Compact a database, or merge and compact several. More...

#include <compactor.h>

Inheritance diagram for Xapian::Compactor:

Collaboration diagram for Xapian::Compactor:

Classes
class	Internal

Public Types
enum	compaction_level { STANDARD = 0, FULL = 1, FULLER = 2 }
	Compaction level. More...

Public Member Functions
	Compactor ()

virtual	~Compactor ()

void	set_block_size (size_t block_size)
	Set the block size to use for tables in the output database. More...

void	set_renumber (bool renumber)
	Set whether to preserve existing document id values. More...

void	set_multipass (bool multipass)
	Set whether to merge postlists in multiple passes. More...

void	set_compaction_level (compaction_level compaction)
	Set the compaction level. More...

void	set_destdir (const std::string &destdir)
	Set where to write the output. More...

void	add_source (const std::string &srcdir)
	Add a source database. More...

void	compact ()
	Perform the actual compaction/merging operation. More...

virtual void	set_status (const std::string &table, const std::string &status)
	Update progress. More...

virtual std::string	resolve_duplicate_metadata (const std::string &key, size_t num_tags, const std::string tags[])
	Resolve multiple user metadata entries with the same key. More...

Private Member Functions
void	set_flags_ (unsigned flags, unsigned mask=0)

Private Attributes
Xapian::Internal::intrusive_ptr< Internal >	internal

Detailed Description

Compact a database, or merge and compact several.

Definition at line 42 of file compactor.h.

Member Enumeration Documentation

◆ compaction_level

enum Xapian::Compactor::compaction_level

Compaction level.

Enumerator
STANDARD	Don't split items unnecessarily.
FULL	Split items whenever it saves space (the default).
FULLER	Allow oversize items to save more space (not recommended if you ever plan to update the compacted database).

Definition at line 48 of file compactor.h.

Constructor & Destructor Documentation

◆ Compactor()

Xapian::Compactor::Compactor ( )

Definition at line 95 of file compactor.cc.

◆ ~Compactor()

Xapian::Compactor::~Compactor ( )

virtual

Definition at line 97 of file compactor.cc.

Member Function Documentation

◆ add_source()

void Xapian::Compactor::add_source ( const std::string & srcdir )

Add a source database.

Deprecated:: Use Database::compact(destdir[, compactor]) instead.

Parameters

srcdir The path to the source database to add.

Definition at line 118 of file compactor.cc.

Referenced by DEFINE_TESTCASE().

◆ compact()

void Xapian::Compactor::compact ( )

Perform the actual compaction/merging operation.

Deprecated:: Use Database::compact(destdir[, compactor]) instead.

Definition at line 124 of file compactor.cc.

References Xapian::Database::add_database(), Xapian::Database::compact(), internal, and srcdir.

Referenced by DEFINE_TESTCASE().

◆ resolve_duplicate_metadata()

string Xapian::Compactor::resolve_duplicate_metadata	(	const std::string &	key,
		size_t	num_tags,
		const std::string	tags[]
	)

virtual

Resolve multiple user metadata entries with the same key.

When merging, if the same user metadata key is set in more than one input, then this method is called to allow this to be resolving in an appropriate way.

The default implementation just returns tags[0].

For multipass this will currently get called multiple times for the same key if there are duplicates to resolve in each pass, but this may change in the future.

Since 1.4.6, an implementation of this method can return an empty string to indicate that the appropriate result is to not set a value for this user metadata key in the output database. In older versions, you should not return an empty string.

Parameters

key	The metadata key with duplicate entries.
num_tags	How many tags there are.
tags	An array of num_tags strings containing the tags to merge.

Reimplemented in MyCompactor.

Definition at line 142 of file compactor.cc.

References backend_mismatch().

Referenced by GlassCompact::merge_postlists(), and ChertCompact::merge_postlists().

◆ set_block_size()

void Xapian::Compactor::set_block_size ( size_t block_size )

Set the block size to use for tables in the output database.

Parameters

block_size The block size to use. Valid block sizes are currently powers of two between 2048 and 65536, with the default being 8192, but the valid sizes and default may change in the future.

Definition at line 100 of file compactor.cc.

◆ set_compaction_level()

void Xapian::Compactor::set_compaction_level ( compaction_level compaction )

inline

Set the compaction level.

Parameters

compaction

Available values are:

Xapian::Compactor::STANDARD - Don't split items unnecessarily.
Xapian::Compactor::FULL - Split items whenever it saves space (the default).
Xapian::Compactor::FULLER - Allow oversize items to save more space (not recommended if you ever plan to update the compacted database).

Definition at line 116 of file compactor.h.

References srcdir, tags, and XAPIAN_DEPRECATED.

◆ set_destdir()

void Xapian::Compactor::set_destdir ( const std::string & destdir )

Set where to write the output.

Deprecated:: Use Database::compact(destdir[, compactor]) instead.

Parameters

destdir Output path. This can be the same as an input if that input is a stub database (in which case the database(s) listed in the stub will be compacted to a new database and then the stub will be atomically updated to point to this new database).

Definition at line 112 of file compactor.cc.

Referenced by DEFINE_TESTCASE().

◆ set_flags_()

void Xapian::Compactor::set_flags_	(	unsigned	flags,
		unsigned	mask = `0`
	)

private

Definition at line 106 of file compactor.cc.

◆ set_multipass()

void Xapian::Compactor::set_multipass ( bool multipass )

inline

Set whether to merge postlists in multiple passes.

Parameters

multipass If true and merging more than 3 databases, merge the postlists in multiple passes, which is generally faster but requires more disk space for temporary files. By default we don't do this.

Definition at line 102 of file compactor.h.

References Xapian::DBCOMPACT_MULTIPASS.

Referenced by DEFINE_TESTCASE().

◆ set_renumber()

void Xapian::Compactor::set_renumber ( bool renumber )

inline

Set whether to preserve existing document id values.

Parameters

renumber The default is true, which means that document ids will be renumbered - currently by applying the same offset to all the document ids in a particular source database.

If false, then the document ids must be unique over all source databases. Currently the ranges of document ids in each source must not overlap either, though this restriction may be removed in the future.

Definition at line 90 of file compactor.h.

References Xapian::DBCOMPACT_NO_RENUMBER.

Referenced by DEFINE_TESTCASE().

◆ set_status()

void Xapian::Compactor::set_status	(	const std::string &	table,
		const std::string &	status
	)

virtual

Update progress.

Subclass this method if you want to get progress updates during compaction. This is called for each table first with empty status, And then one or more times with non-empty status.

The default implementation does nothing.

Parameters

table	The table currently being compacted.
status	A status message.

Reimplemented in MyCompactor.

Definition at line 135 of file compactor.cc.

Referenced by GlassDatabase::compact(), and ChertDatabase::compact().

Member Data Documentation

◆ internal

Xapian::Internal::intrusive_ptr<Internal> Xapian::Compactor::internal

private

Reference counted internals.

Definition at line 60 of file compactor.h.

Referenced by compact().

The documentation for this class was generated from the following files:

include/xapian/compactor.h
api/compactor.cc

Classes

Public Types

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Member Enumeration Documentation

◆ compaction_level

Constructor & Destructor Documentation

◆ Compactor()

◆ ~Compactor()

Member Function Documentation

◆ add_source()

◆ compact()

◆ resolve_duplicate_metadata()

◆ set_block_size()

◆ set_compaction_level()

◆ set_destdir()

◆ set_flags_()

◆ set_multipass()

◆ set_renumber()

◆ set_status()

Member Data Documentation

◆ internal