third round of indexing all files

[David Bremner]

It seems noticeably faster (on the order of 30-50% faster) and the
code is quite a bit simpler to adapt the approach in [1] to only
delete the terms we are going to re-add via indexing.

This obsoletes the previous series at [2]. It still has all of the
issues mentioned there UI-wise, and the question of the index options
design probably needs more thought.

This is new in this round

     [rfc patch v3 2/6] lib: add _notmuch_message_remove_indexed_terms

This is has been pretty drastically rewritten compared to daniel's version [3]

     [rfc patch v3 3/6] added notmuch_message_reindex

This is the same, except I added simple performance tests

     [rfc patch v3 4/6] add "notmuch reindex" subcommand


[1]: id:1471178598-9639-1-git-send-email-david at tethera.net
[2]: id:20170402131646.29884-1-david at tethera.net
[3]: id:20170402131646.29884-3-david at tethera.net