From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Schulte Subject: Re: "noweb-ref": Limiting scope of definitions? Date: Wed, 01 Feb 2012 07:16:06 -0700 Message-ID: <87obtilgtb.fsf@gmx.com> References: Mime-Version: 1.0 Content-Type: text/plain Return-path: Received: from eggs.gnu.org ([140.186.70.92]:38392) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Rsc5M-0003Q2-7t for emacs-orgmode@gnu.org; Wed, 01 Feb 2012 10:27:04 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Rsc5H-0000j7-FS for emacs-orgmode@gnu.org; Wed, 01 Feb 2012 10:27:00 -0500 Received: from mailout-us.gmx.com ([74.208.5.67]:60239) by eggs.gnu.org with smtp (Exim 4.71) (envelope-from ) id 1Rsc5H-0000j3-92 for emacs-orgmode@gnu.org; Wed, 01 Feb 2012 10:26:55 -0500 List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org Sender: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org To: Yu Cc: org-mode mailing list Yu writes: > Hello! > > Scenario > ----------------- > > * Maintain a potentially long org file as an appendix. > * In this appendix, many independent scripting tasks will be documented. > * Each such scripting task has similiar partial tasks, e.g. imports, > "settings", ... > > In such a case, many independent scripts may contain a block > "<>" or something or something to the same effect, i.e. an > overlap of definitions could be avoided only by some sort of scoping: > (a) manually, e.g. by prefixing the names with a script-specific > string (e.g. <>. While this is flexible (i.e. > allows reuse of code from other script), it also reduces the > readability of the NOWEB code). > (b) providing some sort of scoping, e.g. declaring definitions section-local. > > > Implementation Idea > ---------------------------------- > > A simple implementation, maintaining the flexibility of the manual > solution, would be to create a header argument like ":noweb-prefix", > that effectively just adds a prefix to the noweb reference names (both > in declaring the block (#+name, :noweb-ref) and in using it > (<>)), unless such a prefix is explicitly specified. This could > then be set as needed, for specific blocks or subtrees (as property) > or any mixture of such. > > A prefix would then be recognized by a delimiter string to be specified. > > This solution would also be downward compatible with existing files, > as the syntax for recognizing a prefix would be relevant only when > deciding whether to apply an explicitly introduced ":noweb-prefix". > This does seem like a good idea and a header argument such as ":noweb-prefix" is certainly the way to implement such functionality. I'll add this to my long term stack. Best, -- Eric Schulte http://cs.unm.edu/~eschulte/