com.itextpdf.pdfocr.OcrPdfCreator Maven / Gradle / Ivy
/*
This file is part of the iText (R) project.
Copyright (c) 1998-2020 iText Group NV
Authors: iText Software.
This program is offered under a commercial and under the AGPL license.
For commercial licensing, contact us at https://itextpdf.com/sales. For AGPL licensing, see below.
AGPL licensing:
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see inputImages,
final PdfWriter pdfWriter,
final PdfOutputIntent pdfOutputIntent)
throws OcrException {
LOGGER.info(MessageFormatUtil.format(
PdfOcrLogMessageConstant.START_OCR_FOR_IMAGES,
inputImages.size()));
IMetaInfo storedMetaInfo = null;
if (ocrEngine instanceof IThreadLocalMetaInfoAware) {
storedMetaInfo = ((IThreadLocalMetaInfoAware)ocrEngine).getThreadLocalMetaInfo();
((IThreadLocalMetaInfoAware)ocrEngine).setThreadLocalMetaInfo(
new OcrPdfCreatorMetaInfo(((IThreadLocalMetaInfoAware)ocrEngine).getThreadLocalMetaInfo(),
UUID.randomUUID(),
null != pdfOutputIntent ? PdfDocumentType.PDFA : PdfDocumentType.PDF));
}
// map contains:
// keys: image files
// values:
// map pageNumber -> retrieved text data(text and its coordinates)
Map>> imagesTextData =
new LinkedHashMap>>();
try {
for (File inputImage : inputImages) {
imagesTextData.put(inputImage,
ocrEngine.doImageOcr(inputImage));
}
} finally {
if (ocrEngine instanceof IThreadLocalMetaInfoAware) {
((IThreadLocalMetaInfoAware)ocrEngine).setThreadLocalMetaInfo(storedMetaInfo);
}
}
// create PdfDocument
return createPdfDocument(pdfWriter, pdfOutputIntent, imagesTextData);
}
/**
* Performs OCR with set parameters using provided {@link IOcrEngine} and
* creates PDF using provided {@link com.itextpdf.kernel.pdf.PdfWriter}.
*
* @param inputImages {@link java.util.List} of images to be OCRed
* @param pdfWriter the {@link com.itextpdf.kernel.pdf.PdfWriter} object
* to write final PDF document to
* @return result {@link com.itextpdf.kernel.pdf.PdfDocument} object
* @throws OcrException if provided font is incorrect
*/
public final PdfDocument createPdf(final List inputImages,
final PdfWriter pdfWriter)
throws OcrException {
return createPdfA(inputImages, pdfWriter, null);
}
/**
* Gets used {@link IOcrEngine}.
*
* Returns {@link IOcrEngine} reader object to perform OCR.
* @return selected {@link IOcrEngine} instance
*/
public final IOcrEngine getOcrEngine() {
return ocrEngine;
}
/**
* Sets {@link IOcrEngine} reader object to perform OCR.
* @param reader selected {@link IOcrEngine} instance
*/
public final void setOcrEngine(final IOcrEngine reader) {
ocrEngine = reader;
}
/**
* Adds image (or its one page) and text that was found there to canvas.
