com.bigdata.rdf.sail.webapp

Class BigdataRDFContext

java.lang.Object

com.bigdata.rdf.sail.BigdataBaseContext

com.bigdata.rdf.sail.webapp.BigdataRDFContext

public class BigdataRDFContext
extends BigdataBaseContext

Class encapsulates state shared by QueryServlet(s) for the same IIndexManager.

Author:

Martyn Cutcher, thompsonbry@users.sourceforge.net

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
class	BigdataRDFContext.AbstractQueryTask Abstract base class for running queries handles the timing, pipe, reporting, obtains the connection, and provides the finally {} semantics for each type of query task.

Field Summary

Fields

Modifier and Type	Field and Description
protected static String	ANALYTIC URL Query parameter used to request the "analytic" query hints.
protected static String	CONTEXT_URI URL query parameter used to specify the default context(s) for INSERT (POST-WITH-URIs, POST-WITH-BODY).
protected static String	DEFAULT_GRAPH_URI URL query parameter used to specify a URI in the default graph for SPARQL query (but not for SPARQL update).
protected static String	DEFAULT_XSL_STYLESHEET The default XSL style sheet.
protected static String	EXPLAIN URL Query parameter used to request the explanation of a query rather than its results.
protected static String	EXPLAIN_DETAILS Optional value for the EXPLAIN URL query parameter that may be used to request more detail in the "EXPLAIN" of a query.
static String	HTTP_HEADER_BIGDATA_MAX_QUERY_MILLIS HTTP header may be used to specify the timeout for a query.
static String	HTTP_HEADER_ECHO_BACK_QUERY HTTP header may be used to echo back the query.
protected static String	MONITOR URL Query parameter used to request an incremental XHTML representation reporting on the progress of a SPARQL UPDATE.
protected static String	NAMED_GRAPH_URI URL query parameter used to specify a URI in the set of named graphs for SPARQL query (but not for SPARQL update).
protected static String	NAMESPACE URL query parameter used to specify a non-default KB namespace (as an alternative to having it in the URL path).
protected static String	RTO URL Query parameter used to request the use of the Runtime Query Optimizer.
protected static String	URI URL query parameter used to specify the URI(s) from which data will be loaded for INSERT (POST-WITH-URIs.
protected static String	USING_GRAPH_URI URL query parameter used to specify a URI in the default graph for SPARQL UPDATE.
protected static String	USING_NAMED_GRAPH_URI URL query parameter used to specify a URI in the set of named graphs for SPARQL UPDATE.
protected static String	XHTML URL Query parameter used to request an XHTML response for SPARQL QUERY or SPARQL UPDATE.
protected static String	XSL_STYLESHEET URL Query parameter used to specify an XSL style sheet to be associated with the response in combination with the XHTML URL query parameter.

Constructor Summary

Constructors

Constructor and Description
BigdataRDFContext(SparqlEndpointConfig config, IIndexManager indexManager)

Method Summary

Methods

Modifier and Type	Method and Description
void	abortTx(long tx) Abort a transaction obtained by newTx(long).
static void	addHtmlHeader(XMLBuilder.Node current, String charset) Utility method to consolidate header into a single location.
void	commitTx(long tx) Commit a transaction obtained by newTx(long)
static String	getBaseURI(javax.servlet.http.HttpServletRequest req, javax.servlet.http.HttpServletResponse resp) Return the effective baseURI for the request.
SparqlEndpointConfig	getConfig()
protected static Boolean	getEffectiveBooleanValue(String s, Boolean defaultValue) Return the effective boolean value of a URL query parameter such as "analytic".
protected static String	getEffectiveStringValue(String s, String defaultValue) Return the effective string value of a URL query parameter.
List<String>	getNamespacesTx(long tx)
AtomicLong	getQueryIdFactory()
BigdataRDFContext.AbstractQueryTask	getQueryTask(BigdataSailRepositoryConnection cxn, String namespace, long timestamp, String queryStr, String baseURI, ASTContainer astContainer, boolean includeInferred, Map<String,org.openrdf.model.Value> bindings, String acceptOverride, javax.servlet.http.HttpServletRequest req, javax.servlet.http.HttpServletResponse resp, OutputStream os) Return the task which will execute the SPARQL Query -or- SPARQL UPDATE.
ThreadPoolExecutorBaseStatisticsTask	getSampleTask()
AbstractTripleStore	getTripleStore(String namespace, long timestamp) Return a read-only view of the AbstractTripleStore for the given namespace will read from the commit point associated with the given timestamp.
long	newTx(long timestamp) Obtain a new transaction to protect operations against the specified view of the database.

Methods inherited from class com.bigdata.rdf.sail.BigdataBaseContext

getIndexManager, isScaleOut

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

EXPLAIN

protected static final String EXPLAIN

URL Query parameter used to request the explanation of a query rather than its results.

See Also:

Constant Field Values

EXPLAIN_DETAILS

protected static final String EXPLAIN_DETAILS

Optional value for the EXPLAIN URL query parameter that may be used to request more detail in the "EXPLAIN" of a query.

See Also:

Constant Field Values

ANALYTIC

protected static final String ANALYTIC

URL Query parameter used to request the "analytic" query hints. MAY be null, in which case we do not set QueryHints.ANALYTIC query hint.

See Also:

Constant Field Values

RTO

protected static final String RTO

URL Query parameter used to request the use of the Runtime Query Optimizer.

See Also:

Constant Field Values

XHTML

protected static final String XHTML

URL Query parameter used to request an XHTML response for SPARQL QUERY or SPARQL UPDATE. For SPARQL QUERY, this provides an XHTML table view of the solutions. For SPARQL UPDATE, this provides an incremental progress report on the UPDATE request.

See Also:

Constant Field Values

XSL_STYLESHEET

protected static final String XSL_STYLESHEET

URL Query parameter used to specify an XSL style sheet to be associated with the response in combination with the XHTML URL query parameter.

See Also:

Constant Field Values

DEFAULT_XSL_STYLESHEET

protected static final String DEFAULT_XSL_STYLESHEET

The default XSL style sheet.

See Also:

XSL_STYLESHEET

MONITOR

protected static final String MONITOR

URL Query parameter used to request an incremental XHTML representation reporting on the progress of a SPARQL UPDATE.

Note: When this option is specified, the SPARQL UPDATE will use an HTTP status code of 200 (Ok) even if the UPDATE fails. In the event of an UPDATE failure, the stack trace will be formatted into the response.

See Also:

Constant Field Values

DEFAULT_GRAPH_URI

protected static final String DEFAULT_GRAPH_URI

URL query parameter used to specify a URI in the default graph for SPARQL query (but not for SPARQL update).

See Also:

Constant Field Values

NAMED_GRAPH_URI

protected static final String NAMED_GRAPH_URI

URL query parameter used to specify a URI in the set of named graphs for SPARQL query (but not for SPARQL update).

See Also:

Constant Field Values

USING_GRAPH_URI

protected static final String USING_GRAPH_URI

URL query parameter used to specify a URI in the default graph for SPARQL UPDATE.

See Also:

Constant Field Values

URI

protected static final String URI

URL query parameter used to specify the URI(s) from which data will be loaded for INSERT (POST-WITH-URIs.

See Also:

InsertServlet, Constant Field Values

CONTEXT_URI

protected static final String CONTEXT_URI

URL query parameter used to specify the default context(s) for INSERT (POST-WITH-URIs, POST-WITH-BODY).

See Also:

InsertServlet, Constant Field Values

USING_NAMED_GRAPH_URI

protected static final String USING_NAMED_GRAPH_URI

URL query parameter used to specify a URI in the set of named graphs for SPARQL UPDATE.

See Also:

Constant Field Values

NAMESPACE

protected static final String NAMESPACE

URL query parameter used to specify a non-default KB namespace (as an alternative to having it in the URL path). The path takes precendence over this query parameter.

See Also:

BigdataRDFServlet.getNamespace(HttpServletRequest), Constant Field Values

HTTP_HEADER_BIGDATA_MAX_QUERY_MILLIS

public static final String HTTP_HEADER_BIGDATA_MAX_QUERY_MILLIS

HTTP header may be used to specify the timeout for a query.

See Also:

(Set timeout on remote query), Constant Field Values

HTTP_HEADER_ECHO_BACK_QUERY

public static final String HTTP_HEADER_ECHO_BACK_QUERY

HTTP header may be used to echo back the query.

See Also:

Constant Field Values

Constructor Detail

BigdataRDFContext

public BigdataRDFContext(SparqlEndpointConfig config,
                 IIndexManager indexManager)

Method Detail

getQueryIdFactory

public final AtomicLong getQueryIdFactory()

getConfig

public SparqlEndpointConfig getConfig()

getSampleTask

public ThreadPoolExecutorBaseStatisticsTask getSampleTask()

getBaseURI

public static String getBaseURI(javax.servlet.http.HttpServletRequest req,
                javax.servlet.http.HttpServletResponse resp)

Return the effective baseURI for the request. This may be set using the URL query parameter. If it is not set, it defaults to the request URL.

Parameters:

req - The request.

resp - The response.

Returns:

The effective baseURI and never null.

getEffectiveBooleanValue

protected static Boolean getEffectiveBooleanValue(String s,
                               Boolean defaultValue)

Return the effective boolean value of a URL query parameter such as "analytic". If the URL query parameter was not given, then the effective boolean value is the defaultValue. If a URL query parameter which is given without an explicit value, such as &analytic, then it will be reported as true.

Parameters:

s - The value of the URL query parameter.

defaultValue - The default value to return if the parameter was not given.

Returns:

The effective boolean value.

getEffectiveStringValue

protected static String getEffectiveStringValue(String s,
                             String defaultValue)

Return the effective string value of a URL query parameter. If the URL query parameter was not given, or if it was given without an explicit value, then the effective string value is the defaultValue.

Parameters:

s - The value of the URL query parameter.

defaultValue - The default value to return if the parameter was not given or given without an explicit value.

Returns:

The effective value.

getQueryTask

public BigdataRDFContext.AbstractQueryTask getQueryTask(BigdataSailRepositoryConnection cxn,
                                               String namespace,
                                               long timestamp,
                                               String queryStr,
                                               String baseURI,
                                               ASTContainer astContainer,
                                               boolean includeInferred,
                                               Map<String,org.openrdf.model.Value> bindings,
                                               String acceptOverride,
                                               javax.servlet.http.HttpServletRequest req,
                                               javax.servlet.http.HttpServletResponse resp,
                                               OutputStream os)
                                                 throws org.openrdf.query.MalformedQueryException,
                                                        IOException

Return the task which will execute the SPARQL Query -or- SPARQL UPDATE.

Note: The OutputStream is passed in rather than the HttpServletResponse in order to permit operations such as "DELETE WITH QUERY" where this method is used in a context which writes onto an internal pipe rather than onto the HttpServletResponse.

Parameters:

namespace - The namespace associated with the AbstractTripleStore view.

timestamp - The timestamp associated with the AbstractTripleStore view.

queryStr - The query (for log messages).

baseURI - The baseURI (since BLZG-2039).

astContainer - The ASTContainer for the already parsed query (since BLZG-2039).

includeInferred -

acceptOverride - Override the Accept header (optional). This is used by UPDATE and DELETE so they can control the RDFFormat of the materialized query results.

req - The request.

resp - The response.

os - Where to write the results. Note: This is NOT always the OutputStream associated with the response! For DELETE-WITH-QUERY and UPDATE-WITH-QUERY this is a PipedOutputStream.

Returns:

The task.

Throws:

IOException

org.openrdf.query.MalformedQueryException

getTripleStore

public AbstractTripleStore getTripleStore(String namespace,
                                 long timestamp)

Return a read-only view of the AbstractTripleStore for the given namespace will read from the commit point associated with the given timestamp.

Parameters:

namespace - The namespace.

timestamp - A timestamp -or- a tx identifier.

Returns:

The AbstractTripleStore -or- null if none is found for that namespace and timestamp. FIXME GROUP_COMMIT: Review all callers. They are suspect. The code will sometimes resolve the KB as of the timestamp, but, given that the default is to read against the lastCommitTime, that does NOT prevent a concurrent destroy or create of a KB that invalidates such a pre-condition test. The main reason for such pre-condition tests is to provide nice HTTP status code responses when an identified namespace does (or does not) exist. The better way to handle this is by pushing the pre-condition test down into the AbstractRestApiTask and then throwning out an appropriate marked exception that gets correctly converted into an HTTP BAD_REQUEST message rather than sending back a stack trace.

getNamespacesTx

public List<String> getNamespacesTx(long tx)

newTx

public long newTx(long timestamp)

Obtain a new transaction to protect operations against the specified view of the database. This uses the transaction mechanisms to prevent recycling during operations NOT OTHERWISE PROTECTED by a BigdataSail.BigdataSailConnection for what would otherwise amount to dirty reads. This is especially critical for reads on the global row store since it can not be protected by the BigdataSail.BigdataSailConnection for cases where the KB instance does not yet exist. The presence of such a tx does NOT prevent concurrent commits. It only prevents recycling during such commits (and even then only on the RWStore backend).

Parameters:

timestamp - The timestamp for the desired view.

Returns:

The transaction identifier -or- timestamp if the IIndexManager is not a Journal.

See Also:

ITransactionService.newTx(long), NSS concurrency problem with list namespaces and create namespace

abortTx

public void abortTx(long tx)

Abort a transaction obtained by newTx(long). This decements the native active transaction counter for the RWStore. Once that counter reaches zero, recycling will occur the next time an unisolated mutation goes through a commit on the journal.

Parameters:

tx - The transaction identifier.

commitTx

public void commitTx(long tx)

Commit a transaction obtained by newTx(long)

Parameters:

tx - The transaction identifier.

See Also:

Support read/write transactions in the REST API

addHtmlHeader

public static void addHtmlHeader(XMLBuilder.Node current,
                 String charset)
                          throws IOException

Utility method to consolidate header into a single location. The post-condition is that the current node is open for writing on the body element.

Parameters:

current -

Throws:

IOException