- All Implemented Interfaces:
Closeable
,AutoCloseable
,MergePolicy.MergeContext
,TwoPhaseCommit
,Accountable
IndexWriter
creates and maintains an index.
The IndexWriterConfig.OpenMode
option on IndexWriterConfig.setOpenMode(OpenMode)
determines
whether a new index is created, or whether an existing index is opened. Note that you can open an
index with IndexWriterConfig.OpenMode.CREATE
even while readers are using the index. The old readers will
continue to search the "point in time" snapshot they had opened, and won't see the newly created
index until they re-open. If IndexWriterConfig.OpenMode.CREATE_OR_APPEND
is used IndexWriter will create a
new index if there is not already an index at the provided path and otherwise open the existing
index.
In either case, documents are added with addDocument
and
removed with deleteDocuments(Term...)
or deleteDocuments(Query...)
. A document
can be updated with updateDocument
(which just deletes
and then adds the entire document). When finished adding, deleting and updating documents, close
should be called.
Each method that changes the index returns a long
sequence number, which expresses the
effective order in which each change was applied. commit()
also returns a sequence number,
describing which changes are in the commit point and which are not. Sequence numbers are
transient (not saved into the index in any way) and only valid within a single
IndexWriter
instance.
These changes are buffered in memory and periodically flushed to the Directory
(during
the above method calls). A flush is triggered when there are enough added documents since the
last flush. Flushing is triggered either by RAM usage of the documents (see IndexWriterConfig.setRAMBufferSizeMB(double)
) or the number of added documents (see IndexWriterConfig.setMaxBufferedDocs(int)
). The default is to flush when RAM usage hits IndexWriterConfig.DEFAULT_RAM_BUFFER_SIZE_MB
MB. For best indexing speed you should flush by RAM
usage with a large RAM buffer. In contrast to the other flush options IndexWriterConfig.setRAMBufferSizeMB(double)
and IndexWriterConfig.setMaxBufferedDocs(int)
,
deleted terms won't trigger a segment flush. Note that flushing just moves the internal buffered
state in IndexWriter into the index, but these changes are not visible to IndexReader until
either commit()
or close()
is called. A flush may also trigger one or more
segment merges which by default run with a background thread so as not to block the addDocument
calls (see below for changing the MergeScheduler
).
Opening an IndexWriter
creates a lock file for the directory in use. Trying to
open another IndexWriter
on the same directory will lead to a LockObtainFailedException
.
Expert: IndexWriter
allows an optional IndexDeletionPolicy
implementation
to be specified. You can use this to control when prior commits are deleted from the index. The
default policy is KeepOnlyLastCommitDeletionPolicy
which removes all prior commits as
soon as a new commit is done. Creating your own policy can allow you to explicitly keep previous
"point in time" commits alive in the index for some time, either because this is useful for your
application, or to give readers enough time to refresh to the new commit without having the old
commit deleted out from under them. The latter is necessary when multiple computers take turns
opening their own IndexWriter
and IndexReader
s against a single shared index
mounted via remote filesystems like NFS which do not support "delete on last close" semantics. A
single computer accessing an index via NFS is fine with the default deletion policy since NFS
clients emulate "delete on last close" locally. That said, accessing an index via NFS will likely
result in poor performance compared to a local IO device.
Expert: IndexWriter
allows you to separately change the MergePolicy
and
the MergeScheduler
. The MergePolicy
is invoked whenever there are changes to the
segments in the index. Its role is to select which merges to do, if any, and return a MergePolicy.MergeSpecification
describing the merges. The default is LogByteSizeMergePolicy
. Then, the MergeScheduler
is invoked with the requested merges
and it decides when and how to run the merges. The default is ConcurrentMergeScheduler
.
NOTE: if you hit a VirtualMachineError, or disaster strikes during a checkpoint then IndexWriter will close itself. This is a defensive measure in case any internal state (buffered documents, deletions, reference counts) were corrupted. Any subsequent calls will throw an AlreadyClosedException.
NOTE: IndexWriter
instances are completely thread safe, meaning multiple
threads can call any of its methods, concurrently. If your application requires external
synchronization, you should not synchronize on the IndexWriter
instance as
this may cause deadlock; use your own (non-Lucene) objects instead.
NOTE: If you call Thread.interrupt()
on a thread that's within
IndexWriter, IndexWriter will try to catch this (eg, if it's in a wait() or Thread.sleep()), and
will then throw the unchecked exception ThreadInterruptedException
and clear the
interrupt status on the thread.
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescriptionprivate class
private static interface
static final class
DocStats for this index(package private) static interface
Interface for internal atomic events.(package private) static final class
static interface
IfDirectoryReader.open(IndexWriter)
has been called (ie, this writer is in near real-time mode), then after a merge completes, this class can be invoked to warm the reader on the newly merged segment, before the merge commits.private static class
private class
-
Field Summary
FieldsModifier and TypeFieldDescriptionprivate static int
private final IndexWriter.AddIndexesMergeSource
private final BufferedUpdatesStream
private final AtomicLong
private boolean
private boolean
private final Object
private final LiveIndexWriterConfig
private final IndexFileDeleter
private boolean
private final Directory
private final Directory
(package private) final DocumentsWriter
private final boolean
Used only for testing.private final IndexWriterEventListener
private final IndexWriter.EventQueue
private Collection<String>
private final AtomicInteger
private final AtomicInteger
private final DocumentsWriter.FlushNotifications
private final Object
(package private) final FieldInfos.FieldNumbers
private final InfoStream
If enabled, information about merges will be printed to this.private long
static final int
Hard limit on maximum number of documents that may be added to the index.static final int
Maximum value of the token position in an indexed field.static final int
Maximum length string for a stored field.static final int
Absolute hard maximum length for a term, in bytes once encoded as UTF8.private final AtomicBoolean
private final List<MergePolicy.OneMerge>
private final AtomicLong
Counts how many merges have completed; this is used byforceApply(FrozenBufferedUpdates)
to handle concurrently apply deletes/updates with merges completing.private long
private int
private IndexWriter.Merges
private final MergeScheduler
private final MergeScheduler.MergeSource
private final HashSet<SegmentCommitInfo>
private SegmentInfos
private long
private final Deque<MergePolicy.OneMerge>
private final AtomicLong
How many documents are in the index, or are in the process of being added (reserved).private long
private final ReaderPool
private List<SegmentCommitInfo>
private final Set<SegmentMerger>
private final Set<MergePolicy.OneMerge>
private final SegmentInfos
private final Map<SegmentCommitInfo,
Boolean> private final boolean
static final String
Key for the source of a segment in thediagnostics
.static final String
Source of a segment which results from a call toaddIndexes(CodecReader...)
.static final String
Source of a segment which results from a flush.static final String
Source of a segment which results from a merge of other segments.private long
System.nanoTime() when commit started; used to write an infoStream message about how long commit took.private final AtomicReference<Throwable>
private static final int
static final String
Name of the write lock in the index.private final ReentrantLock
private Lock
Fields inherited from interface org.apache.lucene.util.Accountable
NULL_ACCOUNTABLE
-
Constructor Summary
ConstructorsConstructorDescriptionIndexWriter
(Directory d, IndexWriterConfig conf) Constructs a new IndexWriter per the settings given inconf
. -
Method Summary
Modifier and TypeMethodDescriptionprivate void
_mergeInit
(MergePolicy.OneMerge merge) private void
Aborts running merges.private void
acquireWriteLocks
(Directory... dirs) Acquires write locks on all the directories; be sure to match with a call toIOUtils.close(java.io.Closeable...)
in a finally clause.long
addDocument
(Iterable<? extends IndexableField> doc) Adds a document to this index.long
addDocuments
(Iterable<? extends Iterable<? extends IndexableField>> docs) Atomically adds a block of documents with sequentially assigned document IDs, such that an external reader will see all or none of the documents.long
addIndexes
(CodecReader... readers) Merges the provided indexes into this index.long
addIndexes
(Directory... dirs) Adds all segments from an array of indexes into this index.void
Runs a single merge operation foraddIndexes(CodecReader...)
