File History vs. Path History

[Matt Mackall]

On Sun, 2013-01-20 at 08:41 +0000, Paul Mensonides wrote:
> On Sat, 19 Jan 2013 14:41:59 -0600, Matt Mackall wrote:
> 
> > On Sat, 2013-01-19 at 08:17 +0000, Paul Mensonides wrote:
> 
> > Now replace "president" with "Makefile". From the perspective of your
> > build system, it does not matter at all how a particular file got to be
> > the Makefile, it's _still in charge_. Because files on your filesystem
> > are addressed by name, not by content or history, in a very real way,
> > filenames matter more than file history.
> > 
> > Opinions differ as to whether this is the right point of view for an SCM
> > to take; it's a deeply moot and uninteresting question for Mercurial at
> > this point.
> 
> I understand.
> 
> >> Explicitly, what I actually need (wian) is all changesets which are
> >> descendants of a particular changeset which affect a particular file
> >> (with a --follow variation).  E.g.
> > 
> >> hg wian -r 0 A
> >>    => changesets 0 and 1
> > 
> > Unfortunately, we don't have a convenient way to do this for two
> > reasons:
> 
> Okay, so (thinking out loud) I can get all of the heads in the repository, 
> build a graph of the changesets via parents, run hg stat --changes each 
> changeset, build an entity index with the "history" that I want, and 
> serialize the entire thing.  Later, updating requires only checking last 
> used heads against current heads and only processing the changesets back 
> to those already processed.  Doesn't sound terribly fun, but it's doable.

If you're going to throw some 'real' code at the problem, you might
consider parsing and inverting 'hg debugindex path/to/file'.

-- 
Mathematics is the supreme nostalgia of our time.