)
.simplifiedTextStorageId(SIMPLIFIED_ID)
.sectionGridStorageId(SECTION_GRID_ID)
.build();
}
@@ -124,13 +133,16 @@ public class BaseTest {
storageService.storeObject(TenantContext.getTenantId(), ORIGIN_FILE_ID, fileStream);
return LayoutParsingRequest.builder()
.imagesFileStorageId(Optional.of(IMAGE_FILE_ID))
.layoutParsingType(LayoutParsingType.REDACT_MANAGER)
.originFileStorageId(ORIGIN_FILE_ID)
.tablesFileStorageId(Optional.of(TABLE_FILE_ID))
.pageFileStorageId(PAGES_FILE_ID)
.positionBlockFileStorageId(POSITION_FILE_ID)
.imagesFileStorageId(Optional.of(IMAGE_FILE_ID))
.structureFileStorageId(STRUCTURE_FILE_ID)
.textBlockFileStorageId(TEXT_FILE_ID)
.positionBlockFileStorageId(POSITION_FILE_ID)
.pageFileStorageId(PAGES_FILE_ID)
.simplifiedTextStorageId(SIMPLIFIED_ID)
.sectionGridStorageId(SECTION_GRID_ID)
.build();
}