// *
public final ImageRecord getMetadata(BufferedImage bi) throws DjatokaException {
if (bi == null) throw new DjatokaException("Image Does Not Exist");
logger.debug("getMetadata(BufferedImage): " + bi.getWidth());
try {
ImageRecord r = new ImageRecord();
r.setWidth(bi.getWidth());
r.setHeight(bi.getHeight());
r.setDWTLevels(DEFAULT_LEVELS);
r.setLevels(DEFAULT_LEVELS);
r.setBitDepth(bi.getColorModel().getPixelSize());
r.setNumChannels(bi.getColorModel().getNumColorComponents());
// r.setCompositingLayerCount(getNumberOfPages(r)); // 'bi' refers to just one page extracted
// from the PDF file.
// logger.debug("r2: "+r.toString());
// TODO
return r;
} catch (Exception e) {
throw new DjatokaException(e);
}
}
@Override
// TODO
// FIXME
public final String[] getXMLBox(ImageRecord r) throws DjatokaException {
String[] xml = null;
try {
if (r.getImageFile() == null
&& r.getObject() != null
&& r.getObject() instanceof InputStream) {
// xml = new JP2ImageInfo((InputStream) r.getObject()).getXmlDocs();
} else {
// xml = new JP2ImageInfo(new File(r.getImageFile())).getXmlDocs();
}
} catch (Exception e) {
logger.error(e, e);
}
return xml;
}
/**
* Extracts region defined in DjatokaDecodeParam as BufferedImage
*
* @param input ImageRecord wrapper containing file reference, inputstream, etc.
* @param params DjatokaDecodeParam instance containing region and transform settings.
* @return extracted region as a BufferedImage
* @throws DjatokaException
*/
@Override
public BufferedImage process(ImageRecord input, DjatokaDecodeParam params)
throws DjatokaException {
logger.debug("in imagerecord;");
if (input.getImageFile() != null) return process(input.getImageFile(), params);
else if (input.getObject() != null && (input.getObject() instanceof InputStream))
return process((InputStream) input.getObject(), params);
else
throw new DjatokaException(
"File not defined and Input Object Type "
+ input.getObject().getClass().getName()
+ " is not supported");
}
/**
* Get PDF information with pdfinfo: - "Pages: X": number of pages; - "Page X size: www.ww
* hhh.hh": size of each page, in pts.
*
* @returns a map: - [Pages][n] - [Page 1][111.11 222.22] - [Page i][www.ww hhh.hh] - [Page
* n][999.99 1000.00]
*/
private static Map<String, String> getPDFProperties(ImageRecord input) throws DjatokaException {
logger.debug("Getting PDF info");
try {
setPDFCommandsPath();
} catch (IllegalStateException e) {
logger.error("Failed to set PDF commands path: ", e);
throw e;
}
HashMap<String, String> pdfProperties = new HashMap<String, String>();
String sourcePath = null;
if (input.getImageFile() != null) {
logger.debug("PDFInfo image file: " + input.getImageFile());
sourcePath = input.getImageFile();
} else if (input.getObject() != null && (input.getObject() instanceof InputStream)) {
FileInputStream fis = null;
fis = (FileInputStream) input.getObject();
File in;
// Copy to tmp file
try {
String cacheDir = OpenURLJP2KService.getCacheDir();
if (cacheDir != null) {
in = File.createTempFile("tmp", ".pdf", new File(cacheDir));
} else {
in = File.createTempFile("tmp", ".pdf");
}
in.deleteOnExit();
FileOutputStream fos = new FileOutputStream(in);
IOUtils.copyStream(fis, fos);
} catch (IOException e) {
logger.error(e, e);
throw new DjatokaException(e);
}
sourcePath = in.getAbsolutePath();
} else {
throw new DjatokaException(
"File not defined and Input Object Type "
+ input // .getObject().getClass().getName()
+ " is not supported");
}
String pdfinfoCmd[] = PDFINFO_COMMAND.clone();
pdfinfoCmd[PDFINFO_COMMAND_POSITION_BIN] = pdfinfoPath;
pdfinfoCmd[PDFINFO_COMMAND_POSITION_FIRSTPAGE] = "1";
pdfinfoCmd[PDFINFO_COMMAND_POSITION_LASTPAGE] =
"-1"; // Last page even we not knowing its number.
pdfinfoCmd[PDFINFO_COMMAND_POSITION_FILE] = sourcePath;
Process pdfProc = null;
try {
ArrayList<MatchResult> pageSizes = new ArrayList<MatchResult>();
MatchResult pages = null;
pdfProc = Runtime.getRuntime().exec(pdfinfoCmd);
BufferedReader lr = new BufferedReader(new InputStreamReader(pdfProc.getInputStream()));
String line;
for (line = lr.readLine(); line != null; line = lr.readLine()) {
Matcher mm1 = PAGES_PATT.matcher(line);
if (mm1.matches()) pages = mm1.toMatchResult();
Matcher mm2 = MEDIABOX_PATT.matcher(line);
if (mm2.matches()) pageSizes.add(mm2.toMatchResult());
}
int istatus = pdfProc.waitFor();
if (istatus != 0)
logger.error("pdfinfo proc failed, exit status=" + istatus + ", file=" + sourcePath);
if (pages == null) {
logger.error(
"Did not find 'Pages' line in output of pdfinfo command: "
+ Arrays.deepToString(pdfinfoCmd));
pdfProperties.put("Pages", "0");
} else {
// int n = Integer.parseInteger(pages.group(1));
pdfProperties.put("Pages", pages.group(1));
}
if (pageSizes.isEmpty()) {
logger.error(
"Did not find \"Page X size\" lines in output of pdfinfo command: "
+ Arrays.deepToString(pdfinfoCmd));
throw new IllegalArgumentException("Failed to get pages size of PDF with pdfinfo.");
} else {
for (MatchResult mr : pageSizes) {
String page = mr.group(1);
float x0 = Float.parseFloat(mr.group(2));
float y0 = Float.parseFloat(mr.group(3));
float x1 = Float.parseFloat(mr.group(4));
float y1 = Float.parseFloat(mr.group(5));
float w = Math.abs(x1 - x0);
float h = Math.abs(y1 - y0);
// Have to scale page sizes by max dpi (MAX_DPI / DEFAULT_DENSITY). Otherwise,
// BookReader.js will request the wrong zoom level (svc.level).
float ws = w * MAX_DPI / DEFAULT_DENSITY;
float hs = h * MAX_DPI / DEFAULT_DENSITY;
String width = "" + ws; // mr.group(2);
String height = "" + hs; // mr.group(3);
pdfProperties.put("Page " + page, width + " " + height);
}
}
} catch (Exception e) {
logger.error("Failed getting PDF information: ", e);
throw new DjatokaException("Failed getting PDF information: ", e);
} finally {
// Our exec() should just consume one of the streams, but we want to stay safe.
// http://mark.koli.ch/2011/01/leaky-pipes-remember-to-close-your-streams-when-using-javas-runtimegetruntimeexec.html
org.apache.commons.io.IOUtils.closeQuietly(pdfProc.getOutputStream());
org.apache.commons.io.IOUtils.closeQuietly(pdfProc.getInputStream());
org.apache.commons.io.IOUtils.closeQuietly(pdfProc.getErrorStream());
}
return pdfProperties;
}
/**
* Returns PDF props in ImageRecord
*
* @param r ImageRecord containing absolute file path of PDF file.
* @return a populated ImageRecord object
* @throws DjatokaException
*/
@Override
public final ImageRecord getMetadata(ImageRecord r) throws DjatokaException {
if ((r.getImageFile() == null || !new File(r.getImageFile()).exists()) && r.getObject() == null)
throw new DjatokaException("Image Does Not Exist: " + r.toString());
logger.debug("Get metadata: " + r.toString());
try {
DjatokaDecodeParam params = new DjatokaDecodeParam();
BufferedImage bi = process(r, params);
r.setWidth(bi.getWidth());
r.setHeight(bi.getHeight());
r.setDWTLevels(DEFAULT_LEVELS);
r.setLevels(DEFAULT_LEVELS);
r.setBitDepth(bi.getColorModel().getPixelSize());
r.setNumChannels(bi.getColorModel().getNumColorComponents());
// r.setCompositingLayerCount(getNumberOfPages(r)); // Semantics: number of pages in the PDF
// file.
HashMap<String, String> pdfProps = (HashMap<String, String>) getPDFProperties(r);
int n = Integer.parseInt(pdfProps.remove("Pages"));
r.setCompositingLayerCount(n);
// Since it is not possible for the viewer to query about a specific page's width and height
// (because in Djatoka's point of view a PDF is just one image with various compositing
// layers, which are the pages),
// at this point right here we query the PDF file about the size of all pages and store this
// information in a Map. This map can be returned by getMetadata by setting it as the
// instProps member of the
// ImageRecord class, which Djatoka already implements and which is returned as JSON to the
// viewer JS.
// The viewer then has to store this information and later query it instead of asking Djatoka
// (getMetadata) again.
// Map<String, String> instProps = getPagesSizes(r);
r.setInstProps(pdfProps);
logger.debug("instProps: " + r.getInstProps());
logger.debug("Get metadata: " + r.toString());
} catch (Exception e) {
throw new DjatokaException(e);
}
return r;
}
