Skip Menu |

This queue is for tickets about the Sepia CPAN distribution.

Report information
The Basics
Id: 73721
Status: open
Priority: 0/
Queue: Sepia

People
Owner: Nobody in particular
Requestors: user42 [...] zip.com.au
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)



Subject: sepia.el following to L</Foo> link
Date: Wed, 04 Jan 2012 11:20:52 +1100
To: bug-Sepia [...] rt.cpan.org
From: Kevin Ryde <user42 [...] zip.com.au>
In a foo.pod buffer containing =head2 Foo L</Foo> and point on the L<, M-x sepia-pod-follow-link-at-point gives an error Can't find anchor for /Foo. It looks like it's using `search-forward' where re-search-forward may be intended, and there's also a `match-beginning' without a "0" arg. I get some joy from changing (unless (search-forward (concat "^=.*" sec) nil t) (error "Can't find anchor for %s." str)) (setq target (match-beginning))) to (unless (re-search-forward (concat "^=.*" (regexp-quote sec)) nil t) (error "Can't find anchor for %s." str)) (setq target (match-beginning 0))) Maybe the L<name/sec> part would be something similar too. I wondered if the search should let-bind case-fold-search. Case-insensitive might be friendly. When writing pod you'd probably prefer to have the case in the link the same as the target, but some of the pod converters etc might be forgiving. -- The sigfile English idioms series: "Putting all your eggs in one basket" -- protecting your assets by putting them in one place you can watch carefully.
On Tue Jan 03 19:20:58 2012, user42@zip.com.au wrote: Show quoted text
> In a foo.pod buffer containing > > =head2 Foo > > L</Foo> > > and point on the L<, M-x sepia-pod-follow-link-at-point gives an error > > Can't find anchor for /Foo. > > It looks like it's using `search-forward' where re-search-forward may be > intended, and there's also a `match-beginning' without a "0" arg.
I think you're correct -- at least, making your suggested changes Works for Me. Show quoted text
> I wondered if the search should let-bind case-fold-search. > Case-insensitive might be friendly. When writing pod you'd probably > prefer to have the case in the link the same as the target, but some of > the pod converters etc might be forgiving.
Right -- case-insensitive when following a link, and case-sensitive if I ever add link completion (possibly very useful with long headings). Anyone who has multiple headings differing only by case probably deserves to lose. Anyways, I believe this is fixed (along with another bug identifying links at EOF. I pushed the changes latest to repo.or.cz
Subject: Re: [rt.cpan.org #73721] sepia.el following to L</Foo> link
Date: Fri, 13 Jan 2012 08:57:23 +1100
To: bug-sepia [...] rt.cpan.org
From: Kevin Ryde <user42 [...] zip.com.au>
"Sean O'Rourke via RT" <bug-sepia@rt.cpan.org> writes: Show quoted text
> > and case-sensitive if I ever > add link completion (possibly very useful with long headings).
Yep. Or let the user's completion-ignore-case setting take effect, if that was close enough to a sensible default. pod-mode.el might end up that way in its pod-link-section (though it uses icomplete or something by default). Either way yes I couldn't see much merit in there ever being two headings of different case.