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

[Michal Orzel <michal.orzel@xxxxxxx>]

Hi Luca,

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.

~Michal