Package cz.cvut.kbss.ontodriver.rdf4j.config

Class Rdf4jOntoDriverProperties

java.lang.Object
cz.cvut.kbss.ontodriver.rdf4j.config.Rdf4jOntoDriverProperties
public abstract class Rdf4jOntoDriverProperties extends Object

  • Field Details

    • USE_VOLATILE_STORAGE

      public static final String USE_VOLATILE_STORAGE
      Specifies whether an in-memory storage should be used for local RDF4J repositories.

      When set to true, any local RDF4J repositories that are created by the driver are created as only MemoryStores without any persistent backend. Repositories accessed over the Internet or already existing locally are not affected by this setting.

      Boolean value expected, default is false.

      See Also:

    • USE_INFERENCE

      public static final String USE_INFERENCE
      Specifies whether RDF4J inference (RDFS, forward chaining) should be used.

      Note that this setting applies only to local storages (in memory or native), remote storages use their own inference settings.

      Boolean value expected, default is false.

      See Also:

    • LOAD_ALL_THRESHOLD

      public static final String LOAD_ALL_THRESHOLD
      Specifies how many requested assertions suffice to perform load all.

      More specifically, if the number of assertions requested by an AxiomDescriptor is low enough, the statements for them will be loaded by asking RDF4J for statements with subject and property bound.

      Otherwise, statements will be loaded using only subject bound and will be filtered according to the assertions. This will in most cases have better performance than loading with bound property.

      See Also:

    • REPOSITORY_CONFIG

      public static final String REPOSITORY_CONFIG
      Path to repository configuration file.

      The configuration file is a Turtle file corresponding to the repository config file schema defined by RDF4J. See for example http://docs.rdf4j.org/server-workbench-console/#_repository_configuration_templates_advanced

      Sample configuration files are available in the RDF4J GitHub repo: https://github.com/eclipse/rdf4j/tree/master/repository/api/src/main/resources/org/eclipse/rdf4j/repository/config

      The config file can be present on the classpath (if the value starts with the classpath: prefix) or can be loaded using the value as file path. The loaded configuration supersedes relevant properties passed to the driver, e.g., USE_VOLATILE_STORAGE or USE_INFERENCE.

      Note that the config applies only to embedded repositories created by the driver, repositories on a RDF4J server to which the driver just connects must preexist and the configuration does not apply to them. The physical URI specified in configuration must correspond to the URI of the repository in the config file, i.e., for memory store, the repository ID must be the same. For a native store, the physical URI must be in the form /local-path/repositories/repository-id, where local-path will be used for initialization of a local RepositoryManager and repository-id must again correspond to the repository ID in the configuration file.

      See Also:

    • RECONNECT_ATTEMPTS

      public static final String RECONNECT_ATTEMPTS
      Number of attempts to reconnect to the repository in case of connection issues.

      This setting applies only when using a remote rdf4j-server repository (including GraphDB). Applies also when initial connection on startup is attempted.

      Note that once connection is successful, the counter of failed attempts is reset.

      See Also:

    • INFERENCE_IN_DEFAULT_CONTEXT

      public static final String INFERENCE_IN_DEFAULT_CONTEXT
      Whether inferred assertion values are expected to exist in the default repository context.

      Normally, RDF4J stores inferred statements in the context of the statement from which they were inferred. However, if one uses SPIN rules or custom inference rules https://rdf4j.org/documentation/programming/repository/#custom-inferencing, their results will be stored in the default context. Setting this parameter to true will cause the statement loading to use the default context for loading inferred assertions.

      Note that GraphDB inference is handled automatically and does not need this parameter. However, setting it will override the default automatic inference context resolution for GraphDB.

      Defaults to false, i.e., inferred statements are expected in the same context as their causes.

      See Also:

    • CONNECTION_REQUEST_TIMEOUT

      public static final String CONNECTION_REQUEST_TIMEOUT
      Timeout for the underlying HTTP client to request a connection to the remote repository from the connection pool.

      RDF4J internally uses the Apache HTTP Client with a connection pool that is able to grow up to a configured limit. When the pool is exhausted and the driver requests another connection, it may happen that no connection is available will become available, possibly leading to a deadlock. This parameter sets a timeout for the connection acquisition requests from the pool, so that the application is able to fail gracefully in case of too much load and does not get stuck.

      The value should be an integer corresponding to the number of milliseconds.

      Applies only to instances connected to a RDF4J server.

      See Also:

    • MAX_CONNECTION_POOL_SIZE

      public static final String MAX_CONNECTION_POOL_SIZE
      Maximum size of the underlying HTTP client connection pool.

      RDF4J internally uses the Apache HTTP Client with a connection pool that is able to grow up to a configured limit. This parameter allows to set this limit.

      See Also:

    • TRANSACTION_ISOLATION_LEVEL

      public static final String TRANSACTION_ISOLATION_LEVEL
      RDF4J transaction isolation level.

      See IsolationLevels for acceptable values.

      See Also: