/// <summary>
/// Initialises the class to map the content item URL.
/// </summary>
/// <param name="context">An IConnectionContext object that contains the content item URL to map.</param>
/// <remarks>
/// The following URL's are supported:
/// glyma://[Repository Name] - Defines a crawl start address for a Glyma repository defined in the BCS model file.
/// glyma://[Repository Name]/[Domain Guid] - Defines a crawl start adddress for the specified domain in a Glyma repository.
/// glyma://[Repository Name]/[Domain GUID]/[Root Map GUID]/[Parent Map GUID]/[Node GUID]/ - Defines the URL for a Glyma Map.
/// glyma://[Repository Name]/[Domain GUID]/[Root Map GUID]/[Parent Map GUID]/[Node GUID] - Defines the URL for a Glyma Node.
/// </remarks>
public override void Initialize(Microsoft.Office.Server.Search.Connector.IConnectionContext context)
{
_sourceUri = context.Path;
if (!_sourceUri.Scheme.Equals(GlymaModelConstants.Protocol, StringComparison.OrdinalIgnoreCase))
{
throw new ArgumentException("Invalid Glyma URL provided. The URL must start with: glyma://.");
}
_lobSystem = this.Catalog.GetLobSystem(GlymaModelConstants.LobSystemName);
// Get the LOB system instance based on the host specified in the context path.
// The LOB system instance is passed as a parameter to the ExecuteStatic method in the GlymaRepositoryUtility class.
if (string.IsNullOrEmpty(_sourceUri.Host))
{
throw new ArgumentException("Invalid Glyma URL provided. The URL must specify a repository name in the form: glyma://[Repository Name] where [Repository Name] is the name of a LOB system instance in the BCS model file.");
}
INamedLobSystemInstanceDictionary lobInstances = _lobSystem.GetLobSystemInstances();
_lobSystemInstance = _lobSystem.GetLobSystemInstances().GetByKey(_sourceUri.Host, StringComparison.OrdinalIgnoreCase);
if (_lobSystemInstance == null)
{
throw new ArgumentException("Invalid Glyma URL provided. The repository name specified does not exist in the BCS model file.");
}
// Set the entity to be a Glyma Map by default. When the URL is a crawl start address, it won't contain the identity of an entity instance.
// By setting the entity but not the entity instance, the Finder method of the entity is executed to return entity instances to crawl.
_entity = this.Catalog.GetEntity(GlymaModelConstants.Namespace, GlymaModelConstants.MapEntityName);
// Extract the entity instance identity from the context path. The entity instance could either be a Glyma Map (which contains a trailing '/' in its URL) or a Glyma Node.
// If an entity instance identity can be extracted, the SpecificFinder method of the entity is executed to return details for that entity instance.
if (_sourceUri.Segments.Length == 5)
{
if (!_sourceUri.Segments[4].EndsWith("/"))
{
_entity = this.Catalog.GetEntity(GlymaModelConstants.Namespace, GlymaModelConstants.NodeEntityName);
}
string domainIdString = _sourceUri.Segments[1].TrimEnd('/');
string rootMapIdString = _sourceUri.Segments[2].TrimEnd('/');
string mapIdString = _sourceUri.Segments[3].TrimEnd('/');
string nodeIdString = _sourceUri.Segments[4].TrimEnd('/');
Guid domainId = new Guid(domainIdString);
Guid rootMapId = new Guid(rootMapIdString);
Guid mapId = new Guid(mapIdString);
Guid nodeId = new Guid(nodeIdString);
Object[] identifier = { _lobSystemInstance.Name, domainId, rootMapId, mapId, nodeId };
_identity = new Microsoft.BusinessData.Runtime.Identity(identifier);
}
}
/// <summary>
/// Initialises the class to map the content item URL.
/// </summary>
/// <param name="context">An IConnectionContext object that contains the content item URL to map.</param>
/// <remarks>
/// The following URL's are supported:
/// glyma://[Repository Name] - Defines a crawl start address for a Glyma repository defined in the BCS model file.
/// glyma://[Repository Name]/[Domain Guid] - Defines a crawl start adddress for the specified domain in a Glyma repository.
/// glyma://[Repository Name]/[Domain GUID]/[Root Map GUID]/[Parent Map GUID]/[Node GUID]/ - Defines the URL for a Glyma Map.
/// glyma://[Repository Name]/[Domain GUID]/[Root Map GUID]/[Parent Map GUID]/[Node GUID] - Defines the URL for a Glyma Node.
