Re: [PATCH v2] post-receive-email: explicitly set Content-Type header

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



> Johannes Sixt wrote:
> > Am 9/20/2011 12:42, schrieb Shumkin Alexey:
> 
> >> 1. post-send-mail uses description file of a repo
> >> 2. gitweb also uses this file and AFAIK it assumes one to be in
> >> UTF-8 (I do not know whether it can be changed there but I tested
> >> gitweb once long time ago)
> >> 3. So if i18n.logoutputencoding is not UTF-8 we get a message
> >> composed with mixed encodings. This fact oblidge us to encode
> >> headers (as quoted printable at least) and synchronize body
> >> message that contain repo description (in UTF-8) and diffstat (in
> >> i18n.logoutputencoding).
> [...]
> > In this case, it may make sense to have a separate setting, but you
> > should call git like this:
> >
> >    git -c "i18n.logoutputencoding=$emailcharset" show ...
> >    git -c "i18n.logoutputencoding=$emailcharset" rev-list
> > --pretty ...
> 
> Something like this, I suppose?
> 
> This teaches post-receive-email to use plumbing where possible and to
> explicitly declare what encoding it expects output to use.  Completely
> untested --- basic sanity checking, testing, and tweaks to allow
> overriding the choice of encoding left as an exercise to the reader.
> 
> Based on patches by Gerrit Pape and Jeff King and advice from Johannes
> Sixt.
> 
> Signed-off-by: Jonathan Nieder <jrnieder@xxxxxxxxx>
> ---
> More background:
> http://thread.gmane.org/gmane.comp.version-control.git/124350/focus=124355
> 
>  contrib/hooks/post-receive-email |   15 +++++++++------
>  1 files changed, 9 insertions(+), 6 deletions(-)
> 
> diff --git a/contrib/hooks/post-receive-email
> b/contrib/hooks/post-receive-email index ba077c13..bc603b02 100755
> --- a/contrib/hooks/post-receive-email
> +++ b/contrib/hooks/post-receive-email
> @@ -234,6 +234,9 @@ generate_email_header()
>  	cat <<-EOF
>  	To: $recipients
>  	Subject: ${emailprefix}$projectdesc $refname_type
> $short_refname ${change_type}d. $describe
> +	MIME-Version: 1.0
> +	Content-Type: text/plain; charset=utf-8
> +	Content-Transfer-Encoding: 8bit
>  	X-Git-Refname: $refname
>  	X-Git-Reftype: $refname_type
>  	X-Git-Oldrev: $oldrev
> @@ -462,7 +465,7 @@ generate_delete_branch_email()
>  	echo "       was  $oldrev"
>  	echo ""
>  	echo $LOGEND
> -	git show -s --pretty=oneline $oldrev
> +	git diff-tree --encoding=UTF-8 -s --pretty=oneline $oldrev
>  	echo $LOGEND
>  }
>  
> @@ -538,11 +541,11 @@ generate_atag_email()
>  		# performed on them
>  		if [ -n "$prevtag" ]; then
>  			# Show changes since the previous release
> -			git rev-list --pretty=short
> "$prevtag..$newrev" | git shortlog
> +			git shortlog --encoding=UTF-8
> "$prevtag..$newrev" else
>  			# No previous tag, show all the changes
> since time # began
> -			git rev-list --pretty=short $newrev | git
> shortlog
> +			git shortlog --encoding=UTF-8 "$newrev"
>  		fi
>  		;;
>  	*)
> @@ -562,7 +565,7 @@ generate_delete_atag_email()
>  	echo "       was  $oldrev"
>  	echo ""
>  	echo $LOGEND
> -	git show -s --pretty=oneline $oldrev
> +	git diff-tree --encoding=UTF-8 -s --pretty=oneline $oldrev
>  	echo $LOGEND
>  }
>  
> @@ -608,7 +611,7 @@ generate_general_email()
>  	echo ""
>  	if [ "$newrev_type" = "commit" ]; then
>  		echo $LOGBEGIN
> -		git show --no-color --root -s --pretty=medium $newrev
> +		git diff-tree --encoding=UTF-8 --root -s
> --pretty=oneline $newrev echo $LOGEND
>  	else
>  		# What can we do here?  The tag marks an object that
> is not @@ -627,7 +630,7 @@ generate_delete_general_email()
>  	echo "       was  $oldrev"
>  	echo ""
>  	echo $LOGEND
> -	git show -s --pretty=oneline $oldrev
> +	git diff-tree --encoding=UTF-8 -s --pretty=oneline $oldrev
>  	echo $LOGEND
>  }
>  

As I understand, this patches make email message with explicitly set
Content-Type header (it's ok) and UTF-8 encoding (that is a subject
to discuss).
My proposition was in to send email message in explicitly defined
custom encoding. Why? In development process under Windows non-UTF-8
encoding is used (cp1251 in my case). So, filenames have this encoding,
and as we know Git stores their names as is - in cp1251 - without a
conversion. And filenames are also used in diff-stat (with
core.quotepath= false, BTW, I did not take into account this config)
without any conversion. So if we'll make all text in UTF-8 but
filenames are still non-UTF-8, email would look corrupted.

--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Linux Kernel Development]     [Gcc Help]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [V4L]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Fedora Users]