private void scan(ByteArrayInputStream in, String path, SVNDirEntry dirEntry) {
try {
Metadata metadata = new Metadata();
metadata.set(Metadata.RESOURCE_NAME_KEY, path);
// The following code part is from an proposal of the Authors of
// Tika:
// https://issues.apache.org/jira/browse/TIKA-232
TikaConfig config = TikaConfig.getDefaultConfig(); // without a
// delegate
// parser
Parser parser = new AutoDetectParser(config);
DefaultHandler handler = new BodyContentHandler();
parser.parse(in, handler, metadata);
getDocument().addTokenizedField(FieldNames.CONTENTS, handler.toString());
} catch (Exception e) {
LOGGER.error("We had an exception " + path + " (r" + dirEntry.getRevision() + ")", e);
} finally {
try {
in.close();
} catch (Exception e) {
LOGGER.error("We had an exception " + path + " (r" + dirEntry.getRevision() + ")", e);
}
}
}
@Override
protected ParseContext buildParseContext(
Metadata metadata, String targetMimeType, TransformationOptions options) {
ParseContext context = super.buildParseContext(metadata, targetMimeType, options);
boolean recurse = includeContents;
if (options.getIncludeEmbedded() != null) {
recurse = options.getIncludeEmbedded();
}
if (recurse) {
// Use an auto detect parser to handle the contents
if (tikaConfig == null) {
tikaConfig = TikaConfig.getDefaultConfig();
}
context.set(Parser.class, new AutoDetectParser(tikaConfig));
}
return context;
}
private class FileEmbeddedDocumentExtractor implements EmbeddedDocumentExtractor {
private int count = 0;
private final TikaConfig config = TikaConfig.getDefaultConfig();
public boolean shouldParseEmbedded(Metadata metadata) {
return true;
}
public void parseEmbedded(
InputStream inputStream,
ContentHandler contentHandler,
Metadata metadata,
boolean outputHtml)
throws SAXException, IOException {
String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
if (name == null) {
name = "file" + count++;
}
MediaType contentType = detector.detect(inputStream, metadata);
if (name.indexOf('.') == -1 && contentType != null) {
try {
name += config.getMimeRepository().forName(contentType.toString()).getExtension();
} catch (MimeTypeException e) {
e.printStackTrace();
}
}
String relID = metadata.get(Metadata.EMBEDDED_RELATIONSHIP_ID);
if (relID != null && !name.startsWith(relID)) {
name = relID + "_" + name;
}
File outputFile = new File(extractDir, name);
File parent = outputFile.getParentFile();
if (!parent.exists()) {
if (!parent.mkdirs()) {
throw new IOException("unable to create directory \"" + parent + "\"");
}
}
System.out.println("Extracting '" + name + "' (" + contentType + ") to " + outputFile);
FileOutputStream os = null;
try {
os = new FileOutputStream(outputFile);
if (inputStream instanceof TikaInputStream) {
TikaInputStream tin = (TikaInputStream) inputStream;
if (tin.getOpenContainer() != null && tin.getOpenContainer() instanceof DirectoryEntry) {
POIFSFileSystem fs = new POIFSFileSystem();
copy((DirectoryEntry) tin.getOpenContainer(), fs.getRoot());
fs.writeFilesystem(os);
} else {
IOUtils.copy(inputStream, os);
}
} else {
IOUtils.copy(inputStream, os);
}
} catch (Exception e) {
logger.warn("Ignoring unexpected exception trying to save embedded file " + name, e);
} finally {
if (os != null) {
os.close();
}
}
}
protected void copy(DirectoryEntry sourceDir, DirectoryEntry destDir) throws IOException {
for (org.apache.poi.poifs.filesystem.Entry entry : sourceDir) {
if (entry instanceof DirectoryEntry) {
// Need to recurse
DirectoryEntry newDir = destDir.createDirectory(entry.getName());
copy((DirectoryEntry) entry, newDir);
} else {
// Copy entry
InputStream contents = new DocumentInputStream((DocumentEntry) entry);
try {
destDir.createDocument(entry.getName(), contents);
} finally {
contents.close();
}
}
}
}
}