Re: [PATCH v2 1/3] python/sepolicy: Fix sepolicy manpage -w ...

[Ondrej Mosnacek <omosnace@redhat.com>]

On Tue, Nov 15, 2022 at 8:49 PM Petr Lautrbach <plautrba@redhat.com> wrote:
>
> From: Petr Lautrbach <lautrbach@redhat.com>
>
> Commit 7494bb1298b3 ("sepolicy: generate man pages in parallel")
> improved sepolicy performance but broke `sepolicy manpage -w ...` as it
> didn't collect data about domains and roles from ManPage() and so
> HTMLManPages() generated only empty page. This is fixed now, domains
> and roles are being collected and used for HTML pages.
>
> Signed-off-by: Petr Lautrbach <lautrbach@redhat.com>
> ---
>  python/sepolicy/sepolicy.py         | 13 +++++++++++--
>  python/sepolicy/sepolicy/manpage.py | 12 +++++-------
>  2 files changed, 16 insertions(+), 9 deletions(-)
>
> diff --git a/python/sepolicy/sepolicy.py b/python/sepolicy/sepolicy.py
> index 733d40484709..2ca02ee9a0cf 100755
> --- a/python/sepolicy/sepolicy.py
> +++ b/python/sepolicy/sepolicy.py
[...]
> @@ -347,9 +348,17 @@ def manpage(args):
>      else:
>          test_domains = args.domain
>
> +    manpage_domains = set()
> +    manpage_roles = set()
>      p = Pool()
> +    async_results = []
>      for domain in test_domains:
> -        p.apply_async(manpage_work, [domain, path, args.root, args.source_files, args.web])
> +        async_results.append(p.apply_async(manpage_work, [domain, path, args.root, args.source_files, args.web]))

> +    results = map(lambda x: x.get(), async_results)
> +    for result in results:
> +        manpage_domains.update(set(result[0]))
> +        manpage_roles.update(set(result[1]))

The above four lines can be written a bit more nicely as follows:

for result in async_results:
    domains, roles = result.get()
    manpage_domains.update(domains)
    manpage_roles.update(roles)

Note that set.update() accepts any iterable, it doesn't need to be a set.

> +
>      p.close()
>      p.join()
>
[...]


--
Ondrej Mosnacek
Senior Software Engineer, Linux Security - SELinux kernel
Red Hat, Inc.