how to convert/normalize plain footnotes into [fn:] ones

how to convert/normalize plain footnotes into [fn:] ones

[Gerald Wildgruber]

Hi,

I'm trying to convert older orgmode documents which have plain,
"[1]"-like footnotes into the newer "[fn:]" format.

Can this be done automatically?

In

https://lists.gnu.org/archive/html/emacs-orgmode/2015-12/msg00435.html

Nicolas mentions "org-footnote-normalize" which would turn old footnotes
in new ones, but it does nothing in my case except adding a "*
Footnotes" entry.

Thanks

Gerald.

Re: how to convert/normalize plain footnotes into [fn:] ones

[Nicolas Goaziou]

Hello,

Gerald Wildgruber <Gerald.Wildgruber@unibas.ch> writes:

> I'm trying to convert older orgmode documents which have plain,
> "[1]"-like footnotes into the newer "[fn:]" format.
>
> Can this be done automatically?

You could use `query-replace-regexp' a.k.a. C-M-%.

> Nicolas mentions "org-footnote-normalize" which would turn old footnotes
> in new ones, but it does nothing in my case except adding a "*
> Footnotes" entry.

You misunderstood. `org-footnote-normalize' used to turn every footnote
into [1], [2], ... whereas it now turns them into [fn:1], [fn:2],... IOW
it cannot be used to convert from old syntax to new one.

Regards,

-- 
Nicolas Goaziou

Re: how to convert/normalize plain footnotes into [fn:] ones

[Gerald Wildgruber]

Hi Nicolas,

thanks:

I had actually tried that already and came up with the following regex:

\([^ ]\)\(\[\)\([0-9]\) --> \1\2fn:\3

i.e. three groupings: 1) no whitespace, 2) "[" and 3) a number, and then
introducing "fn:" between "[" and the number: but it is not perfect, as
it still gives false positives; do you have something better?

The problem unfortunately arises also by pandoc still using the now
obsolete syntax of plain footnotes when converting (e.g. latex) to org
mode!


Gerald.


On Mi, Dez 21 2016, Nicolas Goaziou <mail@nicolasgoaziou.fr> wrote:

> Hello,
>
> Gerald Wildgruber <Gerald.Wildgruber@unibas.ch> writes:
>
>> I'm trying to convert older orgmode documents which have plain,
>> "[1]"-like footnotes into the newer "[fn:]" format.
>>
>> Can this be done automatically?
>
> You could use `query-replace-regexp' a.k.a. C-M-%.
>
>> Nicolas mentions "org-footnote-normalize" which would turn old footnotes
>> in new ones, but it does nothing in my case except adding a "*
>> Footnotes" entry.
>
> You misunderstood. `org-footnote-normalize' used to turn every footnote
> into [1], [2], ... whereas it now turns them into [fn:1], [fn:2],... IOW
> it cannot be used to convert from old syntax to new one.
>
> Regards,


--
Dr.Gerald Wildgruber
eikones NFS Bildkritik
NCCR Iconic Criticism
Universität Basel
Rheinsprung 11
CH-4051 Basel
www.eikones.ch
T. +41 (0)61 267 18 02
F. +41 (0)61 267 18 11
---------------------
Sent with mu4e

Re: how to convert/normalize plain footnotes into [fn:] ones

[Nicolas Goaziou]

Hello,

Gerald Wildgruber <Gerald.Wildgruber@unibas.ch> writes:

> I had actually tried that already and came up with the following regex:
>
> \([^ ]\)\(\[\)\([0-9]\) --> \1\2fn:\3
>
> i.e. three groupings: 1) no whitespace, 2) "[" and 3) a number, and then
> introducing "fn:" between "[" and the number: but it is not perfect, as
> it still gives false positives; do you have something better?

I would use (and, in fact, used)

 \\[\\([0-9]+\\)\\] --> \\[fn:\\1\\]

and eyeballing the results.

> The problem unfortunately arises also by pandoc still using the now
> obsolete syntax of plain footnotes when converting (e.g. latex) to org
> mode!

I'm certain Pandoc importer will ultimately catch-up.

Regards,

-- 
Nicolas Goaziou

Re: how to convert/normalize plain footnotes into [fn:] ones

[Gerald Wildgruber]

Hi Nicolas,

thanks again!

On Mi, Dez 21 2016, Nicolas Goaziou <mail@nicolasgoaziou.fr> wrote:

> I would use (and, in fact, used)
>
>  \\[\\([0-9]+\\)\\] --> \\[fn:\\1\\]
>
> and eyeballing the results.

That works quite well; but as I'm trying to convert a whole book I also
have lots of "\cite[NUMBER]{my_ref}" commands from TeX; and these all
give false positives with the regex above. Actually I couldn't figure
out what a regex would have to look like that excludes occurences of the
mentioned expression if they are preceded by "cite".

>> The problem unfortunately arises also by pandoc still using the now
>> obsolete syntax of plain footnotes when converting (e.g. latex) to org
>> mode!
>
> I'm certain Pandoc importer will ultimately catch-up.

Yes, I posted the problem on the pandoc mailing list and Albert
Krewinkel kindly fixed it immediately and pushed the code to the dev
version of pandoc:

https://github.com/jgm/pandoc/commit/d27188ad309feb9a76f6dc3d816ad78722078fe1

I checked out the haskell source, compiled it and it works perfectly
now! The org writer produces the correct footnote format.

Gerald.