public abstract class StatementWrapper extends Statement
Statement
implementations that wrap another statement.
This is intended for use with a custom RetryPolicy
, LoadBalancingPolicy
or
SpeculativeExecutionPolicy
. The client code can wrap a statement to "mark" it, or add
information that will lead to special handling in the policy.
Example:
// Define your own subclass
public class MyCustomStatement extends StatementWrapper {
public MyCustomStatement(Statement wrapped) {
super(wrapped);
}
}
// In your load balancing policy, add a special case for that new type
public class MyLoadBalancingPolicy implements LoadBalancingPolicy {
public Iterator<Host> newQueryPlan(String loggedKeyspace, Statement statement) {
if (statement instanceof MyCustomStatement) {
// return specially crafted plan
} else {
// return default plan
}
}
}
// The client wraps whenever it wants to trigger the special plan
Statement s = new SimpleStatement("...");
session.execute(s); // will use default plan
session.execute(new MyCustomStatement(s)); // will use special plan
idempotent, NULL_PAYLOAD_VALUE
Modifier | Constructor and Description |
---|---|
protected |
StatementWrapper(Statement wrapped)
Builds a new instance.
|
Modifier and Type | Method and Description |
---|---|
Statement |
disableTracing()
Disables tracing for this query.
|
Statement |
enableTracing()
Enables tracing for this query.
|
ConsistencyLevel |
getConsistencyLevel()
The consistency level for this query.
|
long |
getDefaultTimestamp()
The default timestamp for this query.
|
int |
getFetchSize()
The fetch size for this query.
|
Host |
getHost() |
String |
getKeyspace()
Returns the keyspace this query operates on.
|
Map<String,ByteBuffer> |
getOutgoingPayload()
Returns this statement's outgoing payload.
|
ByteBuffer |
getPagingState() |
int |
getReadTimeoutMillis()
Return the per-host read timeout that was set for this statement.
|
RetryPolicy |
getRetryPolicy()
Returns the retry policy sets for this query, if any.
|
ByteBuffer |
getRoutingKey(ProtocolVersion protocolVersion,
CodecRegistry codecRegistry)
Returns the routing key (in binary raw form) to use for token aware routing of this query.
|
ConsistencyLevel |
getSerialConsistencyLevel()
The serial consistency level for this query.
|
Boolean |
isIdempotent()
Whether this statement is idempotent, i.e.
|
boolean |
isIdempotentWithDefault(QueryOptions queryOptions) |
boolean |
isTracing()
Returns whether tracing is enabled for this query or not.
|
int |
requestSizeInBytes(ProtocolVersion protocolVersion,
CodecRegistry codecRegistry)
Returns the number of bytes required to encode this statement.
|
Statement |
setConsistencyLevel(ConsistencyLevel consistency)
Sets the consistency level for the query.
|
Statement |
setDefaultTimestamp(long defaultTimestamp)
Sets the default timestamp for this query (in microseconds since the epoch).
|
Statement |
setFetchSize(int fetchSize)
Sets the query fetch size.
|
Statement |
setHost(Host host)
Sets the
Host that should handle this query. |
Statement |
setIdempotent(boolean idempotent)
Sets whether this statement is idempotent.
|
Statement |
setOutgoingPayload(Map<String,ByteBuffer> payload)
Set the given outgoing payload on this statement.
|
Statement |
setPagingState(PagingState pagingState)
Sets the paging state.
|
Statement |
setPagingState(PagingState pagingState,
CodecRegistry codecRegistry)
Sets the paging state.
|
Statement |
setPagingStateUnsafe(byte[] pagingState)
Sets the paging state.
|
Statement |
setReadTimeoutMillis(int readTimeoutMillis)
Overrides the default per-host read timeout (
SocketOptions.getReadTimeoutMillis() ) for
this statement. |
Statement |
setRetryPolicy(RetryPolicy policy)
Sets the retry policy to use for this query.
|
Statement |
setSerialConsistencyLevel(ConsistencyLevel serialConsistency)
Sets the serial consistency level for the query.
