取出pdf文档文字
String sourceFolder2 = "E:\picture2\租赁合同2.pdf";
PdfDocument doc = new PdfDocument(new PdfReader(sourceFolder2));
float height = doc.getPage(1).getPageSize().getHeight();
float width = doc.getPage(1).getPageSize().getWidth();
Rectangle rect = new Rectangle(width,height);
FilteredTextEventListener filterListener = new FilteredTextEventListener(new LocationTextExtractionStrategy(), new TextRegionEventFilter(rect));
String extractedText = PdfTextExtractor.getTextFromPage(doc.getPage(1), filterListener);
System.out.println(extractedText);- 上面的例子就可以取出第一页所有的文字，如果需要取出某些文字需要知道文字的具体方位，画个矩形就可以取出
- 以上代码依赖com.itextpdf.kernel
- 取出多个位置的文字
@Test
public void testWithMultiFilteredRenderListener() throws IOException {
PdfDocument pdfDocument = new PdfDocument(new PdfReader(sourceFolder "test.pdf"));
float x1, y1, x2, y2;
FilteredEventListener listener = new FilteredEventListener();
x1 = 122;
x2 = 22;
y1 = 678.9f;
y2 = 12;
ITextExtractionStrategy region1Listener = listener.attachEventListener(new LocationTextExtractionStrategy(),
new TextRegionEventFilter(new Rectangle(x1, y1, x2, y2)));
x1 = 156;
x2 = 13;
y1 = 678.9f;
y2 = 12;
ITextExtractionStrategy region2Listener = listener.attachEventListener(new LocationTextExtractionStrategy(),
new TextRegionEventFilter(new Rectangle(x1, y1, x2, y2)));
PdfCanvasProcessor parser = new PdfCanvasProcessor(new GlyphEventListener(listener));
parser.processPageContent(pdfDocument.getPage(1));
Assert.assertEquals("Your", region1Listener.getResultantText());
Assert.assertEquals("dju", region2Listener.getResultantText());
}- 遍历pdf每个字符
- 之前一直以为Listen监听遍历pdf文本只能一段一段遍历，现在发现他实际上提供了遍历字符的方法
- 两个监听器，一个监听的段落，一个监听每个字符
static class MyEventListener implements IEventListener { private List<Rectangle> rectangles = new ArrayList<>(); @Override public void eventOccurred(IEventData data, EventType type) { if (type == EventType.RENDER_TEXT) { TextRenderInfo renderInfo = (TextRenderInfo) data; Vector startPoint = renderInfo.getDescentLine().getStartPoint(); Vector endPoint = renderInfo.getAscentLine().getEndPoint(); float x1 = Math.min(startPoint.get(0), endPoint.get(0)); float x2 = Math.max(startPoint.get(0), endPoint.get(0)); float y1 = Math.min(startPoint.get(1), endPoint.get(1)); float y2 = Math.max(startPoint.get(1), endPoint.get(1)); rectangles.add(new Rectangle(x1, y1, x2 - x1, y2 - y1)); } } @Override public Set<EventType> getSupportedEvents() { return new LinkedHashSet<>(Collections.singletonList(EventType.RENDER_TEXT)); } public List<Rectangle> getRectangles() { return rectangles; } public void clear() { rectangles.clear(); } } static class MyCharacterEventListener extends MyEventListener { @Override public void eventOccurred(IEventData data, EventType type) { if (type == EventType.RENDER_TEXT) { TextRenderInfo renderInfo = (TextRenderInfo) data; for (TextRenderInfo tri : renderInfo.getCharacterRenderInfos()) { super.eventOccurred(tri, type); } } } }
- 标记每个字符，提供了这样的方法，可以发挥想象做更多的事，给个图片更清楚点
private void parseAndHighlight(String input, String output, boolean singleCharacters) throws IOException { PdfDocument pdfDocument = new PdfDocument(new PdfReader(input), new PdfWriter(output)); MyEventListener myEventListener = singleCharacters ? new MyCharacterEventListener() : new MyEventListener(); PdfDocumentContentParser parser = new PdfDocumentContentParser(pdfDocument); for (int pageNum = 1; pageNum <= pdfDocument.getNumberOfPages(); pageNum ) { parser.processContent(pageNum, myEventListener); List<Rectangle> rectangles = myEventListener.getRectangles(); PdfCanvas canvas = new PdfCanvas(pdfDocument.getPage(pageNum)); canvas.setLineWidth(0.5f); canvas.setStrokeColor(ColorConstants.RED); for (Rectangle rectangle : rectangles) { canvas.rectangle(rectangle); canvas.stroke(); } myEventListener.clear(); } pdfDocument.close(); }
- 要实现上面的效果，只要调用上面的方法即可
