app-admin/collectd: Bumped to v5.5.1

[Whissi]

• New upstream version (Closes #570766, #517060)
• Non-functional mongodb plugin disabled in ebuild (see upstream
  issue kde-apps/kdeartwork*: Port to kf5 #492, closes #524400)
• Multiple dependency problems fixed
• linux-info eclass usage fixed (Closes #570778)
• Support for new plugins added (amqp, genericjmx, ipvs, modbus, sigrok,
  tail_csv, vserver, write_riemann)
• PYTHON_SINGLE_TARGET support for Python 3.4 and 3.5 enabled
• Documentation improved (USE flag description, elogs)
• Linux capabilities usage expanded to improve running collectd as
  unprivileged user (disable via filecaps USE flag)
• Runscript rewritten
  • Runscript now supports OpenRC's default status function
  • Runscript now guarantees service shutdown
  • Validating configuration before (re)starting daemon

[Whissi]

@jbergstroem ping

[jbergstroem]

Thanks for your work. Will review over the weekend.

[gktrk]

Is there a reason why "epatch ${PATCHES[@]}" is called on line 265? I believe the EAPI6 equivalent is eapply, unless the patches are not in p1 format.

[jbergstroem]

@gktrk afaik its easier to call default in eapi6 to get PATCHES() applied (among other things).

[Whissi]

Yes, because we provide our own SRC_PREPARE() function so we have to call epatch and eapply_user on our own.

In theory we could call default instead however if we ever have to apply a patch with different patch level we would have to rewrite so I chose this style on purpose.

[Whissi]

However I re-read EAPI6's src_prepare() definition and will update the PR to switch to default usage and also eliminate eutils eclass usage.

[jbergstroem]

@Whissi yeah, sorry for not being clear enough; I meant invoking default to get parent functionality just like you implemented.

[monsieurp]

@jbergstroem let us know when the PR's good go to!

[jbergstroem]

LGTM