Was looking at what was output for <pkgname>-package.Rd
and wondered if any there was any means (via macro, etc)
to merge some of the same information with a template
for my package manpage? As much (all?) of the generated
information was already provided in the DESCRIPTION, I'd
prefer not to have to update the information in multiple
places. I'm thinking here that I could provide a template
file "<pkgname>-package.Rd.in" and during build, the
DCF information could be substituted appropriately and
"<pkgname>-package.Rd" would be output.
see also:
promptPackage method
----------------------------------------------------------
SIGSIG -- signature too long (core dumped)
Package manpage DCF hooks
6 messages · Paul Roebuck, Gabor Grothendieck, Martin Maechler
What I do is make my whatever-package.Rd page be the central page where one can get a list of all the other places one can look for info (rather than placing the info itself there). See, for example, library(dyn) package?dyn
On 11/14/05, Paul Roebuck <roebuck at mdanderson.org> wrote:
Was looking at what was output for <pkgname>-package.Rd and wondered if any there was any means (via macro, etc) to merge some of the same information with a template for my package manpage? As much (all?) of the generated information was already provided in the DESCRIPTION, I'd prefer not to have to update the information in multiple places. I'm thinking here that I could provide a template file "<pkgname>-package.Rd.in" and during build, the DCF information could be substituted appropriately and "<pkgname>-package.Rd" would be output. see also: promptPackage method ---------------------------------------------------------- SIGSIG -- signature too long (core dumped)
______________________________________________ R-devel at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
On Mon, 14 Nov 2005, Gabor Grothendieck wrote:
On 11/14/05, Paul Roebuck <roebuck at mdanderson.org> wrote:
Was looking at what was output for <pkgname>-package.Rd and wondered if any there was any means (via macro, etc) to merge some of the same information with a template for my package manpage? As much (all?) of the generated information was already provided in the DESCRIPTION, I'd prefer not to have to update the information in multiple places. I'm thinking here that I could provide a template file "<pkgname>-package.Rd.in" and during build, the DCF information could be substituted appropriately and "<pkgname>-package.Rd" would be output. see also: promptPackage method
What I do is make my whatever-package.Rd page be the central page where one can get a list of all the other places one can look for info (rather than placing the info itself there). See, for example, library(dyn) package?dyn
Thanks for your reply. That gives me some additional ideas but still think being able to display DCF information and public function listing would be a nice thing to have. For example, 'dyn-package.Rd' repeats its DCF description. ---------------------------------------------------------- SIGSIG -- signature too long (core dumped)
"Paul" == Paul Roebuck <roebuck at mdanderson.org>
on Mon, 14 Nov 2005 15:57:04 -0600 (CST) writes:
Paul> On Mon, 14 Nov 2005, Gabor Grothendieck wrote:
>> On 11/14/05, Paul Roebuck <roebuck at mdanderson.org> wrote:
>>
>> > Was looking at what was output for <pkgname>-package.Rd
>> > and wondered if any there was any means (via macro, etc)
>> > to merge some of the same information with a template
>> > for my package manpage? As much (all?) of the generated
>> > information was already provided in the DESCRIPTION, I'd
>> > prefer not to have to update the information in multiple
>> > places. I'm thinking here that I could provide a template
>> > file "<pkgname>-package.Rd.in" and during build, the
>> > DCF information could be substituted appropriately and
>> > "<pkgname>-package.Rd" would be output.
>> >
>> > see also:
>> > promptPackage method
>>
>> What I do is make my whatever-package.Rd page be
>> the central page where one can get a list of all
>> the other places one can look for info (rather than
>> placing the info itself there). See, for example,
>>
>> library(dyn)
>> package?dyn
Paul> Thanks for your reply. That gives me some additional
Paul> ideas but still think being able to display DCF
Paul> information and public function listing would be a nice
Paul> thing to have. For example, 'dyn-package.Rd' repeats its
Paul> DCF description.
which I agree is not ideal. I agree that such information
should in principle reside in one place and be
``auto-distributed'' to other places during package installation
and maybe also package load time.
Note that packageDescription("dyn")
returns an object that contains (and may print if you want) the
DCF information.
One possibility I see would be the convention that the
'generated' (text, html, tex) help files for 'package-<name>'
would combine both the packageDescription() and
the contents of <name>-package.Rd.