.private void
private long
adjustPendingNumDocs
(long numDocs) void
advanceSegmentInfosVersion
(long newVersion) IfSegmentInfos.getVersion()
is belownewVersion
then update it to this value.private void
private boolean
assertSoftDeletesCount
(CodecReader reader, int expectedCount) private DocValuesUpdate[]
buildDocValuesUpdate
(Term term, Field[] updates) private static void
carryOverHardDeletes
(ReadersAndUpdates mergedReadersAndUpdates, int maxDoc, Bits prevHardLiveDocs, Bits currentHardLiveDocs, MergeState.DocMap segDocMap) This method carries over hard-deleted documents that are applied to the source segment during a merge.private void
changed()
Called internally if any index state has changed.private void
Called whenever the SegmentInfos has been updated and the index files referenced exist (correctly) in the index directory.private void
Checkpoints with IndexFileDeleter, so it's aware of new files, and increments changeCount, so on close/commit we will write a new segments file, but does NOT bump segmentInfos.version.(package private) final SegmentInfos
Tests should use this method to snapshot the current segmentInfos to have a consistent viewvoid
close()
Closes all open resources and releases the write lock.private void
closeMergeReaders
(MergePolicy.OneMerge merge, boolean suppressExceptions, boolean droppedSegment) closeSegmentStates
(BufferedUpdatesStream.SegmentState[] segStates, boolean success) Close segment states previously opened with openSegmentStates.final long
commit()
Commits all pending changes (added and deleted documents, segment merges, added indexes, etc.) to the index, and syncs all referenced index files, such that a reader will see the changes and the index updates will survive an OS or machine crash or power loss.private long
commitInternal
(MergePolicy mergePolicy) private boolean
commitMerge
(MergePolicy.OneMerge merge, MergeState.DocMap[] docMaps) private ReadersAndUpdates
commitMergedDeletesAndUpdates
(MergePolicy.OneMerge merge, MergeState.DocMap[] docMaps) Carefully merges deletes and updates for the segments we just merged.private SegmentCommitInfo
copySegmentAsIs
(SegmentCommitInfo info, String segName, IOContext context) Copies the segment files as-is into the IndexWriter's directory.private void
countSoftDeletes
(CodecReader reader, Bits wrappedLiveDocs, Bits hardLiveDocs, Counter softDeleteCounter, Counter hardDeleteCounter) (package private) static void
createCompoundFile
(InfoStream infoStream, TrackingDirectoryWrapper directory, SegmentInfo info, IOContext context, IOConsumer<Collection<String>> deleteFiles) NOTE: this method creates a compound file for all files returned by info.files().void
decRefDeleter
(SegmentInfos segmentInfos) Record that the files referenced by thisSegmentInfos
are no longer in use.long
Delete all documents in the index.long
deleteDocuments
(Term... terms) Deletes the document(s) containing any of the terms.long
deleteDocuments
(Query... queries) Deletes the document(s) matching any of the provided queries.private void
deleteNewFiles
(Collection<String> files) Tries to delete the given files if unreferencedvoid
Expert: remove any index files that are no longer used.protected void
A hook for extending classes to execute operations after pending added and deleted documents have been flushed to the Directory but before the change is committed (new segments_N file written).protected void
A hook for extending classes to execute operations before pending added and deleted documents are flushed to the Directory.private boolean
doFlush
(boolean applyAllDeletes) Returns true a segment was flushed or deletes were applied.private void
doWait()
private void
Drops a segment that has 100% deleted documents.protected final void
Used internally to throw anAlreadyClosedException
if this IndexWriter has been closed (closed=true
) or is in the process of closing (closing=true
).protected final void
ensureOpen
(boolean failIfClosing) Used internally to throw anAlreadyClosedException
if this IndexWriter has been closed or is in the process of closing.private void
(package private) final void
executeMerge
(MergeTrigger trigger) private boolean
filesExist
(SegmentInfos toSync) private void
finishApply
(BufferedUpdatesStream.SegmentState[] segStates, boolean success, Set<String> delFiles) private void
private StandardDirectoryReader
finishGetReaderMerge
(AtomicBoolean stopCollectingMergedReaders, Map<String, SegmentReader> mergedReaders, Map<String, SegmentReader> openedReadOnlyClones, SegmentInfos openingSegmentInfos, boolean applyAllDeletes, boolean writeAllDeletes, MergePolicy.MergeSpecification pointInTimeMerges, long maxCommitMergeWaitMillis) final void
flush()
(package private) final void
flush
(boolean triggerMerge, boolean applyAllDeletes) Flush all in-memory buffered updates (adds and deletes) to the Directory.private void
flushFailed
(SegmentInfo info) Cleans up residuals from a segment that could not be entirely flushed due to an errorfinal boolean
Expert: Flushes the next pending writer per thread buffer if available or the largest active non-pending writer per thread buffer in the calling thread.(package private) final void
forceApply
(FrozenBufferedUpdates updates) Translates a frozen packet of delete term/query, or doc values updates, into their actual docIDs in the index, and applies the change.void
forceMerge
(int maxNumSegments) Forces merge policy to merge segments until there are<= maxNumSegments
.void
forceMerge
(int maxNumSegments, boolean doWait) Just likeforceMerge(int)
, except you can specify whether the call should block until all merging completes.void
Forces merging of all segments that have deleted documents.void
forceMergeDeletes
(boolean doWait) Just likeforceMergeDeletes()
, except you can specify whether the call should block until the operation completes.(package private) static int
Returns the analyzer used by this index.(package private) final int
Returns aLiveIndexWriterConfig
, which can be used to query the IndexWriter current settings, as well as modify "live" ones.Returns the Directory used by this index.Returns accurateIndexWriter.DocStats
for this writer.(package private) DocumentsWriter
Return an unmodifiable set of all field names as visible from this IndexWriter, across all segments of the index.private FieldInfos.FieldNumbers
Loads or returns the already loaded the global field number map for thisSegmentInfos
.(package private) final int
(package private) final int
final long
Returns the number of bytes currently being flushedprivate List<SegmentCommitInfo>
getInfosToApply
(FrozenBufferedUpdates updates) Returns theSegmentCommitInfo
that this packet is supposed to apply its deletes to, or null if the private segment was already merged away.Returns the info stream that can be used to log messagesReturns the commit user data iterable previously set withsetLiveCommitData(Iterable)
, or null if nothing has been set yet.long
Returns the highest sequence number across all completed operations, or 0 if no operations have finished yet.Expert: to be used by aMergePolicy
to avoid selecting merges for segments already being merged.private MergePolicy.OneMerge
Expert: theMergeScheduler
calls this method to retrieve the next merge requested by the MergePolicy(package private) final int
long
Returns the number of documents in the index including documents are being added (i.e., reserved).(package private) ReadersAndUpdates
getPooledInstance
(SegmentCommitInfo info, boolean create) (package private) DirectoryReader
getReader
(boolean applyAllDeletes, boolean writeAllDeletes) Expert: returns a readonly reader, covering all committed as well as un-committed changes to the index.(package private) final int
If thisIndexWriter
was closed as a side-effect of a tragic exception, e.g.private void
(package private) boolean
Returns true if there are any changes or deletes that are not flushed or applied.boolean
Returns true if this index has deletions (including buffered deletions).boolean
Expert: returns true if there are merges waiting to be scheduled.final boolean
Returns true if there may be changes that have not been committed.void
incRefDeleter
(SegmentInfos segmentInfos) Record that the files referenced by thisSegmentInfos
are still in use.(package private) boolean
isClosed()
(package private) static boolean
isCongruentSort
(Sort indexSort, Sort otherSort) Returns true ifindexSort
is a prefix ofotherSort
.(package private) boolean
protected boolean
Tests should override this to enable test points.(package private) final boolean
isFullyDeleted
(ReadersAndUpdates readersAndUpdates) boolean
isOpen()
Returnstrue
if thisIndexWriter
is still open.(package private) final int
maxDoc
(int i) private boolean
Returns true if any merges in pendingMerges or runningMerges are maxNumSegments merges.private void
final void
Expert: asks the mergePolicy whether any merges are necessary now and if so, runs the requested merges and then iterate (test again if merges are needed) until no more merges are returned by the mergePolicy.private final void
maybeMerge
(MergePolicy mergePolicy, MergeTrigger trigger, int maxNumSegments) private long
