Sigmah has a lot of database text content that can be indexed and searched by keywords, providing a useful capability to users to fetch the actual entities (orgunits, projects) that correspond to such keywords.
Users also upload files (Word, PDF, others) that have a lot of text. Here the search results may point to both the files themselves, and also to the entities owning those files where the search term was found.
Entity access (view) permissions have to be enforced in both cases. Another challenge is to integrate both database search and file search under the same UI, provide users with helpful search features (similar terms, relevance rankings).