*
* @param pdfDocument result {@link com.itextpdf.kernel.pdf.PdfDocument}
* @param imageSize size of the image according to the selected
* {@link ScaleMode}
* @param pageText text that was found on this image (or on this page)
* @param imageData input image if it is a single page or its one page if
* this is a multi-page image
* @param createPdfA3u true if PDF/A3u document is being created
* @throws OcrException if PDF/A3u document is being created and provided
* font contains notdef glyphs
*/
private void addToCanvas(final PdfDocument pdfDocument,
final Rectangle imageSize,
final List pageText, final ImageData imageData,
final boolean createPdfA3u) throws OcrException {
final Rectangle rectangleSize =
ocrPdfCreatorProperties.getPageSize() == null
? imageSize : ocrPdfCreatorProperties.getPageSize();
PageSize size = new PageSize(rectangleSize);
PdfPage pdfPage = pdfDocument.addNewPage(size);
PdfCanvas canvas = new NotDefCheckingPdfCanvas(pdfPage, createPdfA3u);
PdfLayer[] layers = createPdfLayers(ocrPdfCreatorProperties.getImageLayerName(),
ocrPdfCreatorProperties.getTextLayerName(),
pdfDocument);
if (layers[0] != null) {
canvas.beginLayer(layers[0]);
}
addImageToCanvas(imageData, imageSize, canvas);
if (layers[0] != null && layers[0] != layers[1]) {
canvas.endLayer();
}
// how much the original image size changed
float multiplier = imageData == null
? 1 : imageSize.getWidth()
/ PdfCreatorUtil.getPoints(imageData.getWidth());
if (layers[1] != null && layers[0] != layers[1]) {
canvas.beginLayer(layers[1]);
}
try {
addTextToCanvas(imageSize, pageText, canvas, multiplier,
pdfPage.getMediaBox());
} catch (OcrException e) {
LOGGER.error(MessageFormatUtil.format(
OcrException.CANNOT_CREATE_PDF_DOCUMENT,
e.getMessage()));
throw new OcrException(OcrException.CANNOT_CREATE_PDF_DOCUMENT)
.setMessageParams(e.getMessage());
}
if (layers[1] != null) {
canvas.endLayer();
}
}
/**
* Creates a new PDF document using provided properties, adds images with
* recognized text.
*
* @param pdfWriter the {@link com.itextpdf.kernel.pdf.PdfWriter} object
* to write final PDF document to
* @param pdfOutputIntent {@link com.itextpdf.kernel.pdf.PdfOutputIntent}
* for PDF/A-3u document
* @param imagesTextData map that contains input image files as keys,
* and as value: map pageNumber -> text for the page
* @return result {@link com.itextpdf.kernel.pdf.PdfDocument} object
*/
private PdfDocument createPdfDocument(final PdfWriter pdfWriter,
final PdfOutputIntent pdfOutputIntent,
final Map>> imagesTextData) {
PdfDocument pdfDocument;
boolean createPdfA3u = pdfOutputIntent != null;
if (createPdfA3u) {
pdfDocument = new PdfADocument(pdfWriter,
PdfAConformanceLevel.PDF_A_3U, pdfOutputIntent,
new DocumentProperties().setEventCountingMetaInfo(new PdfOcrMetaInfo()));
} else {
pdfDocument = new PdfDocument(pdfWriter,
new DocumentProperties().setEventCountingMetaInfo(new PdfOcrMetaInfo()));
}
// pdfLang should be set in PDF/A mode
boolean hasPdfLangProperty = ocrPdfCreatorProperties.getPdfLang() != null
&& !ocrPdfCreatorProperties.getPdfLang().equals("");
if (createPdfA3u && !hasPdfLangProperty) {
LOGGER.error(MessageFormatUtil.format(
OcrException.CANNOT_CREATE_PDF_DOCUMENT,
PdfOcrLogMessageConstant.PDF_LANGUAGE_PROPERTY_IS_NOT_SET));
throw new OcrException(OcrException.CANNOT_CREATE_PDF_DOCUMENT)
.setMessageParams(PdfOcrLogMessageConstant.PDF_LANGUAGE_PROPERTY_IS_NOT_SET);
}
// add metadata
if (hasPdfLangProperty) {
pdfDocument.getCatalog()
.setLang(new PdfString(ocrPdfCreatorProperties.getPdfLang()));
}
// set title if it is not empty
if (ocrPdfCreatorProperties.getTitle() != null) {
pdfDocument.getCatalog().setViewerPreferences(
new PdfViewerPreferences().setDisplayDocTitle(true));
PdfDocumentInfo info = pdfDocument.getDocumentInfo();
info.setTitle(ocrPdfCreatorProperties.getTitle());
}
// reset passed font provider
ocrPdfCreatorProperties.getFontProvider().reset();
addDataToPdfDocument(imagesTextData, pdfDocument, createPdfA3u);
return pdfDocument;
}
/**
* Places provided images and recognized text to the result PDF document.