maybeProcessEvents
(long seqNo) Processes all events and might trigger a merge if the given seqNo is negativeprivate StandardDirectoryReader
maybeReopenMergedNRTReader
(Map<String, SegmentReader> mergedReaders, Map<String, SegmentReader> openedReadOnlyClones, SegmentInfos openingSegmentInfos, boolean applyAllDeletes, boolean writeAllDeletes) protected void
merge
(MergePolicy.OneMerge merge) Merges the indicated segments, replacing them in the stack with a single segment.private void
mergeFinish
(MergePolicy.OneMerge merge) Does finishing for a merge, which is fast but holds the synchronized lock on IndexWriter instance.(package private) final void
mergeInit
(MergePolicy.OneMerge merge) Does initial setup for a merge, which is fast but holds the synchronized lock on IndexWriter instance.private int
mergeMiddle
(MergePolicy.OneMerge merge, MergePolicy mergePolicy) Does the actual (time-consuming) work of the merge, but without holding synchronized lock on IndexWriter instanceprotected void
mergeSuccess
(MergePolicy.OneMerge merge) Hook that's called when the specified merge is complete.private void
(package private) SegmentCommitInfo
private final String
private void
(package private) boolean
nrtIsCurrent
(SegmentInfos infos) int
Obtain the number of deleted docs for a pooled reader.final int
Returns the number of deletes a merge would claim back if the given segment is merged.final int
Expert: Return the number of documents currently buffered in RAM.void
onTragicEvent
(Throwable tragedy, String location) This method should be called on a tragic event ie.private BufferedUpdatesStream.SegmentState[]
openSegmentStates
(List<SegmentCommitInfo> infos, Set<SegmentCommitInfo> alreadySeenSegments, long delGen) Opens SegmentReader and inits SegmentState for each segment.final long
Expert: prepare for commit.private long
private MergePolicy.MergeSpecification
preparePointInTimeMerge
(SegmentInfos mergingSegmentInfos, BooleanSupplier stopCollectingMergeResults, MergeTrigger trigger, IOConsumer<SegmentCommitInfo> mergeFinished) This optimization allows a commit/getReader to wait for merges on smallish segments to reduce the eventual number of tiny segments in the commit point / NRT Reader.private void
processEvents
(boolean triggerMerge) private void
publishFlushedSegment
(SegmentCommitInfo newSegment, FieldInfos fieldInfos, FrozenBufferedUpdates packet, FrozenBufferedUpdates globalPacket, Sorter.DocMap sortMap) Atomically adds the segment private delete packet and publishes the flushed segments SegmentInfo to the index writer.private void
publishFlushedSegments
(boolean forced) Publishes the flushed segment, segment-private deletes (if any) and its associated global delete (if present) to IndexWriter.private long
final long
Return the memory usage of this object in bytes.(package private) static FieldInfos
private boolean
Checks whether this merge involves any segments already participating in a merge.(package private) void
release
(ReadersAndUpdates readersAndUpdates) private void
release
(ReadersAndUpdates readersAndUpdates, boolean assertLiveInfo) private void
reserveDocs
(long addedNumDocs) Anything that will add N docs to the index should reserve first to make sure it's allowed.private void
void
rollback()
Close theIndexWriter
without committing any changes that have occurred since the last commit (or since it was opened, if commit hasn't been called).private void
private void
(package private) String
Returns a string description of all segments, for debugging.(package private) String
segString
(Iterable<SegmentCommitInfo> infos) private String
segString
(SegmentCommitInfo info) Returns a string description of the specified segment, for debugging.(package private) static void
setDiagnostics
(SegmentInfo info, String source) private static void
setDiagnostics
(SegmentInfo info, String source, Map<String, String> details) final void
setLiveCommitData
(Iterable<Map.Entry<String, String>> commitUserData) Sets the iterator to provide the commit user data map at commit time.final void
setLiveCommitData
(Iterable<Map.Entry<String, String>> commitUserData, boolean doIncrementVersion) Sets the commit user data iterator, controlling whether to advance theSegmentInfos.getVersion()
.(package private) static void
setMaxDocs
(int maxDocs) Used only for testing.private boolean
shouldClose
(boolean waitForClose) private void
shutdown()
Gracefully closes (commits, waits for merges), but calls rollback if there's an exc so the IndexWriter is always closed.long
softUpdateDocument
(Term term, Iterable<? extends IndexableField> doc, Field... softDeletes) Expert: Updates a document by first updating the document(s) containingterm
with the given doc-values fields and then adding the new document.long
softUpdateDocuments
(Term term, Iterable<? extends Iterable<? extends IndexableField>> docs, Field... softDeletes) Expert: Atomically updates documents matching the provided term with the given doc-values fields and adds a block of documents with sequentially assigned document IDs, such that an external reader will see all or none of the documents.private void
startCommit
(SegmentInfos toSync) Walk through all files referenced by the current segmentInfos and ask the Directory to sync each file, if it wasn't already.private void
private void
testReserveDocs
(long addedNumDocs) Does a best-effort check, that the current index would accept this many additional docs, but does not actually reserve them.(package private) SegmentInfos
toLiveInfos
(SegmentInfos sis) private void
tooManyDocs
(long addedNumDocs) private void
tragicEvent
(Throwable tragedy, String location) This method set the tragic exception unless it's already set and closes the writer if necessary.(package private) final boolean
tryApply
(FrozenBufferedUpdates updates) Translates a frozen packet of delete term/query, or doc values updates, into their actual docIDs in the index, and applies the change.long
tryDeleteDocument
(IndexReader readerIn, int docID) Expert: attempts to delete by document ID, as long as the provided reader is a near-real-time reader (fromDirectoryReader.open(IndexWriter)
).private long
tryModifyDocument
(IndexReader readerIn, int docID, IndexWriter.DocModifier toApply) long
tryUpdateDocValue
(IndexReader readerIn, int docID, Field... fields) Expert: attempts to update doc values by document ID, as long as the provided reader is a near-real-time reader (fromDirectoryReader.open(IndexWriter)
).long
updateBinaryDocValue
(Term term, String field, BytesRef value) long
updateDocument
(Term term, Iterable<? extends IndexableField> doc) Updates a document by first deleting the document(s) containingterm
and then adding the new document.private long
updateDocuments
(DocumentsWriterDeleteQueue.Node<?> delNode, Iterable<? extends Iterable<? extends IndexableField>> docs) long
updateDocuments
(Term delTerm, Iterable<? extends Iterable<? extends IndexableField>> docs) Atomically deletes documents matching the provided delTerm and adds a block of documents with sequentially assigned document IDs, such that an external reader will see all or none of the documents.long
updateDocuments
(Query delQuery, Iterable<? extends Iterable<? extends IndexableField>> docs) Similar toupdateDocuments(Term, Iterable)
, but take a query instead of a term to identify the documents to be updatedlong
updateDocValues
(Term term, Field... updates) Updates documents' DocValues fields to the given values.long
updateNumericDocValue
(Term term, String field, long value) private MergePolicy.MergeSpecification
updatePendingMerges
(MergePolicy mergePolicy, MergeTrigger trigger, int maxNumSegments) private void
validate
(SegmentCommitInfo info) private void
Confirms that the incoming index sort (if any) matches the existing index sort (if any).private void
(package private) void
Wait for any currently outstanding merges to finish.private void
writeReaderPool
(boolean writeDeletes) Ensures that all changes in the reader-pool are written to disk.(package private) final void
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
Methods inherited from interface org.apache.lucene.util.Accountable
getChildResources
-
Field Details
-
MAX_DOCS
public static final int MAX_DOCSHard limit on maximum number of documents that may be added to the index. If you try to add more than this you'll hitIllegalArgumentException
.- See Also:
-
MAX_POSITION
public static final int MAX_POSITIONMaximum value of the token position in an indexed field.- See Also:
-
actualMaxDocs
private static int actualMaxDocs -
enableTestPoints
private final boolean enableTestPointsUsed only for testing. -
UNBOUNDED_MAX_MERGE_SEGMENTS
private static final int UNBOUNDED_MAX_MERGE_SEGMENTS- See Also:
-
WRITE_LOCK_NAME
Name of the write lock in the index.- See Also:
-
SOURCE
Key for the source of a segment in thediagnostics
.- See Also:
-
SOURCE_MERGE
Source of a segment which results from a merge of other segments.- See Also:
-
SOURCE_FLUSH
Source of a segment which results from a flush.- See Also:
-
SOURCE_ADDINDEXES_READERS
Source of a segment which results from a call toaddIndexes(CodecReader...)
