GCube Document Library
Contents
Design Overview
The gCube Document Library (gDL) is a library that works on top of the Content Manager. This library offers a simpler way for client to access documents in the d4science system. The major change in the ContentManager is the document model (more generic, more extensible), the gDL reduces the development costs for using this new feature.
Document Model
Implementation Overview
The gDL is composed by tree types of access interfaces:
- the CollectionManager is used to manage the Collections in the system.
- the CMReader is used to read document from a collection. it offers methods to get a document or part of it (Metadata, Annotations, Parts, Alternatives);
- the CMWriter is used to add document or update it or a part of it;
- the CMView is used to get documents from a specific view of a collection.
Accessing a collection
The GCube Document Library offers a set of utilities to help the client to interact with the Content Manager.
Reading a collection
The CMReader
let the client to read elements from a collection. There are methods to read single document or part of it and methods to read multiple documents returned as GCubeDocument iterator.
Method to access GCubeDocument lets filter and project the result. Both operations are achieved through the Projection system.
Projection
The Projection lets filter and project the results of a "query".
There is a Projection class for each element of model.
For each element is possible to define which fields get back through the includeXXX(boolean include)
method.
For each element is also possible to set a constraint through the constrainXXX(Constraint constraint)
method, or remove it with resetXXXConstraint()
one.
The ProjectionsFactory
class provide a set of utility method for fast retrieving element dedicated projection.
Writing into a collection
The CMWriter
class offers method to write, update and delete document from a collection.
The write methods accept new document and returns the new document id.
The update method accept a preceding retrieved document on which updates are made.
Finally the document remotion requires only the document id.
Each method have both a single element operation than a multiple element operation using an iterator as input and as output.
Managing a collection
The CollectionManager
class is an utility to manage a collection. It can return the collection Reader and Writer for respectively read and write the collection. It can also manage the collection Views with facilities method to access and manage it.
Configuring the environment
Supporting the SMS protocol
The Storage Management Service introduces a new protocol, called sms, with an handler implementation.
To let know the JVM about the existence of a new protocol handler you can use one of this solutions:
- Specify a new java property at JVM startup: -Djava.protocol.handler.pkgs=org.gcube.contentmanagement.storagelayer.storagemanagementservice.stubs.protocol
- if you are developing a service and are you using the last distribution you don't have to change anything because the new GHN distribution already do it.
- if you are developing a portlet or a portal library check if your portal instance have the option specified in the environment variable CATALINA_OPTS.
- if you are running your client locally on Eclipse you can specify the options on run configuration, Arguments tab, VM arguments section.
- Use this sentence in your client testing program:
System.setProperty("java.protocol.handler.pkgs", "org.gcube.contentmanagement.storagelayer.storagemanagementservice.stubs.protocol");
Examples
Here some examples on gDL use. More examples are on examples folder in gDL project on SVN.
Managing Collections examples
We shows some examples on managing Collections.
List all collections
This example shows how to list all collection on a scope.
GCUBEScope scope = GCUBEScope.getScope("/gcube"); System.out.println("Collections in "+scope+" :"); for (CMCollection collection:Collections.list(scope)){ System.out.println(collection); }
Find collection by name
This example shows how to find a collection using his name.
GCUBEScope scope = GCUBEScope.getScope("/gcube"); String collectionName = "My Test Collection"; System.out.println("Collections in "+scope+" with name \""+collectionName+"\":"); for (CMCollection collection:Collections.findByName(scope, collectionName)){ System.out.println(collection); }
Create a collection using SMS-Plugin
This code create a new collection with the given name and description.
The create method returns a list of collection references because a create command can produce more than one collection (for example OAI-Plugin).
To run this example the SMS-Plugin jar is required.