|
isBatchIdempotent
protected StatementWrapper(Statement wrapped)
wrapped
- the wrapped statement.public Statement setConsistencyLevel(ConsistencyLevel consistency)
Statement
setConsistencyLevel
in class Statement
consistency
- the consistency level to set.Statement
object.public ConsistencyLevel getConsistencyLevel()
Statement
getConsistencyLevel
in class Statement
null
if no consistency level has been
specified (through setConsistencyLevel
). In the latter case, the default
consistency level will be used.public Statement setSerialConsistencyLevel(ConsistencyLevel serialConsistency)
Statement
The serial consistency level is only used by conditional updates (INSERT
, UPDATE
or DELETE
statements with an IF
condition). For those, the serial
consistency level defines the consistency level of the serial phase (or "paxos" phase) while
the normal consistency level defines the consistency for the "learn" phase, i.e. what type of
reads will be guaranteed to see the update right away. For instance, if a conditional write has
a regular consistency of QUORUM (and is successful), then a QUORUM read is guaranteed to see
that write. But if the regular consistency of that write is ANY, then only a read with a
consistency of SERIAL is guaranteed to see it (even a read with consistency ALL is not
guaranteed to be enough).
The serial consistency can only be one of ConsistencyLevel.SERIAL
or ConsistencyLevel.LOCAL_SERIAL
. While ConsistencyLevel.SERIAL
guarantees full
linearizability (with other SERIAL updates), ConsistencyLevel.LOCAL_SERIAL
only
guarantees it in the local data center.
The serial consistency level is ignored for any query that is not a conditional update (serial reads should use the regular consistency level for instance).
setSerialConsistencyLevel
in class Statement
serialConsistency
- the serial consistency level to set.Statement
object.public ConsistencyLevel getSerialConsistencyLevel()
Statement
See Statement.setSerialConsistencyLevel(ConsistencyLevel)
for more detail on the serial
consistency level.
getSerialConsistencyLevel
in class Statement
null
if no serial consistency
level has been specified (through Statement.setSerialConsistencyLevel(ConsistencyLevel)
). In
the latter case, the default serial consistency level will be used.public Statement enableTracing()
Statement
By default (that is unless you call this method), tracing is not enabled.
enableTracing
in class Statement
Statement
object.public Statement disableTracing()
Statement
disableTracing
in class Statement
Statement
object.public boolean isTracing()
Statement
public ByteBuffer getRoutingKey(ProtocolVersion protocolVersion, CodecRegistry codecRegistry)
Statement
The routing key is optional in that implementers are free to return null
. The
routing key is an hint used for token-aware routing (see TokenAwarePolicy
), and if provided should correspond to the
binary value for the query partition key. However, not providing a routing key never causes a
query to fail and if the load balancing policy used is not token aware, then the routing key
can be safely ignored.
getRoutingKey
in class Statement
protocolVersion
- the protocol version that will be used if the actual implementation
needs to serialize something to compute the key.codecRegistry
- the codec registry that will be used if the actual implementation needs to
serialize something to compute this key.null
.public String getKeyspace()
Statement
Note that not all query specify on which keyspace they operate on, and so this method can
always return null
. Firstly, some queries do not operate inside a keyspace: keyspace
creation, USE
queries, user creation, etc. Secondly, even query that operate within a
keyspace do not have to specify said keyspace directly, in which case the currently logged in
keyspace (the one set through a USE
query (or through the use of Cluster.connect(String)
)). Lastly, as for the routing key, this keyspace information is only a
hint for token-aware routing (since replica placement depend on the replication strategy in use
which is a per-keyspace property) and having this method return null
(or even a bogus
keyspace name) will never cause the query to fail.
getKeyspace
in class Statement
null
.public Statement setRetryPolicy(RetryPolicy policy)
Statement
The default retry policy, if this method is not called, is the one returned by Policies.getRetryPolicy()
in the cluster configuration. This
method is thus only useful in case you want to punctually override the default policy for this
request.
setRetryPolicy
in class Statement
policy
- the retry policy to use for this query.Statement
object.public RetryPolicy getRetryPolicy()
Statement
getRetryPolicy
in class Statement
null
if no query specific
retry policy has been set through Statement.setRetryPolicy(com.datastax.driver.core.policies.RetryPolicy)
(in which case the Cluster retry
policy will apply if necessary).public Statement setFetchSize(int fetchSize)
Statement
The fetch size controls how much resulting rows will be retrieved simultaneously (the goal being to avoid loading too much results in memory for queries yielding large results). Please note that while value as low as 1 can be used, it is *highly* discouraged to use such a low value in practice as it will yield very poor performance. If in doubt, leaving the default is probably a good idea.
Only SELECT
queries only ever make use of this setting.
Note that unlike other configuration, when this statement is prepared BoundStatement
s created off of PreparedStatement
do not inherit this configuration.
Note: Paging is not supported with the native protocol version 1. If you call this method
with fetchSize > 0
and fetchSize != Integer.MAX_VALUE
and the protocol
version is in use (i.e. if you've force version 1 through Cluster.Builder.withProtocolVersion(com.datastax.driver.core.ProtocolVersion)
or you use Cassandra 1.2), you will get UnsupportedProtocolVersionException
when submitting this statement for execution.
setFetchSize
in class Statement
fetchSize
- the fetch size to use. If fetchSize <e; 0
, the default fetch size
will be used. To disable paging of the result set, use fetchSize ==
Integer.MAX_VALUE
.Statement
object.public int getFetchSize()
Statement
getFetchSize
in class Statement
Statement.setFetchSize(int)
is used), the default fetch size will be used.public Statement setDefaultTimestamp(long defaultTimestamp)
Statement
This feature is only available when version V3
or higher of the
native protocol is in use. With earlier versions, calling this method has no effect.
The actual timestamp that will be used for this query is, in order of preference:
USING
TIMESTAMP
syntax);
Long.MIN_VALUE
;
TimestampGenerator
currently in use, if different
from Long.MIN_VALUE
.
Note that unlike other configuration, when this statement is prepared BoundStatement
s created off of PreparedStatement
do not inherit this configuration.
setDefaultTimestamp
in class Statement
defaultTimestamp
- the default timestamp for this query (must be strictly positive).Statement
object.Cluster.Builder.withTimestampGenerator(TimestampGenerator)
public long getDefaultTimestamp()
Statement
getDefaultTimestamp
in class Statement
public Statement setReadTimeoutMillis(int readTimeoutMillis)
Statement
SocketOptions.getReadTimeoutMillis()
) for
this statement.
You should override this only for statements for which the coordinator may allow a longer server-side timeout (for example aggregation queries).
Note that unlike other configuration, when this statement is prepared BoundStatement
s created off of PreparedStatement
do not inherit this configuration.
setReadTimeoutMillis
in class Statement
readTimeoutMillis
- the timeout to set. Negative values are not allowed. If it is 0, the
read timeout will be disabled for this statement.Statement
object.public int getReadTimeoutMillis()
Statement
getReadTimeoutMillis
in class Statement
SocketOptions.getReadTimeoutMillis()
will be used.public Statement setPagingState(PagingState pagingState, CodecRegistry codecRegistry)
Statement
This will cause the next execution of this statement to fetch results from a given page, rather than restarting from the beginning.
You get the paging state from a previous execution of the statement (see ExecutionInfo.getPagingState()
. This is typically used to iterate in a "stateless" manner
(e.g. across HTTP requests):
Statement st = new SimpleStatement("your query");
ResultSet rs = session.execute(st.setFetchSize(20));
int available = rs.getAvailableWithoutFetching();
for (int i = 0; i < available; i++) {
Row row = rs.one();
// Do something with row (e.g. display it to the user...)
}
// Get state and serialize as string or byte[] to store it for the next execution
// (e.g. pass it as a parameter in the "next page" URI)
PagingState pagingState = rs.getExecutionInfo().getPagingState();
String savedState = pagingState.toString();
// Next execution:
// Get serialized state back (e.g. get URI parameter)
String savedState = ...
Statement st = new SimpleStatement("your query");
st.setPagingState(PagingState.fromString(savedState));
ResultSet rs = session.execute(st.setFetchSize(20));
int available = rs.getAvailableWithoutFetching();
for (int i = 0; i < available; i++) {
...
}
The paging state can only be reused between perfectly identical statements (same query string, same bound parameters). Altering the contents of the paging state or trying to set it on a different statement will cause this method to fail.
Note that, due to internal implementation details, the paging state is not portable across
native protocol versions (see the online documentation
for more explanations about the native protocol). This means that PagingState
instances
generated with an old version won't work with a higher version. If that is a problem for you,
consider using the "unsafe" API (see Statement.setPagingStateUnsafe(byte[])
).
setPagingState
in class Statement
pagingState
- the paging state to set, or null
to remove any state that was
previously set on this statement.codecRegistry
- the codec registry that will be used if this method needs to serialize the
statement's values in order to check that the paging state matches.Statement
object.Statement.setPagingState(PagingState)
public Statement setPagingState(PagingState pagingState)
Statement
This method calls Statement.setPagingState(PagingState, CodecRegistry)
with CodecRegistry.DEFAULT_INSTANCE
. Whether you should use this or the other variant depends on
the type of statement this is called on:
BoundStatement
, the codec registry isn't actually needed, so it's always
safe to use this method;
SimpleStatement
or BuiltStatement
, you can use this method if you
use no custom codecs, or if your custom codecs are registered with the default registry.
