Global Options
dotNetRDF has a set of global static options provided by the Options class. These are used to configure various aspects of dotNetRDF behaviour, some act as defaults for certain functionality while others toggle functionality/features on/off as desired.
Here are the available options:
Core Options
FullTripleIndexing
Controls whether implementations of BaseTripleCollection that support indexing will create full triple indexes, defaults to true. See Minimizing Memory Usage for more discussions on triple indexing.
InternUris
Controls the URI Interning feature of the library which is used to speed up URI equality checks and reduce memory usage. However depending on your usage of the library it may actually inflate memory usage, see the linked page for more discussion of this option.
Literal Options
LiteralEqualityMode
Controls the behaviour when you call Equals() on instances of ILiteralNode. This takes a value from the LiteralEqualityMode enumeration, the default is Strict. In Strict mode literal equality follows RDF term equality semantics i.e. the lexical values must be exactly equal and so must the language tag/datatype URI if present.
If you set this to Loose instead then value equality semantics are used i.e. if the literals represent the same value then they are considered equal.
Please see Equality and Comparison for more information.
LiteralValueNormalization
Controls whether literal values have Unicode normalization applied to them when they are created, this defaults to false. May be useful to enable if you are ingesting non-normalized data though may increase data load times.
DefaultCulture
Controls the culture used for comparing the lexical values of literals when lexical sorting is used. The default is to use the invariant culture to preserve behavioural compatibility with past releases. It may be useful to customize this setting if you want to more naturally sort string data.
ComparisonOptions
Controls how the lexical values of literals are compared when lexical sorting is used. The default is to use Ordinal comparison which preserves behavioural compatibility with past releases. It may be useful to customize this setting if you want to more naturally sort string data.
HTTP Options
ForceHttpBasicAuth
Controls whether to always send HTTP Basic Authentication headers in HTTP requests when appropriate, defaults to false.
Normally dotNetRDF relies on the standard .Net HTTP support which uses the standard HTTP challenge response pattern for establishing authenticated connections. However some badly behaved servers may not carry out the challenge response correctly in which case it may be necessary to enable this option. This only works for servers that use HTTP Basic Authentication, other authentication types cannot be forced.
HttpDebugging
Controls whether to print debug information about HTTP requests and responses to the console, defaults to false. See Debugging HTTP Communication for more information.
HttpFullDebugging
Controls whether the full response stream for HTTP responses is printed to the console for debugging, defaults to false. See Debugging HTTP Communication for more information.
Parsing Options
ForceBlockingIO
Controls whether parsers will rely on fully blocking buffered IO to read in data, defaults to false.
Normally the library chooses whether to use this based on the type of data being read and avoids fully blocking buffered IO when possible. However they may be occasions where the chosen option is incorrect if you use an unusual data source or the data source is wrapped. Therefore sometimes it may be necessary to enforce blocking IO by setting this option.
UriLoaderCaching
Controls whether dotNetRDF will maintain a cache of RDF retrieved from files for future use, defaults to true. This cache can speed up subsequent loads from the web but may not be desirable if the retrieved data is being auto-generated or frequently update. Thus in some cases you may wish to disable this functionality.
UriLoaderTimeout
Sets the timeout for retrieving RDF data from the web in milliseconds, defaults to 15,000 i.e. 15 seconds. This timeout reflects the time to establish a connection to the HTTP server, it is not a timeout on how long parsing the retrieved data takes.
It may be useful to increase this if you have a slow internet connection or are accessing RDF from a slow service.
Query Options
QueryExecutionTimeout
Global timeout setting for in-memory query execution given in milliseconds, defaults to 180,000 i.e. 3 minutes. This setting has no-effect for other forms of query execution.
This can be overridden as a per-query level by the Timeout property of a SparqlQuery instance.
QueryOptimisation
Controls whether query pattern optimization is applied, defaults to true. May be useful to disable if the default optimizer is optimizing your query poorly.
AlgebraOptimisation
Controls whether query algebra optimization is applied, defaults to true. May be useful to disable if the standard optimizers are optimizing your query poorly or if you want to execute the basic algebra only.
UnsafeOptimisation
Controls whether certain optimizations which may make queries run faster but change the semantics of the query can be applied, defaults to false.
QueryDefaultSyntax
Sets the default syntax used for parsing SPARQL queries, takes a value from the SparqlQuerySyntax enumeration. Defaults to Sparql_1_1 i.e. SPARQL 1.1
QueryAllowUnknownFunctions
Controls whether unknown SPARQL extension functions should be allowed, defaults to true. When enabled unknown extension functions are simply treated as functions that return unbound for every input. When disabled an unknown extension function will give a parser error.
RigorousEvaluation
Controls whether the in-memory SPARQL engine rigorously checks the provided matches from the underlying ISparqlDataset, defaults to false as this is usually unnecessary.
You may need to enable this if your ISparqlDataset implementation cannot guarantee that a call to GetTriplesWithX() only returns triples matching the given arguments.
StrictOperations
Controls whether SPARQL operators are restricted to only the implementations required by the SPARQL specifications, defaults to false.
When disabled certain additional operators will be permitted such as date time computations. See SPARQL Operators for discussion of this feature.
UsePLinqEvaluation
Controls whether the in-memory SPARQL engine uses PLINQ evaluation to improve SPARQL performance, defaults to true. This is only available on platforms that support PLINQ i.e. .Net 4.0 non-Silverlight builds.
May be useful to disable if you have a multi-core machine but only want to use a single core for query evaluation.
Update Options
UpdateExecutionTimeout
Sets the timeout used for in-memory update execution in milliseconds, defaults to 180,000 i.e. 3 minutes. This setting has no-effect for other forms of update execution.
This can be overridden as a per-update level by the Timeout property of a Sparql
Writing Options
AllowMultiThreadedWriting
Controls whether writers that support multi-threading writing will use this to improve write performance, defaults to false. This is disabled by default because while it can improve performance it may cause error information to propagate incorrectly resulting in a failed write with no user indication of this.
DefaultCompressionLevel
Sets the default compression level that writers will use if they support varying compression levels.
UseBomForUtf8
Controls whether the optional UTF-8 BOM is included in UTF-8 output, defaults to false.
Generally speaking it is best to leave this setting untouched since creating files with UTF-8 BOM may render those files unusable by non .Net tools since most UTF-8 aware tools don't expect to see the optional BOM.
UseDtd
Controls whether writers that may use DTDs in their output will actually use a DTD, defaults to false.
Many places have restrictions on the use of DTDs in documents because these can be used to do DoS attacks. Enable this option only if using dotNetRDF as part of a system where all components are trusted to exchange data containing DTDs.