Work around being unable to query bookmarks API with invalid URI schemes

[wincohi]

Continuing from #16.

Notes are associated by ID, so using bookmarks.search() is the simplest way to get the ID of a note from a file that doesn't include IDs. However, this throws an exception whenever the url search parameter is deemed "invalid," and this may prevent some users from recovering certain notes on entries that link to these "invalid" schemes.

A possible workaround might search within the already-printed bookmarks tree the extension generates on startup, since I can do this my own way without being limited by the bookmarks API.