Otherwise, use the other method and provide the registry that contains your codecs.
setPagingState
in class Statement
pagingState
- the paging state to set, or null
to remove any state that was
previously set on this statement.public Statement setPagingStateUnsafe(byte[] pagingState)
Statement
Contrary to Statement.setPagingState(PagingState)
, this method takes the "raw" form of the
paging state (previously extracted with ExecutionInfo.getPagingStateUnsafe()
. It won't
validate that this statement matches the one that the paging state was extracted from. If the
paging state was altered in any way, you will get unpredictable behavior from Cassandra
(ranging from wrong results to a query failure). If you decide to use this variant, it is
strongly recommended to add your own validation (for example, signing the raw state with a
private key).
setPagingStateUnsafe
in class Statement
pagingState
- the paging state to set, or null
to remove any state that was
previously set on this statement.Statement
object.public ByteBuffer getPagingState()
public Statement setIdempotent(boolean idempotent)
Statement
See Statement.isIdempotent()
for more explanations about this property.
setIdempotent
in class Statement
idempotent
- the new value.Statement
object.public Boolean isIdempotent()
Statement
If a statement is not idempotent, the driver will ensure that it never gets executed more than once, which means:
retries
on write timeouts or request errors;
speculative executions
.
Note that this method can return null
, in which case the driver will default to
QueryOptions.getDefaultIdempotence()
.
By default, this method returns null
for all statements, except for
BuiltStatement
- value will be inferred from the query: if it updates counters,
prepends/appends to a list, or uses a function call or QueryBuilder.raw(String)
anywhere in an inserted
value, the result will be false
; otherwise it will be true
.
Batch
and BatchStatement
:
Statement.setIdempotent(boolean)
forces a value that overrides calculated
value.
Note that when a statement is prepared (Session.prepare(String)
), its idempotence
flag will be propagated to all PreparedStatement
s created from it.
isIdempotent
in class Statement
null
to use QueryOptions.getDefaultIdempotence()
.public boolean isIdempotentWithDefault(QueryOptions queryOptions)
public Map<String,ByteBuffer> getOutgoingPayload()
Statement
This method returns null
if no payload has been set, otherwise it always returns
immutable maps.
This feature is only available with ProtocolVersion.V4
or above. Trying to include
custom payloads in requests sent by the driver under lower protocol versions will result in an
UnsupportedFeatureException
(wrapped in a NoHostAvailableException
).
getOutgoingPayload
in class Statement
null
if no payload has
been set.public Statement setOutgoingPayload(Map<String,ByteBuffer> payload)
Statement
This method makes a defensive copy of the given map, but its values remain inherently mutable. Care should be taken not to modify the original map once it is passed to this method.
This feature is only available with ProtocolVersion.V4
or above. Trying to include
custom payloads in requests sent by the driver under lower protocol versions will result in an
UnsupportedFeatureException
(wrapped in a NoHostAvailableException
).
setOutgoingPayload
in class Statement
payload
- the outgoing payload to include with this statement, or null
to clear
any previously entered payload.Statement
object.public int requestSizeInBytes(ProtocolVersion protocolVersion, CodecRegistry codecRegistry)
Statement
The calculated size may be overestimated by a few bytes, but is never underestimated. If the size cannot be calculated, this method returns -1.
Note that the returned value is not cached, but instead recomputed at every method call.
requestSizeInBytes
in class Statement
public Host getHost()
getHost
in class Statement
Statement.setHost(Host)
public Statement setHost(Host host)
Statement
Host
that should handle this query.
In the general case, use of this method is heavily discouraged and should only be used in the following cases:
system
and system_views
keyspaces.
Configuring a specific host causes the configured LoadBalancingPolicy
to be
completely bypassed. However, if the load balancing policy dictates that the host is at
distance HostDistance.IGNORED
or there is no active connectivity to the host, the
request will fail with a NoHostAvailableException
.
Note that unlike other configuration, when this statement is prepared BoundStatement
s created off of PreparedStatement
do not inherit this configuration.
Copyright © 2012–2018. All rights reserved.