/// </remarks>
public override void Initialize(Microsoft.Office.Server.Search.Connector.IConnectionContext context)
{
_sourceUri = context.Path;
if (!_sourceUri.Scheme.Equals(GlymaModelConstants.Protocol, StringComparison.OrdinalIgnoreCase))
{
throw new ArgumentException("Invalid Glyma URL provided. The URL must start with: glyma://.");
}
_lobSystem = this.Catalog.GetLobSystem(GlymaModelConstants.LobSystemName);
// Get the LOB system instance based on the host specified in the context path.
// The LOB system instance is passed as a parameter to the ExecuteStatic method in the GlymaRepositoryUtility class.
if (string.IsNullOrEmpty(_sourceUri.Host))
{
throw new ArgumentException("Invalid Glyma URL provided. The URL must specify a repository name in the form: glyma://[Repository Name] where [Repository Name] is the name of a LOB system instance in the BCS model file.");
}
INamedLobSystemInstanceDictionary lobInstances = _lobSystem.GetLobSystemInstances();
_lobSystemInstance = _lobSystem.GetLobSystemInstances().GetByKey(_sourceUri.Host, StringComparison.OrdinalIgnoreCase);
if (_lobSystemInstance == null)
{
throw new ArgumentException("Invalid Glyma URL provided. The repository name specified does not exist in the BCS model file.");
}
// Set the entity to be a Glyma Map by default. When the URL is a crawl start address, it won't contain the identity of an entity instance.
// By setting the entity but not the entity instance, the Finder method of the entity is executed to return entity instances to crawl.
_entity = this.Catalog.GetEntity(GlymaModelConstants.Namespace, GlymaModelConstants.MapEntityName);
// Extract the entity instance identity from the context path. The entity instance could either be a Glyma Map (which contains a trailing '/' in its URL) or a Glyma Node.
// If an entity instance identity can be extracted, the SpecificFinder method of the entity is executed to return details for that entity instance.
if (_sourceUri.Segments.Length == 5)
{
if (!_sourceUri.Segments[4].EndsWith("/"))
{
_entity = this.Catalog.GetEntity(GlymaModelConstants.Namespace, GlymaModelConstants.NodeEntityName);
}
string domainIdString = _sourceUri.Segments[1].TrimEnd('/');
string rootMapIdString = _sourceUri.Segments[2].TrimEnd('/');
string mapIdString = _sourceUri.Segments[3].TrimEnd('/');
string nodeIdString = _sourceUri.Segments[4].TrimEnd('/');
Guid domainId = new Guid(domainIdString);
Guid rootMapId = new Guid(rootMapIdString);
Guid mapId = new Guid(mapIdString);
Guid nodeId = new Guid(nodeIdString);
Object[] identifier = { _lobSystemInstance.Name, domainId, rootMapId, mapId, nodeId };
_identity = new Microsoft.BusinessData.Runtime.Identity(identifier);
}
}
public override void Initialize(Microsoft.Office.Server.Search.Connector.IConnectionContext context)
{
//connector.UriParser uri = new connector.UriParser(context.Path);
//Uri sourceUri = uri.ToUri(false); // No case encoding
Uri sourceUri = context.Path;
this.lobSystemInstance =
this.lobSystem.GetLobSystemInstances()[0].Value;
string filepath = @"\\" + sourceUri.Host + sourceUri.AbsolutePath.Replace('/', '\\');
///
///To decode spaces and other characters in file names that are encoded by search
///
filepath = SPHttpUtility.UrlPathDecode(filepath, false);
if (Directory.Exists(filepath))
{
///
///Only set the entity for the start address so that idEnumerator is called on the entity
///
this.entity = this.Catalog.GetEntity("MyFileConnector", "MyFolder");
///
///If the Uri belongs to a folder, set the identity so that
///associations, specificifnder and streamaccessors are called
///
this.identity = new Microsoft.BusinessData.Runtime.Identity(filepath);
}
else if (File.Exists(filepath))
{
///
///If it is a file, switch the entity name to MyFile
///
this.entity = this.Catalog.GetEntity("MyFileConnector", "MyFile");
///
///If the Uri belongs to a file, set the identity so that findSpecific is called
///
this.identity = new Microsoft.BusinessData.Runtime.Identity(filepath);
}
}
/// <summary>
/// This method does the processing of incoming URLs. The intent here is to receive a URL via this context
/// object, parse it to determine what metadata object it refers to, and then populate members of this class
/// with this information.
