Marmelator/pdftron-logic-commons
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 32 Wiki Activity

gradle draft

Browse Source
This commit is contained in:
Andrei Isvoran 2024-02-06 16:03:03 +02:00
parent 3bbd13daca
commit 46c9bed663
17 changed files with 454 additions and 190 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
.gitignore vendored
View File

@ -27,3 +27,11 @@
**/classpath-data.json
**/dependencies-and-licenses-overview.txt
git.tag
gradle.properties
gradlew
gradlew.bat
gradle/
**/.gradle
**/build

19
.gitlab-ci.yml
View File

@ -1,4 +1,21 @@
include:
- project: 'gitlab/gitlab'
ref: 'main'
file: 'ci-templates/maven_deps.yml'
file: 'ci-templates/gradle_java.yml'
deploy:
stage: deploy
tags:
- dind
script:
- echo "Building with gradle version ${BUILDVERSION}"
- gradle -Pversion=${BUILDVERSION} publish
- gradle bootBuildImage --cleanCache --publishImage -PbuildbootDockerHostNetwork=true -Pversion=${BUILDVERSION}
- echo "BUILDVERSION=$BUILDVERSION" >> version.env
artifacts:
reports:
dotenv: version.env
rules:
- if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH
- if: $CI_COMMIT_BRANCH =~ /^release/
- if: $CI_COMMIT_TAG

136
build.gradle.kts Normal file
View File

@ -0,0 +1,136 @@
plugins {
`java-library`
`maven-publish`
`kotlin-dsl`
pmd
checkstyle
jacoco
id("io.freefair.lombok") version "8.4"
}
java.sourceCompatibility = JavaVersion.VERSION_17
java.targetCompatibility = JavaVersion.VERSION_17
repositories {
mavenLocal()
maven {
url = uri("https://nexus.knecon.com/repository/gindev/")
}
maven {
url = uri("https://pdftron.com/maven/release")
}
maven {
url = uri("https://repo.spring.io/milestone")
}
maven {
url = uri("http://dist.wso2.org/maven2/")
isAllowInsecureProtocol = true
}
maven {
url = uri("https://projectlombok.org/edge-releases")
}
maven {
url = uri("https://repo.maven.apache.org/maven2/")
}
maven {
url = uri("https://pdftron.com/maven/release")
}
maven {
url = uri("https://nexus.knecon.com/repository/red-platform-releases/")
credentials {
username = providers.gradleProperty("mavenUser").getOrNull();
password = providers.gradleProperty("mavenPassword").getOrNull();
}
}
}
dependencies {
api(libs.org.projectlombok.lombok)
api(libs.org.springframework.boot.spring.boot.configuration.processor)
api(libs.com.google.guava.guava)
api(libs.org.projectlombok.lombok)
api(libs.org.springframework.boot.spring.boot.configuration.processor)
implementation("com.pdftron:PDFNet:10.5.0")
testImplementation(libs.org.junit.jupiter.junit.jupiter)
testImplementation(libs.org.assertj.assertj.core)
testImplementation(libs.org.mockito.mockito.core)
testImplementation(libs.org.apache.logging.log4j.log4j.slf4j2.impl)
testImplementation(libs.org.junit.jupiter.junit.jupiter)
testImplementation(libs.org.assertj.assertj.core)
testImplementation(libs.org.mockito.mockito.core)
compileOnly(libs.org.slf4j.slf4j.api)
compileOnly(libs.com.pdftron.pdfnet)
}
group = "com.iqser.red.commons"
version = "2.0-SNAPSHOT"
description = "pdftron-logic-commons"
java.sourceCompatibility = JavaVersion.VERSION_17
val testsJar by tasks.registering(Jar::class) {
archiveClassifier.set("tests")
from(sourceSets["test"].output)
}
java {
withSourcesJar()
withJavadocJar()
}
publishing {
publications.create<MavenPublication>("maven") {
from(components["java"])
artifact(testsJar)
}
}
tasks.withType<JavaCompile>() {
options.encoding = "UTF-8"
}
tasks.withType<Javadoc>() {
options.encoding = "UTF-8"
}
pmd {
isConsoleOutput = true
}
tasks.pmdMain {
pmd.ruleSetFiles = files("${rootDir}/config/pmd/pmd.xml")
}
tasks.pmdTest {
pmd.ruleSetFiles = files("${rootDir}/config/pmd/test_pmd.xml")
}
tasks.named<Test>("test") {
useJUnitPlatform()
reports {
junitXml.outputLocation.set(layout.buildDirectory.dir("reports/junit"))
}
}
tasks.test {
finalizedBy(tasks.jacocoTestReport) // report is always generated after tests run
}
tasks.jacocoTestReport {
dependsOn(tasks.test) // tests are required to run before generating the report
reports {
xml.required.set(true)
csv.required.set(false)
html.outputLocation.set(layout.buildDirectory.dir("jacocoHtml"))
}
}
java {
withJavadocJar()
}

