[PATCH] Fix renumbering for footnotes at BOL.

[PATCH] Fix renumbering for footnotes at BOL.

[Tassilo Horn]

Hi all,

org-footnote-renumber-fn:N invalidates footnotes that appear as the
first thing on a line.  Here's a test file including the recipe.

--8<---------------cut here---------------start------------->8---
* Bla

  This is a line in which the following footnote that is inserted exactly
  [fn:1] is the first thing in that line.  And now invoke
  org-footnote-renumber-fn:N and see it breaking.

* Footnotes

[fn:1] Test
--8<---------------cut here---------------end--------------->8---

Note, that in larger org files, it's likely that you don't even notice
that the footnotes break.

Once I added a patch on patchwork that fixed this issue, and that's
shown as "accepted".

  http://patchwork.newartisans.com/patch/705/

I think, I've marked it as deferred or so myself because I've thought
the issue was fixed in the meantime, but that doesn't seem to be true.

Here's an updated patch.
---
 lisp/org-footnote.el |   11 ++++-------
 1 files changed, 4 insertions(+), 7 deletions(-)

diff --git a/lisp/org-footnote.el b/lisp/org-footnote.el
index cce0921..931f7dd 100644
--- a/lisp/org-footnote.el
+++ b/lisp/org-footnote.el
@@ -893,16 +893,13 @@ If LABEL is non-nil, delete that footnote instead."
 	(goto-char (point-min))
 	(while (re-search-forward "\\[fn:\\([0-9]+\\)[]:]" nil t)
 	  (setq i (string-to-number (match-string 1)))
-	  (when (and (string-match "\\S-" (buffer-substring
-					   (point-at-bol) (match-beginning 0)))
-		     (not (assq i map)))
+	  (when (not (assq i map))
 	    (push (cons i (number-to-string (incf n))) map)))
 	(goto-char (point-min))
 	(while (re-search-forward "\\(\\[fn:\\)\\([0-9]+\\)\\([]:]\\)" nil t)
-	  (replace-match (concat "\\1"
-				 (cdr (assq (string-to-number (match-string 2))
-					    map))
-				 "\\3")))))))
+	  (setq i (cdr (assq (string-to-number (match-string 2)) map)))
+	  (assert (progn i) t "Footnote has no number. Better undo renumbering!")
+	  (replace-match (concat "\\1" i "\\3")))))))
 
 (defun org-footnote-auto-adjust-maybe ()
   "Renumber and/or sort footnotes according to user settings."
-- 
1.7.6.1

Re: [PATCH] Fix renumbering for footnotes at BOL.

[Nicolas Goaziou]

Hello,

Tassilo Horn <tassilo@member.fsf.org> writes:

> org-footnote-renumber-fn:N invalidates footnotes that appear as the
> first thing on a line.  Here's a test file including the recipe.
>
> * Bla
>
>   This is a line in which the following footnote that is inserted exactly
>   [fn:1] is the first thing in that line.  And now invoke
>   org-footnote-renumber-fn:N and see it breaking.
>
> * Footnotes
>
> [fn:1] Test
>
> Note, that in larger org files, it's likely that you don't even notice
> that the footnotes break.

It looks like the original function isn't right in more than one way (it
doesn't even make sure the matched string is really a footnote). But
that's another topic.

About your patch, I have but one question: I don't get a situation in
which the assert would be triggered, may you give me such an example?

Regards,

-- 
Nicolas Goaziou

Re: [PATCH] Fix renumbering for footnotes at BOL.

[Tassilo Horn]

Nicolas Goaziou <n.goaziou@gmail.com> writes:

> It looks like the original function isn't right in more than one way
> (it doesn't even make sure the matched string is really a
> footnote). But that's another topic.

Yes, indeed.

> About your patch, I have but one question: I don't get a situation in
> which the assert would be triggered, may you give me such an example?

I don't either, and that's a good thing.  I added that just as some
extra paranoia so that you get an error if a footnote gets broken.

Bye,
Tassilo

Re: [PATCH] Fix renumbering for footnotes at BOL.

[Nicolas Goaziou]

Tassilo Horn <tassilo@member.fsf.org> writes:

> Nicolas Goaziou <n.goaziou@gmail.com> writes:

>> About your patch, I have but one question: I don't get a situation in
>> which the assert would be triggered, may you give me such an example?

> I don't either, and that's a good thing.  I added that just as some
> extra paranoia so that you get an error if a footnote gets broken.

Well, provided the first part of the function doesn't modify the buffer,
I still don't see how the assert could fail, even with a broken
footnote. Now, I may also be less paranoid than you are in that case.

Do you still think that assert should be a stayer? If so, I'll apply
your patch.

Regards,

-- 
Nicolas Goaziou

Re: [PATCH] Fix renumbering for footnotes at BOL.

[Tassilo Horn]

Nicolas Goaziou <n.goaziou@gmail.com> writes:

Hi!

>> I don't either, and that's a good thing.  I added that just as some
>> extra paranoia so that you get an error if a footnote gets broken.
>
> Well, provided the first part of the function doesn't modify the
> buffer, I still don't see how the assert could fail, even with a
> broken footnote.  Now, I may also be less paranoid than you are in
> that case.

You are right.  And the regex doesn't even match broken footnotes, that
is, footnotes with missing numbers.

> Do you still think that assert should be a stayer?

Nope, feel free to remove that line.

Bye,
Tassilo

Re: [PATCH] Fix renumbering for footnotes at BOL.

[Nicolas Goaziou]

Patch applied. Thank you.

Now, onto the other problems in that function...

Regards,

-- 
Nicolas Goaziou