[PATCH 05/10] parsemail: Handle series sent "in-reply-to"

[Andy Doan]

On 06/13/2016 05:41 AM, Stephen Finucane wrote:
> Take a series, "v2", sent as a reply to another, "v1", like so:
>
>      [PATCH 0/3] A cover letter
>        [PATCH 1/3] The first patch
>        ...
>        [PATCH v2 0/3] A cover letter
>          [PATCH v2 1/3] The first patch
>          ...
>
> The current behavior for traversing references is oldest first. For
> example, for "PATCH v2 1/3" above, the references field will look like
> so:
>
>      ["0/3", "v2 0/3", "v2 1/3"]
>
> Where "0/3" corresponds to the message-id of "PATCH 0/3".
>
> By traversing this way, patches sent in-reply-to an existing series
> will always be linked to that series, instead of creating a new series,
> as expected. Flip things around, and instead attempt to find series
> using the most recent message-id first. This will ensure the most
> recent series is always used.
>
> Signed-off-by: Stephen Finucane <stephen.finucane at intel.com>
> ---
>   patchwork/bin/parsemail.py |   10 +++++++++-
>   1 files changed, 9 insertions(+), 1 deletions(-)
>
> diff --git a/patchwork/bin/parsemail.py b/patchwork/bin/parsemail.py
> index f52776e..8baacf6 100755
> --- a/patchwork/bin/parsemail.py
> +++ b/patchwork/bin/parsemail.py
> @@ -137,7 +137,7 @@ def find_series(mail):
>       """
>       series = None
>
> -    for ref in find_references(mail) + [mail.get('Message-ID').strip()]:
> +    for ref in [mail.get('Message-ID').strip()] + find_references(mail)[::-1]:
>           # try parsing by RFC5322 fields first
>           try:
>               series_ref = SeriesReference.objects.get(msgid=ref)
> @@ -575,6 +575,10 @@ def parse_mail(mail, list_id=None):
>               series.save()
>
>               for ref in refs + [msgid]:  # save references for series

I think this can be simplified and perhaps be less racy:
> +                # prevent duplication
> +                if SeriesReference.objects.filter(msgid=ref).exists():
> +                    continue
> +
>                   series_ref = SeriesReference(series=series,
>                                                msgid=ref)
>                   series_ref.save()

    try:
        SeriesReference.objects.create(series=series, msgid=ref)
    except IntegrityError:
         pass # already exists

> @@ -621,6 +625,10 @@ def parse_mail(mail, list_id=None):
>                   series.save()
>
>                   for ref in refs + [msgid]:  # save references for series
> +                    # prevent duplication
> +                    if SeriesReference.objects.filter(msgid=ref).exists():
> +                        continue
> +
>                       series_ref = SeriesReference(series=series,
>                                                    msgid=ref)
>                       series_ref.save()
>