*
* @param imagesTextData map that contains input image
* files as keys, and as value:
* map pageNumber -> text for the page
* @param pdfDocument result {@link com.itextpdf.kernel.pdf.PdfDocument}
* @param createPdfA3u true if PDF/A3u document is being created
* @throws OcrException if input image cannot be read or provided font
* contains NOTDEF glyphs
*/
private void addDataToPdfDocument(
final Map>> imagesTextData,
final PdfDocument pdfDocument,
final boolean createPdfA3u) throws OcrException {
for (Map.Entry>> entry
: imagesTextData.entrySet()) {
try {
File inputImage = entry.getKey();
List imageDataList =
PdfCreatorUtil.getImageData(inputImage,
ocrPdfCreatorProperties.getImageRotationHandler());
LOGGER.info(MessageFormatUtil.format(
PdfOcrLogMessageConstant.NUMBER_OF_PAGES_IN_IMAGE,
inputImage.toString(), imageDataList.size()));
Map> imageTextData = entry.getValue();
if (imageTextData.keySet().size() > 0) {
for (int page = 0; page < imageDataList.size(); ++page) {
ImageData imageData = imageDataList.get(page);
final Rectangle imageSize =
PdfCreatorUtil.calculateImageSize(
imageData,
ocrPdfCreatorProperties.getScaleMode(),
ocrPdfCreatorProperties.getPageSize());
if (imageTextData.containsKey(page + 1)) {
addToCanvas(pdfDocument, imageSize,
imageTextData.get(page + 1),
imageData, createPdfA3u);
}
}
}
} catch (IOException e) {
LOGGER.error(MessageFormatUtil.format(
PdfOcrLogMessageConstant.CANNOT_ADD_DATA_TO_PDF_DOCUMENT,
e.getMessage()));
}
}
}
/**
* Places given image to canvas to background to a separate layer.
*
* @param imageData input image as {@link java.io.File}
* @param imageSize size of the image according to the selected
* {@link ScaleMode}
* @param pdfCanvas canvas to place the image
*/
private void addImageToCanvas(final ImageData imageData,
final Rectangle imageSize,
final PdfCanvas pdfCanvas) {
if (imageData != null) {
if (ocrPdfCreatorProperties.getPageSize() == null) {
pdfCanvas.addImage(imageData, imageSize, false);
} else {
final Point coordinates =
PdfCreatorUtil.calculateImageCoordinates(
ocrPdfCreatorProperties.getPageSize(), imageSize);
final Rectangle rect =
new Rectangle(
(float)coordinates.x, (float)coordinates.y,
imageSize.getWidth(), imageSize.getHeight());
pdfCanvas.addImage(imageData, rect, false);
}
}
}
/**
* Places retrieved text to canvas to a separate layer.
*
* @param imageSize size of the image according to the selected
* {@link ScaleMode}
* @param pageText text that was found on this image (or on this page)
* @param pdfCanvas canvas to place the text
* @param multiplier coefficient to adjust text placing on canvas
* @param pageMediaBox page parameters
* @throws OcrException if PDF/A3u document is being created and provided
* font contains notdef glyphs
*/
private void addTextToCanvas(
final Rectangle imageSize,
final List pageText,
final PdfCanvas pdfCanvas,
final float multiplier,
final Rectangle pageMediaBox)
throws OcrException {
if (pageText != null && pageText.size() > 0) {
final Point imageCoordinates =
PdfCreatorUtil.calculateImageCoordinates(
ocrPdfCreatorProperties.getPageSize(), imageSize);
for (TextInfo item : pageText) {
String line = item.getText();
final float bboxWidthPt = getWidthPt(item, multiplier);
final float bboxHeightPt = getHeightPt(item, multiplier);
FontProvider fontProvider = getOcrPdfCreatorProperties()
.getFontProvider();
String fontFamily = getOcrPdfCreatorProperties()
.getDefaultFontFamily();
if (lineNotEmpty(line, bboxHeightPt, bboxWidthPt)) {
Document document = new Document(pdfCanvas.getDocument());
document.setFontProvider(fontProvider);
// Scale the text width to fit the OCR bbox
final float fontSize = PdfCreatorUtil.calculateFontSize(
