* changed Document.addField(): instead of defining field options as booleans (still supported for backwards compatibility), they can now contain a string defining the storage, indexing and term vector options for this field. this way all options supported by lucene can be used. see apidoc for accepted values. in addition addField() now accepts an instance of helma.Search.Document.Field as single argument too.
* added Document.Field() constructor wrapping a lucene document field instance. The constructor can be called with the same options as Document.addField()
* Document.getField() now returns an instance of helma.Search.Document.Field instead of a plain JS object. For backwards compatibility name and value can still be accessed as properties, but to get detailed information about the field the appropriate methods needs to be called (eg. isTokenized(), isStored())
* removed Index.getModifier() as all but one method of IndexModifier is now available in IndexWriter too
* removed private getWriterModifier() method as waiting for index unlock is now done by Lucene itself (defaults to 1 second)
* getWriter() now supports a second argument enabling or disabling auto commit (defaults to false)
* changed default boolean clause in BooleanQuery.addTerm() and BooleanQuery.addQuery() from "and" (MUST) to SHOULD
* removed Document.getDateField(): the value of a field converted into a date is now accessible via the "dateValue" property of the field
* added Document.removeField(name)
it is now possible..
* to specify a maximum response size
* to define a custom response handler (e.g. for directly writing the response to the disk)
* to specify credentials directly in the URL
additionallz the returned result-object now contains a map with all header-fields.
Although the global variable "groups" still contains the manager object to handle all group objects, the inner structure was streamlined to conform with Helma's library structure (helma namespace) and to generally prevent namespace cluttering.
* Move static methods down to the bottom of the file to circumvent
a strange bug that prevents documentation from being generated.
* Some JSDoc comment fixes
* Use and support helma.File as Mime part input source
* Remove code examples to make jsdoc actually render the comments
* Implement get/setMimepartType() methods
