Re: [O] Making the output of %% functions linkable in agenda views
On 7 Apr 2011, Charles Sebold wrote: Here's a test. Take the line below and put it in an agenda file, then view the agenda. %%(format %s (concat [[elisp:(info)] [Link to info]])) The results as I see them give me a link that I can mouse-click on, but I can't move point to it and hit C-c C-o. What do I need to do to make this a normal org link in every way? Or is this a bug? (Obviously this is just a demo problem, I'm doing something much more hideously complicated in real life, but this was good enough to share to show what the problem was.) *bump* Anybody care to take this on? -- Charles Sebold 3rd of May, 2011 GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600) | No Gnus v0.18 | org-mode 7.5
Re: [O] Making the output of %% functions linkable in agenda views
charles.seb...@lcms.org wrote: On 7 Apr 2011, Charles Sebold wrote: Here's a test. Take the line below and put it in an agenda file, then view the agenda. %%(format %s (concat [[elisp:(info)] [Link to info]])) The results as I see them give me a link that I can mouse-click on, but I can't move point to it and hit C-c C-o. What do I need to do to make this a normal org link in every way? Or is this a bug? (Obviously this is just a demo problem, I'm doing something much more hideously complicated in real life, but this was good enough to share to show what the problem was.) Well, this feels like a bug: when I C-c C-o on it, it says No links, but if I press RET to go to the agenda file and then C-x k to kill the buffer, and then do C-c C-o again I get the Execute (info) as elisp? prompt (and if I say yes, it does indeed open Info). So there is something funky going on, but I'm not sure what. Nick
Re: [O] Making the output of %% functions linkable in agenda views
charles.seb...@lcms.org writes: On 7 Apr 2011, Charles Sebold wrote: Here's a test. Take the line below and put it in an agenda file, then view the agenda. %%(format %s (concat [[elisp:(info)] [Link to info]])) The results as I see them give me a link that I can mouse-click on, but I can't move point to it and hit C-c C-o. What do I need to do to make this a normal org link in every way? Or is this a bug? Here's what seems to be going on. Clicking with the mouse works because it calls org-open-at-mouse, which (by calling org-open-at-point) simply looks at the text surrounding the click point to find a link. The function org-agenda-open-link, on the other hand, grabs part of the headline as a string (the prefix), then jumps to the location in the original buffer and calls org-offer-links-in-entry to look for links in both the entry and the prefix of agenda text. Obviously, it won't find your link in the original entry since it looks like this: %%(format %s (concat [[elisp:(info)] [Link to info]])) I think it shouldn't be a problem to pass the whole agenda line to org-offer-links-in-entry. This will not result in duplicate links, since org-offer-links-in-entry uniquifies the links. I've attached a patch. The bigger question, however, is why the expression above is even showing up in the agenda, since it contains no scheduling information. :) Best, Matt From 58c7621c0c84b9c1930098a098e4559aa516eec0 Mon Sep 17 00:00:00 2001 From: Matt Lundin m...@imapmail.org Date: Tue, 3 May 2011 15:54:35 -0400 Subject: [PATCH] Pass entire text of agenda line to org-offer-links-in-entry. * lisp/org-agenda.el (org-agenda-open-link): Pass entire text of agenda line to org-offer-links-in-entry. This fixes bug noticed by Charles Sebold, in which links that are dynamically formatted for the agenda view are ignored. http://permalink.gmane.org/gmane.emacs.orgmode/40673 --- lisp/org-agenda.el |4 +--- 1 files changed, 1 insertions(+), 3 deletions(-) diff --git a/lisp/org-agenda.el b/lisp/org-agenda.el index 65f7cea..609edd4 100644 --- a/lisp/org-agenda.el +++ b/lisp/org-agenda.el @@ -6879,9 +6879,7 @@ at the text of the entry itself. (org-get-at-bol 'org-marker))) (buffer (and marker (marker-buffer marker))) (prefix (buffer-substring - (point-at-bol) - (or (text-property-any (point-at-bol) (point-at-eol) 'org-heading t) - (point-at-bol) + (point-at-bol) (point-at-eol (cond (buffer (with-current-buffer buffer -- 1.7.5
Re: [O] Making the output of %% functions linkable in agenda views
On 3 May 2011, Matt Lundin wrote: The bigger question, however, is why the expression above is even showing up in the agenda, since it contains no scheduling information. :) Hm. I didn't think about whether anybody else could see it, come to think of it; my real problem (for which this was just a mock-up) shows up in a heavily customized agenda I set up, which isn't schedule-based. -- Charles Sebold 3rd of May, 2011 GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600) No Gnus v0.18 | Org-mode version 7.5 (release_7.5.246.ga35b4)
Re: [O] Making the output of %% functions linkable in agenda views
charles.seb...@lcms.org (Sebold, Charles) writes: On 3 May 2011, Matt Lundin wrote: The bigger question, however, is why the expression above is even showing up in the agenda, since it contains no scheduling information. :) Hm. I didn't think about whether anybody else could see it, come to think of it; my real problem (for which this was just a mock-up) shows up in a heavily customized agenda I set up, which isn't schedule-based. I should clarify. I, too, can see the link in the agenda. I had been under the impression that only diary sexps will appear in the agenda, but it seems that a generic sexp that returns a string will also show up. I'm not sure whether this is a feature or a bug. Best, Matt
Re: [O] Making the output of %% functions linkable in agenda views
On 3 May 2011, Matt Lundin wrote: I should clarify. I, too, can see the link in the agenda. I had been under the impression that only diary sexps will appear in the agenda, but it seems that a generic sexp that returns a string will also show up. I'm not sure whether this is a feature or a bug. OK, I see now. I built this based on the behavior of google-weather-el, which includes a function for inclusion in agendas. I think it's a feature; in my (real-world) case, my function runs based on the value of the local variable `date' (apparently set when the agenda is built) and provides a correspondingly different result for every day's agenda. That is legitimately useful. Google-weather-el works the same way. See: (info (org)Weekly/daily agenda) ...for examples of org-mode itself using this (e.g., for org-calendar-holiday, org-bbdb-anniversary, etc.). -- Charles Sebold 3rd of May, 2011 GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600) No Gnus v0.18 | Org-mode version 7.5 (release_7.5.246.ga35b4)
Re: [O] Making the output of %% functions linkable in agenda views
Matt Lundin m...@imapmail.org wrote: I should clarify. I, too, can see the link in the agenda. I had been under the impression that only diary sexps will appear in the agenda, but it seems that a generic sexp that returns a string will also show up. I'm not sure whether this is a feature or a bug. I *think* it's a feature but I'm really not sure: the regexp is just ^?%%( so it does not seem to care what comes after it. Various things (like the sunrise/sunset hacks) use it, so if it is deemed to be a bug-to-be-fixed, they will break when it *is* fixed :-) So is a diary sexp just something that matches ^?%%( *diary ? Nick
Re: [O] Making the output of %% functions linkable in agenda views
Nick Dokos nicholas.do...@hp.com writes: Matt Lundin m...@imapmail.org wrote: I should clarify. I, too, can see the link in the agenda. I had been under the impression that only diary sexps will appear in the agenda, but it seems that a generic sexp that returns a string will also show up. I'm not sure whether this is a feature or a bug. I *think* it's a feature but I'm really not sure: the regexp is just ^?%%( so it does not seem to care what comes after it. Various things (like the sunrise/sunset hacks) use it, so if it is deemed to be a bug-to-be-fixed, they will break when it *is* fixed :-) Thanks! Yes this indeed a feature. I just wanted to make sure I wasn't sending a patch that enhanced a bug. :) Best, Matt