.- See Also:
-
MAX_TERM_LENGTH
public static final int MAX_TERM_LENGTHAbsolute hard maximum length for a term, in bytes once encoded as UTF8. If a term arrives from the analyzer longer than this length, anIllegalArgumentException
is thrown and a message is printed to infoStream, if set (seeIndexWriterConfig.setInfoStream(InfoStream)
).- See Also:
-
MAX_STORED_STRING_LENGTH
public static final int MAX_STORED_STRING_LENGTHMaximum length string for a stored field. -
tragedy
-
directoryOrig
-
directory
-
changeCount
-
lastCommitChangeCount
private volatile long lastCommitChangeCount -
rollbackSegments
-
pendingCommit
-
pendingSeqNo
private volatile long pendingSeqNo -
pendingCommitChangeCount
private volatile long pendingCommitChangeCount -
filesToCommit
-
segmentInfos
-
globalFieldNumberMap
-
docWriter
-
eventQueue
-
mergeSource
-
addIndexesMergeSource
-
writeDocValuesLock
-
deleter
-
segmentsToMerge
-
mergeMaxNumSegments
private int mergeMaxNumSegments -
writeLock
-
closed
private volatile boolean closed -
closing
private volatile boolean closing -
maybeMerge
-
commitUserData
-
mergingSegments
-
mergeScheduler
-
runningAddIndexesMerges
-
pendingMerges
-
runningMerges
-
mergeExceptions
-
mergeGen
private long mergeGen -
merges
-
didMessageState
private boolean didMessageState -
flushCount
-
flushDeletesCount
-
readerPool
-
bufferedUpdatesStream
-
eventListener
-
mergeFinishedGen
Counts how many merges have completed; this is used byforceApply(FrozenBufferedUpdates)
to handle concurrently apply deletes/updates with merges completing. -
config
-
startCommitTime
private long startCommitTimeSystem.nanoTime() when commit started; used to write an infoStream message about how long commit took. -
pendingNumDocs
How many documents are in the index, or are in the process of being added (reserved). E.g., operations like addIndexes will first reserve the right to add N docs, before they actually change the index, much like how hotels place an "authorization hold" on your credit card to make sure they can later charge you when you check out. -
softDeletesEnabled
private final boolean softDeletesEnabled -
flushNotifications
-
infoStream
If enabled, information about merges will be printed to this. -
commitLock
-
fullFlushLock
-
-
Constructor Details
-
IndexWriter
Constructs a new IndexWriter per the settings given inconf
. If you want to make "live" changes to this writer instance, usegetConfig()
.NOTE: after ths writer is created, the given configuration instance cannot be passed to another writer.
- Parameters:
d
- the index directory. The index is either created or appended accordingconf.getOpenMode()
.conf
- the configuration settings according to which IndexWriter should be initialized.- Throws:
IOException
- if the directory cannot be read/written to, or if it does not exist andconf.getOpenMode()
isOpenMode.APPEND
or if there is any other low-level IO error
-
-
Method Details
-
setMaxDocs
static void setMaxDocs(int maxDocs) Used only for testing. -
getActualMaxDocs
static int getActualMaxDocs() -
getReader
Expert: returns a readonly reader, covering all committed as well as un-committed changes to the index. This provides "near real-time" searching, in that changes made during an IndexWriter session can be quickly made available for searching without closing the writer nor callingcommit()
.Note that this is functionally equivalent to calling {#flush} and then opening a new reader. But the turnaround time of this method should be faster since it avoids the potentially costly
commit()
.You must close the
IndexReader
returned by this method once you are done using it.It's near real-time because there is no hard guarantee on how quickly you can get a new reader after making changes with IndexWriter. You'll have to experiment in your situation to determine if it's fast enough. As this is a new and experimental feature, please report back on your findings so we can learn, improve and iterate.
The resulting reader supports
DirectoryReader.openIfChanged(org.apache.lucene.index.DirectoryReader)
, but that call will simply forward back to this method (though this may change in the future).The very first time this method is called, this writer instance will make every effort to pool the readers that it opens for doing merges, applying deletes, etc. This means additional resources (RAM, file descriptors, CPU time) will be consumed.
For lower latency on reopening a reader, you should call
IndexWriterConfig.setMergedSegmentWarmer(org.apache.lucene.index.IndexWriter.IndexReaderWarmer)
to pre-warm a newly merged segment before it's committed to the index. This is important for minimizing index-to-search delay after a large merge.If an addIndexes* call is running in another thread, then this reader will only search those segments from the foreign index that have been successfully copied over, so far.
NOTE: Once the writer is closed, any outstanding readers may continue to be used. However, if you attempt to reopen any of those readers, you'll hit an
AlreadyClosedException
.- Returns:
- IndexReader that covers entire index plus all changes made so far by this IndexWriter instance
- Throws:
IOException
- If there is a low-level I/O error
-
finishGetReaderMerge
private StandardDirectoryReader finishGetReaderMerge(AtomicBoolean stopCollectingMergedReaders, Map<String, SegmentReader> mergedReaders, Map<String, throws IOExceptionSegmentReader> openedReadOnlyClones, SegmentInfos openingSegmentInfos, boolean applyAllDeletes, boolean writeAllDeletes, MergePolicy.MergeSpecification pointInTimeMerges, long maxCommitMergeWaitMillis) - Throws:
IOException
-
maybeReopenMergedNRTReader
private StandardDirectoryReader maybeReopenMergedNRTReader(Map<String, SegmentReader> mergedReaders, Map<String, throws IOExceptionSegmentReader> openedReadOnlyClones, SegmentInfos openingSegmentInfos, boolean applyAllDeletes, boolean writeAllDeletes) - Throws:
IOException
-
ramBytesUsed
public final long ramBytesUsed()Description copied from interface:Accountable
Return the memory usage of this object in bytes. Negative values are illegal.- Specified by:
ramBytesUsed
in interfaceAccountable
-
getFlushingBytes
public final long getFlushingBytes()Returns the number of bytes currently being flushed -
writeSomeDocValuesUpdates
- Throws:
IOException
-
numDeletedDocs
Obtain the number of deleted docs for a pooled reader. If the reader isn't being pooled, the segmentInfo's delCount is returned.- Specified by:
numDeletedDocs
in interfaceMergePolicy.MergeContext
-
ensureOpen
Used internally to throw anAlreadyClosedException
if this IndexWriter has been closed or is in the process of closing.- Parameters:
failIfClosing
- if true, also fail whenIndexWriter
is in the process of closing (closing=true
) but not yet done closing (closed=false
)- Throws:
AlreadyClosedException
- if this IndexWriter is closed or in the process of closing
-
ensureOpen
Used internally to throw anAlreadyClosedException
if this IndexWriter has been closed (closed=true
) or is in the process of closing (closing=true
).Calls
ensureOpen(true)
.- Throws:
AlreadyClosedException
- if this IndexWriter is closed
-
validateIndexSort
private void validateIndexSort()Confirms that the incoming index sort (if any) matches the existing index sort (if any). -
isCongruentSort
Returns true ifindexSort
is a prefix ofotherSort
. -
readFieldInfos
- Throws:
IOException
-
getFieldNumberMap
Loads or returns the already loaded the global field number map for thisSegmentInfos
. If thisSegmentInfos
has no global field number map the returned instance is empty- Throws:
IOException
-
getConfig
Returns aLiveIndexWriterConfig
, which can be used to query the IndexWriter current settings, as well as modify "live" ones. -
messageState
private void messageState() -
shutdown
Gracefully closes (commits, waits for merges), but calls rollback if there's an exc so the IndexWriter is always closed. This is called fromclose()
whenLiveIndexWriterConfig.commitOnClose
istrue
.- Throws:
IOException
-
close
Closes all open resources and releases the write lock.If
LiveIndexWriterConfig.commitOnClose
istrue
, this will attempt to gracefully shut down by writing any changes, waiting for any running merges, committing, and closing. In this case, note that:- If you called prepareCommit but failed to call commit, this method will throw
IllegalStateException
and theIndexWriter
will not be closed. - If this method throws any other exception, the
IndexWriter
will be closed, but changes may have been lost.
Note that this may be a costly operation, so, try to re-use a single writer instead of closing and opening a new one. See
commit()
for caveats about write caching done by some IO devices.NOTE: You must ensure no other threads are still making changes at the same time that this method is invoked.
- Specified by:
close
in interfaceAutoCloseable
- Specified by:
close
in interfaceCloseable
- Throws:
IOException
- If you called prepareCommit but failed to call commit, this method will throw
-
shouldClose
private boolean shouldClose(boolean waitForClose) -
getDirectory
Returns the Directory used by this index. -
getInfoStream
Description copied from interface:MergePolicy.MergeContext
Returns the info stream that can be used to log messages- Specified by:
getInfoStream
in interfaceMergePolicy.MergeContext
-
getAnalyzer
Returns the analyzer used by this index. -
advanceSegmentInfosVersion
public void advanceSegmentInfosVersion(long newVersion) IfSegmentInfos.getVersion()
is belownewVersion
then update it to this value. -
hasDeletions
public boolean hasDeletions()Returns true if this index has deletions (including buffered deletions). Note that this will return true if there are buffered Term/Query deletions, even if it turns out those buffered deletions don't match any documents. -
addDocument
Adds a document to this index.Note that if an Exception is hit (for example disk full) then the index will be consistent, but this document may not have been added. Furthermore, it's possible the index will have one segment in non-compound format even when using compound files (when a merge has partially succeeded).
This method periodically flushes pending documents to the Directory (see above), and also periodically triggers segment merges in the index according to the
MergePolicy
in use.Merges temporarily consume space in the directory. The amount of space required is up to 1X the size of all segments being merged, when no readers/searchers are open against the index, and up to 2X the size of all segments being merged when readers/searchers are open against the index (see
forceMerge(int)
for details). The sequence of primitive merge operations performed is governed by the merge policy.Note that each term in the document can be no longer than
MAX_TERM_LENGTH
in bytes, otherwise an IllegalArgumentException will be thrown.Note that it's possible to create an invalid Unicode string in java if a UTF16 surrogate pair is malformed. In this case, the invalid characters are silently replaced with the Unicode replacement character U+FFFD.
- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
addDocuments
public long addDocuments(Iterable<? extends Iterable<? extends IndexableField>> docs) throws IOException Atomically adds a block of documents with sequentially assigned document IDs, such that an external reader will see all or none of the documents.WARNING: the index does not currently record which documents were added as a block. Today this is fine, because merging will preserve a block. The order of documents within a segment will be preserved, even when child documents within a block are deleted. Most search features (like result grouping and block joining) require you to mark documents; when these documents are deleted these search features will not work as expected. Obviously adding documents to an existing block will require you the reindex the entire block.
However it's possible that in the future Lucene may merge more aggressively re-order documents (for example, perhaps to obtain better index compression), in which case you may need to fully re-index your documents at that time.
See
addDocument(Iterable)
for details on index and IndexWriter state after an Exception, and flushing/merging temporary free space requirements.NOTE: tools that do offline splitting of an index (for example, IndexSplitter in contrib) or re-sorting of documents (for example, IndexSorter in contrib) are not aware of these atomically added documents and will likely break them up. Use such tools at your own risk!
- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
updateDocuments
public long updateDocuments(Term delTerm, Iterable<? extends Iterable<? extends IndexableField>> docs) throws IOException Atomically deletes documents matching the provided delTerm and adds a block of documents with sequentially assigned document IDs, such that an external reader will see all or none of the documents.- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
updateDocuments
public long updateDocuments(Query delQuery, Iterable<? extends Iterable<? extends IndexableField>> docs) throws IOException Similar toupdateDocuments(Term, Iterable)
, but take a query instead of a term to identify the documents to be updated- Throws:
IOException
-
updateDocuments
private long updateDocuments(DocumentsWriterDeleteQueue.Node<?> delNode, Iterable<? extends Iterable<? extends IndexableField>> docs) throws IOException - Throws:
IOException
-
softUpdateDocuments
public long softUpdateDocuments(Term term, Iterable<? extends Iterable<? extends IndexableField>> docs, Field... softDeletes) throws IOException Expert: Atomically updates documents matching the provided term with the given doc-values fields and adds a block of documents with sequentially assigned document IDs, such that an external reader will see all or none of the documents.One use of this API is to retain older versions of documents instead of replacing them. The existing documents can be updated to reflect they are no longer current while atomically adding new documents at the same time.
In contrast to
updateDocuments(Term, Iterable)
this method will not delete documents in the index matching the given term but instead update them with the given doc-values fields which can be used as a soft-delete mechanism.See
addDocuments(Iterable)
andupdateDocuments(Term, Iterable)
.- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
tryDeleteDocument
Expert: attempts to delete by document ID, as long as the provided reader is a near-real-time reader (fromDirectoryReader.open(IndexWriter)
). If the provided reader is an NRT reader obtained from this writer, and its segment has not been merged away, then the delete succeeds and this method returns a valid (> 0) sequence number; else, it returns -1 and the caller must then separately delete by Term or Query.NOTE: this method can only delete documents visible to the currently open NRT reader. If you need to delete documents indexed after opening the NRT reader you must use
deleteDocuments(Term...)
).- Throws:
IOException
-
tryUpdateDocValue
Expert: attempts to update doc values by document ID, as long as the provided reader is a near-real-time reader (fromDirectoryReader.open(IndexWriter)
). If the provided reader is an NRT reader obtained from this writer, and its segment has not been merged away, then the update succeeds and this method returns a valid (> 0) sequence number; else, it returns -1 and the caller must then either retry the update and resolve the document again. If a doc values fields data isnull
the existing value is removed from all documents matching the term. This can be used to un-delete a soft-deleted document since this method will apply the field update even if the document is marked as deleted.NOTE: this method can only updates documents visible to the currently open NRT reader. If you need to update documents indexed after opening the NRT reader you must use
updateDocValues(Term, Field...)
.- Throws:
IOException
-
tryModifyDocument
private long tryModifyDocument(IndexReader readerIn, int docID, IndexWriter.DocModifier toApply) throws IOException - Throws:
IOException
-
dropDeletedSegment
Drops a segment that has 100% deleted documents.- Throws:
IOException
-
deleteDocuments
Deletes the document(s) containing any of the terms. All given deletes are applied and flushed atomically at the same time.- Parameters:
terms
- array of terms to identify the documents to be deleted- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
deleteDocuments
Deletes the document(s) matching any of the provided queries. All given deletes are applied and flushed atomically at the same time.- Parameters:
queries
- array of queries to identify the documents to be deleted- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
updateDocument
Updates a document by first deleting the document(s) containingterm
and then adding the new document. The delete and then add are atomic as seen by a reader on the same index (flush may happen only after the add).- Parameters:
term
- the term to identify the document(s) to be deleteddoc
- the document to be added- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
softUpdateDocument
public long softUpdateDocument(Term term, Iterable<? extends IndexableField> doc, Field... softDeletes) throws IOException Expert: Updates a document by first updating the document(s) containingterm
with the given doc-values fields and then adding the new document. The doc-values update and then add are atomic as seen by a reader on the same index (flush may happen only after the add).One use of this API is to retain older versions of documents instead of replacing them. The existing documents can be updated to reflect they are no longer current while atomically adding new documents at the same time.
In contrast to
updateDocument(Term, Iterable)
this method will not delete documents in the index matching the given term but instead update them with the given doc-values fields which can be used as a soft-delete mechanism.See
addDocuments(Iterable)
andupdateDocuments(Term, Iterable)
.- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
updateNumericDocValue
Updates a document'sNumericDocValues
forfield
to the givenvalue
. You can only update fields that already exist in the index, not add new fields through this method. You can only update fields that were indexed with doc values only.- Parameters:
term
- the term to identify the document(s) to be updatedfield
- field name of theNumericDocValues
fieldvalue
- new value for the field- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
updateBinaryDocValue
Updates a document'sBinaryDocValues
forfield
to the givenvalue
. You can only update fields that already exist in the index, not add new fields through this method. You can only update fields that were indexed only with doc values.NOTE: this method currently replaces the existing value of all affected documents with the new value.
