org.nuxeo.ecm.platform.semanticentities

Interface SemanticAnalysisService

All Known Implementing Classes:

SemanticAnalysisServiceImpl

public interface SemanticAnalysisService

Service to asynchronously launch and monitor the semantic analysis of Nuxeo documents.

Method Summary

Methods

Modifier and Type	Method and Description
List<OccurrenceGroup>	analyze(CoreSession session, DocumentModel doc) Synchronous analysis of a document (without linking).
List<OccurrenceGroup>	analyze(CoreSession session, String textContent) Synchronous analysis of pre-extracted text content (without linking).
void	clearProgressStatus(String repositoryName, DocumentRef docRef) Mark the status as complete (even though the document might still be under processing).
void	createLinks(DocumentModel document, CoreSession session, List<OccurrenceGroup> occurrenceGroups) Save semantic links in the repository to materialize the occurrence relationships between documents and entities
ProgressStatus	getProgressStatus(String repositoryName, DocumentRef docRef) Return a description of the state of the analysis of the document or null if no analysis is in progress.
boolean	isActive()
void	launchAnalysis(String repositoryName, DocumentRef docRef) Launch an asynchronous analysis of a document.
void	launchSynchronousAnalysis(DocumentModel doc, CoreSession session) Launch a analysis of a document and wait for the result before returning.
void	scheduleSerializationTask(SerializationTask task) Asynchronously save the result of the analyze using a dedicated sequential thread.

Method Detail

isActive

boolean isActive()

Returns:

true if the service could be correctly initialized

analyze

List<OccurrenceGroup> analyze(CoreSession session,
                            String textContent)
                              throws IOException,
                                     ClientException

Synchronous analysis of pre-extracted text content (without linking).

Parameters:

session - used to create in-memory prefetched entity suggestions if available in the analysis engine response.

textContent - the text to send to the engines

Returns:

Occurrence suggestions

Throws:

IOException - if the engine is not reachable or fails

ClientException - if the in-memory creation of prefetched entity suggestions fails (very unlikely)

analyze

List<OccurrenceGroup> analyze(CoreSession session,
                            DocumentModel doc)
                              throws IOException,
                                     ClientException

Synchronous analysis of a document (without linking).

Parameters:

session - used to create in-memory prefetched entity suggestions if available in the analysis engine response.

doc - the document to analyze (must be attached to an active core session to extract the text content)

Returns:

Occurrence suggestions

Throws:

ClientException - if the text extraction fails

IOException - if the engine is not reachable or fails

scheduleSerializationTask

void scheduleSerializationTask(SerializationTask task)

Asynchronously save the result of the analyze using a dedicated sequential thread.

Parameters:

task - a DTO that contains the reference of a document and the occurrence groups to link to it

launchAnalysis

void launchAnalysis(String repositoryName,
                  DocumentRef docRef)
                    throws ClientException

Launch an asynchronous analysis of a document. The result of the analysis is stored directly in the Nuxeo repository, usually as relationship to semantic entities, other documents or updated properties.

Parameters:

repositoryName - the repository where the document is stored

docRef - the reference of the document to analyze

Throws:

ClientException - if a property of the document to analyze is not available due to a database connection issue for instance.

launchSynchronousAnalysis

void launchSynchronousAnalysis(DocumentModel doc,
                             CoreSession session)
                               throws ClientException,
                                      DereferencingException,
                                      IOException

Launch a analysis of a document and wait for the result before returning. The result of the analysis is stored directly in the Nuxeo repository, usually as relationship to semantic entities, other documents or updated properties.

Parameters:

doc - the document to analyze

Throws:

ClientException - if a property of the document to analyze is not available due to a database connection issue for instance.

IOException - if the connection to the analysis engine fails, or if the engine it-self fails

DereferencingException - if the dereferencing process fails (e.g. due to a network failure to a remote knowledge base).

getProgressStatus

ProgressStatus getProgressStatus(String repositoryName,
                               DocumentRef docRef)

Return a description of the state of the analysis of the document or null if no analysis is in progress.

Parameters:

docRef -

Returns:

state description or null

clearProgressStatus

void clearProgressStatus(String repositoryName,
                       DocumentRef docRef)

Mark the status as complete (even though the document might still be under processing).

createLinks

void createLinks(DocumentModel document,
               CoreSession session,
               List<OccurrenceGroup> occurrenceGroups)
                 throws ClientException,
                        IOException

Save semantic links in the repository to materialize the occurrence relationships between documents and entities

Throws:

IOException - if the remote entity sources fail during lookups

ClientException - if the document repository fails during local lookups or saving entities and occurrences