GCUBEScope scope = GCUBEScope.getScope("/gcube"); //collection informations String collectionName = "MyTestCollection"; String collectionDescription = "This is my test collection"; //we don't want to propagate the request to others CM boolean propagateRequest = false; //we want the collection to be readable and writable boolean readable = true; boolean writable = true; //finally we create the collection List<CollectionReference> collectionReferences =SMSUtil.createCollection(scope, propagateRequest, collectionName, collectionDescription, readable, writable); System.out.println("Created collections:"); for (CollectionReference collectionReference:collectionReferences){ System.out.println("Collection id: "+collectionReference.getCollectionID()); if (collectionReference.getReader()!=null) System.out.println("Reader EPR: "+collectionReference.getReader()); if (collectionReference.getWriter()!=null) System.out.println("Writer EPR: "+collectionReference.getWriter()); System.out.println(); }
Wrap a collection using SMS-Plugin
This example shows how to wrap an existing collection on SMS. To run this example the SMS-Plugin jar is required.
GCUBEScope scope = GCUBEScope.getScope("/gcube"); //the collection id on SMS (the id of collection IO on the SMS) String collectionId = "74fe4120-8f5b-11df-8083-c574cd534edf"; //we don't want to propagate the request to others CM boolean propagateRequest = false; //we want the collection to be readable and writable boolean readable = true; boolean writable = true; //finally we wrap the collection List<CollectionReference> collectionReferences = SMSUtil.wrapCollection(scope, propagateRequest, collectionId, readable, writable); System.out.println("Collections wrapped:"); for (CollectionReference collectionReference:collectionReferences){ System.out.println("Collection id: "+collectionReference.getCollectionID()); if (collectionReference.getReader()!=null) System.out.println("Reader EPR: "+collectionReference.getReader()); if (collectionReference.getWriter()!=null) System.out.println("Writer EPR: "+collectionReference.getWriter()); System.out.println(); }
Reading examples
We shows some examples on reading documents.
Get a single document
This example shows how to get a single document using his id.
//we instantiate the CMReader CMReader cmReader = new CMReader(collectionId, scope); //then we retrieve the document, we are not specifying projections, therefore the entire document is returned. GCubeDocument document = cmReader.getDocument(documentId); System.out.println("Retrieved document:"); System.out.println("id: " + document.getId()); System.out.println("name: " + document.getName()); System.out.println("mimeType: " + document.getMimeType());
Get documents by name
This example shows how to get documents from a collection using a name constraint.
String collectionId = "74fe4120-8f5b-11df-8083-c574cd534edf"; System.out.println("collectionId: "+collectionId); System.out.println(); String documentName = "Column dedication by Eumachos and Amias"; //we instantiate the CMReader CMReader cmReader = new CMReader(collectionId, scope); //we define a gcube main document projection GCubeMainDocumentProjection projection = ProjectionsFactory.MAIN_DOCUMENT(); //we set a constraint over the name field projection.constrainName(is(documentName)); //we are asking for all documents having the specified name Iterator<GCubeDocument> documentIterator = cmReader.getDocuments(projection); int i = 0; System.out.println("Retrieved documents:"); //finally we iterate over the collection while(documentIterator.hasNext()){ GCubeDocument document = documentIterator.next(); System.out.println(i+") "+document); i++; } System.out.println(); System.out.println("Found "+i+" documents");
To use the is
function remember to add this import:
import static org.gcube.contentmanagement.contentmanager.stubs.model.constraints.Constraints.*;
Get all documents in a collection
This example shows how to get all documents from a collection.
//we instantiate the CMReader CMReader cmReader = new CMReader(collectionId, scope); //then we retrieve all collection documents //we are not specifying projections, therefore the entire document is returned. Iterator<GCubeDocument> documentIterator = cmReader.getDocuments(); int i = 0; System.out.println("Collection documents:"); //finally we iterate over the collection while(documentIterator.hasNext()){ GCubeDocument document = documentIterator.next(); String id = document.getId(); int parts = document.getParts().size(); int alternatives = document.getAlternatives().size(); int metadata = document.getMetadata().size(); int annotations = document.getAnnotations().size(); System.out.printf("%d id: %s parts: %d alternatives: %d metadata: %d annotations: %d %n", i, id, parts, alternatives, metadata, annotations); i++; }
Get a document using projections
This example shows how to use projections during document retrieving.
