Message Title

During some testing I found that if you use an AnalyzerDiscriminator for class level analyzer selection, you can't seem to override this at field level by specifying a specific @Analyzer.

In our case, we use an AnalyzerDiscriminator that selects a language based stemming Analyzer but for one specific field, we are trying to use whitespace analyzer.

@AnalyzerDef(name = "autocompleteAnalyzer", tokenizer = @TokenizerDef(factory = WhitespaceTokenizerFactory.class), filters = {
		@TokenFilterDef(factory = LowerCaseFilterFactory.class),
		@TokenFilterDef(factory = StopFilterFactory.class, params = {
	      @Parameter(name="words", value= "nl/msw/compraventa/dao/hibernate/search/stopwords/stopwords-combined.txt" ),
	      @Parameter(name="ignoreCase", value="true")
	    }),
		@TokenFilterDef(factory = ASCIIFoldingFilterFactory.class),
		@TokenFilterDef(factory = EdgeNGramFilterFactory.class, params = {
			@org.hibernate.search.annotations.Parameter(name = "minGramSize", value = "2"),
			@org.hibernate.search.annotations.Parameter(name = "maxGramSize", value = "20") 
		})
})

@Field(name="indexKeywords",index=Index.YES,analyze=Analyze.YES,analyzer=@Analyzer(definition="autocompleteAnalyzer"),store=Store.NO)