/**
* Performs OCR operation.
*
* @param imageList a list of <code>IIOImage</code> objects
* @param rect the bounding rectangle defines the region of the image to be recognized. A
* rectangle of zero dimension or <code>null</code> indicates the whole image.
* @return the recognized text
* @throws TesseractException
*/
public String doOCR(List<IIOImage> imageList, Rectangle rect) throws TesseractException {
init();
setTessVariables();
try {
StringBuilder sb = new StringBuilder();
for (IIOImage oimage : imageList) {
pageNum++;
try {
setImage(oimage.getRenderedImage(), rect);
sb.append(getOCRText());
} catch (IOException ioe) {
// skip the problematic image
logger.log(Level.SEVERE, ioe.getMessage(), ioe);
}
}
if (hocr) {
sb.insert(0, htmlBeginTag).append(htmlEndTag);
}
return sb.toString();
} finally {
dispose();
}
}
// tokenizes a string for PBM and PGM headers and plain PBM and PGM data. If oneChar is true,
// then
// rather than parsing a whole string, a single character is read (not including whitespace or
// comments)
static String tokenizeString(InputStream stream, boolean oneChar) throws IOException {
final int EOF = -1;
StringBuilder b = new StringBuilder();
int c;
boolean inComment = false;
while (true) {
c = stream.read();
if (c == EOF)
throw new IOException(
"Stream ended prematurely, before table reading was completed."); // no more tokens
else if (inComment) {
if (c == '\r' || c == '\n') // escape the comment
inComment = false;
else {
} // do nothing
} else if (Character.isWhitespace((char) c)) {
} // do nothing
else if (c == '#') // start of a comment
{
inComment = true;
} else // start of a string
{
b.append((char) c);
break;
}
}
if (oneChar) return b.toString();
// at this point we have a valid string. We read until whitespace or a #
while (true) {
c = stream.read();
if (c == EOF) break;
else if (c == '#') // start of comment, read until a '\n'
{
while (true) {
c = stream.read(); // could hit EOF, which is fine
if (c == EOF) break;
else if (c == '\r' || c == '\n') break;
}
// break; // comments are not delimiters
} else if (Character.isWhitespace((char) c)) break;
else b.append((char) c);
}
return b.toString();
}
/**
* Writes the <code>shape</code>, <code>coords</code>, <code>href</code>,
* <code>nohref</code> Attribute for the specified figure and shape.
*
* @return Returns true, if the polygon is inside of the image bounds.
*/
private boolean writePolyAttributes(IXMLElement elem, SVGFigure f, Shape shape) {
AffineTransform t = TRANSFORM.getClone(f);
if (t == null) {
t = drawingTransform;
} else {
t.preConcatenate(drawingTransform);
}
StringBuilder buf = new StringBuilder();
float[] coords = new float[6];
GeneralPath path = new GeneralPath();
for (PathIterator i = shape.getPathIterator(t, 1.5f);
! i.isDone(); i.next()) {
switch (i.currentSegment(coords)) {
case PathIterator.SEG_MOVETO :
if (buf.length() != 0) {
throw new IllegalArgumentException("Illegal shape "+shape);
}
if (buf.length() != 0) {
buf.append(',');
}
buf.append((int) coords[0]);
buf.append(',');
buf.append((int) coords[1]);
path.moveTo(coords[0], coords[1]);
break;
case PathIterator.SEG_LINETO :
if (buf.length() != 0) {
buf.append(',');
}
buf.append((int) coords[0]);
buf.append(',');
buf.append((int) coords[1]);
path.lineTo(coords[0], coords[1]);
break;
case PathIterator.SEG_CLOSE :
path.closePath();
break;
default :
throw new InternalError("Illegal segment type "+i.currentSegment(coords));
}
}
elem.setAttribute("shape", "poly");
elem.setAttribute("coords", buf.toString());
writeHrefAttribute(elem, f);
return path.intersects(new Rectangle2D.Float(bounds.x, bounds.y, bounds.width, bounds.height));
}
public void draw(BufferedImage buf) {
WritableRaster dst = buf.getRaster();
blit(dst, bg.getRaster(), Coord.z);
alphablit(dst, rmeter(sbars[0].getRaster(), lmax[0], max), mc[0]);
alphablit(
dst, lmeter(sbars[1].getRaster(), lmax[1], max), mc[1].sub(bars[1].getWidth() - 1, 0));
alphablit(
dst, lmeter(sbars[2].getRaster(), lmax[2], max), mc[2].sub(bars[2].getWidth() - 1, 0));
alphablit(dst, rmeter(sbars[3].getRaster(), lmax[3], max), mc[3]);
if (lfood != null) {
double e = foodeff(lfood);
alphablit(dst, rgmeter(lfood, e, 0), mc[0]);
alphablit(dst, lgmeter(lfood, e, 1), mc[1].sub(bars[1].getWidth() - 1, 0));
alphablit(dst, lgmeter(lfood, e, 2), mc[2].sub(bars[1].getWidth() - 1, 0));
alphablit(dst, rgmeter(lfood, e, 3), mc[3]);
}
alphablit(dst, rmeter(bars[0].getRaster(), lev[0], max), mc[0]);
alphablit(dst, lmeter(bars[1].getRaster(), lev[1], max), mc[1].sub(bars[1].getWidth() - 1, 0));
alphablit(dst, lmeter(bars[2].getRaster(), lev[2], max), mc[2].sub(bars[2].getWidth() - 1, 0));
alphablit(dst, rmeter(bars[3].getRaster(), lev[3], max), mc[3]);
