[Piuparts-devel] faster-rdep-calc/rdep-chain-len

[Andreas Beckmann]

On 2013-04-14 00:36, Dave Steele wrote:
> With that, I think I've addressed the open issues related to
> faster-rdep-calc and rdep-chain-len (BTW, chain-len is now useful -
> resolve-virtual was the fix to cause the waiting-for-dependency count
> to be driven quickly to zero). Some rework remains after the chdir()
> elimination goes to the front of the line, followed by testing. There
> are still a few issues remaining with the original
> detect_well_known_errors submission that I expect to address.

* "packagesdb - Refactor out get_rdep_dict() from calc_rrdep_counts()."
  since is get_rdep_dict() is an internal interface, it should be called
  _get_rdep_dict() for consistency - or do you expect anything outside
PackagesDB to call this?

* "packagesdb.py - Use get_package() in get_rdep_dict() to resolve
virtual pkgs."
  that completely is borked - you probably want to use

  for pkg in self.get_all_packages():
      # use the Packages dependencies() method for a conservative count
      for dep in pkg.dependencies():
         dep_pkg = self.get_package(dep, resolve_virtual=True):
         if dep_pkg is not None:
            dep = dep_pkg["Package"]
         if not dep in self._rdeps:
            self._rdeps[dep] = []
         self._rdeps[dep].append(pkg["Package"])

  and drop the _find_all_packages() call (already done by
  get_all_packages()

* "packagesdb - Add routine to calc rdep metrics by package as needed."
  NACK until you insert a5348699 "lib/db: Package: rename count member
variables" before this patch (and adjust the variable names as I did in
my fa97aaac variant of your patch)
This currently (c8118e1) introduces really bad code due to the name
clash between the Package methods and member variables