- Parameters:
term
- the term to identify the document(s) to be updatedfield
- field name of theBinaryDocValues
fieldvalue
- new value for the field- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
updateDocValues
Updates documents' DocValues fields to the given values. Each field update is applied to the set of documents that are associated with theTerm
to the same value. All updates are atomically applied and flushed together. If a doc values fields data isnull
the existing value is removed from all documents matching the term.- Parameters:
updates
- the updates to apply- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error
-
buildDocValuesUpdate
-
getFieldNames
Return an unmodifiable set of all field names as visible from this IndexWriter, across all segments of the index. -
getSegmentCount
final int getSegmentCount() -
getNumBufferedDocuments
final int getNumBufferedDocuments() -
maxDoc
final int maxDoc(int i) -
getFlushCount
final int getFlushCount() -
getFlushDeletesCount
final int getFlushDeletesCount() -
newSegmentName
-
forceMerge
Forces merge policy to merge segments until there are<= maxNumSegments
. The actual merges to be executed are determined by theMergePolicy
.This is a horribly costly operation, especially when you pass a small
maxNumSegments
; usually you should only call this if the index is static (will no longer be changed).Note that this requires free space that is proportional to the size of the index in your Directory: 2X if you are not using compound file format, and 3X if you are. For example, if your index size is 10 MB then you need an additional 20 MB free for this to complete (30 MB if you're using compound file format). This is also affected by the
Codec
that is used to execute the merge, and may result in even a bigger index. Also, it's best to callcommit()
afterwards, to allow IndexWriter to free up disk space.If some but not all readers re-open while merging is underway, this will cause
> 2X
temporary space to be consumed as those new readers will then hold open the temporary segments at that time. It is best not to re-open readers while merging is running.The actual temporary usage could be much less than these figures (it depends on many factors).
In general, once this completes, the total size of the index will be less than the size of the starting index. It could be quite a bit smaller (if there were many pending deletes) or just slightly smaller.
If an Exception is hit, for example due to disk full, the index will not be corrupted and no documents will be lost. However, it may have been partially merged (some segments were merged but not all), and it's possible that one of the segments in the index will be in non-compound format even when using compound file format. This will occur when the Exception is hit during conversion of the segment into compound format.
This call will merge those segments present in the index when the call started. If other threads are still adding documents and flushing segments, those newly created segments will not be merged unless you call forceMerge again.
- Parameters:
maxNumSegments
- maximum number of segments left in the index after merging finishes- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO error- See Also:
-
forceMerge
Just likeforceMerge(int)
, except you can specify whether the call should block until all merging completes. This is only meaningful with aMergeScheduler
that is able to run merges in background threads.- Throws:
IOException
-
maxNumSegmentsMergesPending
private boolean maxNumSegmentsMergesPending()Returns true if any merges in pendingMerges or runningMerges are maxNumSegments merges. -
forceMergeDeletes
Just likeforceMergeDeletes()
, except you can specify whether the call should block until the operation completes. This is only meaningful with aMergeScheduler
that is able to run merges in background threads.- Throws:
IOException
-
forceMergeDeletes
Forces merging of all segments that have deleted documents. The actual merges to be executed are determined by theMergePolicy
. For example, the defaultTieredMergePolicy
will only pick a segment if the percentage of deleted docs is over 10%.This is often a horribly costly operation; rarely is it warranted.
To see how many deletions you have pending in your index, call
IndexReader.numDeletedDocs()
.NOTE: this method first flushes a new segment (if there are indexed documents), and applies all buffered deletes.
- Throws:
IOException
-
maybeMerge
Expert: asks the mergePolicy whether any merges are necessary now and if so, runs the requested merges and then iterate (test again if merges are needed) until no more merges are returned by the mergePolicy.Explicit calls to maybeMerge() are usually not necessary. The most common case is when merge policy parameters have changed.
This method will call the
MergePolicy
withMergeTrigger.EXPLICIT
.- Throws:
IOException
-
maybeMerge
private final void maybeMerge(MergePolicy mergePolicy, MergeTrigger trigger, int maxNumSegments) throws IOException - Throws:
IOException
-
executeMerge
- Throws:
IOException
-
updatePendingMerges
private MergePolicy.MergeSpecification updatePendingMerges(MergePolicy mergePolicy, MergeTrigger trigger, int maxNumSegments) throws IOException - Throws:
IOException
-
getMergingSegments
Expert: to be used by aMergePolicy
to avoid selecting merges for segments already being merged. The returned collection is not cloned, and thus is only safe to access if you hold IndexWriter's lock (which you do when IndexWriter invokes the MergePolicy).The Set is unmodifiable.
- Specified by:
getMergingSegments
in interfaceMergePolicy.MergeContext
-
getNextMerge
Expert: theMergeScheduler
calls this method to retrieve the next merge requested by the MergePolicy -
hasPendingMerges
public boolean hasPendingMerges()Expert: returns true if there are merges waiting to be scheduled. -
rollback
Close theIndexWriter
without committing any changes that have occurred since the last commit (or since it was opened, if commit hasn't been called). This removes any temporary files that had been created, after which the state of the index will be the same as it was when commit() was last called or when this writer was first opened. This also clears a previous call toprepareCommit()
.- Specified by:
rollback
in interfaceTwoPhaseCommit
- Throws:
IOException
- if there is a low-level IO error
-
rollbackInternal
- Throws:
IOException
-
rollbackInternalNoCommit
- Throws:
IOException
-
deleteAll
Delete all documents in the index.This method will drop all buffered documents and will remove all segments from the index. This change will not be visible until a
commit()
has been called. This method can be rolled back usingrollback()
.NOTE: this method is much faster than using deleteDocuments( new MatchAllDocsQuery() ). Yet, this method also has different semantics compared to
deleteDocuments(Query...)
since internal data-structures are cleared as well as all segment information is forcefully dropped anti-viral semantics like omitting norms are reset or doc value types are cleared. Essentially a call todeleteAll()
is equivalent to creating a newIndexWriter
withIndexWriterConfig.OpenMode.CREATE
which a delete query only marks documents as deleted.NOTE: this method will forcefully abort all merges in progress. If other threads are running
forceMerge(int)
,addIndexes(CodecReader[])
orforceMergeDeletes(boolean)
methods, they may receiveMergePolicy.MergeAbortedException
s.- Returns:
- The sequence number for this operation
- Throws:
IOException
-
abortMerges
Aborts running merges. Be careful when using this method: when you abort a long-running merge, you lose a lot of work that must later be redone.- Throws:
IOException
-
waitForMerges
Wait for any currently outstanding merges to finish.It is guaranteed that any merges started prior to calling this method will have completed once this method completes.
- Throws:
IOException
-
checkpoint
Called whenever the SegmentInfos has been updated and the index files referenced exist (correctly) in the index directory.- Throws:
IOException
-
checkpointNoSIS
Checkpoints with IndexFileDeleter, so it's aware of new files, and increments changeCount, so on close/commit we will write a new segments file, but does NOT bump segmentInfos.version.- Throws:
IOException
-
changed
private void changed()Called internally if any index state has changed. -
publishFrozenUpdates
-
publishFlushedSegment
private void publishFlushedSegment(SegmentCommitInfo newSegment, FieldInfos fieldInfos, FrozenBufferedUpdates packet, FrozenBufferedUpdates globalPacket, Sorter.DocMap sortMap) throws IOException Atomically adds the segment private delete packet and publishes the flushed segments SegmentInfo to the index writer.- Throws:
IOException
-
resetMergeExceptions
private void resetMergeExceptions() -
noDupDirs
-
acquireWriteLocks
Acquires write locks on all the directories; be sure to match with a call toIOUtils.close(java.io.Closeable...)
in a finally clause.- Throws:
IOException
-
addIndexes
Adds all segments from an array of indexes into this index.This may be used to parallelize batch indexing. A large document collection can be broken into sub-collections. Each sub-collection can be indexed in parallel, on a different thread, process or machine. The complete index can then be created by merging sub-collection indexes with this method.
NOTE: this method acquires the write lock in each directory, to ensure that no
IndexWriter
is currently open or tries to open while this is running.This method is transactional in how Exceptions are handled: it does not commit a new segments_N file until all indexes are added. This means if an Exception occurs (for example disk full), then either no indexes will have been added or they all will have been.
Note that this requires temporary free space in the
Directory
up to 2X the sum of all input indexes (including the starting index). If readers/searchers are open against the starting index, then temporary free space required will be higher by the size of the starting index (seeforceMerge(int)
for details).This requires this index not be among those to be added.
All added indexes must have been created by the same Lucene version as this index.
- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO errorIllegalArgumentException
- if addIndexes would cause the index to exceedMAX_DOCS
, or if the indoming index sort does not match this index's index sort
-
validateMergeReader
-
addIndexes
Merges the provided indexes into this index.The provided IndexReaders are not closed.
See
addIndexes(org.apache.lucene.store.Directory...)
for details on transactional semantics, temporary free space required in the Directory, and non-CFS segments on an Exception.NOTE: empty segments are dropped by this method and not added to this index.
NOTE: provided
LeafReader
s are merged as specified by theMergePolicy.findMerges(CodecReader...)
