Right now the full text search uses an index field, where all texts are concatenated. To keep this field small we only use fields with less than 1000 characters,which is obviously a problem for texts like blog posts, articles and so on.
The easy solution would be to use a wildcard text index
I want to bump this enhancement request, as it’s one of a couple key features my client is looking for.
Yes, I know. I have to check the wildcard solution, because it would also scan stuff like GUIDs and the index could grow very fast.
I am still looking for a solution. It is not an easy task, because texts could have different language and meaning, e.g. it could be an base64 blob file. I am not sure if there is a generic solution for this problem.
It will be solved in 2.0 with better support for language dependent stop words 