On Wed, Jun 17, 2015 at 5:18 PM, Junio C Hamano <gits...@pobox.com> wrote: > Mike Hommey <m...@glandium.org> writes: > > I'm tempted to make init_notes itself do the check, based on the value > > it is given for a "read_only" argument. > > Yeah, that would be one sensible way to go after making sure that > everything goes thru this interface.
Agreed. Furthermore, consider adding the read_only flag (or however you choose to encode it internally) to struct notes_tree, so that the API functions that _manipulate_ notes trees can immediately bail out when used on a read-only tree (i.e. we want them to fail as early as possible). > > On the other hand, some commands > > do their ref resolving themselves already. > > Again, as long as they do not bypass the "read-only" safety you are > suggesting to add to init_notes(), that is OK. Agreed. An alternative to adding a simple read_only flag argument is to modify the const char *notes_ref argument into two separate arguments: const char *notes_treeish, and const char *update_ref, where the latter should be NULL for read-only trees. That said, currently the logic for actually updating notes ref lives outside the notes.h API (see commit_notes() in notes-utils.h/c), so there might be room for more consolidation/refactoring here... ...Johan -- Johan Herland, <jo...@herland.net> www.herland.net -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
