I know this is kind of minor, but it's been bugging me. I'm using Org-mode for a project and I tend to export to either PDF or HTML rather frequently and it leaves my directory littered with PDF, Tex, and HTML files. Is there a way to have Org-mode export to another location, perhaps a subdirectory called ./exports?
In addition to the use of publishing by modifying your org-publish-project-alist variable as #user1248256 suggested, you can directly specify the org-export-publishing-directory variable within your file:
#+bind: org-export-publishing-directory "./exports"
* This is a test headline
Some text here. This should be exported to the "./exports" directory.
Upon export it will be placed in the "exports" directory, but only if that directory exists. If it does not exist, you will get an error message in the console.
The original question referred to exporting of org-files, while most answers above actually have to do with publishing, which is a different concept.
I believe the best way to solve the problem posed by the OP is to add the following to your emacs initialization file (.emacs):
(defadvice org-export-output-file-name (before org-add-export-dir activate)
"Modifies org-export to place exported files in a different directory"
(when (not pub-dir)
(setq pub-dir "exported-org-files")
(when (not (file-directory-p pub-dir))
(make-directory pub-dir))))
PS:
I realize a 5 year old question might no longer be relevant to the OP, but hopefully people searching for similar stuff will benefit from this answer.
This is a slight modification of a code snippet found in http://rwx.io/posts/org-export-configurations/
The original solution found in the above blog allows for setting up different directories for each exported format. However, if the goal is to avoid having
one's directory "littered with PDF, Tex, and HTML files", I think it is best to have only one directory containing exported files of all formats, which is the essence of the modification I offered above.
Edit: The emacs manual (https://www.gnu.org/software/emacs/manual/html_node/elisp/Porting-old-advice.html#Porting-old-advice) states that the defadvice mechanism was made obsolete by the new advice-add. So here is a code snipet with the same effect, using the recommended advice-add:
(defun org-export-output-file-name-modified (orig-fun extension &optional subtreep pub-dir)
(unless pub-dir
(setq pub-dir "exported-org-files")
(unless (file-directory-p pub-dir)
(make-directory pub-dir)))
(apply orig-fun extension subtreep pub-dir nil))
(advice-add 'org-export-output-file-name :around #'org-export-output-file-name-modified)
As before, this should be placed in your .emacs file.
This probably wasn't possible when the question was first asked, but the simplest solution would be to add the directory to the :EXPORT_FILE_NAME: property:
:PROPERTIES:
:EXPORT_FILE_NAME: exports/<filename>
:END:
Just as in the accepted answer, the directory must exist in order for this to work.
You have to put the following line at the beginning of your org file :
#+EXPORT_FILE_NAME: PATH/filename
Where PATH is the path to the folder where you want your file to be exported (e.g. ~/exports) and filename the name you want to give to your exported file (e.g. tutorial.html).
I believe you can get that with org-publish.
Add to you emacs configuration file something like that:
(setq org-publish-project-alist
'(("html"
:base-directory "~/org/"
:base-extension "org"
:publishing-directory "~/org/exports"
:publishing-function org-publish-org-to-html)
("pdf"
:base-directory "~/org/"
:base-extension "org"
:publishing-directory "~/org/exports"
:publishing-function org-publish-org-to-pdf)
("all" :components ("html" "pdf"))))
Eval this expression (or restart emacs), press C-c C-e X at org-mode, then choose a project from a list.
You can see more information at http://orgmode.org/worg/org-tutorials/org-publish-html-tutorial.html and http://orgmode.org/manual/Publishing.html#Publishing
As stated in the section "Export settings", we can use the EXPORT_FILE_NAME within a file in order to set the output directory. The quote shown below is the relevant part of the documentation
‘EXPORT_FILE_NAME’
The name of the output file to be generated. Otherwise, Org generates the file name based on the buffer name and the extension based on the back-end format.
Related
I would like to setup a personal wiki in org-mode.
So far, I have tried this two ways. My first attempt was using a single "Scientific Notebook.org" file. In this went all my notes - each with its own headline, date and tags (same format as a blog post). I then turned on org-velocity to quickly navigate the file.
But this created two problems: first, I use a lot of math in my notes (LaTeX previews are one of the reasons I want to us org). But these take sooooo long to load, I can't images trying to open a file with several thousand entries (all filled with math!!)
The other problem I have is with tags. I like to use a lot of multi-word tags to cross-reference my notes. But the way org-mode wraps these in the buffer, makes my headings completely unintelligible. Also (maybe it's just me but) I find CamelCase really hard to read, especially when faced with something like:
:monotonicTransformations:homogeneousFunctions:orderedSets:proofs:
Now my second attempt is with Deft. Here, I have broken up each note into its own .org file and creating a dedicated ~/org/ folder to hold my wiki. But this is where I am stuck:
1) How do you setup auto linking, so that typing say "foo bar" in one note, creates a link to "foo bar.org"? Can this be done with radio target? If not, can this syntax [[foo bar]] be overridden to search for headlines in all files in the ~/org/ directory? I tried adding Wiki.el as a minor mode but no dice...
2) How do you tag individual files? And how can you then pull up a list of all tags and use this to filter your list of notes? I have read that bookmark+ lets you do file tagging. But I got so lost in the online docs...
I would love to hear how others have solved these problem, what minor modes you are using, workflows and keyboard shortcuts or other mods!
Thanks!
-Adam
I'm using a simplistic code for a wiki.
Because that's what a wiki is for me: a quick way to categorize things.
The structure is a following:
each subject has its own org file
each topic of subject has its own heading
all org files are in single directory
you can jump to file, or create a new file with helm
That's it. I've found to need to link anything to anything, jump-to-subject functionality
is more than enough. Here's how this looks:
And once within a subject, I can jump across topics with worf.
Here's how this looks:
Having only one note file is, in my opinion, more flexible and compatible. It tends, however, to get slow for even moderatly sized files. Multiple small files are quick, but they need more effort to set up, and they only work within that setup.
Single File Solution
To speed things up, consider setting org-startup-with-latex-preview to nil (or add #+STARTUP: nolatexpreview to your file).
Tags not only get messy when used for keywords, using them also gets rather slow as your file grows. I've played around with some custom functions, but now avoid tags most of the time. Instead I use flat hierarchies, categories and otherwise rely on occur and org-occur (e.g. M-x org-occur begin_proof).
Multiple Files
The org syntax for linking to other files is rather simple: [[./this idea.org][this idea]]. If that is too much hassle, it should be easy to write a function that replaces the active region with an appropriate link.
If you want to link [[this idea]] to a file "this idea.org", you could add a function to org-open-at-point-functions and handle it yourself.
As for tags, you don't tag a file itself, but rather a single top level headline. This of course means that all your troubles with tags a back as well.
Again, I would recommend not using tags. Just make sure the file contains the right keywords at the right places and use occur and friends.
Edit: An Example `org-open-at-point-function'
If you want to search for a fuzzy link in all files in a directory instead of only the current buffer, you can do this by using the org-open-at-point-functions hook. Here is an example:
(defvar my-link-search-directory "/my/notes/directory/")
(defun my-open-link-function ()
"Open link, interpreting it a the name of a headline."
(let* ((el (org-element-context))
(type (first el))
(link-type (plist-get (cadr el) :type))
(path (let ((path-1 (plist-get (cadr el) :path)))
(when (stringp path-1)
(org-link-unescape path-1)))))
(when (and (eql type 'link)
path
(string= link-type "fuzzy"))
(let* ((path (regexp-quote path))
(result
(delq nil
(org-map-entries
(lambda ()
(when (string-match
path
(org-get-heading))
(list (buffer-file-name) (point))))
nil
;; Here we set the scope.
;; 'agenda would search in all agenda files.
;; We want a list of all org files in `my-link-search-directory'.
(directory-files
my-link-search-directory
t "[.]org\\'")))))
(when result
(when (> (length result) 1)
(message "Warning: multiple search results for %s" path))
(let ((file (caar result))
(pos (cadar result)))
(find-file file)
(goto-char pos)))))))
(add-hook
'org-open-at-point-functions
'my-open-link-function)
Note that I haven't tested this much.
Actually, I would recommend against using this unless you really need it. While making fancy extensions is tempting, keeping your notes as simple as possible is preferably. If you have everything in one file, you could edit your notes with notepad or google docs or whatever, should you ever need to.
Currently, I'm able to redirect Emacs backup files (those ending in '~' and those starting in '#') to a central location using the code below (it renames them nicely also, encoding the file path in the file name (e.g., ~/tmp/emacs_autosaves/#!home!cbalz!.bashrc#).
But how to preserve the functionality in that code, while also moving the files that Emacs makes that start with '.#' (those are often or always symlinks)?
Current working code - needs to be modified or augmented to work with files/symlinks starting with '.#' :
(defvar autosave-dir
(concat "~/tmp/emacs_autosaves/" (user-login-name) "/"))
(make-directory autosave-dir t)
(setq auto-save-file-name-transforms `(("\\(?:[^/]*/\\)*\\(.*\\)" ,(concat
autosave-dir "\\1") t))
The dot-hash-files are created by Emacs as a lock to avoid concurrent modification of a file. Unfortunately, the documentation does not mention the possibility of changing the default location of those files:
The file lock is really a file, a symbolic link with a special name,
stored in the same directory as the file you are editing.
I recently found org-annotate-file. I would like to use it to annotate pdf documents or music files or any other files on my computer, and write my annotations in a file annotations.org. I am not looking to include annotations IN the pdf. But what I cannot figure out is what it means to "visit a file"? Does it have to be a file that emacs can open?
But more generally, is there a package that can do something like this: I visit a directory in dired mode, mark a bunch of files on some topic of my interest, and with one command I send links to the files to my annotations.org file (maybe as subheadings under a heading, which may be the directory name), and then I can write the annotations in the annotations file. Then with one command, I should be able to reach any of the files (which org-mode will allow) or open it in an external program. Is this possible in some package?
Thanks.
