Hi, On 7/3/06, Marcel Reutegger <[EMAIL PROTECTED]> wrote:
I agree with all your comments and I would keep the getContentTypes() method in the TextExtractor interface. I think that way it requires less configuration because a TextExtractor knows the mime types it can handle anyway.
Good point, agreed.
How should we handle backward compatibility? Do you think we should keep the existing TextFilter interface?
Keeping a deprecated TextFilter until Jackrabbit 1.2 or 2.0 is probably the best solution. Internally we probably need to continue using the TextFilter interface, as it is possible to adapt TextExtractor to TextFilter but not the other way around. BR, Jukka Zitting
