eZSphinx full-text search engine plugin

It's Sphinx powered search engine for eZ Publish. Currently it supports these searching features.

For integer type attributes

• >=,> operators (they both acts the same)
• <=,< operators (they both acts the same)
• = operator
• Between operator
• IN operator
• Include, Exclude ranges on integer values.

General filtering options

• ClassID or array of classID's
• SectionID or array of SectionID's
• IgnoreVisibility
• PublishDate, identical to default eZSearchEgnine behaviour
• PublishTimestamp, identical to default eZSearchEgnine behaviour
• Searching only in custom attributes
• Multiple sorting
• Subtree fetching
• Parent node fetching

Sphinx related features

• Custom index usage
• Custom field weight. Set title higher priority than content. In other words - relevance ranking.
• Match modes

General features

• Implemented custom fetch function
• Delay auto index update through cronjob
• Live index remove
• Multilanguage support

Unsupported features and some notices.

• At this moment no permissions checking is done.You can just set section identifier and class identifier. This will limit search result to section and class.
• For maximum performance fetching by subtree you can limit subtree directly in sphinx source queries. You can make as many as you want sphinx indexes and use them in fetch function providing index_name variable.

Usage recommendation

For permission checking I prefer implement it in sphinx source queries or just limit fetching by section or classID.

General fetch function attributes

fetch( ezsphinx, search,
hash( query, text,
[ offset, offset, ]
[ limit, limit, ]
[ section_id, id, ]
[ publish_timestamp, time, ]
[ publish_date, date, ]
[ class_id, mixed, ]
[ class_attribute_id, mixed, ]
[ sort_by, sort_by, ]
[ field_weight, array, ]
[ parent_node_id, mixed, ]
[ filter, mixed, ]
[ as_objects, bool, ]
[ index_name, string, ]
[ match_type, string, ]
[ subtree_array, mixed, ]
[ ignore_visibility, ignore_visibility, ] ) )