D4797: storageutil: implement file identifier resolution method (BC)

[indygreg (Gregory Szorc)]

indygreg added a comment.


  In https://phab.mercurial-scm.org/D4797#73451, @martinvonz wrote:
  
  > > with the exception of partial hex node matching, we may want to consider changing revlog.lookup() as well
  >
  > Maybe even without the exception? I thought we used partial nodeids only for debug commands and maybe we can move the partial matching to a higher level.
  
  
  I support moving partial matching to a higher level. Although some storage backends may support a more efficient "native" partial matcher. e.g. SQLite could use various string matching functions. But I'm not sure it is worth it though: I think I'd rather have a generic object holding DAG/index data [that can be shared across all storage backends] and that exposes a partial match API. I dunno. I don't plan to do anything in this area at this time. If you want to hack on things, go for it.

REPOSITORY
  rHG Mercurial

REVISION DETAIL
  https://phab.mercurial-scm.org/D4797

To: indygreg, #hg-reviewers
Cc: martinvonz, mercurial-devel