@Test public void highlightNotDefTest() throws IOException, InterruptedException { String input = sourceFolder "page229.pdf"; String output = outputPath "page229.pdf"; //false 表示短语单词为单位 true表示每个字符都遍历 parseAndHighlight(input, output, false); }
- false的效果
- 定位某些单词
@Test
public void findPosition() throws Exception {
String sourceFolder2 = "E:\picture2\租赁合同2.pdf";
String output = "E:\picture2\租赁合同2_stroke.pdf";
PdfReader reader = new PdfReader(sourceFolder2);
PdfDocument pdfDocument = new PdfDocument(reader, new PdfWriter(output));
PdfPage lastPage = pdfDocument.getLastPage();
RegexBasedLocationExtractionStrategy strategy = new RegexBasedLocationExtractionStrategy("甲方");
PdfCanvasProcessor canvasProcessor = new PdfCanvasProcessor(strategy);
canvasProcessor.processPageContent(lastPage);
Collection<IPdfTextLocation> resultantLocations = strategy.getResultantLocations();
PdfCanvas pdfCanvas = new PdfCanvas(lastPage);
pdfCanvas.setLineWidth(0.5f);
List<IPdfTextLocation> sets = new ArrayList<>();
for (IPdfTextLocation location : resultantLocations) {
Rectangle rectangle = location.getRectangle();
pdfCanvas.rectangle(rectangle);
pdfCanvas.setStrokeColor(ColorConstants.RED);
pdfCanvas.stroke();
System.out.println(rectangle.getX() "," rectangle.getY() "," rectangle.getLeft() ","
rectangle.getRight() "," rectangle.getTop() "," rectangle.getBottom() ","
rectangle.getWidth() "," rectangle.getHeight());
System.out.println(location.getText());
sets.add(location);
}
Collections.sort(sets, new Comparator<IPdfTextLocation>() {
@Override
public int compare(IPdfTextLocation o1, IPdfTextLocation o2) {
return o1.getRectangle().getY() - o2.getRectangle().getY() > 0 ? 1 : o1.getRectangle().getY() - o2.getRectangle().getY() == 0 ? 0 : -1;
}
});
System.out.println(sets.get(0).getRectangle().getY());
pdfDocument.close();
}- 以下是输出
88.0,297.53,88.0,115.72,311.53,297.53,27.720001,14.0
甲方
213.0,674.176,213.0,241.0,688.176,674.176,28.0,14.0
甲方
227.75,767.7765,227.75,254.75,781.2765,767.7765,27.0,13.5
甲方
322.25,767.7765,322.25,349.25,781.2765,767.7765,27.0,13.5
甲方
297.53- 上面的方法用来合同签章定位上，已经可以做到定位最后某个特定单词
添加文字和图片
代码语言：javascript复制@Test
public void imagesWithDifferentDepth() throws IOException, InterruptedException {
String outFileName = destinationFolder "transparencyTest01.pdf";
String cmpFileName = sourceFolder "cmp_transparencyTest01.pdf";
PdfDocument pdfDocument = new PdfDocument(new PdfWriter(outFileName, new WriterProperties()
.setCompressionLevel(CompressionConstants.NO_COMPRESSION)));
PdfPage page = pdfDocument.addNewPage(PageSize.A3);//默认添加A4
PdfCanvas canvas = new PdfCanvas(page);
canvas.setFillColor(ColorConstants.LIGHT_GRAY).fill();//设置填充背景色
canvas.rectangle(80, 0, 700, 1200).fill();
//开始添加文字
canvas
.saveState()
.beginText()
.moveText(116, 1150) //从哪里开始写
.setFontAndSize(PdfFontFactory.createFont(StandardFonts.HELVETICA), 14) //字体和大小
.setFillColor(ColorConstants.MAGENTA) //字体颜色
.showText("8 bit depth PNG") //具体展示的文字
.endText()
.restoreState();
//读取并添加图片到指定位置
ImageData img = ImageDataFactory.create(sourceFolder "manualTransparency_8bit.png");
canvas.addImage(img, 100, 780, 200, false);
//收尾步骤，关闭画布和pdf，否则pdf打开错误
canvas.release();
pdfDocument.close();
}覆盖原来的文字
- 由于itext没提供替换pdf文字的接口，只能通过覆盖文字的形式完成
public void manipulatePdf(String src, String dest) throws IOException, DocumentException {
PdfReader reader = new PdfReader(src);
PdfStamper stamper = new PdfStamper(reader, new FileOutputStream(dest));
PdfContentByte canvas = stamper.getUnderContent(1);
canvas.saveState();
canvas.setColorFill(BaseColor.YELLOW);
canvas.rectangle(36, 786, 66, 16);
canvas.fill();
canvas.restoreState();
//开始写入文本
canvas.beginText();
for (Entry<String, ReplaceRegion> entry : entrys) {
ReplaceRegion val = entry.getValue();
//设置字体
canvas.setFontAndSize(font.getBaseFont(), getFontSize());
canvas.setTextMatrix(val.getX(),val.getY() 2/*修正背景与文本的相对位置*/);
canvas.showText((String) replaceTextMap.get(value.getAliasName()));
}
canvas.endText();
stamper.close();
reader.close();
}