//we instantiate the CMReader CMReader cmReader = new CMReader(collectionId, scope); //then we retrieve the document, //we are specifying the MAIN_DOCUMENT and the METADATA projections, the document returned will have main document data and all related metadata GCubeDocument document = cmReader.getDocument(documentId, ProjectionsFactory.MAIN_DOCUMENT(), ProjectionsFactory.METADATA()); System.out.println("Retrieved document:"); System.out.println("id: " + document.getId()); System.out.println("name: " + document.getName()); System.out.println("mimeType: " + document.getMimeType()); System.out.println("length: " + document.getLength()); System.out.println("contentURL: " + document.getContentURL()); System.out.println(); for (GCubeMetadata metadata:document.getMetadata()){ System.out.println("Metadata:"); System.out.println(" id: " + metadata.getId()); System.out.println(" mimeType: " + metadata.getMimeType()); System.out.println(" length: " + metadata.getLength()); System.out.println(" Metadata info:"); System.out.println(" name: " + metadata.getMetadataName()); System.out.println(" schema: " + metadata.getMetadataSchema()); System.out.println(" language: " + metadata.getMetadataLanguage()); System.out.println(); }
Get a document content
This example shows how to get a document content.
//we instantiate the CMReader CMReader cmReader = new CMReader(collectionId, scope); //then we retrieve the document, we are requesting only the main document GCubeDocument document = cmReader.getDocument(documentId, ProjectionsFactory.MAIN_DOCUMENT()); System.out.println("Retrieved document:"); System.out.println("id: " + document.getId()); System.out.println("name: " + document.getName()); System.out.println("contentURL: " + document.getContentURL()); System.out.println(); System.out.println("Retrieving content..."); //the content is retrieved by the CML library through the content URL InputStream is = document.getContent(); //finally we copy the content to a tmp file File tmpFile = File.createTempFile("mycontent", ".tmp"); IOUtils.copy(is, new FileOutputStream(tmpFile)); System.out.println("Content saved in tmp file: " + tmpFile.getAbsolutePath());
Get Collection dynamic properties
This example shows how to get dynamic properties for a given collection.
String collectionId = "18611ed0-f162-11dd-96f7-b87cc0f0b075"; System.out.println("collectionId: "+collectionId); CollectionManager collectionManager = new CollectionManager(collectionId, scope); System.out.println("Cardinality: "+collectionManager.getCardinality()); SimpleDateFormat sdf = new SimpleDateFormat(); System.out.println("Last Update: "+sdf.format(collectionManager.getLatUpdate().getTime())); System.out.println();
Writing examples
We shows some examples on manipulating collections.
Add a new document to a collection
Add a new document to a collection.
//we instantiate the CMWriter CMWriter cmWriter = new CMWriter(collectionId, scope); //we create an instance of gcube document GCubeDocument myNewDocument = new GCubeDocument("My Test Document", "application/octet-stream"); //then we ask for adding the new document String myNewDocumentId = cmWriter.add(myNewDocument); System.out.println("document created, id: " + myNewDocumentId); //finally we retrieve the created document CMReader cmReader = new CMReader(collectionId, scope); GCubeDocument document = cmReader.getDocument(myNewDocumentId); System.out.println("Created document:"); System.out.println("id: " + document.getId()); System.out.println("name: " + document.getName()); System.out.println("mimeType: " + document.getMimeType()); System.out.println("length: " + document.getLength()); System.out.println("contentURL: " + document.getContentURL()); System.out.println("parts: " + document.getParts().size()); System.out.println("alternatives:" + document.getAlternatives().size()); System.out.println("metadata: " + document.getMetadata().size()); System.out.println("annotations: " + document.getAnnotations().size());
Remove a document from a collection
Remove a document from a collection.
//we instantiate the CMWriter CMWriter cmWriter = new CMWriter(collectionId, scope); writer.remove(documentId); System.out.println("document removed.");
Views examples
We shows some examples on handling views.
Create a Metadata View
Shows how to create a metadata view.
String collectionId = "1c583800-d5d8-11df-9cd6-b72b661eb83e"; System.out.println("collectionId: "+collectionId); System.out.println(); MetadataView myView = new MetadataView(); myView.setName("MyView"); myView.setDescription("This is my test view"); myView.setEditable(false); myView.setIndexable(false); myView.setUserCollection(false); myView.setMetadaName("aquamaps"); myView.setMetadataLanguage("en"); myView.setMetadataSchema("http://aquamaps.xsd"); CollectionManager manager = new CollectionManager(collectionId, scope); manager.createView(myView); System.out.println("View created: "+myView); MetadataViewReader reader = new MetadataViewReader(myView, scope); System.out.println("Metadata:"); Iterator<GCubeMetadata> iterator = reader.getAll(); while(iterator.hasNext()) System.out.println(iterator.next());
List collection views
List all the views for a certain collection.
//we instantiate a collection manager. CollectionManager collectionManager = new CollectionManager(collectionId, scope); //then we ask for all the available views List<CMView<?>> views = collectionManager.findViews(); System.out.println("found "+views.size()+" views:"); for (CMView<?> view:views) System.out.println(view);
Get Metadata elements
Get a metadata view for a collection and the all the elements of the view.
//first we instantiate a collection manager CollectionManager collectionManager = new CollectionManager(collectionId, scope); //then we create a view kind, in this case a MetadataView without specifying any property MetadataView myMedataView = new MetadataView(); //finally we ask to the collection manager to list all views of the kind specified. List<MetadataView> views = collectionManager.findViews(myMedataView); System.out.println("Found views: "+views.size()); //in case we not found views if (views.size()==0) throw new Exception("No views found"); //we get the first of the list MetadataView view = views.get(0); //we instantiate the view reader MetadataViewReader metadataReader = new MetadataViewReader(view, scope); //we request all the elements Iterator<GCubeMetadata> metadata = metadataReader.getAll(); //and print it while (metadata.hasNext()) System.out.println(" >"+metadata.next());
Get Metadata elements using only the metadata view ID
Get a metadata view for a collection and the all the elements of the view. Without knowing the collection ID.
//Use the metadata view ID to find the view List<MetadataView> views = CMViews.findViews(scope, new MetadataView(metadataViewID)); if (views.isEmpty()){ System.out.println("Metadata view not found. View ID: " + metadataViewID); } //Just get the first view MetadataView myMedataView = views.get(0); MetadataViewReader metadataReader = new MetadataViewReader(myMedataView, scope);
F.A.Q.
- Running the examples I get the following exception:
[main] FATAL contexts.GHNClientContext - [0.0s] GHNClientContext: gHN could not complete initialisation java.io.FileNotFoundException: null/config/GHNConfig.xml (No such file or directory) at java.io.FileInputStream.open(Native Method) at java.io.FileInputStream.<init>(FileInputStream.java:106) at org.gcube.common.core.contexts.GHNContext.configureGHN(GHNContext.java:314) at org.gcube.common.core.contexts.GHNClientContext.configureGHN(GHNClientContext.java:35) at org.gcube.common.core.contexts.GHNClientContext.initialise(GHNClientContext.java:27) at org.gcube.common.core.contexts.GHNContext.<clinit>(GHNContext.java:252) at org.gcube.common.core.utils.calls.WSCall.getInitQuery(WSCall.java:55) at org.gcube.common.core.utils.calls.WSCall.getInitQuery(WSCall.java:25) at org.gcube.common.core.utils.calls.GCUBECall.<init>(GCUBECall.java:72) at org.gcube.common.core.utils.calls.GCUBECall.<init>(GCUBECall.java:84) at org.gcube.common.core.utils.calls.WSCall.<init>(WSCall.java:43) at org.gcube.contentmanagement.contentmanager.stubs.calls.BaseCall.<init>(BaseCall.java:25) at org.gcube.contentmanagement.contentmanager.stubs.calls.ManagerCall.<init>(ManagerCall.java:24) at org.gcube.contentmanagement.contentmanager.stubs.calls.ManagerCall.<init>(ManagerCall.java:35) at org.gcube.contentmanagement.contentmanager.stubs.calls.ReadManagerCall.<init>(ReadManagerCall.java:97) at org.gcube.contentmanagement.contentmanagerlibrary.CMReader.<init>(CMReader.java:62) at org.gcube.contentmanager.contentmanagerlibrary.GetDocumentsWithProjections.main(GetDocumentsWithProjections.java:34) Exception in thread "main" java.lang.NullPointerException at org.gcube.common.core.utils.calls.WSCall.getInitQuery(WSCall.java:55) at org.gcube.common.core.utils.calls.WSCall.getInitQuery(WSCall.java:25) at org.gcube.common.core.utils.calls.GCUBECall.<init>(GCUBECall.java:72) at org.gcube.common.core.utils.calls.GCUBECall.<init>(GCUBECall.java:84) at org.gcube.common.core.utils.calls.WSCall.<init>(WSCall.java:43) at org.gcube.contentmanagement.contentmanager.stubs.calls.BaseCall.<init>(BaseCall.java:25) at org.gcube.contentmanagement.contentmanager.stubs.calls.ManagerCall.<init>(ManagerCall.java:24) at org.gcube.contentmanagement.contentmanager.stubs.calls.ManagerCall.<init>(ManagerCall.java:35) at org.gcube.contentmanagement.contentmanager.stubs.calls.ReadManagerCall.<init>(ReadManagerCall.java:97) at org.gcube.contentmanagement.contentmanagerlibrary.CMReader.<init>(CMReader.java:62) ...
In this case you have to configure your eclipse running configuration properly [1].
- Trying to get the document content I get this exception:
Exception in thread "main" java.net.MalformedURLException: unknown protocol: sms at java.net.URL.<init>(URL.java:574) at java.net.URL.<init>(URL.java:464) at java.net.URL.<init>(URL.java:413) at org.gcube.contentmanagement.contentmanagerlibrary.model.GCubeDocument.getContent(GCubeDocument.java:227) ...
Check your configuration about SMS protocol
- Eclipse don't find some functions in the example code (like is, matches, all, either, not, more, less, after, before).
Those are helper function from the CM and are defined on Constraints class. You can statically import it with this statement:
import static org.gcube.contentmanagement.contentmanager.stubs.model.constraints.Constraints.*;
- Trying to interact with the CM I get this exception:
org.apache.commons.discovery.DiscoveryException: org.gcube.common.core.utils.handlers.GCUBEServiceHandler$NoQueryResultException ***** org.gcube.common.core.utils.handlers.GCUBEServiceHandler$NoQueryResultException at org.gcube.common.core.utils.handlers.GCUBEStatefulServiceHandler.findInstances(GCUBEStatefulServiceHandler.java:38) at org.gcube.common.core.utils.handlers.GCUBEServiceHandler.getInstances(GCUBEServiceHandler.java:140) at org.gcube.common.core.utils.handlers.GCUBEServiceHandler.run(GCUBEServiceHandler.java:110) at org.gcube.contentmanagement.contentmanager.stubs.calls.BaseCall$CallHandler.run(BaseCall.java:93) at org.gcube.contentmanagement.contentmanager.stubs.calls.WriteManagerCall.add(WriteManagerCall.java:98) at org.gcube.contentmanagement.contentmanager.stubs.calls.WriteManagerCall.add(WriteManagerCall.java:75) at org.gcube.contentmanagement.gcubedocumentlibrary.CMWriter.add(CMWriter.java:84) at org.gcube.personalization.userprofileaccess.impl.CMSUtils.createDocument(CMSUtils.java:67) at org.gcube.personalization.userprofileaccess.impl.UserProfileAccessFactoryService.createResource(UserProfileAccessFactoryService.java:331)
The exception means that there is not plugin instantiated for the specified collection. Check if the collection id is correct. Otherwise contact the support team.