org.semanticdesktop.aperture.crawler.mail
Class DataObjectFactory

java.lang.Object
  org.semanticdesktop.aperture.crawler.mail.DataObjectFactory

public class DataObjectFactory
extends Object

Creates a set of DataObjects from a MimeMessage.

DataObjectFactory interprets the structure of a MimeMessage and creates a list of DataObjects that model its contents in a way that is most natural to users. Practically this means that the DataObject list should be as similar as possible to how mail readers present the mail.

For example, a multipart/alternative message may have a rather complex object structure (a Part with a MultiPart content, on its turn containing two BodyParts), but this is translated to a single DataObject holding all the mail metadata (sender, receiver, etc) as well as an InputStream accessing the simplest of the two body parts (typically the text/plain part).

Known bugs/features:

• The DataObjectFactory constructs person URIs that begin with the emailperson: prefix and have the person name in them. This means that if the sender and the receiver have the same name, but different addresses, they will be mapped to the same resource with two different addresses. In this way we lose information about which address was the sending address and which address was the receiving address.
• Each email has one nie:contentCreated triple. It is either the date the message was sent, or received or the date of the message retrieval. We loose information what kind of date it is, and we loose the timezone information that is in the original email, but gets swallowed during the crawling process.
• The messageId property is stored in the message metadata with the angle brackets, whereas they probably should be removed.
• Each message part is equipped with the nie:byteSize property. It is against the definition of NIE mime type because it marks the length of the message content only, not of the message as a whole (i.e. excluding the message headers). In multipart/alternative emails the byte size is the size of both the html and plain parts combined.

Nested Class Summary

static interface

DataObjectFactory.PartStreamFactory Obtains InputStreams from Part instances.

Constructor Summary

DataObjectFactory(javax.mail.internet.MimeMessage message, RDFContainerFactory containerFactory, DataObjectFactory.PartStreamFactory streamFactory, DataSource dataSource, URI messageUri, URI folderUri)
A simplified constructor that implies the default part uri delimiter

DataObjectFactory(javax.mail.internet.MimeMessage message, RDFContainerFactory containerFactory, DataObjectFactory.PartStreamFactory streamFactory, DataSource dataSource, URI messageUri, URI folderUri, String partUriDelimiter)
A little simplified constructor that implies the attachment uri prefix to be the same as the message URI.

DataObjectFactory(javax.mail.internet.MimeMessage message, RDFContainerFactory containerFactory, DataObjectFactory.PartStreamFactory streamFactory, DataSource dataSource, URI messageUri, URI folderUri, String partUriDelimiter, String attachmentUriPrefix)
Constructs a data object factory for the given message

Method Summary

void	disposeRemainingObjects() Disposes of the data objects remaining on the list of objects to return.
Map<URI,DataObject>	getAllDataObjects() Returns a map containing all data objects, indexed by their URIs (from DataObject.getID()
DataObject	getObject() Returns a DataObject representing a single message.
DataObject	getObject(String url) Returns a data object with the given url, the data object will be removed from the list of data objects to remove.
DataObject	getObjectAndDisposeAllOtherObjects(String url) Returns a data object with the given url chosen from the set of objects created from the MimeMessage instance passed to the constructor of this DataObjectFactory.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DataObjectFactory

public DataObjectFactory(javax.mail.internet.MimeMessage message,
                         RDFContainerFactory containerFactory,
                         DataObjectFactory.PartStreamFactory streamFactory,
                         DataSource dataSource,
                         URI messageUri,
                         URI folderUri,
                         String partUriDelimiter,
                         String attachmentUriPrefix)
                  throws IOException,
                         javax.mail.MessagingException

Constructs a data object factory for the given message

Parameters:

message -

containerFactory -

executorService -

streamFactory - a factory of streams for mail parts, this parameter can be null, in which case a default implementation will be used, that simply calls Part.getInputStream()

dataSource -

messageUri -

folderUri -

partUriDelimiter -

attachmentUriPrefix -

Throws:

IOException

javax.mail.MessagingException

DataObjectFactory

public DataObjectFactory(javax.mail.internet.MimeMessage message,
                         RDFContainerFactory containerFactory,
                         DataObjectFactory.PartStreamFactory streamFactory,
                         DataSource dataSource,
                         URI messageUri,
                         URI folderUri,
                         String partUriDelimiter)
                  throws IOException,
                         javax.mail.MessagingException

A little simplified constructor that implies the attachment uri prefix to be the same as the message URI.

Parameters:

message -

containerFactory -

streamFactory -

dataSource -

messageUri -

folderUri -

partUriDelimiter -

Throws:

IOException

javax.mail.MessagingException

DataObjectFactory

public DataObjectFactory(javax.mail.internet.MimeMessage message,
                         RDFContainerFactory containerFactory,
                         DataObjectFactory.PartStreamFactory streamFactory,
                         DataSource dataSource,
                         URI messageUri,
                         URI folderUri)
                  throws IOException,
                         javax.mail.MessagingException

A simplified constructor that implies the default part uri delimiter

Parameters:

message -

containerFactory -

executorService -

streamFactory -

dataSource -

messageUri -

folderUri -

Throws:

IOException

javax.mail.MessagingException

Method Detail

getObject

public DataObject getObject()
                     throws javax.mail.MessagingException,
                            IOException

Returns a DataObject representing a single message. This data object contains a flattened version of the (arbitrary complex) tree-like MIME structure of the message. This method is called repeatedly for a single MimeMessage. At the first call it creates a cachedDataObjectsMap of all dataObjects that are to be returned from this message. On all subsequent calls DataObjects from this map are returned.

Returns:

a DataObject instance for the given message

Throws:

javax.mail.MessagingException

IOException

disposeRemainingObjects

public void disposeRemainingObjects()

Disposes of the data objects remaining on the list of objects to return. This method must be called if the user stopped calling the getObject() method BEFORE it returned null, thus marking the fact that all objects have been read.

getObject

public DataObject getObject(String url)

Returns a data object with the given url, the data object will be removed from the list of data objects to remove. It will not be disposed by the disposeRemainingObjects() method and therefore must be disposed manually by the user.

Parameters:

url -

Returns:

a data object with the given url, chosen from among the objects created from the MimeMessage instance passed to the constructor of this DataObjectFactory

getObjectAndDisposeAllOtherObjects

public DataObject getObjectAndDisposeAllOtherObjects(String url)

Returns a data object with the given url chosen from the set of objects created from the MimeMessage instance passed to the constructor of this DataObjectFactory. All other data objects from that set are disposed. If there is no data object with the given url - all data objects are disposed.

Parameters:

url -

Returns:

a data object with the given url, or null if no such data object exists

getAllDataObjects

public Map<URI,DataObject> getAllDataObjects()
                                      throws javax.mail.MessagingException,
                                             IOException

Returns a map containing all data objects, indexed by their URIs (from DataObject.getID()

Returns:

a map containing all data objects, indexed by their URIs (from DataObject.getID()

Throws:

javax.mail.MessagingException

IOException