document, line, fontFamily,
bboxHeightPt, bboxWidthPt);
final float lineWidth = PdfCreatorUtil.getRealLineWidth(document,
line, fontFamily, fontSize);
final float xOffset = getXOffsetPt(item, multiplier);
final float yOffset = getYOffsetPt(item, multiplier, imageSize);
Canvas canvas = new Canvas(pdfCanvas, pageMediaBox);
canvas.setFontProvider(fontProvider);
Text text = new Text(line)
.setHorizontalScaling(bboxWidthPt / lineWidth);
Paragraph paragraph = new Paragraph(text)
.setMargin(0)
.setMultipliedLeading(1.2f);
paragraph.setFontFamily(fontFamily)
.setFontSize(fontSize);
paragraph.setWidth(bboxWidthPt * 1.5f);
if (ocrPdfCreatorProperties.getTextColor() != null) {
paragraph.setFontColor(ocrPdfCreatorProperties.getTextColor());
} else {
paragraph.setTextRenderingMode(TextRenderingMode.INVISIBLE);
}
canvas.showTextAligned(paragraph,
xOffset + (float)imageCoordinates.x,
yOffset + (float)imageCoordinates.y,
TextAlignment.LEFT);
canvas.close();
}
}
}
}
/**
* Creates layers for image and text according rules set in {@link OcrPdfCreatorProperties}.
*
* @param imageLayerName name of the image layer
* @param textLayerName name of the text layer
* @param pdfDocument document to add layers to
*
* @return array of two layers: first layer is for image, second layer is for text.
* Elements may be null meaning that layer creation is not requested
*/
private static PdfLayer[] createPdfLayers(
String imageLayerName,
String textLayerName,
PdfDocument pdfDocument) {
if (imageLayerName == null && textLayerName == null) {
return new PdfLayer[] {null, null};
} else if (imageLayerName == null) {
return new PdfLayer[]{null, new PdfLayer(textLayerName, pdfDocument)};
} else if (textLayerName == null) {
return new PdfLayer[]{new PdfLayer(imageLayerName, pdfDocument), null};
} else if (imageLayerName.equals(textLayerName)) {
PdfLayer pdfLayer = new PdfLayer(imageLayerName, pdfDocument);
return new PdfLayer[] {pdfLayer, pdfLayer};
} else {
return new PdfLayer[] {new PdfLayer(imageLayerName, pdfDocument), new PdfLayer(textLayerName, pdfDocument)};
}
}
/**
* Get left bound of text chunk.
*/
private static float getLeft(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return textInfo.getBbox().get(LEFT_IDX) * multiplier;
} else {
return textInfo.getBboxRect().getLeft() * multiplier;
}
}
/**
* Get right bound of text chunk.
*/
private static float getRight(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return (textInfo.getBbox().get(RIGHT_IDX) + 1) * multiplier - 1;
} else {
return (textInfo.getBboxRect().getRight() + 1) * multiplier - 1;
}
}
/**
* Get top bound of text chunk.
*/
private static float getTop(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return textInfo.getBbox().get(TOP_IDX) * multiplier;
} else {
return textInfo.getBboxRect().getTop() * multiplier;
}
}
/**
* Get bottom bound of text chunk.
*/
private static float getBottom(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return (textInfo.getBbox().get(BOTTOM_IDX) + 1) * multiplier - 1;
} else {
return (textInfo.getBboxRect().getBottom() + 1) * multiplier - 1;
}
}
/**
* Check if line is not empty.
*/
private static boolean lineNotEmpty(String line, float bboxHeightPt, float bboxWidthPt) {
return !line.isEmpty() && bboxHeightPt > 0 && bboxWidthPt > 0;
}
/**
* Get width of text chunk in points.
*/
private static float getWidthPt(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return PdfCreatorUtil.getPoints(
getRight(textInfo, multiplier) - getLeft(textInfo, multiplier));
} else {
return getRight(textInfo, multiplier) - getLeft(textInfo, multiplier);
}
}
/**
* Get height of text chunk in points.
*/
private static float getHeightPt(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return PdfCreatorUtil.getPoints(
getBottom(textInfo, multiplier) - getTop(textInfo, multiplier));
} else {
return getTop(textInfo, multiplier) - getBottom(textInfo, multiplier);
}
}
/**
* Get horizontal text offset in points.
*/
private static float getXOffsetPt(TextInfo textInfo, float multiplier) {
if (textInfo.getBboxRect() == null) {
return PdfCreatorUtil.getPoints(getLeft(textInfo, multiplier));
} else {
return getLeft(textInfo, multiplier);
}
}
/**
* Get vertical text offset in points.
*/
private static float getYOffsetPt(TextInfo textInfo, float multiplier, Rectangle imageSize) {
if (textInfo.getBboxRect() == null) {
return imageSize.getHeight() - PdfCreatorUtil.getPoints(getBottom(textInfo, multiplier));
} else {
return getBottom(textInfo, multiplier);
}
}
/**
* A handler for PDF canvas that validates existing glyphs.
*/
private static class NotDefCheckingPdfCanvas extends PdfCanvas {
private static final long serialVersionUID = 708713860707664107L;
private final boolean createPdfA3u;
public NotDefCheckingPdfCanvas(PdfPage page, boolean createPdfA3u) {
super(page);
this.createPdfA3u = createPdfA3u;
}
@Override
public PdfCanvas showText(GlyphLine text) {
ActualTextCheckingGlyphLine glyphLine =
new ActualTextCheckingGlyphLine(text);
PdfFont currentFont = getGraphicsState().getFont();
boolean notDefGlyphsExists = false;
// default value for error message, it'll be updated with the
// unicode of the not found glyph
String message = PdfOcrLogMessageConstant
.COULD_NOT_FIND_CORRESPONDING_GLYPH_TO_UNICODE_CHARACTER;
for (int i = glyphLine.start; i < glyphLine.end; i++) {
if (isNotDefGlyph(currentFont, glyphLine.get(i))) {
notDefGlyphsExists = true;
message = MessageFormatUtil.format(PdfOcrLogMessageConstant
.COULD_NOT_FIND_CORRESPONDING_GLYPH_TO_UNICODE_CHARACTER,
glyphLine.get(i).getUnicode());
if (this.createPdfA3u) {
// exception is thrown only if PDF/A document is
// being created
throw new OcrException(message);
}
// setting actual text to NotDef glyph
glyphLine.setActualTextToGlyph(i,
glyphLine.toUnicodeString(i, i + 1));
// setting a fake unicode deliberately to pass further
// checks for actual text necessity during iterating over
// glyphline chunks with ActualTextIterator
Glyph glyph = new Glyph(glyphLine.get(i));
glyph.setUnicode(-1);
glyphLine.set(i, glyph);
}
}
// Warning is logged if not PDF/A document is being created
if (notDefGlyphsExists) {
LOGGER.warn(message);
}
return this.showText(glyphLine, new ActualTextIterator(glyphLine));
}
private static boolean isNotDefGlyph(PdfFont font, Glyph glyph) {
if (font instanceof PdfType0Font
|| font instanceof PdfTrueTypeFont) {
return glyph.getCode() == 0;
} else if (font instanceof PdfType1Font
|| font instanceof PdfType3Font) {
return glyph.getCode() == -1;
}
return false;
}
}
/**
* A handler for GlyphLine that checks existing actual text not to
* overwrite it.
*/
private static class ActualTextCheckingGlyphLine extends GlyphLine {
private static final long serialVersionUID = -946356392098459518L;
public ActualTextCheckingGlyphLine(GlyphLine other) {
super(other);
}
public void setActualTextToGlyph(int i, String text) {
// set actual text if it doesn't exist for i-th glyph
if ((this.actualText == null || this.actualText.size() <= i
|| this.actualText.get(i) == null)) {
super.setActualText(i, i + 1, text);
}
}
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy