Package org.eclipse.jdt.core.search

Class SearchParticipant

java.lang.Object
org.eclipse.jdt.core.search.SearchParticipant
public abstract class SearchParticipant extends Object
A search participant describes a particular extension to a generic search mechanism, permitting combined search actions which will involve all required participants.

A search participant is involved in the indexing phase and in the search phase. The indexing phase consists in taking one or more search documents, parse them, and add index entries in an index chosen by the participant. An index is identified by a path on disk. The search phase consists in selecting the indexes corresponding to a search pattern and a search scope, from these indexes the search infrastructure extracts the document paths that match the search pattern asking the search participant for the corresponding document, finally the search participant is asked to locate the matches precisely in these search documents.

This class is intended to be subclassed by clients. During the indexing phase, a subclass will be called with the following requests in order:

During the search phase, a subclass will be called with the following requests in order:

Since:
3.0

  • Constructor Details

    • SearchParticipant

      protected SearchParticipant()
      Creates a new search participant.

  • Method Details

    • beginSearching

      public void beginSearching()
      Notification that this participant's help is needed in a search.

      This method should be re-implemented in subclasses that need to do something when the participant is needed in a search.

    • doneSearching

      public void doneSearching()
      Notification that this participant's help is no longer needed.

      This method should be re-implemented in subclasses that need to do something when the participant is no longer needed in a search.

    • getDescription

      public String getDescription()
      Returns a displayable name of this search participant.

      This method should be re-implemented in subclasses that need to display a meaningful name.

      Returns:
      the displayable name of this search participant

    • getDocument

      public abstract SearchDocument getDocument(String documentPath)
      Returns a search document for the given path. The given document path is a string that uniquely identifies the document. Most of the time it is a workspace-relative path, but it can also be a file system path, or a path inside a zip file.

      Implementors of this method can either create an instance of their own subclass of SearchDocument or return an existing instance of such a subclass.

      Parameters:
      documentPath - the path of the document.
      Returns:
      a search document

    • indexDocument

      public abstract void indexDocument(SearchDocument document, org.eclipse.core.runtime.IPath indexLocation)
      Indexes the given document in the given index. A search participant asked to index a document should parse it and call SearchDocument.addIndexEntry(char[], char[]) as many times as needed to add index entries to the index. If delegating to another participant, it should use the original index location (and not the delegatee's one). In the particular case of delegating to the default search participant (see SearchEngine.getDefaultSearchParticipant()), the provided document's path must be a path ending with one of the Java-like extensions or with '.class'.

      The given index location must represent a path in the file system to a file that either already exists or is going to be created. If it exists, it must be an index file, otherwise its data might be overwritten.

      Clients are not expected to call this method.

      Parameters:
      document - the document to index
      indexLocation - the location in the file system to the index

    • indexResolvedDocument

      public void indexResolvedDocument(SearchDocument document, org.eclipse.core.runtime.IPath indexLocation)
      Indexes the given resolved document in the given index. A search participant asked to index a resolved document should process it and call SearchDocument.addIndexEntry(char[], char[]) as many times as needed to add only those additional index entries which could not have been originally added to the index during a call to indexDocument(org.eclipse.jdt.core.search.SearchDocument, org.eclipse.core.runtime.IPath). If delegating to another participant, it should use the original index location (and not the delegatee's one). In the particular case of delegating to the default search participant (see SearchEngine.getDefaultSearchParticipant()), the provided document's path must be a path ending with one of the Java-like extensions or with '.class'.

      The given index location must represent a path in the file system to a file that either already exists or is going to be created. If it exists, it must be an index file, otherwise its data might be overwritten.

      Clients are not expected to call this method.

      Parameters:
      document - the document to index
      indexLocation - the location in the file system to the index
      Since:
      3.10

    • locateMatches

      public abstract void locateMatches(SearchDocument[] documents, SearchPattern pattern, IJavaSearchScope scope, SearchRequestor requestor, org.eclipse.core.runtime.IProgressMonitor monitor) throws org.eclipse.core.runtime.CoreException
      Locates the matches in the given documents using the given search pattern and search scope, and reports them to the given search requestor. This method is called by the search engine once it has search documents matching the given pattern in the given search scope.

      Note that a participant (e.g. a JSP participant) can pre-process the contents of the given documents, create its own documents whose contents are Java compilation units and delegate the match location to the default participant (see SearchEngine.getDefaultSearchParticipant()). Passing its own SearchRequestor this participant can then map the match positions back to the original contents, create its own matches and report them to the original requestor.

      Implementors of this method should check the progress monitor for cancelation when it is safe and appropriate to do so. The cancelation request should be propagated to the caller by throwing OperationCanceledException.

      Parameters:
      documents - the documents to locate matches in
      pattern - the search pattern to use when locating matches
      scope - the scope to limit the search to
      requestor - the requestor to report matches to
      monitor - the progress monitor to report progress to, or null if no progress should be reported
      Throws:
      org.eclipse.core.runtime.CoreException - if the requestor had problem accepting one of the matches

    • removeIndex

      public void removeIndex(org.eclipse.core.runtime.IPath indexLocation)
      Removes the index for a given path.

      The given index location must represent a path in the file system to a file that already exists and must be an index file, otherwise nothing will be done.

      It is strongly recommended to use this method instead of deleting file directly otherwise cached index will not be removed.

      Parameters:
      indexLocation - the location in the file system to the index
      Since:
      3.2

    • resolveDocument

      public void resolveDocument(SearchDocument document)
      Resolves the given document. A search participant asked to resolve a document should parse it and resolve the types and preserve enough state to be able to tend to a indexResolvedDocument call subsequently. This API is invoked without holding any index related locks or monitors.

      Clients are not expected to call this method.

      Parameters:
      document - the document to resolve
      Since:
      3.10
      See Also:

    • scheduleDocumentIndexing

      public final void scheduleDocumentIndexing(SearchDocument document, org.eclipse.core.runtime.IPath indexPath)
      Schedules the indexing of the given document. Once the document is ready to be indexed, indexDocument(document, indexPath) will be called in a different thread than the caller's thread.

      The given index location must represent a path in the file system to a file that either already exists or is going to be created. If it exists, it must be an index file, otherwise its data might be overwritten.

      When the index is no longer needed, clients should use removeIndex(IPath) to discard it.

      Parameters:
      document - the document to index
      indexPath - the location on the file system of the index

    • selectIndexes

      public abstract org.eclipse.core.runtime.IPath[] selectIndexes(SearchPattern query, IJavaSearchScope scope)
      Returns the collection of index locations to consider when performing the given search query in the given scope. The search engine calls this method before locating matches.

      An index location represents a path in the file system to a file that holds index information.

      Clients are not expected to call this method.

      Parameters:
      query - the search pattern to consider
      scope - the given search scope
      Returns:
      the collection of index paths to consider