// private void getPart(ZipFile zf, Base source,
// Package pkg, String resolvedPartUri, String relationshipType)
private void getPart(
ZipFile zf, OpcPackage pkg, RelationshipsPart rp, Relationship r, ContentTypeManager ctm)
throws Docx4JException, InvalidFormatException, URISyntaxException {
Base source = null;
String resolvedPartUri = null;
if (r.getType().equals(Namespaces.HYPERLINK)) {
// Could be Internal or External
// Example of Internal is w:drawing/wp:inline/wp:docPr/a:hlinkClick
log.info("Encountered (but not loading) hyperlink " + r.getTarget());
return;
} else if (r.getTargetMode() == null || !r.getTargetMode().equals("External")) {
// Usual case
source = rp.getSourceP();
resolvedPartUri =
URIHelper.resolvePartUri(rp.getSourceURI(), new URI(r.getTarget())).toString();
// Now drop leading "/'
resolvedPartUri = resolvedPartUri.substring(1);
// Now normalise it .. ie abc/def/../ghi
// becomes abc/ghi
// Maybe this isn't necessary with a zip file,
// - ZipFile class may be smart enough to do it.
// But it is certainly necessary in the JCR case.
// resolvedPartUri = (new java.net.URI(resolvedPartUri)).normalize().toString();
// log.info("Normalised, it is " + resolvedPartUri );
} else {
// EXTERNAL
if (loadExternalTargets && r.getType().equals(Namespaces.IMAGE)) {
// It could instead be, for example, of type hyperlink,
// and we don't want to try to fetch that
log.warn("Loading external resource " + r.getTarget() + " of type " + r.getType());
BinaryPart bp = ExternalResourceUtils.getExternalResource(r.getTarget());
pkg.getExternalResources().put(bp.getExternalTarget(), bp);
} else {
log.warn(
"Encountered (but not loading) external resource "
+ r.getTarget()
+ " of type "
+ r.getType());
}
return;
}
if (pkg.handled.get(resolvedPartUri) != null) return;
String relationshipType = r.getType();
Part part = getRawPart(zf, ctm, resolvedPartUri, r);
if (part instanceof BinaryPart || part instanceof DefaultXmlPart) {
// The constructors of other parts should take care of this...
part.setRelationshipType(relationshipType);
}
rp.loadPart(part, r);
pkg.handled.put(resolvedPartUri, resolvedPartUri);
// The source Part (or Package) might have a convenience
// method for this
if (source.setPartShortcut(part, relationshipType)) {
log.info(
"Convenience method established from "
+ source.getPartName()
+ " to "
+ part.getPartName());
}
// unusedZipEntries.put(resolvedPartUri, new Boolean(false));
// log.info(".. added." );
RelationshipsPart rrp = getRelationshipsPart(zf, part);
if (rrp != null) {
// recurse via this parts relationships, if it has any
addPartsFromRelationships(zf, part, rrp, ctm);
String relPart = PartName.getRelationshipsPartName(part.getPartName().getName().substring(1));
// unusedZipEntries.put(relPart, new Boolean(false));
}
}
public OpcPackage get(File f) throws Docx4JException {
log.info("Filepath = " + f.getPath());
ZipFile zf = null;
try {
if (!f.exists()) {
log.info("Couldn't find " + f.getPath());
}
zf = new ZipFile(f);
} catch (IOException ioe) {
ioe.printStackTrace();
throw new Docx4JException("Couldn't get ZipFile", ioe);
}
// dumpZipFileContents(zf);
// 1. The idea is to walk the tree of relationships, getting
// everything we need from the zip file. But I'd like to know
// whether there are any orphans, so first we make
// a HashMap containing the names of all the zip file
// entries, so we can tick them off.
// unusedZipEntries = new HashMap();
Enumeration entries = zf.entries();
while (entries.hasMoreElements()) {
ZipEntry entry = (ZipEntry) entries.nextElement();
// unusedZipEntries.put(entry.getName(), new Boolean(true) );
}
// 2. Create a new Package
// Eventually, you'll also be able to create an Excel package etc
// but only the WordML package exists at present
ContentTypeManager ctm = new ContentTypeManager();
try {
InputStream is = getInputStreamFromZippedPart(zf, "[Content_Types].xml");
ctm.parseContentTypesFile(is);
} catch (IOException e) {
throw new Docx4JException("Couldn't get [Content_Types].xml from ZipFile", e);
}
OpcPackage p = ctm.createPackage();
// 3. Get [Content_Types].xml
// Once we've got this, then we can look up the content type for
// each PartName, and use it in the Part constructor.
// p.setContentTypeManager(ctm); - 20080111 - done by ctm.createPackage();
// unusedZipEntries.put("[Content_Types].xml", new Boolean(false));
// 4. Start with _rels/.rels
// <Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
// <Relationship Id="rId3"
// Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/extended-properties" Target="docProps/app.xml"/>
// <Relationship Id="rId2"
// Type="http://schemas.openxmlformats.org/package/2006/relationships/metadata/core-properties"
// Target="docProps/core.xml"/>
// <Relationship Id="rId1"
// Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument"
// Target="word/document.xml"/>
// </Relationships>
String partName = "_rels/.rels";
RelationshipsPart rp = getRelationshipsPartFromZip(p, zf, partName);
p.setRelationships(rp);
// rp.setPackageRelationshipPart(true);
// unusedZipEntries.put(partName, new Boolean(false));
log.info("Object created for: " + partName);
// log.info( rp.toString());
// 5. Now recursively
// (i) create new Parts for each thing listed
// in the relationships
// (ii) add the new Part to the package
// (iii) cross the PartName off unusedZipEntries
addPartsFromRelationships(zf, p, rp, ctm);
// 6. Check unusedZipEntries is empty
// if (log.isDebugEnabled()) {
// Iterator myVeryOwnIterator = unusedZipEntries.keySet().iterator();
// while(myVeryOwnIterator.hasNext()) {
// String key = (String)myVeryOwnIterator.next();
// log.info( key + " " + unusedZipEntries.get(key));
// }
// }
try {
zf.close();
} catch (IOException exc) {
exc.printStackTrace();
}
registerCustomXmlDataStorageParts(p);
return p;
}