39
config/checkstyle/checkstyle.xml Normal file
View File

@ -0,0 +1,39 @@
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE module PUBLIC "-//Puppy Crawl//DTD Check Configuration 1.3//EN"
"http://www.puppycrawl.com/dtds/configuration_1_3.dtd">
<module name="Checker">
<property
name="severity"
value="error"/>
<module name="TreeWalker">
<module name="SuppressWarningsHolder"/>
<module name="MissingDeprecated"/>
<module name="MissingOverride"/>
<module name="AnnotationLocation"/>
<module name="JavadocStyle"/>
<module name="NonEmptyAtclauseDescription"/>
<module name="IllegalImport"/>
<module name="RedundantImport"/>
<module name="RedundantModifier"/>
<module name="EmptyBlock"/>
<module name="DefaultComesLast"/>
<module name="EmptyStatement"/>
<module name="EqualsHashCode"/>
<module name="ExplicitInitialization"/>
<module name="IllegalInstantiation"/>
<module name="ModifiedControlVariable"/>
<module name="MultipleVariableDeclarations"/>
<module name="PackageDeclaration"/>
<module name="ParameterAssignment"/>
<module name="SimplifyBooleanExpression"/>
<module name="SimplifyBooleanReturn"/>
<module name="StringLiteralEquality"/>
<module name="OneStatementPerLine"/>
<module name="FinalClass"/>
<module name="ArrayTypeStyle"/>
<module name="UpperEll"/>
<module name="OuterTypeFilename"/>
</module>
<module name="FileTabCharacter"/>
<module name="SuppressWarningsFilter"/>
</module>

20
config/pmd/pmd.xml Normal file
View File

@ -0,0 +1,20 @@
<?xml version="1.0"?>
<ruleset name="Custom ruleset"
xmlns="http://pmd.sourceforge.net/ruleset/2.0.0"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://pmd.sourceforge.net/ruleset/2.0.0 http://pmd.sourceforge.net/ruleset_2_0_0.xsd">
<description>
Knecon ruleset checks the code for bad stuff
</description>
<rule ref="category/java/errorprone.xml">
<exclude name="MissingSerialVersionUID"/>
<exclude name="AvoidLiteralsInIfCondition"/>
<exclude name="AvoidDuplicateLiterals"/>
<exclude name="NullAssignment"/>
<exclude name="AssignmentInOperand"/>
<exclude name="BeanMembersShouldSerialize"/>
</rule>
</ruleset>

22
config/pmd/test_pmd.xml Normal file
View File

@ -0,0 +1,22 @@
<?xml version="1.0"?>
<ruleset name="Custom ruleset"
xmlns="http://pmd.sourceforge.net/ruleset/2.0.0"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://pmd.sourceforge.net/ruleset/2.0.0 http://pmd.sourceforge.net/ruleset_2_0_0.xsd">
<description>
Knecon test ruleset checks the code for bad stuff
</description>
<rule ref="category/java/errorprone.xml">
<exclude name="MissingSerialVersionUID"/>
<exclude name="AvoidLiteralsInIfCondition"/>
<exclude name="AvoidDuplicateLiterals"/>
<exclude name="NullAssignment"/>
<exclude name="AssignmentInOperand"/>
<exclude name="TestClassWithoutTestCases"/>
<exclude name="BeanMembersShouldSerialize"/>
</rule>
</ruleset>

1
settings.gradle.kts Normal file
View File

@ -0,0 +1 @@
rootProject.name = "pdftron-logic-commons"

9
src/main/java/com/iqser/red/pdftronlogic/commons/Converter.java
View File

@ -62,10 +62,11 @@ public class Converter {
@SneakyThrows
public static Color convertColor(ColorSpace colorSpace, ColorPt colorPt) {
ColorPt rgbColor = colorSpace.convert2RGB(colorPt);
Color color = new Color((float) rgbColor.get(0), (float) rgbColor.get(1), (float) rgbColor.get(2));
rgbColor.destroy();
return color;
try (ColorPt rgbColor = colorSpace.convert2RGB(colorPt)) {
Color color = new Color((float) rgbColor.get(0), (float) rgbColor.get(1), (float) rgbColor.get(2));
rgbColor.destroy();
return color;
}
}

1
src/main/java/com/iqser/red/pdftronlogic/commons/ElementFeatures.java
View File

@ -103,6 +103,7 @@ public class ElementFeatures {
@Getter
@SuperBuilder
@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
@SuppressWarnings("PMD")
public static class Text extends ElementFeatures {
String text;

20
src/main/java/com/iqser/red/pdftronlogic/commons/ImageHashFactory.java
View File

@ -35,17 +35,19 @@ public class ImageHashFactory {
@SneakyThrows
private byte[] getBytesOfImage(com.pdftron.pdf.Image inputImage) {
// 0 because the memory filter determines the size
var memFilter = new MemoryFilter(0, false);
var filterWriter = new FilterWriter(memFilter);
try(var memFilter = new MemoryFilter(0, false)) {
try(var filterWriter = new FilterWriter(memFilter)) {
inputImage.export(filterWriter);
filterWriter.flushAll();
byte[] res = memFilter.getBuffer();
inputImage.export(filterWriter);
filterWriter.flushAll();
byte[] res = memFilter.getBuffer();
memFilter.flushAll();
memFilter.destroy();
filterWriter.destroy();
return res;
memFilter.flushAll();
memFilter.destroy();
filterWriter.destroy();
return res;
}
}
}

205
src/main/java/com/iqser/red/pdftronlogic/commons/InvisibleElementRemovalService.java
View File

@ -151,39 +151,40 @@ public class InvisibleElementRemovalService {
private void execute(PDFDoc pdfDoc, boolean delta, boolean removePaths, Set<String> markedContentToIgnore) {
log.info("Start removing invisible Elements");
ElementWriter writer = new ElementWriter();
ElementReader reader = new ElementReader();
Set<Long> visitedXObjIds = new TreeSet<>();
try(ElementWriter writer = new ElementWriter()) {
try(ElementReader reader = new ElementReader()) {
Set<Long> visitedXObjIds = new TreeSet<>();
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
Page page = iterator.next();
Page page = iterator.next();
visitedXObjIds.add(page.getSDFObj().getObjNum());
visitedXObjIds.add(page.getSDFObj().getObjNum());
InvisibleElementRemovalContext context = InvisibleElementRemovalContext.builder()
.reader(reader)
.clippingPathStack(new ClippingPathStack(page.getMediaBox()))
.markedContentStack(new MarkedContentStack())
.removePaths(removePaths)
.delta(delta)
.overlappedElements(new ArrayList<>())
.visibleElements(new ArrayList<>())
.visitedXObjIds(visitedXObjIds)
.markedContentToIgnore(markedContentToIgnore)
.build();
InvisibleElementRemovalContext context = InvisibleElementRemovalContext.builder()
.reader(reader)
.clippingPathStack(new ClippingPathStack(page.getMediaBox()))
.markedContentStack(new MarkedContentStack())
.removePaths(removePaths)
.delta(delta)
.overlappedElements(new ArrayList<>())
.visibleElements(new ArrayList<>())
.visitedXObjIds(visitedXObjIds)
.markedContentToIgnore(markedContentToIgnore)
.build();
removeClippedElementsAndInvisibleTextAndRememberOverlappedElements(page, writer, context);
removeClippedElementsAndInvisibleTextAndRememberOverlappedElements(page, writer, context);
context.visitedXObjIds().clear();
context.markedContentStack().clear();
context.visitedXObjIds().clear();
context.markedContentStack().clear();
removeOverlappedElements(page, writer, context);
removeOverlappedElements(page, writer, context);
}
writer.destroy();
reader.destroy();
}
}
writer.destroy();
reader.destroy();
log.info("Finished removing invisible Elements");
}
@ -239,68 +240,70 @@ public class InvisibleElementRemovalService {
private void processImages(Element imageElement, ElementWriter writer, InvisibleElementRemovalContext context) throws PDFNetException {
Rect rect = imageElement.getBBox();
try(Rect rect = imageElement.getBBox()) {
if (rect == null) {
return;
}
if (rect == null) {
return;
}
boolean inClippingPath = context.clippingPathStack().almostIntersects(rect.getX1(), rect.getY1(), rect.getWidth(), rect.getHeight());
boolean inClippingPath = context.clippingPathStack().almostIntersects(rect.getX1(), rect.getY1(), rect.getWidth(), rect.getHeight());
if (!context.delta() && inClippingPath) {
context.visibleElements().add(ElementFeatureFactory.extractFeatures(imageElement));
}
if (!context.delta() && inClippingPath) {
context.visibleElements().add(ElementFeatureFactory.extractFeatures(imageElement));
}
if (context.delta() ^ inClippingPath) {
writer.writeElement(imageElement);
if (context.delta() ^ inClippingPath) {
writer.writeElement(imageElement);
}
}
}
private void processText(Element textElement, ElementWriter writer, InvisibleElementRemovalContext context) throws PDFNetException {
Rect textBBox = textElement.getBBox();
try(Rect textBBox = textElement.getBBox()) {
if (textBBox == null) {
writer.writeElement(textElement);
return;
}
GState gState = textElement.getGState();
boolean inClippingPath = context.clippingPathStack().almostIntersects(textBBox.getX1(), textBBox.getY1(), textBBox.getWidth(), textBBox.getHeight());
boolean isTextVisible = isTextRenderedVisibly(gState, textBBox, context);
if (inClippingPath && isTextVisible) {
context.visibleElements().add(ElementFeatureFactory.extractFeatures(textElement));
}
if (!context.delta()) {
if (inClippingPath && isTextVisible) {
if (textBBox == null) {
writer.writeElement(textElement);
} else if (textElement.hasTextMatrix()) {
return;
}
GState gState = textElement.getGState();
boolean inClippingPath = context.clippingPathStack().almostIntersects(textBBox.getX1(), textBBox.getY1(), textBBox.getWidth(), textBBox.getHeight());
boolean isTextVisible = isTextRenderedVisibly(gState, textBBox, context);
if (inClippingPath && isTextVisible) {
context.visibleElements().add(ElementFeatureFactory.extractFeatures(textElement));
}
if (!context.delta()) {
if (inClippingPath && isTextVisible) {
writer.writeElement(textElement);
} else if (textElement.hasTextMatrix()) {
/*
PDFTron Element with type "text" refers to a Tj command. If a Tm command is just above it in the pdf file, PDFTron will join the two commands and treat them as one Element.
hasTextMatrix() checks for this case specifically. Also, Tm changes the position for a whole BT/ET segment, possibly containing multiple Tj commands.
Therefore, the position of a following Tj is affected by not writing the first Element.
This is why, we write only the Tm command:
*/
writer.writeGStateChanges(textElement);
}
} else {
if (!inClippingPath) {
gState.setFillColorSpace(ColorSpace.createDeviceRGB());
// red for elements removed by clipping path
gState.setFillColor(new ColorPt(1, 0, 0));
writer.writeElement(textElement);
}
if (!isTextVisible) {
gState.setFillColorSpace(ColorSpace.createDeviceRGB());
// blue for elements removed due to transparency or not rendered or same color as background
gState.setFillColor(new ColorPt(0, 0, 1));
gState.setTextRenderMode(GState.e_fill_text);
gState.setFillOpacity(1);
writer.writeElement(textElement);
writer.writeGStateChanges(textElement);
}
} else {
if (!inClippingPath) {
gState.setFillColorSpace(ColorSpace.createDeviceRGB());
// red for elements removed by clipping path
gState.setFillColor(new ColorPt(1, 0, 0));
writer.writeElement(textElement);
}
if (!isTextVisible) {
gState.setFillColorSpace(ColorSpace.createDeviceRGB());
// blue for elements removed due to transparency or not rendered or same color as background
gState.setFillColor(new ColorPt(0, 0, 1));
gState.setTextRenderMode(GState.e_fill_text);
gState.setFillOpacity(1);
writer.writeElement(textElement);
}
}
}
}
@ -315,20 +318,21 @@ public class InvisibleElementRemovalService {
context.visitedXObjIds().add(formObj.getObjNum());
// writer needs to be newly initialized when entering a new content stream
// see ElementEditTest in PDFTron (https://www.pdftron.com/documentation/samples/android/java/ElementEditTest)
ElementWriter formWriter = new ElementWriter();
context.clippingPathStack().enterNewGState();
context.clippingPathStack().intersectClippingPath(new GeneralPath(Converter.toRectangle2D(formElement.getBBox())));
context.reader().formBegin();
formWriter.begin(formObj);
try(ElementWriter formWriter = new ElementWriter()) {
context.clippingPathStack().enterNewGState();
context.clippingPathStack().intersectClippingPath(new GeneralPath(Converter.toRectangle2D(formElement.getBBox())));
context.reader().formBegin();
formWriter.begin(formObj);
context.reader().clearChangeList();
formWriter.setDefaultGState(context.reader());
context.reader().clearChangeList();
formWriter.setDefaultGState(context.reader());
processElements(formWriter, context);
formWriter.end();
formWriter.destroy();
context.reader().end();
context.clippingPathStack().leaveGState();
processElements(formWriter, context);
formWriter.end();
formWriter.destroy();
context.reader().end();
context.clippingPathStack().leaveGState();
}
}
}
@ -481,17 +485,18 @@ public class InvisibleElementRemovalService {
context.visitedXObjIds().add(formObj.getObjNum());
// writer needs to be newly initialized when entering a new content stream
// see ElementEditTest in PDFTron (https://www.pdftron.com/documentation/samples/android/java/ElementEditTest)
ElementWriter formWriter = new ElementWriter();
context.reader().formBegin();
formWriter.begin(formObj);
try(ElementWriter formWriter = new ElementWriter()) {
context.reader().formBegin();
formWriter.begin(formObj);
context.reader().clearChangeList();
formWriter.setDefaultGState(context.reader());
context.reader().clearChangeList();
formWriter.setDefaultGState(context.reader());
processOverlappedElements(formWriter, context);
formWriter.end();
formWriter.destroy();
context.reader().end();
processOverlappedElements(formWriter, context);
formWriter.end();
formWriter.destroy();
context.reader().end();
}
}
}
@ -582,18 +587,20 @@ public class InvisibleElementRemovalService {
@SneakyThrows
private void drawBBox(ElementWriter writer, Rectangle2D r, String hexcolor) {
ColorPt colorPt = new ColorPt(Integer.valueOf(hexcolor.substring(1, 3), 16) / 255d,
try(ColorPt colorPt = new ColorPt(Integer.valueOf(hexcolor.substring(1, 3), 16) / 255d,
Integer.valueOf(hexcolor.substring(3, 5), 16) / 255d,
Integer.valueOf(hexcolor.substring(5, 7), 16) / 255d);
ElementBuilder eb = new ElementBuilder();
Element rect = eb.createRect(r.getX(), r.getY(), r.getWidth(), r.getHeight());
rect.setPathStroke(true);
rect.getGState().setStrokeColorSpace(ColorSpace.createDeviceRGB());
rect.getGState().setStrokeColor(colorPt);
writer.writePlacedElement(rect);
Integer.valueOf(hexcolor.substring(5, 7), 16) / 255d)) {
try(ElementBuilder eb = new ElementBuilder()) {
Element rect = eb.createRect(r.getX(), r.getY(), r.getWidth(), r.getHeight());
rect.setPathStroke(true);
rect.getGState().setStrokeColorSpace(ColorSpace.createDeviceRGB());
rect.getGState().setStrokeColor(colorPt);
writer.writePlacedElement(rect);
colorPt.destroy();
eb.destroy();
colorPt.destroy();
eb.destroy();
}
}
}

18
src/main/java/com/iqser/red/pdftronlogic/commons/MarkedContentStack.java
View File

@ -7,33 +7,33 @@ import java.util.Set;
public class MarkedContentStack {
Deque<MarkedContent> markedContentStack = new LinkedList<>();
Deque<MarkedContent> markedContentQueue = new LinkedList<>();
public void enterMarkedContent(String name) {
markedContentStack.push(new MarkedContent(name));
markedContentQueue.push(new MarkedContent(name));
}
public void leaveMarkedContent() {
markedContentStack.pop();
markedContentQueue.pop();
}
public String currentMarkedContent() {
if (markedContentStack.isEmpty()) {
if (markedContentQueue.isEmpty()) {
return "";
}
return markedContentStack.peek().name();
return markedContentQueue.peek().name();
}
public boolean currentMarkedContentContains(String name) {
Iterator<MarkedContent> markedContentIterator = markedContentStack.descendingIterator();
Iterator<MarkedContent> markedContentIterator = markedContentQueue.descendingIterator();
while (markedContentIterator.hasNext()) {
var markedContent = markedContentIterator.next();
if (markedContent.name().equals(name)) {
@ -46,10 +46,10 @@ public class MarkedContentStack {
public boolean currentMarkedContentContainsAny(Set<String> names) {
if (markedContentStack.isEmpty()) {
if (markedContentQueue.isEmpty()) {
return false;
}
Iterator<MarkedContent> markedContentIterator = markedContentStack.descendingIterator();
Iterator<MarkedContent> markedContentIterator = markedContentQueue.descendingIterator();
while (markedContentIterator.hasNext()) {
var markedContent = markedContentIterator.next();
if (names.contains(markedContent.name())) {
@ -62,7 +62,7 @@ public class MarkedContentStack {
public void clear() {
markedContentStack.clear();
markedContentQueue.clear();
}

39
src/main/java/com/iqser/red/pdftronlogic/commons/OCGWatermarkRemovalService.java
View File

@ -43,18 +43,20 @@ public class OCGWatermarkRemovalService {
@SneakyThrows
private void removeOCGWatermarks(PDFDoc pdfDoc) {
ElementReader reader = new ElementReader();
ElementWriter writer = new ElementWriter();
Set<Long> visitedXObjIds = new TreeSet<>();
try(ElementReader reader = new ElementReader()) {
try(ElementWriter writer = new ElementWriter()) {
Set<Long> visitedXObjIds = new TreeSet<>();
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
Page page = iterator.next();
writeAllElementsExceptWatermarks(page, reader, writer, visitedXObjIds);
Page page = iterator.next();
writeAllElementsExceptWatermarks(page, reader, writer, visitedXObjIds);
}
reader.destroy();
writer.destroy();
}
}
reader.destroy();
writer.destroy();
}
@ -124,17 +126,18 @@ public class OCGWatermarkRemovalService {
visitedXObjIds.add(element.getXObject().getObjNum());
// writer needs to be newly initialized when entering a new content stream
// see ElementEditTest in PDFTron (https://www.pdftron.com/documentation/samples/android/java/ElementEditTest)
ElementWriter formWriter = new ElementWriter();
reader.formBegin();
formWriter.begin(element.getXObject());
try(ElementWriter formWriter = new ElementWriter()) {
reader.formBegin();
formWriter.begin(element.getXObject());
reader.clearChangeList();
formWriter.setDefaultGState(reader);
reader.clearChangeList();
formWriter.setDefaultGState(reader);
processElements(page, reader, formWriter, visitedXObjIds);
formWriter.end();
formWriter.destroy();
reader.end();
processElements(page, reader, formWriter, visitedXObjIds);
formWriter.end();
formWriter.destroy();
reader.end();
}
}
}

25
src/main/java/com/iqser/red/pdftronlogic/commons/PdfTextExtraction.java
View File

@ -16,20 +16,21 @@ import lombok.experimental.UtilityClass;
@UtilityClass
public class PdfTextExtraction {
private static String execute(PDFDoc pdfDoc) throws IOException, PDFNetException{
TextExtractor extractor = new TextExtractor();
List<String> texts = new ArrayList<>();
private static String execute(PDFDoc pdfDoc) throws PDFNetException{
try(TextExtractor extractor = new TextExtractor()) {
List<String> texts = new ArrayList<>();
PageIterator iterator = pdfDoc.getPageIterator();
while (iterator.hasNext()) {
Page page = iterator.next();
extractor.begin(page);
texts.add(extractor.getAsText());
PageIterator iterator = pdfDoc.getPageIterator();
while (iterator.hasNext()) {
Page page = iterator.next();
extractor.begin(page);
texts.add(extractor.getAsText());
}
extractor.destroy();
pdfDoc.close();
return String.join("\n", texts);
}
extractor.destroy();
pdfDoc.close();
return String.join("\n", texts);
}
public static String extractAllTextFromDocument(InputStream fileStream) throws IOException, PDFNetException {

80
src/main/java/com/iqser/red/pdftronlogic/commons/WatermarkRemovalService.java
View File