API. Default behavior is to merge all provided readers into a single segment. You can modify this by overriding thefindMerge
API in your custom merge policy.- Returns:
- The sequence number for this operation
- Throws:
CorruptIndexException
- if the index is corruptIOException
- if there is a low-level IO errorIllegalArgumentException
- if addIndexes would cause the index to exceedMAX_DOCS
-
addIndexesReaderMerge
Runs a single merge operation foraddIndexes(CodecReader...)
.Merges and creates a SegmentInfo, for the readers grouped together in provided OneMerge.
- Parameters:
merge
- OneMerge object initialized from readers.- Throws:
IOException
- if there is a low-level IO error
-
copySegmentAsIs
private SegmentCommitInfo copySegmentAsIs(SegmentCommitInfo info, String segName, IOContext context) throws IOException Copies the segment files as-is into the IndexWriter's directory.- Throws:
IOException
-
doAfterFlush
A hook for extending classes to execute operations after pending added and deleted documents have been flushed to the Directory but before the change is committed (new segments_N file written).- Throws:
IOException
-
doBeforeFlush
A hook for extending classes to execute operations before pending added and deleted documents are flushed to the Directory.- Throws:
IOException
-
prepareCommit
Expert: prepare for commit. This does the first phase of 2-phase commit. This method does all steps necessary to commit changes since this writer was opened: flushes pending added and deleted docs, syncs the index files, writes most of next segments_N file. After calling this you must call eithercommit()
to finish the commit, orrollback()
to revert the commit and undo all changes done since the writer was opened.You can also just call
commit()
directly without prepareCommit first in which case that method will internally call prepareCommit.- Specified by:
prepareCommit
in interfaceTwoPhaseCommit
- Returns:
- The sequence number of the last operation in the commit. All sequence numbers <= this value will be reflected in the commit, and all others will not.
- Throws:
IOException
-
flushNextBuffer
Expert: Flushes the next pending writer per thread buffer if available or the largest active non-pending writer per thread buffer in the calling thread. This can be used to flush documents to disk outside of an indexing thread. In contrast toflush()
this won't mark all currently active indexing buffers as flush-pending.Note: this method is best-effort and might not flush any segments to disk. If there is a full flush happening concurrently multiple segments might have been flushed. Users of this API can access the IndexWriters current memory consumption via
ramBytesUsed()
- Returns:
true
iff this method flushed at least on segment to disk.- Throws:
IOException
-
prepareCommitInternal
- Throws:
IOException
-
preparePointInTimeMerge
private MergePolicy.MergeSpecification preparePointInTimeMerge(SegmentInfos mergingSegmentInfos, BooleanSupplier stopCollectingMergeResults, MergeTrigger trigger, IOConsumer<SegmentCommitInfo> mergeFinished) throws IOException This optimization allows a commit/getReader to wait for merges on smallish segments to reduce the eventual number of tiny segments in the commit point / NRT Reader. We wrap aOneMerge
to update themergingSegmentInfos
once the merge has finished. We replace the source segments in the SIS that we are going to commit / open the reader on with the freshly merged segment, but ignore all deletions and updates that are made to documents in the merged segment while it was merging. The updates that are made do not belong to the point-in-time commit point / NRT READER and should therefore not be included. See the clone call inonMergeComplete
below. We also ensure that we pull the merge readers while holdingIndexWriter
's lock. Otherwise we could see concurrent deletions/updates applied that do not belong to the segment.- Throws:
IOException
-
writeReaderPool
Ensures that all changes in the reader-pool are written to disk.- Parameters:
writeDeletes
- iftrue
if deletes should be written to disk too.- Throws:
IOException
-
setLiveCommitData
Sets the iterator to provide the commit user data map at commit time. Calling this method is considered a committable change and will becommitted
even if there are no other changes this writer. Note that you must call this method beforeprepareCommit()
. Otherwise it won't be included in the follow-oncommit()
.NOTE: the iterator is late-binding: it is only visited once all documents for the commit have been written to their segments, before the next segments_N file is written
-
setLiveCommitData
public final void setLiveCommitData(Iterable<Map.Entry<String, String>> commitUserData, boolean doIncrementVersion) Sets the commit user data iterator, controlling whether to advance theSegmentInfos.getVersion()
.- See Also:
-
getLiveCommitData
Returns the commit user data iterable previously set withsetLiveCommitData(Iterable)
, or null if nothing has been set yet. -
commit
Commits all pending changes (added and deleted documents, segment merges, added indexes, etc.) to the index, and syncs all referenced index files, such that a reader will see the changes and the index updates will survive an OS or machine crash or power loss. Note that this does not wait for any running background merges to finish. This may be a costly operation, so you should test the cost in your application and do it only when really necessary.Note that this operation calls Directory.sync on the index files. That call should not return until the file contents and metadata are on stable storage. For FSDirectory, this calls the OS's fsync. But, beware: some hardware devices may in fact cache writes even during fsync, and return before the bits are actually on stable storage, to give the appearance of faster performance. If you have such a device, and it does not have a battery backup (for example) then on power loss it may still lose data. Lucene cannot guarantee consistency on such devices.
If nothing was committed, because there were no pending changes, this returns -1. Otherwise, it returns the sequence number such that all indexing operations prior to this sequence will be included in the commit point, and all other operations will not.
- Specified by:
commit
in interfaceTwoPhaseCommit
- Returns:
- The sequence number of the last operation in the commit. All sequence numbers <= this value will be reflected in the commit, and all others will not.
- Throws:
IOException
- See Also:
-
hasUncommittedChanges
public final boolean hasUncommittedChanges()Returns true if there may be changes that have not been committed. There are cases where this may return true when there are no actual "real" changes to the index, for example if you've deleted by Term or Query but that Term or Query does not match any documents. Also, if a merge kicked off as a result of flushing a new segment duringcommit()
, or a concurrent merged finished, this method may return true right after you had just calledcommit()
. -
hasChangesInRam
boolean hasChangesInRam()Returns true if there are any changes or deletes that are not flushed or applied. -
commitInternal
- Throws:
IOException
-
finishCommit
- Throws:
IOException
-
flush
Moves all in-memory segments to theDirectory
, but does not commit (fsync) them (callcommit()
for that).- Throws:
IOException
-
flush
Flush all in-memory buffered updates (adds and deletes) to the Directory.- Parameters:
triggerMerge
- if true, we may merge segments (if deletes or docs were flushed) if necessaryapplyAllDeletes
- whether pending deletes should also- Throws:
IOException
-
doFlush
Returns true a segment was flushed or deletes were applied.- Throws:
IOException
-
applyAllDeletesAndUpdates
- Throws:
IOException
-
getDocsWriter
DocumentsWriter getDocsWriter() -
numRamDocs
public final int numRamDocs()Expert: Return the number of documents currently buffered in RAM. -
ensureValidMerge
-
commitMergedDeletesAndUpdates
private ReadersAndUpdates commitMergedDeletesAndUpdates(MergePolicy.OneMerge merge, MergeState.DocMap[] docMaps) throws IOException Carefully merges deletes and updates for the segments we just merged. This is tricky because, although merging will clear all deletes (compacts the documents) and compact all the updates, new deletes and updates may have been flushed to the segments since the merge was started. This method "carries over" such new deletes and updates onto the newly merged segment, and saves the resulting deletes and updates files (incrementing the delete and DV generations for merge.info). If no deletes were flushed, no new deletes file is saved.- Throws:
IOException
-
carryOverHardDeletes
private static void carryOverHardDeletes(ReadersAndUpdates mergedReadersAndUpdates, int maxDoc, Bits prevHardLiveDocs, Bits currentHardLiveDocs, MergeState.DocMap segDocMap) throws IOException This method carries over hard-deleted documents that are applied to the source segment during a merge.- Throws:
IOException
-
commitMerge
private boolean commitMerge(MergePolicy.OneMerge merge, MergeState.DocMap[] docMaps) throws IOException - Throws:
IOException
-
handleMergeException
- Throws:
IOException
-
merge
Merges the indicated segments, replacing them in the stack with a single segment.- Throws:
IOException
-
mergeSuccess
Hook that's called when the specified merge is complete. -