/// </summary>
/// <param name="context">
/// Supplies the connection context. The context contains (most importantly) the URL of the item being crawled,
/// and also other information about the current crawl.
/// </param>
public override void Initialize(
Microsoft.Office.Server.Search.Connector.IConnectionContext context)
{
Uri sourceUri = context.Path;
//
// A URL can point to a LobSystem, a LobSystemInstance, an Entity, or a specific instance of an Entity.
// A specific instance of an Entity for MyFileConnector would be, of course, an actual file or folder on
// disk. The other BCS metadata objects exist to provide you more flexibility in designing your connector
// by providing slightly different crawling behaviors (which we'll detail below).
//
// Your job in this method is to parse the URL and decide what it's pointing to. Then, populate the
// appropriate class members to indicate what it's pointing to.
//
// To indicate that the supplied URL refers to a(n): Populate only these class members:
//
// LobSystem this.LobSystem
//
// LobSystemInstance this.LobSystem
// this.LobSystemInstance
//
// Entity this.LobSystem
// this.LobSystemInstance
// this.Entity
//
// Entity instance this.LobSystem
// this.LobSystemInstance
// this.Entity
// this.Identity
//
// What happens when these BCS metadata objects are crawled? Obviously, if you've got an entity instance,
// you're crawling an actual item in your repository. But what about the others? In general, there's no real
// data or metadata associated with any of them until you get down to an actual Entity instance (i.e., a
// file or folder on disk, in the case of this connector). So the only real thing that happens when these
// objects are crawled is that they emit their children, which will then subsequently be crawled. So here's
// what the "children" of each of these metadata objects are:
//
// Metadata object type Children that are emitted when crawled
//
// LobSystem All LobSystemInstances in the model file that are marked with the
// 'ShowInSearchUI' property.
//
// LobSystemInstance If at least one Entity defined in this LobSystemInstance has a Finder method
// with the 'RootFinder' property set, the LobSystemInstance will emit all Entities
// that have Finder methods marked with the 'RootFinder' property. If an Entity
// does not have a RootFinder, it will not be emitted.
//
// If no Entities in the LobSystemInstance with a Finder marked with the
// 'RootFinder' property, the LobSystemInstance will emit any Entities that have
// both IdEnumerator and SpecificFinder methods defined in the model file. Note: If
// you define only one of the two (either IdEnumerator or SpecificFinder) in this
// case, an exception will be thrown at crawl time. You must define both to be
// crawled in this manner.
//
// (Up to this point, with the previous two metadata objects, we had not yet called into the custom
// connector shim. Instead, we were able to figure out what to emit from the model file alone. Starting with
// the Entity, we will be calling into the custom connector shim, and also the full and incremental crawl
// behavior can be different...)
//
// Metadata object type Children that are emitted when crawled
//
// Entity Full crawl:
// If the Entity has a Finder with the 'RootFinder' property set, that method is
// called. Whatever entity instances your shim emits are the child items of the
// Entity. If there is no Finder with the 'RootFinder' property set, but there
// is an IdEnumerator defined for the entity, that method will be called, and
// whatever entity instances your shim emits are the child items of the Entity.
//
// Incremental crawl:
// If the Entity is the source for any AssociationNavigator methods that are marked
// with the 'DirectoryLink' property, the behavior is the same as in the full
// crawl. If there are no AssociationNavigator methods where the Entity is the
// source, marked with the 'DirectoryLink' property, and the Entity has both a
// ChangedIdEnumerator and DeletedIdEnumerator defined, then both of those methods
// will be called in an incremental crawl. If none of the above is true, the
// incremental crawl behavior is the same as full crawl.
//
// (Now, for an actual Entity instance (e.g., a file or folder on disk, in this case), it will be helpful to
// define the concept of 'container'-type Entities. A 'container' type Entity is an Entity that is defined
// in your model file as the SourceEntity of any AssociationNavigator marked with the 'DirectoryLink'
// property.)
//
// Metadata object type Children that are emitted when crawled
//
// Container-type Full crawl:
// Entity instance The association navigators for which the Entity type is the SourceEntity are
// called.
