Namespace Lucene.Net.Benchmarks.ByTask.Tasks

Extendable benchmark tasks.

Classes

AddDocTask

Add a document, optionally of a certain size.

Other side effects: none.

Takes optional param: document size.

Adds an input index to an existing index, using AddIndexes(Directory[]) or AddIndexes(IndexReader[]). The location of the input index is specified by the parameter ADDINDEXES_INPUT_DIR and is assumed to be a directory on the file system.

Takes optional parameter Lucene.Net.Benchmarks.ByTask.Tasks.AddIndexesTask.useAddIndexesDir which specifies which AddIndexes variant to use (defaults to true, to use AddIndexes(Directory)).

AnalyzerFactoryTask

Analyzer factory construction task. The name given to the constructed factory may be given to NewAnalyzerTask, which will call Create().

BenchmarkHighlighter

Abstract class for benchmarking highlighting performance

ClearStatsTask

Clear statistics data.

Other side effects: None.

CloseIndexTask

Close index writer.

Other side effects: index writer object in perfRunData is nullified.

Takes optional param "doWait": if false, then close(false) is called.

CloseReaderTask

Close index reader.

Other side effects: index reader in perfRunData is nullified.

This would cause read related tasks to reopen their own reader.

CloseTaxonomyIndexTask

Close taxonomy index.

Other side effects: taxonomy writer object in perfRunData is nullified.

CloseTaxonomyReaderTask

Close taxonomy reader.

Other side effects: taxonomy reader in perfRunData is nullified.

CommitIndexTask

Commits the IndexWriter.

CommitTaxonomyIndexTask

Commits the Taxonomy Index.

ConsumeContentSourceTask

Consumes a ContentSource.

CreateIndexTask

Creates an index.

CreateTaxonomyIndexTask

Create a taxonomy index.

Other side effects: taxonomy writer object in perfRunData is set.

ForceMergeTask

Runs forceMerge on the index.

Other side effects: none.

ImplementationExtensions

LUCENENET specific extension methods for the NewCollationAnalyzerTask.Implementation enumeration.

NearRealtimeReaderTask

Spawns a BG thread that periodically (defaults to 3.0 seconds, but accepts param in seconds) wakes up and asks IndexWriter for a near real-time reader. Then runs a single query (body: 1) sorted by docdate, and prints time to reopen and time to run the search.

@lucene.experimental It's also not generally usable, eg you cannot change which query is executed.

NewAnalyzerTask

Create a new Analyzer and set it it in the getRunData() for use by all future tasks.

NewCollationAnalyzerTask

NewLocaleTask

Set a for use in benchmarking.

NewRoundTask

Increment the counter for properties maintained by Round Number.

Other side effects: if there are props by round number, log value change.

OpenIndexTask

Open an index writer.

OpenReaderTask

Open an index reader.

Other side effects: index reader object in perfRunData is set.

Optional params commitUserData eg. OpenReader(false,commit1)

OpenTaxonomyIndexTask

Open a taxonomy index.

Other side effects: taxonomy writer object in perfRunData is set.

OpenTaxonomyReaderTask

Open a taxonomy index reader.

Other side effects: taxonomy reader object in perfRunData is set.

PerfTask

An abstract task to be tested for performance.

PrintReaderTask

Opens a reader and prints basic statistics.

ReadTask

Read index (abstract) task. Sub classes implement WithSearch, WithWarm, WithTraverse and WithRetrieve

ReadTokensTask

Simple task to test performance of tokenizers. It just creates a token stream for each field of the document and read all tokens out of that stream.

ReopenReaderTask

Reopens IndexReader and closes old IndexReader.

RepAllTask

Report all statistics with no aggregations.

Other side effects: None.

ReportTask

Report (abstract) task - all report tasks extend this task.

RepSelectByPrefTask

Report by-name-prefix statistics with no aggregations.

Other side effects: None.

RepSumByNameRoundTask

Report all statistics grouped/aggregated by name and round.

Other side effects: None.

RepSumByNameTask

Report all statistics aggregated by name.

Other side effects: None.

RepSumByPrefRoundTask

Report all prefix matching statistics grouped/aggregated by name and round.

Other side effects: None.

RepSumByPrefTask

Report by-name-prefix statistics aggregated by name.

Other side effects: None.

ResetInputsTask

Reset inputs so that the test run would behave, input wise, as if it just started. This affects e.g. the generation of docs and queries.

ResetSystemEraseTask

Reset all index and input data and call gc, erase index and dir, does NOT clear statistics.

This contains ResetInputs.

Other side effects: writers/readers nullified, deleted, closed. Index is erased. Directory is erased.

ResetSystemSoftTask

Reset all index and input data and call gc, does NOT erase index/dir, does NOT clear statistics. This contains ResetInputs.

Other side effects: writers/readers nullified, closed. Index is NOT erased. Directory is NOT erased.

RollbackIndexTask

Rollback the index writer.

SearchTask

Search task.

Note: This task reuses the reader if it is already open. Otherwise a reader is opened at start and closed at the end.

SearchTravRetHighlightTask

Search and Traverse and Retrieve docs task. Highlight the fields in the retrieved documents.

SearchTravRetLoadFieldSelectorTask

Search and Traverse and Retrieve docs task using a FieldVisitor loading only the requested fields.

SearchTravRetTask

Search and Traverse and Retrieve docs task.

SearchTravRetVectorHighlightTask

Search and Traverse and Retrieve docs task. Highlight the fields in the retrieved documents by using FastVectorHighlighter.

SearchTravTask

Search and Traverse task.

SearchWithCollectorTask

Does search w/ a custom collector

SearchWithSortTask

Does sort search on specified field.

SetPropTask

Set a performance test configuration property. A property may have a single value, or a sequence of values, separated by ":". If a sequence of values is specified, each time a new round starts, the next (cyclic) value is taken.

Other side effects: none.

Takes mandatory param: "name,value" pair.

TaskSequence

Sequence of parallel or sequential tasks.

UpdateDocTask

Update a document, using , optionally with of a certain size.

Other side effects: none.

Takes optional param: document size.

WaitForMergesTask

Waits for merges to finish.

WaitTask

Simply waits for the specified (via the parameter) amount of time. For example Wait(30s) waits for 30 seconds. This is useful with background tasks to control how long the tasks run.

You can specify h, m, or s (hours, minutes, seconds) as the trailing time unit. No unit is interpreted as seconds.

WarmTask

Warm reader task: retrieve all reader documents.

WriteEnwikiLineDocTask

A WriteLineDocTask which for Wikipedia input, will write category pages to another file, while remaining pages will be written to the original file. The categories file is derived from the original file, by adding a prefix "categories-".

WriteLineDocTask

A task which writes documents, one line per document. Each line is in the following format: title <TAB> date <TAB> body. The output of this task can be consumed by LineDocSource and is intended to save the IO overhead of opening a file per document to be indexed.

Enums

NewCollationAnalyzerTask.Implementation

Different Collation implementations: currently limited to what is provided in ICU.

See Comparison of implementations