Of course, it can be done. However, it seems the actual code of org-annotate-file.el, that I found
here, doesn't seem to accept annotating a file that has not been opened (visited means here opened), because the function to annotate uses the current open file as a source for the name. The current implementation of org-annotate-file is this:
(defun org-annotate-file ()
"Put a section for the current file into your annotation file"
(interactive)
(error-if-no-file)
(org-annotate-file-show-section))
At least you could modify it to accept an arbitrary file (if you provide it):
(defun org-annotate-file (&optional filename)
"Put a section for the current file into your annotation file"
(interactive "FFile to tag: ")
; if a file is specified, bypass the check for error when no file
(if filename
(org-annotate-file-show-section filename)
(progn
(error-if-no-file)
(org-annotate-file-show-section))))
This ask you for a file name whenever you do M-xorg-annotate-file.
You also have to change the org-annotate-file-show-section to accept either a file name or a buffer. The first let should be like this:
(defun org-annotate-file-show-section (&optional buffer-or-file)
"Visit the buffer named `org-annotate-file-storage-file' and
show the relevant section"
(let* ((line (buffer-substring-no-properties (point-at-bol) (point-at-eol)))
(filename (if (stringp buffer-or-file)
buffer-or-file
(get-filename buffer-or-file (buffer-file-name))))
(link (get-link filename))
(search-link (org-make-link-string
(concat "file:" filename "::" line)
(org-annotate-file-prettyfy-desc line))))
(show-annotations filename link)
.... rest of the code....
dired integration can be started from here, but I'm still not familiar with the dired API...
EDIT: I'm creating a branch in bitbucket for that modifications. I find the utility very useful and might use it myself. I'll post the link here. And here it is: https://bitbucket.org/dsevilla/org-annotate-file/src
BACKGROUND: In org-mode, the variable org-archive-location is set to "%s_archive::" by default, so that a file "toto.org" archives into a file "toto.org_archive". I would like it to archive to "toto.ref" instead. I am using org-mode version 7.4 (out of the git server).
I would have thought it to be as simple as
(setq org-archive-location
`(replace-regexp-in-string ".org" ".ref" %s)
)
But I was pointed out that this was not proper in LISP (plus, it did not work). My final solution is as follow, you should be able to adapt to most clever configurations of org-archive-location:
(setq org-archive-location "%s::* ARCHIVES")
(defadvice org-extract-archive-file (after org-to-ref activate)
(setq ad-return-value
(replace-regexp-in-string "\\.org" ".ref" ad-return-value)
)
)
Note that:
1) I voluntarily did not add a $ at the end of ".org" so that it would properly alter "test.org.gpg" into "test.ref.gpg".
2) It seems that one should use the regular expression "\.org" (rather than, say, ".org") (longer explanation below in the answers).
You can't define a variable in Emacs such that its value is obtained by running code; variables have simple, static values.
You can achieve the effect you described by advising the function org-extract-archive-file, which is the one that generates an archive location from org-archive-location:
(defadvice org-extract-archive-file (after org-to-ref activate)
(setq ad-return-value
(replace-regexp-in-string "\\.org" ".ref" ad-return-value)))
This works for me now, but of course the internals of org-mode are subject to change and this solution may not work forever.
You should not quote an expression that you want to evaluate. Note also that in a regular expression, . matches any character.
Here is an example of how to set the file, the location (e.g., main heading) in the file, and whether or not to include additional archive information:
(let* (
(org-archive-location "~/todo.org::* TASKS")
(org-archive-save-context-info nil))
...)
You can try this: #+ARCHIVE: %s.ref:: at the beginning of your org file.
Read more about it here.
Also, other interesting option is to set inside your headtree the following, for instance:
* Main Header of tree
:PROPERTIES:
:ARCHIVE: toto.ref:: * Main Header of tree in archive file
:END:
** sub tree of main header and so on
The latter I took from this video.
I would like to set up a project to be published as HTML using org-mode.
I don't want to litter my .emacs with project definitions, and I was wondering where I could put the (setq org-publish-project-alist) variable.
Can I somehow put it in the same dir?
Ryan McGeary describes what I think is a good way to organize emacs startup files.
Update:
The domain emacsblog.org expired :(
You can look at the cached copy of the originally linked page.
You could just add a new file in your .emacs.d (or whereever) and do a load-file in your .emacs file.
-- EDIT --
For example, you could have the following in your .emacs
(load (expand-file-name "~/.emacs.d/lisp/personal-org-mode-stuff.el"))
and then put all of your customization stuff in ~/.emacs.d/lisp/personal-org-mode-stuff.el and it will load that file and import all of your .emacs
Another poster also posted a link to a description of how to add your lisp files to the load path and require them.
If you don't set it manually at all, but rather use Emacs' customize mechanism to control the value of this variable, your .emacs file will not be cluttered if you add the following two lines to your .emacs:
(setq custom-file "~/.emacs-custom.el")
(load custom-file 'noerror)
Although some might claim that it's kind of ironic that you have to add two lines two your .emacs file to declutter it that way...