NAME

SWISH::Prog::Index - handle Swish-e indexing

SYNOPSIS

 use SWISH::Prog::Index
 my $indexer = SWISH::Prog::Index->new(
                format      => 'native2' || 'btree2',
                config      => SWISH::Prog::Config->new,
                exe         => 'path/to/swish-e',
                verbose     => 0|1|2|3,
                warnings    => 0|1|2|3,
                name        => 'path/to/myindex'
                );
                
 $indexer->run;
 
 print "index files: $_" for $indexer->files;
 
 # or from the command line
 
 DirTree.pl some/path | perl -MSWISH::Prog::Index -e go
 

 

DESCRIPTION

SWISH::Prog::Index performs Swish-e indexing. For version 2.x of Swish-e this is simply a convenience wrapper around the swish-e binary executable.

FUNCTIONS

go

Magic method to run the indexer with all defaults. See SYNOPSIS for an example.

You might object that the perl line is harder to remember than 'swish-e -v0 -W0 -S prog -i stdin' and it takes 0.1 second longer to run.

True enough. But go() was easy to code in two lines, so I did. And it requires one less keystroke. Laziness is a virtue!

go() is the only export of this module.

METHODS

new

Create indexer object. All the following parameters are also accessor methods.

name

The name of the index. Should be a file path.

config

A SWISH::Prog::Config object.

exe

The path to the swish-e executable. If empty, will just look in $ENV{PATH}.

verbose

Takes same args as swish-e -v option.

warnings

Takes same args as swish-e -W option.

format

Tell the indexer what kind of index to expect. Options are native2 (default) or btree (for the experimental incremental feature in version 2.4).

The format param API is subject to change as Swish3 is developed.

fh

Get or set the open() filehandle for the swish-e process. CAUTION: don't set unless you know what you're doing.

You can print() to the filehandle using the SWISH::Prog index() method. Or do it directly like:

 print { $indexer->fh } "your headers and body here";
 

The filehandle is close()'d by the DESTROY magic method in this class. So when $indexer goes undef, the indexing process is closed automatically.

rm

Remove the index (all the associated index files). Useful if creating a temp index for merging, etc.

Returns 0 and carps if there was a problem unlink()ing any file. Returns 1 otherwise.

mv( new_name )

Rename the index. Useful if creating temp indexes, etc.

Returns 1 on success, 0 on failure.

files

Returns a list of all the associated files for the index.

run( [cmd] )

Start the indexer on its merry way. Stores the filehandle with the fh method for later access via SWISH::Prog or other methods TBD.

Returns the $indexer object.

You likely don't want to pass cmd in but let run() construct it for you.

merge( @list_of_indexes )

merge() will merge @list_of_indexes together with the index named in the calling object.

Returns the $indexer object on success, 0 on failure.

add( swish_prog_doc_object )

add() will merge swish_prog_doc_object with the index named in the calling object. If the existing index uses the btree2 format (incremental mode), that API will be used. Otherwise, swish_prog_doc_object is indexed as a temporary index and then merged.

Returns $indexer object on success, 0 on failure.

AUTHOR

Peter Karman, <perl@peknet.com>

COPYRIGHT AND LICENSE

Copyright 2006 by Peter Karman

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

cpanm SWISH::Prog

perl -MCPAN -e shell
install SWISH::Prog