abortOneMerge
- Throws:
IOException
-
registerMerge
Checks whether this merge involves any segments already participating in a merge. If not, this merge is "registered", meaning we record that its segments are now participating in a merge, and true is returned. Else (the merge conflicts) false is returned.- Throws:
IOException
-
mergeInit
Does initial setup for a merge, which is fast but holds the synchronized lock on IndexWriter instance.- Throws:
IOException
-
_mergeInit
- Throws:
IOException
-
setDiagnostics
-
setDiagnostics
-
mergeFinish
Does finishing for a merge, which is fast but holds the synchronized lock on IndexWriter instance. -
closeMergeReaders
private void closeMergeReaders(MergePolicy.OneMerge merge, boolean suppressExceptions, boolean droppedSegment) throws IOException - Throws:
IOException
-
countSoftDeletes
private void countSoftDeletes(CodecReader reader, Bits wrappedLiveDocs, Bits hardLiveDocs, Counter softDeleteCounter, Counter hardDeleteCounter) throws IOException - Throws:
IOException
-
assertSoftDeletesCount
- Throws:
IOException
-
mergeMiddle
Does the actual (time-consuming) work of the merge, but without holding synchronized lock on IndexWriter instance- Throws:
IOException
-
addMergeException
-
getBufferedDeleteTermsSize
final int getBufferedDeleteTermsSize() -
newestSegment
SegmentCommitInfo newestSegment() -
segString
String segString()Returns a string description of all segments, for debugging. -
segString
-
segString
Returns a string description of the specified segment, for debugging. -
doWait
private void doWait() -
filesExist
- Throws:
IOException
-
toLiveInfos
-
startCommit
Walk through all files referenced by the current segmentInfos and ask the Directory to sync each file, if it wasn't already. If that succeeds, then we prepare a new segments_N file but do not fully commit it.- Throws:
IOException
-
onTragicEvent
This method should be called on a tragic event ie. if a downstream class of the writer hits an unrecoverable exception. This method does not rethrow the tragic event exception.Note: This method will not close the writer but can be called from any location without respecting any lock order
-
tragicEvent
This method set the tragic exception unless it's already set and closes the writer if necessary. Note this method will not rethrow the throwable passed to it.- Throws:
IOException
-
maybeCloseOnTragicEvent
- Throws:
IOException
-
getTragicException
If thisIndexWriter
was closed as a side-effect of a tragic exception, e.g. disk full while flushing a new segment, this returns the root cause exception. Otherwise (no tragic exception has occurred) it returns null. -
isOpen
public boolean isOpen()Returnstrue
if thisIndexWriter
is still open. -
testPoint
-
nrtIsCurrent
-
isClosed
boolean isClosed() -
isDeleterClosed
boolean isDeleterClosed() -
deleteUnusedFiles
Expert: remove any index files that are no longer used.IndexWriter normally deletes unused files itself, during indexing. However, on Windows, which disallows deletion of open files, if there is a reader open on the index then those files cannot be deleted. This is fine, because IndexWriter will periodically retry the deletion.
However, IndexWriter doesn't try that often: only on open, close, flushing a new segment, and finishing a merge. If you don't do any of these actions with your IndexWriter, you'll see the unused files linger. If that's a problem, call this method to delete them (once you've closed the open readers that were preventing their deletion).
In addition, you can call this method to delete unreferenced index commits. This might be useful if you are using an
IndexDeletionPolicy
which holds onto index commits until some criteria are met, but those commits are no longer needed. Otherwise, those commits will be deleted the next time commit() is called.- Throws:
IOException
-
createCompoundFile
static void createCompoundFile(InfoStream infoStream, TrackingDirectoryWrapper directory, SegmentInfo info, IOContext context, IOConsumer<Collection<String>> deleteFiles) throws IOException NOTE: this method creates a compound file for all files returned by info.files(). While, generally, this may include separate norms and deletion files, this SegmentInfo must not reference such files when this method is called, because they are not allowed within a compound file.- Throws:
IOException
-
deleteNewFiles
Tries to delete the given files if unreferenced- Parameters:
files
- the files to delete- Throws:
IOException
- if anIOException
occurs- See Also:
-
flushFailed
Cleans up residuals from a segment that could not be entirely flushed due to an error- Throws:
IOException
-
publishFlushedSegments
Publishes the flushed segment, segment-private deletes (if any) and its associated global delete (if present) to IndexWriter. The actual publishing operation is synced onIW -> BDS
so that theSegmentInfo
's delete generation is always GlobalPacket_deleteGeneration + 1- Parameters:
forced
- iftrue
this call will block on the ticket queue if the lock is held by another thread. iffalse
the call will try to acquire the queue lock and exits if it's held by another thread.- Throws:
IOException
-
incRefDeleter
Record that the files referenced by thisSegmentInfos
are still in use.- Throws:
IOException
-
decRefDeleter
Record that the files referenced by thisSegmentInfos
are no longer in use. Only call this if you are sure you previously calledincRefDeleter(org.apache.lucene.index.SegmentInfos)
.- Throws:
IOException
-
maybeProcessEvents
Processes all events and might trigger a merge if the given seqNo is negative- Parameters:
seqNo
- if the seqNo is less than 0 this method will process events otherwise it's a no-op.- Returns:
- the given seqId inverted if negative.
- Throws:
IOException
-
processEvents
- Throws:
IOException
-
reserveDocs
private void reserveDocs(long addedNumDocs) Anything that will add N docs to the index should reserve first to make sure it's allowed. This will throwIllegalArgumentException
if it's not allowed. -
testReserveDocs
private void testReserveDocs(long addedNumDocs) Does a best-effort check, that the current index would accept this many additional docs, but does not actually reserve them.- Throws:
IllegalArgumentException
- if there would be too many docs
-
tooManyDocs
private void tooManyDocs(long addedNumDocs) -
getPendingNumDocs
public long getPendingNumDocs()Returns the number of documents in the index including documents are being added (i.e., reserved). -
getMaxCompletedSequenceNumber
public long getMaxCompletedSequenceNumber()Returns the highest sequence number across all completed operations, or 0 if no operations have finished yet. Still in-flight operations (in other threads) are not counted until they finish. -
adjustPendingNumDocs
private long adjustPendingNumDocs(long numDocs) -
isFullyDeleted
- Throws:
IOException
-
numDeletesToMerge
Returns the number of deletes a merge would claim back if the given segment is merged.- Specified by:
numDeletesToMerge
in interfaceMergePolicy.MergeContext
- Parameters:
info
- the segment to get the number of deletes for- Throws:
IOException
- See Also:
-
release
- Throws:
IOException
-
release
private void release(ReadersAndUpdates readersAndUpdates, boolean assertLiveInfo) throws IOException - Throws:
IOException
-
getPooledInstance
-
tryApply
Translates a frozen packet of delete term/query, or doc values updates, into their actual docIDs in the index, and applies the change. This is a heavy operation and is done concurrently by incoming indexing threads. This method will return immediately without blocking if another thread is currently applying the package. In order to ensure the packet has been applied,forceApply(FrozenBufferedUpdates)
must be called.- Throws:
IOException
-
forceApply
Translates a frozen packet of delete term/query, or doc values updates, into their actual docIDs in the index, and applies the change. This is a heavy operation and is done concurrently by incoming indexing threads.- Throws:
IOException
-
getInfosToApply
Returns theSegmentCommitInfo
that this packet is supposed to apply its deletes to, or null if the private segment was already merged away. -
finishApply
private void finishApply(BufferedUpdatesStream.SegmentState[] segStates, boolean success, Set<String> delFiles) throws IOException - Throws:
IOException
-
closeSegmentStates
private BufferedUpdatesStream.ApplyDeletesResult closeSegmentStates(BufferedUpdatesStream.SegmentState[] segStates, boolean success) throws IOException Close segment states previously opened with openSegmentStates.- Throws:
IOException
-
openSegmentStates
private BufferedUpdatesStream.SegmentState[] openSegmentStates(List<SegmentCommitInfo> infos, Set<SegmentCommitInfo> alreadySeenSegments, long delGen) throws IOException Opens SegmentReader and inits SegmentState for each segment.- Throws:
IOException
-
isEnableTestPoints
protected boolean isEnableTestPoints()Tests should override this to enable test points. Default isfalse
. -
validate
-
cloneSegmentInfos
Tests should use this method to snapshot the current segmentInfos to have a consistent view -
getDocStats
Returns accurateIndexWriter.DocStats
for this writer. The numDoc for instance can change after maxDoc is fetched that causes numDocs to be greater than maxDoc which makes it hard to get accurate document stats from IndexWriter.
-