StringBuilder tbuf = new StringBuilder();
for (int i = 0; i < 4; i++)
tbuf.append(
String.format(
"%s: %s/%s\n", rnm[i], Utils.fpformat(lev[i], 3, 1), Utils.fpformat(lmax[i], 3, 1)));
tooltip = RichText.render(tbuf.toString(), 0).tex();
}
public static String getLines(String s) {
StringBuilder sb = new StringBuilder();
int index = 0;
for (int i = 1; index != -1; i++) {
sb.append(i + "\n");
index = s.indexOf("\n", index + 1);
}
return sb.toString();
}
public static String inputStreamReaderToString(InputStreamReader in) throws Exception {
StringBuilder sb = new StringBuilder();
char[] buf = new char[1025];
while (in.ready()) {
sb.append((new String(buf, 0, in.read(buf, 0, 1024))).replaceAll("\r\n", "\n"));
}
in.close();
String tmp = sb.toString();
while (tmp.contains("\r\n")) {
tmp = tmp.replaceAll("\r\n", "\n");
}
return tmp;
}
private StringBuilder createOutpuFile(
int numOfTags,
int[] tags,
int tagStartIndexForFilename,
boolean multiple,
int subImageIndex) {
final StringBuilder filenameBuilder = new StringBuilder();
int i = tagStartIndexForFilename;
while (true) {
filenameBuilder.append(names[tags[i]]);
if (++i == numOfTags) {
break;
} else {
filenameBuilder.append('.');
}
}
if (multiple) {
filenameBuilder.append('-').append(subImageIndex);
}
return filenameBuilder.append(".png");
}
public void decode() throws IOException {
readTagDescriptors();
final MessageDigest md;
try {
md = MessageDigest.getInstance("SHA-512");
} catch (NoSuchAlgorithmException e) {
e.printStackTrace();
return;
}
final ComponentColorModel colorModel =
new ComponentColorModel(
ColorSpace.getInstance(ColorSpace.CS_sRGB),
true,
false,
Transparency.TRANSLUCENT,
DataBuffer.TYPE_BYTE);
final int[] bandOffsets = {2, 1, 0, 3};
final Map<String, String> duplicateGuard = new HashMap<String, String>();
final byte[] bytes = data.array();
final int[] tags = new int[8];
final List<String> symLinkCommand = new ArrayList<String>(4);
symLinkCommand.add("ln");
symLinkCommand.add("-s");
symLinkCommand.add(null);
symLinkCommand.add(null);
for (int fileIndex = 0; fileIndex < fileCount; fileIndex++) {
data.position(fileDescriptorsOffset + ((4 + 8) * fileIndex));
final int fileDataOffset = data.getInt();
int numOfTags = 8;
for (int i = 0; i < 8; i++) {
int tag = data.get() & 0xff;
if (tag == 0) {
numOfTags = i;
break;
}
tags[i] = tag;
}
data.position(fileDataSectionOffset + fileDataOffset);
final int artRows = data.getShort();
final int artColumns = data.getShort();
// skip unknown
data.position(data.position() + 28);
final int[] subimageOffsets = readNumericArray(9, false);
final int[] subimageWidths = readNumericArray(9, true);
final int[] subimageHeights = readNumericArray(9, true);
assert numOfTags >= 2;
final File dir;
final int tagStartIndexForFilename;
if (numOfTags == 2) {
tagStartIndexForFilename = 1;
dir = new File(outputDir, names[tags[0]].toString());
} else {
tagStartIndexForFilename = 2;
dir = new File(outputDir, names[tags[0]] + "/" + names[tags[1]]);
}
boolean dirCreated = false;
final int imageCount = artRows * artColumns;
for (int subImageIndex = 0; subImageIndex < imageCount; subImageIndex++) {
final int w = subimageWidths[subImageIndex];
final int h = subimageHeights[subImageIndex];
if (w <= 0 || h <= 0) {
continue;
}
final int srcPos = fileDataSectionOffset + fileDataOffset + subimageOffsets[subImageIndex];
final int srcLength = w * h * 4;
md.update(bytes, srcPos, srcLength);
md.update((byte) w);
md.update((byte) h);
final String digest = convertToHex(md.digest());
md.reset();
if (!dirCreated) {
//noinspection ResultOfMethodCallIgnored
dir.mkdirs();
dirCreated = true;
}
final byte[] bgra = new byte[srcLength];
// cannot pass bytes directly, offset in DataBufferByte is not working
System.arraycopy(bytes, srcPos, bgra, 0, bgra.length);
BufferedImage image =
new BufferedImage(
colorModel,
(WritableRaster)
Raster.createRaster(
new PixelInterleavedSampleModel(
DataBuffer.TYPE_BYTE, w, h, 4, w * 4, bandOffsets),
new DataBufferByte(bgra, bgra.length),
null),
false,
null);
final StringBuilder outFilename =
createOutpuFile(
numOfTags, tags, tagStartIndexForFilename, imageCount > 0, subImageIndex);
final String oldOutFilename = duplicateGuard.get(digest);
if (oldOutFilename == null) {
File file = new File(dir, outFilename.toString());
duplicateGuard.put(digest, file.getPath());
ImageIO.write(image, "png", file);
} else {
symLinkCommand.set(2, oldOutFilename);
symLinkCommand.set(3, dir + "/" + outFilename);
Process process = new ProcessBuilder(symLinkCommand).start();
try {
if (process.waitFor() != 0) {
throw new IOException(
"Can't create symlink " + symLinkCommand.get(2) + " " + symLinkCommand.get(3));
}
} catch (InterruptedException e) {
throw new IOException(e);
}
}
}
}
}