@ -93,27 +93,28 @@ public class WatermarkRemovalService {
Map<Long, List<ElementFeatures>> formObjectsAndImagesForPages = new HashMap<>();
Set<Long> visitedXObjIds = new TreeSet<>();
ElementReader reader = new ElementReader();
try(ElementReader reader = new ElementReader()) {
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
Page page = iterator.next();
Page page = iterator.next();
double minAreaCoveringFromPage = AREA_THRESHOLD * page.getPageHeight() * page.getPageWidth();
double minAreaCoveringFromPage = AREA_THRESHOLD * page.getPageHeight() * page.getPageWidth();
LinkedList<ElementFeatures> elementFeaturesLinkedList = new LinkedList<>();
LinkedList<ElementFeatures> elementFeaturesLinkedList = new LinkedList<>();
reader.begin(page);
for (Element element = reader.next(); element != null; element = reader.next()) {
processElement(element, visitedXObjIds, elementFeaturesLinkedList, formObjectsOccuringMoreThanOnceOnAPage, minAreaCoveringFromPage, page);
reader.begin(page);
for (Element element = reader.next(); element != null; element = reader.next()) {
processElement(element, visitedXObjIds, elementFeaturesLinkedList, formObjectsOccuringMoreThanOnceOnAPage, minAreaCoveringFromPage, page);
}
formObjectsAndImagesForPages.put(page.getSDFObj().getObjNum(), elementFeaturesLinkedList);
}
formObjectsAndImagesForPages.put(page.getSDFObj().getObjNum(), elementFeaturesLinkedList);
reader.destroy();
return formObjectsAndImagesForPages;
}
reader.destroy();
return formObjectsAndImagesForPages;
}
@ -220,13 +221,14 @@ public class WatermarkRemovalService {
}
if (visitedXObjIds.add(element.getXObject().getObjNum())) {
ElementReader xObjectReader = new ElementReader();
xObjectReader.begin(element.getXObject());
for (Element element1 = xObjectReader.next(); element1 != null; element1 = xObjectReader.next()) {
processElement(element1, visitedXObjIds, elementFeaturesLinkedList, formObjectsOccuringMoreThanOnceOnAPage, minAreaCoveringPage, page);
try(ElementReader xObjectReader = new ElementReader()) {
xObjectReader.begin(element.getXObject());
for (Element element1 = xObjectReader.next(); element1 != null; element1 = xObjectReader.next()) {
processElement(element1, visitedXObjIds, elementFeaturesLinkedList, formObjectsOccuringMoreThanOnceOnAPage, minAreaCoveringPage, page);
}
elementFeaturesLinkedList.add(ElementFeatureFactory.extractFeatures(element));
xObjectReader.destroy();
}
elementFeaturesLinkedList.add(ElementFeatureFactory.extractFeatures(element));
xObjectReader.destroy();
} else {
elementFeaturesLinkedList.add(ElementFeatureFactory.extractFeatures(element));
}
@ -256,21 +258,22 @@ public class WatermarkRemovalService {
@SneakyThrows
private void removeAllWatermarks(PDFDoc pdfDoc, List<ElementFeatures> watermarksElementFeaturesList) {
ElementReader reader = new ElementReader();
ElementWriter writer = new ElementWriter();
Set<Long> visitedXObjIds = new TreeSet<>();
try(ElementReader reader = new ElementReader()) {
try (ElementWriter writer = new ElementWriter()) {
Set<Long> visitedXObjIds = new TreeSet<>();
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
for (PageIterator iterator = pdfDoc.getPageIterator(); iterator.hasNext(); ) {
Page page = iterator.next();
Page page = iterator.next();
writeAllElementsExceptWatermarks(page, reader, writer, watermarksElementFeaturesList, visitedXObjIds);
writeAllElementsExceptWatermarks(page, reader, writer, watermarksElementFeaturesList, visitedXObjIds);
}
reader.destroy();
writer.destroy();
}
}
reader.destroy();
writer.destroy();
}
@ -390,17 +393,18 @@ public class WatermarkRemovalService {
visitedXObjIds.add(element.getXObject().getObjNum());
// writer needs to be newly initialized when entering a new content stream
// see ElementEditTest in PDFTron (https://www.pdftron.com/documentation/samples/android/java/ElementEditTest)
ElementWriter formWriter = new ElementWriter();
reader.formBegin();
formWriter.begin(element.getXObject());
try(ElementWriter formWriter = new ElementWriter()) {
reader.formBegin();
formWriter.begin(element.getXObject());
reader.clearChangeList();
formWriter.setDefaultGState(reader);
reader.clearChangeList();
formWriter.setDefaultGState(reader);
processElements(page, reader, formWriter, watermarksElementFeaturesList, visitedXObjIds);
formWriter.end();
formWriter.destroy();
reader.end();
processElements(page, reader, formWriter, watermarksElementFeaturesList, visitedXObjIds);
formWriter.end();
formWriter.destroy();
reader.end();
}
}
}

1
src/test/java/com/iqser/red/pdftronlogic/commons/InvisibleElementRemovalServiceTest.java
View File

@ -13,6 +13,7 @@ import com.pdftron.pdf.PDFNet;
import lombok.SneakyThrows;
@SuppressWarnings("PMD")
class InvisibleElementRemovalServiceTest {
InvisibleElementRemovalService invisibleElementRemovalService;

1
src/test/java/com/iqser/red/pdftronlogic/commons/WatermarkRemovalServiceTest.java
View File

@ -13,6 +13,7 @@ import com.pdftron.pdf.PDFNet;
import lombok.SneakyThrows;
@Disabled
@SuppressWarnings("PMD")
class WatermarkRemovalServiceTest {
@SneakyThrows