Class | Ferret::Index::FieldInfos |
In: |
lib/ferret/field_infos.rb
ext/r_index.c |
Parent: | Object |
The FieldInfos class holds all the field descriptors for an index. It is this class that is used to create a new index using the FieldInfos#create_index method. If you are happy with the default properties for FieldInfo then you don‘t need to worry about this class. IndexWriter can create the index for you. Otherwise you should set up the index like in the example;
field_infos = FieldInfos.new(:term_vector => :no) field_infos.add_field(:title, :index => :untokenized, :term_vector => :no, :boost => 10.0) field_infos.add_field(:content) field_infos.add_field(:created_on, :index => :untokenized_omit_norms, :term_vector => :no) field_infos.add_field(:image, :store => :compressed, :index => :no, :term_vector => :no) field_infos.create_index("/path/to/index")
See FieldInfo for the available field property values.
When you create the FieldInfos object you specify the default properties for the fields. Often you‘ll specify all of the fields in the index before you create the index so the default values won‘t come into play. However, it is possible to continue to dynamically add fields as indexing goes along. If you add a document to the index which has fields that the index doesn‘t know about then the default properties are used for the new field.
Load FieldInfos from a YAML file. The YAML file should look something like this: default:
store: :yes index: :yes term_vector: :no
id: index: :untokenized term_vector: :no title: boost: 20.0 term_vector: :no content: term_vector: :with_positions_offsets
Create a new FieldInfos object which uses the default values for fields specified in the default hash parameter. See FieldInfo for available property values.
Add a FieldInfo object. Use the FieldInfos#add_field method where possible.
Add a FieldInfo object. Use the FieldInfos#add_field method where possible.
Add a new field to the FieldInfos object. See FieldInfo for a description of the available properties.
Create a new index in the directory specified. The directory dir can either be a string path representing a directory on the file-system or an actual directory object. Care should be taken when using this method. Any existing index (or other files for that matter) will be deleted from the directory and overwritten by the new index.
Return a list of the field names (as symbols) of all the tokenized fields in the index.