The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

NAME

Lucy::Index::DataWriter - Write data to an index.

SYNOPSIS

    # Abstract base class.

DESCRIPTION

DataWriter is an abstract base class for writing index data, generally in segment-sized chunks. Each component of an index -- e.g. stored fields, lexicon, postings, deletions -- is represented by a DataWriter/DataReader pair.

Components may be specified per index by subclassing Architecture.

CONSTRUCTORS

new( [labeled params] )

    my $writer = MyDataWriter->new(
        snapshot   => $snapshot,      # required
        segment    => $segment,       # required
        polyreader => $polyreader,    # required
    );
  • snapshot - The Snapshot that will be committed at the end of the indexing session.

  • segment - The Segment in progress.

  • polyreader - A PolyReader representing all existing data in the index. (If the index is brand new, the PolyReader will have no sub-readers).

ABSTRACT METHODS

add_inverted_doc( [labeled params] )

Process a document, previously inverted by inverter.

  • inverter - An Inverter wrapping an inverted document.

  • doc_id - Internal number assigned to this document within the segment.

add_segment( [labeled params] )

Add content from an existing segment into the one currently being written.

  • reader - The SegReader containing content to add.

  • doc_map - An array of integers mapping old document ids to new. Deleted documents are mapped to 0, indicating that they should be skipped.

finish()

Complete the segment: close all streams, store metadata, etc.

format()

Every writer must specify a file format revision number, which should increment each time the format changes. Responsibility for revision checking is left to the companion DataReader.

METHODS

delete_segment(reader)

Remove a segment's data. The default implementation is a no-op, as all files within the segment directory will be automatically deleted. Subclasses which manage their own files outside of the segment system should override this method and use it as a trigger for cleaning up obsolete data.

  • reader - The SegReader containing content to merge, which must represent a segment which is part of the the current snapshot.

merge_segment( [labeled params] )

Move content from an existing segment into the one currently being written.

The default implementation calls add_segment() then delete_segment().

  • reader - The SegReader containing content to merge, which must represent a segment which is part of the the current snapshot.

  • doc_map - An array of integers mapping old document ids to new. Deleted documents are mapped to 0, indicating that they should be skipped.

metadata()

Arbitrary metadata to be serialized and stored by the Segment. The default implementation supplies a Hash with a single key-value pair for "format".

get_snapshot()

Accessor for "snapshot" member var.

get_segment()

Accessor for "segment" member var.

get_polyreader()

Accessor for "polyreader" member var.

get_schema()

Accessor for "schema" member var.

get_folder()

Accessor for "folder" member var.

INHERITANCE

Lucy::Index::DataWriter isa Lucy::Object::Obj.