[gentoo-portage-dev] [PATCH] depgraph: Skip atoms that are None (bug 505944)

[gentoo-portage-dev] [PATCH] depgraph: Skip atoms that are None (bug 505944)

[Alexander Berntsen]

Don't try to display atoms that are None when in debug mode.
---
Can I get an ACK on this please? Do share your ideas for refactoring,
if any. But even if this is not how we want it to work in the end, I
still think we should commit it to fix the bug.


 pym/_emerge/depgraph.py | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/pym/_emerge/depgraph.py b/pym/_emerge/depgraph.py
index abb70a7..556cb08 100644
--- a/pym/_emerge/depgraph.py
+++ b/pym/_emerge/depgraph.py
@@ -2211,12 +2211,13 @@ class depgraph(object):
 				# Display the specific atom from SetArg or
 				# Package types.
 				uneval = ""
-				if dep.atom is not dep.atom.unevaluated_atom:
-					uneval = " (%s)" % (dep.atom.unevaluated_atom,)
-				writemsg_level(
-					"%s%s%s required by %s\n" %
-					("Parent Dep:".ljust(15), dep.atom, uneval, myparent),
-					level=logging.DEBUG, noiselevel=-1)
+				if dep.atom is not None:
+					if dep.atom is not dep.atom.unevaluated_atom:
+						uneval = " (%s)" % (dep.atom.unevaluated_atom,)
+					writemsg_level(
+						"%s%s%s required by %s\n" %
+						("Parent Dep:".ljust(15), dep.atom, uneval, myparent),
+						level=logging.DEBUG, noiselevel=-1)
 
 		# Ensure that the dependencies of the same package
 		# are never processed more than once.
-- 
1.8.3.2

Re: [gentoo-portage-dev] [PATCH] depgraph: Skip atoms that are None (bug 505944)

[Brian Dolbec]

On Thu, 27 Mar 2014 13:21:03 +0100
Alexander Berntsen <bernalex@gentoo.org> wrote:

> Don't try to display atoms that are None when in debug mode.
> ---
> Can I get an ACK on this please? Do share your ideas for refactoring,
> if any. But even if this is not how we want it to work in the end, I
> still think we should commit it to fix the bug.
> 
> 

This is not nearly bad enough a bug to commit a panic fix.
It is the result of a typo on the commandline only.

I did some testing on my system.  This bug depends on there being a
resume list.  During testing, I tried dropping the -p option in a root
terminal, with a different change applied to prevent the traceback.

It output a huge long list of debug info, did not merge anything, now
it no longer has that merge list so only outputs:

emerge: It seems we have nothing to resume...


diff --git a/pym/_emerge/depgraph.py b/pym/_emerge/depgraph.py
index 83035c2..61ef31d 100644
--- a/pym/_emerge/depgraph.py
+++ b/pym/_emerge/depgraph.py
@@ -1942,8 +1942,9 @@ class depgraph(object):
                                # Display the specific atom from SetArg or
                                # Package types.
                                uneval = ""
-                               if dep.atom is not dep.atom.unevaluated_atom:
-                                       uneval = " (%s)" % (dep.atom.unevaluated_atom,)
+                               if dep.atom and dep.atom.unevaluated_atom:
+                                       if dep.atom is not dep.atom.unevaluated_atom:
+                                               uneval = " (%s)" % (dep.atom.unevaluated_atom,)
                                writemsg_level(
                                        "%s%s%s required by %s\n" %
                                        ("Parent Dep:".ljust(15), dep.atom, uneval, myparent),
lines 1-16/16 (END)


This takes advantage of the fact that uneval is initialized to ""
Then only reassigns it if there is actually a mismatch.
-- 
Brian Dolbec <dolsen>

Re: [gentoo-portage-dev] [PATCH] depgraph: Skip atoms that are None (bug 505944)

[Alexander Berntsen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 27/03/14 17:16, Brian Dolbec wrote:
> diff --git a/pym/_emerge/depgraph.py b/pym/_emerge/depgraph.py 
> index 83035c2..61ef31d 100644 --- a/pym/_emerge/depgraph.py +++
> b/pym/_emerge/depgraph.py @@ -1942,8 +1942,9 @@ class
> depgraph(object): # Display the specific atom from SetArg or #
> Package types. uneval = "" -                               if
> dep.atom is not dep.atom.unevaluated_atom: -
> uneval = " (%s)" % (dep.atom.unevaluated_atom,) +
> if dep.atom and dep.atom.unevaluated_atom: +
> if dep.atom is not dep.atom.unevaluated_atom: +
> uneval = " (%s)" % (dep.atom.unevaluated_atom,) writemsg_level( 
> "%s%s%s required by %s\n" % ("Parent Dep:".ljust(15), dep.atom,
> uneval, myparent), lines 1-16/16 (END)
> 
> 
> This takes advantage of the fact that uneval is initialized to "" 
> Then only reassigns it if there is actually a mismatch.
Sure. I'd use one "if" instead of two.
- -- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iF4EAREIAAYFAlM1O+UACgkQRtClrXBQc7XL+AD/Wn5sbE/O3QFAl7nBYblVWFAK
LBVwmY3XEOT2blgIPfIA/2a8IspgldZxO2AmLQ/bZNnFcigUPejaZ1V8J97JJuiu
=qc0Y
-----END PGP SIGNATURE-----