Re: copydir prototype - Mailing list pgsql-hackers

From Tom Lane
Subject Re: copydir prototype
Date
Msg-id 2444.1289596415@sss.pgh.pa.us
Whole thread Raw
In response to copydir prototype  (Robert Haas <robertmhaas@gmail.com>)
Responses Re: copydir prototype
List pgsql-hackers
Robert Haas <robertmhaas@gmail.com> writes:
> This patch:
> http://git.postgresql.org/gitweb?p=postgresql.git;a=commit;h=bb0fe9feb9fd75a6aaa960028a9f810c883b0fc4
> ...did not tidy up after itself as well as it might have.  In
> particular, it left the prototype for copydir() in src/include/port.h,
> while moving the source file from src/port/copydir.c to
> src/backend/storage/file/copydir.c.  I think this should probably be
> cleaned up, but I'm a little uncertain where the best place to put
> that prototype is.  I am tempted to just put this in a new header file
> named according to our usual conventions, namely
> src/include/storage/copydir.c, but since there's only one public
> function at present perhaps someone would like to argue for including
> it in some other, already-exstant header.

copydir.h I assume you meant?  Seems reasonable.

> A patch I'm working on needs to expose the copy_file() function, which
> is currently static to copydir.c, so maybe it would be preferable to
> rename copydir.c to copy.c and add the header as
> src/include/storage/copy.h.

-1 for that.  Aside from being generally opposed to inessential file
renamings, I think this will risk confusion with commands/copy.[ch].
        regards, tom lane


pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: copydir prototype
Next
From: Dimitri Fontaine
Date:
Subject: Re: wCTE behaviour