Martin
On Tue, 15 Nov 2005, Martin Maechler wrote:
"Paul" == Paul Roebuck <roebuck at mdanderson.org>
on Mon, 14 Nov 2005 15:57:04 -0600 (CST) writes:
Paul> On Mon, 14 Nov 2005, Gabor Grothendieck wrote:
>> On 11/14/05, Paul Roebuck <roebuck at mdanderson.org> wrote:
>>
>> > Was looking at what was output for <pkgname>-package.Rd
>> > and wondered if any there was any means (via macro, etc)
>> > to merge some of the same information with a template
>> > for my package manpage? As much (all?) of the generated
>> > information was already provided in the DESCRIPTION, I'd
>> > prefer not to have to update the information in multiple
>> > places. I'm thinking here that I could provide a template
>> > file "<pkgname>-package.Rd.in" and during build, the
>> > DCF information could be substituted appropriately and
>> > "<pkgname>-package.Rd" would be output.
>> >
>> > see also:
>> > promptPackage method
>>
>> What I do is make my whatever-package.Rd page be
>> the central page where one can get a list of all
>> the other places one can look for info (rather than
>> placing the info itself there). See, for example,
>>
>> library(dyn)
>> package?dyn
Paul> Thanks for your reply. That gives me some additional
Paul> ideas but still think being able to display DCF
Paul> information and public function listing would be a nice
Paul> thing to have. For example, 'dyn-package.Rd' repeats its
Paul> DCF description.
which I agree is not ideal. I agree that such information
should in principle reside in one place and be
``auto-distributed'' to other places during package installation
and maybe also package load time.
Note that packageDescription("dyn") returns an object that
contains (and may print if you want) the DCF information.
I'm aware of this, having used it in various places. What I don't know is how to access/use it during package installation (if even possible). Using read.dcf and a sed script, I could probably manage to perform the template merge. But I don't know how to invoke such without adding a configure script (overkill for R-only packages), as 'install.R' is meant for something else.
One possibility I see would be the convention that the 'generated' (text, html, tex) help files for 'package-<name>' would combine both the packageDescription() and the contents of <name>-package.Rd.
Well, a system-level approach would be preferable to doing this per-package. R-2.3 then? ---------------------------------------------------------- SIGSIG -- signature too long (core dumped)
"Paul" == Paul Roebuck <roebuck at mdanderson.org>
on Tue, 15 Nov 2005 13:07:47 -0600 (CST) writes:
Paul> On Tue, 15 Nov 2005, Martin Maechler wrote:
>> >>>>> "Paul" == Paul Roebuck <roebuck at mdanderson.org>
>> >>>>> on Mon, 14 Nov 2005 15:57:04 -0600 (CST) writes:
>>
Paul> On Mon, 14 Nov 2005, Gabor Grothendieck wrote:
>> >> On 11/14/05, Paul Roebuck <roebuck at mdanderson.org> wrote:
>> >>
>> >> > Was looking at what was output for <pkgname>-package.Rd
>> >> > and wondered if any there was any means (via macro, etc)
>> >> > to merge some of the same information with a template
>> >> > for my package manpage? As much (all?) of the generated
>> >> > information was already provided in the DESCRIPTION, I'd
>> >> > prefer not to have to update the information in multiple
>> >> > places. I'm thinking here that I could provide a template
>> >> > file "<pkgname>-package.Rd.in" and during build, the
>> >> > DCF information could be substituted appropriately and
>> >> > "<pkgname>-package.Rd" would be output.
>> >> >
>> >> > see also:
>> >> > promptPackage method
>> >>
>> >> What I do is make my whatever-package.Rd page be
>> >> the central page where one can get a list of all
>> >> the other places one can look for info (rather than
>> >> placing the info itself there). See, for example,
>> >>
>> >> library(dyn)
>> >> package?dyn
>>
Paul> Thanks for your reply. That gives me some additional
Paul> ideas but still think being able to display DCF
Paul> information and public function listing would be a nice
Paul> thing to have. For example, 'dyn-package.Rd' repeats its
Paul> DCF description.
>>
>> which I agree is not ideal. I agree that such information
>> should in principle reside in one place and be
>> ``auto-distributed'' to other places during package installation
>> and maybe also package load time.
>>
>> Note that packageDescription("dyn") returns an object that
>> contains (and may print if you want) the DCF information.
Paul> I'm aware of this, having used it in various places. What
Paul> I don't know is how to access/use it during package
Paul> installation (if even possible). Using read.dcf and a sed
Paul> script, I could probably manage to perform the template
Paul> merge. But I don't know how to invoke such without adding
Paul> a configure script (overkill for R-only packages), as
Paul> 'install.R' is meant for something else.
>> One possibility I see would be the convention that the
>> 'generated' (text, html, tex) help files for 'package-<name>'
>> would combine both the packageDescription() and
>> the contents of <name>-package.Rd.
Paul> Well, a system-level approach would be preferable to doing
Paul> this per-package.
Definitely, and actually I was only thinking of the former.
Paul> R-2.3 then?
with help of contributions from smart R-devel
readers/contributors that should be fairly plausible,
otherwise I'm much less confident.
Martin