//
// Incremental crawl:
// This is the same as in a full crawl, unless several things are configured to
// enable your shim to more intelligently decide what to emit. Here's what needs
// to be configured to enable this behavior:
//
// 1. The SpecificFinder of the Entity has to return a field containing the number
// of deleted direct child items in that particular container and this field
// must be identified in the model file by defining the 'DeletedCountField'
// property on the SpecificFinder method instance, where its value must be the
// name of the field that returns the delete count.
//
// 2. The AssociationNavigator for which this Entity is the SourceEntity must
// include in its return type descriptor a last modified time DateTime field
// and this field must be identified in the model file by defining the
// 'LastModifiedTimeStamp' property on the AssociationNavigator method
// instance, where its value must be the name of the fiels that returns the
// last modified DateTime.
//
// 3. The AssociationNavigator for which this Entity is the SourceEntity must
// have an input filter defined, and that input filter must have the
// 'CrawlStartTime' string property defined:
//
// <FilterDescriptor Name="LastModifiedFilter" Type="Input">
// <Properties>
// <Property Name="CrawlStartTime" Type="String">x</Property>
// </Properties>
// </FilterDescriptor>
//
// 4. The AssociationNavigator for this this Entity is the SourceEntity must
// take a DateTime input parameter, and that parameter must be associated with
// the filter described in #3:
//
// <Parameter Name="lastModifiedTime" Direction="In">
// <TypeDescriptor
// Name="lastModifiedTime"
// TypeName="System.DateTime"
// AssociatedFilter="LastModifiedFilter" />
// </Parameter>
//
// If this is the case, then the association navigator is called, and your shim
// will be provided the last modified time to determine what items to return.
//
// Non-container-type Full and incremental crawl behavior is the same. The SpecificFinder of the
// Entity instance non-container Entity is called. You can also enable caching behavior, if the
// method that emitted this Entity instance has the 'UseClientCachingForSearch'
// property defined on its method instance. If caching is enabled in this way,
// the SpecificFinder is not called, and only the data returned by the method
// that originally emitted this Entity instance (i.e., an AssociationNavigator
// or IdEnumerator, etc.) will be used to index the item. Note: If you return a
// security descriptor in the return type descriptor, identified by the
// 'WindowsSecurityDescriptorField' defined on the method instance, then items
// will *not* be cached, regardless of if you set the 'UseClientCachingForSearch'
// property or not. The reason is that the SharePoint Search crawler has a limited
// size for caching an individual item, and security descriptors can regularly
// exceed that size.
//
//
// As we mentioned above, we expect that every URL will resolve to a valid folder or file once we
// translate it. Because of this, we won't really need to crawl any of the BCS metadata objects (LobSystem,
// LobSystemInstance or Entity) - we'll only be crawling Entity instances. So based on the above, we know
// that this method must end up populating all four LobUri properties.
//
//
// We already populated this.lobSystem in the constructor. Now here's this.lobSystemInstance.
// LobSystemInstances are statically defined in your model file. In this sample, we've only defined one.
//
this.lobSystemInstance = this.lobSystem.GetLobSystemInstances()[0].Value;
//
// Next, we need to figure out what Entity to which the URL refers to populate this.Entity. So for starters,
// let's read the first segment to see what entity type we should assign to this.Entity.
//
String entityType = sourceUri.Segments[1].Replace("/", "");
if(entityType.Equals("MyFolder", StringComparison.OrdinalIgnoreCase))
{
this.entity = this.Catalog.GetEntity("MyFileConnector", "MyFolder");
}
else if (entityType.Equals("MyFile", StringComparison.OrdinalIgnoreCase))
{
this.entity = this.Catalog.GetEntity("MyFileConnector", "MyFile");
}
else
{
throw new Microsoft.BusinessData.Runtime.RuntimeException(String.Format(
"Invalid entity type {0} specified in URL {1}",
entityType,
sourceUri.ToString()));
}
//
// Finally, populate this.identity. Generally, the 'identity' is whatever your repository needs to uniquely
// identify an Entity instance. In the case of our repository, an NTFS file system, that would conveniently
// just be the path to the file. So reconstruct the URL as a UNC path, minus the first segment.
//
StringBuilder path = new StringBuilder(@"\\");
path.Append(sourceUri.Host);
path.Append(@"\");
for (int i = 2; i < sourceUri.Segments.Length; ++i)
{
path.Append(SPHttpUtility.UrlPathDecode(sourceUri.Segments[i].Replace('/', '\\'), false));
}
this.identity = new Microsoft.BusinessData.Runtime.Identity(path.ToString());
}
