com.bigdata.sparse

Interface SparseRowStore.Options

Enclosing class:

SparseRowStore

public static interface SparseRowStore.Options

Options for the SparseRowStore.

Author:

Bryan Thompson

Field Summary

Fields

Modifier and Type	Field and Description
static String	DEFAULT_PRIMARY_KEY_UNICODE_CLEAN
static String	DEFAULT_SCHEMA_NAME_UNICODE_CLEAN
static String	PRIMARY_KEY_UNICODE_CLEAN The primary key was originally written using a Unicode sort key.
static String	SCHEMA_NAME_UNICODE_CLEAN The schema name was originally written using a Unicode sort key.

Field Detail

SCHEMA_NAME_UNICODE_CLEAN

static final String SCHEMA_NAME_UNICODE_CLEAN

The schema name was originally written using a Unicode sort key. However, the JDK can generate Unicode sort keys with embedded nuls which in turn will break the logic to detect the end of the schema name in the key. In order to accommodate this behavior, the schema name is now encoded as UTF8 which also has the advantage that we can decode the schema name. Standard prefix compression on the B+Tree should make up for the larger representation of the schema name in the B+Tree.

This change was introduced on 7/29/2010 in the trunk. When this property is true it breaks compatibility with earlier revisions of the SparseRowStore. This flag may be set to false for backward compatibility.

See Also:

DEFAULT_SCHEMA_NAME_UNICODE_CLEAN

DEFAULT_SCHEMA_NAME_UNICODE_CLEAN

static final String DEFAULT_SCHEMA_NAME_UNICODE_CLEAN

See Also:

https://sourceforge.net/apps/trac/bigdata/ticket/107, Constant Field Values

PRIMARY_KEY_UNICODE_CLEAN

static final String PRIMARY_KEY_UNICODE_CLEAN

The primary key was originally written using a Unicode sort key. However, the JDK generates Unicode sort keys with embedded nuls and that broke the logic to detect the end of the Unicode primary keys. In order to accommodate this behavior, the Unicode primary key is now encoded as UTF8 which also has the advantage that we can decode Unicode primary keys. Standard prefix compression on the B+Tree should make up for the larger representation of the Unicode primary key in the B+Tree.

This change was introduced on 7/15/2010 in the trunk and breaks compatibility with earlier revisions of the SparseRowStore. This flag may be set to false for backward compatibility.

See Also:

DEFAULT_PRIMARY_KEY_UNICODE_CLEAN

DEFAULT_PRIMARY_KEY_UNICODE_CLEAN

static final String DEFAULT_PRIMARY_KEY_UNICODE_CLEAN

See Also:

https://sourceforge.net/apps/trac/bigdata/ticket/107, Constant Field Values