Re: [PATCH 1/2] xen/cppcheck: sort alphabetically cppcheck report entries

[Jan Beulich <jbeulich@xxxxxxxx>]

On 09.01.2023 12:15, Michal Orzel wrote:
> On 06/01/2023 11:41, Luca Fancellu wrote:
>> Sort alphabetically cppcheck report entries when producing the text
>> report, this will help comparing different reports and will group
>> together findings from the same file.
>>
>> Signed-off-by: Luca Fancellu <luca.fancellu@xxxxxxx>
>> ---
>>  xen/scripts/xen_analysis/cppcheck_report_utils.py | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/xen/scripts/xen_analysis/cppcheck_report_utils.py 
>> b/xen/scripts/xen_analysis/cppcheck_report_utils.py
>> index 02440aefdfec..f02166ed9d19 100644
>> --- a/xen/scripts/xen_analysis/cppcheck_report_utils.py
>> +++ b/xen/scripts/xen_analysis/cppcheck_report_utils.py
>> @@ -104,6 +104,8 @@ def cppcheck_merge_txt_fragments(fragments_list, 
>> out_txt_file, strip_paths):
>>                  for path in strip_paths:
>>                      text_report_content[i] = text_report_content[i].replace(
>>                                                                  path + "/", 
>> "")
>> +            # sort alphabetically the entries
>> +            text_report_content.sort()
>>              # Write the final text report
>>              outfile.writelines(text_report_content)
>>      except OSError as e:
>> --
>> 2.17.1
>>
>>
> 
> Having the report sorted is certainly a good idea. I am just thinking whether 
> it should be done
> per file or per finding (e.g. rule). When fixing MISRA issues, best approach 
> is to try to fix all
> the issues for a given rule (i.e. a series fixing one rule) rather than all 
> the issues in a file
> from different rules. Having a report sorted per finding would make this 
> process easier. We could
> add a custom key to sort function to take the second element (after splitting 
> with ':' separator)
> which is the name of the finding to achieve this goal. Let me know your 
> thoughts.

+1 - sorting by file name wants to be the 2nd sorting criteria, i.e. only among
all instances of the same finding.

Jan