Hi Jan,

My comments are below. My standard disclaimer is: if a suggestion I make 
is too disruptive to change at this time, please ignore it.

Thanks,
Keith

General: Docstrings don't need to mention the function name, as it is 
available right there, and docstring parsers will know this.

ai_get_manifest.py:
116: "is not None:"
161: "is not None:"
193: "is not None"
412: "is None:"

629-631: This no longer will "return None, -1" - is this intended?
632: bare "except" clause. Please determine what errors specifically 
could occur, or use "except StandardError:". May also be combined with 
previous clause.

787: If this needs to be done, is there a bug on it? If so, just remove 
the line entirely (or reference the bug). If not, can you file a bug on 
this?

878: bare 'except' clause. Can you convert to 'except StandardError'? If 
so, then 876-877 are no longer necessary.

ai_parse_manifest.py:

98: For readability, can you change this to "return None" (a blank 
'return' implicitly returns None)

ai_sd.py:

34: Please move this import to the end of the list of imports

202, 271: "cmd_popen.poll() is None"
280: "is not None:"
285: "if cmd_stderr:"

308: Trailing slash no longer needed

370: "if service_name:"

401-402: This line would be more readable if split up.

430-434: Same comment as line 878 from ai_get_manifest



Non-PEP-8/pylint changes / additional notes:
The following comments are completely optional, and likely out of scope, 
but I can't help but bring them up:

ai_get_manifest:
209: This variable could be removed; by checking for "client_arch is 
(not) None" you can determine the whether or not client_arch has yet 
been set.
216-224: The 'platform' python module provides this functionality 
internally. Alternatively, os.uname() also exposes this data.
Same comments for code in AICriteriaPlatform and AICriteriaCPU



Jan Damborsky wrote:
> Hi,
>
> Could I please ask for reviewing the 2.4->2.6 & PEP8 changes
> for AI client specific files ?
>
> Changes are covered by following bug:
> 12391 Transition from Python2.4 to Python2.6 - AI client portion
>
> Thank you very much,
> Jan
>
>
> * Webrev
> ========
> http://cr.opensolaris.org/~dambi/python-24-26/
>
> * Testing
> =========
> * platforms tested: x86, Sparc
> * modules affected:
>  - AI client service discovery (ai_sd, ai_get_manifest)
>  - wrapper for AI manifest parser (ai_parse_manifest, auto_install)
>
> * procedures
>  - AI images based on build 125 created with changes incorporated
>  - 2.6 changes for AI manifest parser taken from Jack and applied
>    as well in order to be able to test the wrapper
>
> * scenarios tested along with results
>  [a] manifest obtained involving service discovery using mDNS - OK
>  [b] manifest obtained while bypassing mDNS - OK
>  [c] manifest without criteria (default) obtained - OK
>  [d] manifest with criteria (MAC or IPv4) obtained - OK
>  [e] obtained manifest successfully parsed by Automated Installer - OK
>
> Automated Installation then always failed when initiating the
> transfer phase which is expected, since changes for Transfer
> module were not integrated yet.
> SMF log file for scenario [d] attached.
>
>
> * pylint results:
> =================
> http://cr.opensolaris.org/~dambi/pylint/ai_get_manifest.py.pylint.txt
> http://cr.opensolaris.org/~dambi/pylint/ai_parse_manifest.py.pylint.txt
> http://cr.opensolaris.org/~dambi/pylint/ai_sd.py.pylint.txt
>
> ai_get_manifest.py and ai_sd.py are not rated at 10 (8.87 and 9.87 
> respectively),
> as PEP8 changes are not high priority at this point. If there is an easy,
> straightforward, non-invasive way to improve the ratio, the changes 
> are to
> be accepted. Otherwise, separate bug is planned to be filed for 
> further PEP8
> specific changes.
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> caiman-discuss mailing list
> caiman-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss

Reply via email to