Tokenizing words joined by symbols

[jbriales]

I found that when I have a field containing combinations of words like "foo/bar" the tokenizer does not seem to treat this as two separate words ("foo" and "bar"), and instead the search query would find this item only if I searched for "foo/bar", but NOT if I search for "foo" or "bar" alone.
Is this the expected behavior or a known issue?

[jbriales]

I fixed this by setting lunr.tokenizer.seperator = /[\W_]+/ in the pouchdb-quick-search/lib/index.js file in my own fork.