svn commit: r420830 - in head/databases: . neo4j neo4j/files

[Mathieu Arnold]

+--On 25 août 2016 10:11:41 +0100 Chris Rees <crees at physics.org> wrote:
| 
| 
| On 25 August 2016 09:47:07 BST, Mathieu Arnold <mat at FreeBSD.org> wrote:
|> 
|> 
|> +--On 25 août 2016 02:19:15 +0000 Chris Rees <crees at FreeBSD.org> wrote:
|>| +post-patch:
|>| +	${REINPLACE_CMD} ${_SUB_LIST_TEMP} ${WRKSRC}/bin/neo4j-shared.sh \
|>| +					${WRKSRC}/conf/neo4j.conf
|> 
|> _SUB_LIST_TEMP is an undocumented *internal* framework variable.  It
|> should
|> *never* be used in a port's Makefile.
|> With the amount of things I'm rewriting, it may disapear any day, and I
|> will not hunt down the ports that use it where they should not.
| 
| I must say I had a reservation about using it, although I think I've seen
| this done before.  I think this should be exposed as ${SUB_SED_COMMAND}
| or something because it's useful.
| 
| Is there a mechanism for having patchfiles as SUB_FILES?

No, it is a bad idea to use sed on patches because it makes running make
makepatch impossible.

It is why everyone uses sed in post-patch to replace the %%foo%% they put
in their patches.

To regenerate your patches, all you then have to do is:

make clean extract do-patch makepatch

(or if USES=dos2unix)

make clean extract dos2unix do-patch makepatch


-- 
Mathieu Arnold
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 949 bytes
Desc: not available
URL: <http://lists.freebsd.org/pipermail/svn-ports-all/attachments/20160825/07583aca/attachment-0001.sig>