一、引入maven
<dependency><groupId>org.apache.pdfbox</groupId><artifactId>pdfbox</artifactId><version>2.0.25</version>
</dependency>
二、代码工具类
package com.jiayou.peis.utils;
import com.google.common.collect.Lists;
import com.jiayou.peis.entity.ImageObject;
import org.apache.commons.io.FileUtils;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.cos.COSName;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDResources;
import org.apache.pdfbox.pdmodel.common.PDStream;
import org.apache.pdfbox.pdmodel.graphics.PDXObject;
import org.apache.pdfbox.pdmodel.graphics.image.PDImage;
import org.apache.pdfbox.pdmodel.graphics.image.PDImageXObject;
import org.apache.pdfbox.text.PDFTextStripper;import javax.imageio.ImageIO;
import java.awt.image.BufferedImage;
import java.io.*;
import java.util.ArrayList;
import java.util.List;
public class PdfUtils {
public static String text(byte[] data) throws IOException {return PdfUtils.text(data, true);}public static String text(byte[] data, boolean sortByPosition) throws IOException {ByteArrayInputStream inputStream = new ByteArrayInputStream(data);return PdfUtils.text(inputStream, sortByPosition);}public static String text(File file, boolean sortByPosition) throws IOException {InputStream inputStream = new FileInputStream(file);return PdfUtils.text(inputStream, sortByPosition);}public static String text(File file) throws IOException {return PdfUtils.text(file, true);}public static String text(InputStream inputStream) throws IOException {return text(inputStream, true);}public static String text(InputStream inputStream, boolean sortByPosition) throws IOException {PDDocument document = null;try {
document = Loader.loadPDF(inputStream);PDFTextStripper textStripper = new PDFTextStripper();int numberOfPages = document.getNumberOfPages();textStripper.setStartPage(1);textStripper.setEndPage(numberOfPages);textStripper.setSortByPosition(sortByPosition);textStripper.setShouldSeparateByBeads(true);return StrUtils.removeReturnChar(textStripper.getText(document));} finally {CloseUtils.closeQuietly(document, inputStream);}}public static List<ImageObject> images(File file) throws IOException {InputStream inputStream = new FileInputStream(file);return PdfUtils.images(inputStream);}public static List<ImageObject> images(byte[] data) throws IOException {ByteArrayInputStream inputStream = null;try {inputStream = new ByteArrayInputStream(data);return PdfUtils.images(inputStream);} finally {CloseUtils.closeQuietly(inputStream);}}public static List<ImageObject> images(InputStream inputStream) throws IOException {List<ImageObject> imageList = Lists.newArrayList();PDDocument document = null;try {
document = Loader.loadPDF(inputStream);PDResources pdResources = document.getPage(0).getResources();int i = 0;for (COSName csName : pdResources.getXObjectNames()) {
PDXObject pdxObject = pdResources.getXObject(csName);if (pdxObject instanceof PDImageXObject) {
PDStream pdStream = pdxObject.getStream();PDImageXObject image = new PDImageXObject(pdStream, pdResources);String imageSuffix = imageSuffix(image);BufferedImage bufferedImage = image.getImage();ImageObject object = new ImageObject();object.setIndex(i++);object.setImage(bufferedImage);object.setSuffix(imageSuffix);imageList.add(object);}}} finally {CloseUtils.closeQuietly(document, inputStream);}return imageList;}private static String imageSuffix(PDImageXObject pdImage) throws IOException {String suffix = pdImage.getSuffix();if (suffix == null || "jb2".equals(suffix)) {suffix = "png";} else if ("jpx".equals(suffix)) {suffix = "jp2";}if (hasMasks(pdImage)) {suffix = "png";}return suffix;}private static boolean hasMasks(PDImage pdImage) throws IOException {if (pdImage instanceof PDImageXObject) {PDImageXObject ximg = (PDImageXObject) pdImage;return ximg.getMask() != null || ximg.getSoftMask() != null;}return false;}public static void saveImage(List<ImageObject> imageList, String dir, String prefixName) throws IOException {File imgDir = new File(dir);FileUtils.forceMkdir(imgDir);for(ImageObject image:imageList){File imgFile = new File(dir, prefixName+"_"+image.getIndex()+"."+image.getSuffix());ImageIO.write(image.getImage(), image.getSuffix(), imgFile);}}
}