specifying (at the minimum) a title, a permanent link and the content
of the entry. Text-only, HTML and XHTML entries are supported.
-The code for this library is hosted at http://code.fperrin.net/atom.git.
+The code for this library is hosted at http://code.tar-jx.bz/atom.git;
+this manual can be found at http://tar-jx.bz/code/atom.html.
* Installation
The feed is created with =atom-create=, giving it a title and a Web
address at the minimum. Entries may then be added one by one with
-=atom-add-{text,html,xhtml}-entry=.
+=atom-add-{text, html, xhtml}-entry=.
A typical usage would look like this:
(atom-print my-atom-feed))
#+END_SRC
+See the docstrings for the methods above for more details.
+
* Additionnal notes
+** If what you want to do is not possible here
+
The =my-atom-feed= object in the example above is really only an XML
tree as defined by the =xml.el= package. This means you can manipulate
-it, as long as you are careful not to mess up the XML structure. For
+it, as long as you are careful when manipulating the XML structure. For
instance, if you want to add somebody as a contributor to an entry,
-you could say the following:
+and also add an =lang= attribute, you could say the following:
#+BEGIN_SRC elisp
(let ((entry (atom-add-html-entry my-atom-feed
"http://example.org/witty"
"<p>This is <i>clever</i>, isn't it?")))
(atom-modify-entry entry 'contributor
- (atom-massage-author '("John Clever" "jc@example.net"))))
+ (atom-massage-author '("John Clever" "jc@example.net")))
+ (let* ((content (assoc 'content entry))
+ (attrs (xml-node-attributes entry)))
+ (setcar (cdr entry) (cons '(lang . "en") attrs))))
#+END_SRC
+** Conformingness of produced feeds
+
As of now, the library doesn't check whether there are two entries
with the same =id= value (which is illegal), or with the same
=updated= value (which reportedly confuse some readers).
+
+The encoding of the resulting feed is hard-coded to UTF-8.
+
+** Outputting RSS feeds
+
+Use =atom-to-rss-print= and =atom-to-rss-write-file=.
+
+Producing RSS from Atom feeds is not optimal. In particular :
+
+- the =updated= and the =pubDate= in the two standards don't seem to
+ have the same semantics (last meaningfull change VS publication of
+ the entry) ;
+
+- the =description= of the channel is mandatory in RSS. The value for
+ this element is taken from the =subtitle= element of an Atom feed,
+ which is optional, so this library may produce non conforming RSS
+ feeds.
+
+** XHTML entries
+
+According to the w3c, relative links in an Atom feed can confuse feed
+readers. As a result, this library's default behaviour is to translate
+all addresses in the =href= attribute of =a= elements and =src= of
+=img= to absolute links. This can be disabled by setting NOCONVERT to
+t when calling =atom-add-xhtml-entry=.
+
+In the =pre= element, whitespace is significant. However,
+=xml-parse-region= then =xml-print= will add spaces and
+identation. This is not something that can be fixed from =atom=.
+
+If you already have ypur XHTML content in Lisp format (as opposed to
+simply a long string), you can pass it directly, as in:
+
+#+BEGIN_SRC elisp
+ (atom-add-xhtml-entry
+ my-atom-feed
+ "An XHTML example"
+ "http://example.org/emacs-haiku"
+ '((h1 nil "Emacs Haiku")
+ (p nil "The friends chat gaily," (br)
+ "I stand up to join their talk." (br)
+ "My save-excursion." (br))
+ (p ((class . "author-name")) nil "Oliver Scholz")))
+#+END_SRC
+
+This will save a call to =xml-parse-region=.
+
+* License
+
+=atom.el= ---An elisp library for creating Atom feeds.
+Copyright (C) 2011 Frédéric Perrin.
+
+This program is free software: you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation, either version 3 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+GNU General Public License for more details.
+
+The full text of the GNU General Public License can be found at the
+following address: <http://www.gnu.org/licenses/gpl-3.0.txt>
;; permanent link and the content of the entry. Text-only, HTML and
;; XHTML entries are supported.
-;; A feed is really a Lisp structure as used by the `xml.el' package,
-;; without the parent `feed' element.
+;; It is possible to produce both Atom and RSS feeds.
;; A typical usage would look like this:
;; "http://example.org/hello"
;; "Hello the world!")
;;
-;; ; A text-only entry, with all the optional pieces of data
-;; (atom-add-text-entry
-;; my-atom-feed
-;; "Bonjour"
-;; "http://example.org/bonjour"
-;; "Bonjour à tout le monde !"
-;; ;; optional: the last modification time
-;; (date-to-time "2011-01-30 23:40:12")
-;; ;; optional: an identifier for this entry; a common way to generate it is
-;; ;; to use the domain name and the creation date of the entry.
-;; (atom-generate-id "http://example.org"
-;; (date-to-time "2011-01-30 10:01:05"))
-;; ;; optional: a summary for this entry
-;; "Bonjour, monde.")
-;;
;; (atom-add-xhtml-entry
;; my-atom-feed
;; "An XHTML example"
;; "http://example.org/html-example"
;; "<p>One can also use <acronym>XHTML</acronym> in the entries.</p>")
-;; (atom-print my-atom-feed))
+;;
+;; (atom-print my-atom-feed)
+;; ;; If you prefer RSS feeds:
+;; (atom-to-rss-print my-atom-feed))
+
+;; Full documentation is available at <http://tar-jx.bz/code/atom.html>.
;;; Code:
(setcar (cdr guid) (list (cons 'isPermaLink "false"))))
(if (and descr
(equal (xml-get-attribute descr 'type) "xhtml"))
- (setcar (cddr descr) (xml-node-text descr))))
+ (setcar (cddr descr) (xml-node-as-text descr))))
`(item nil ,@item)))
(defun atom-to-rss-translator (source target translations)
(when data
(atom-modify-entry target to data)))))
-(defun xml-node-text (node)
- (with-temp-buffer
- (xml-print (xml-node-children node))
- (buffer-string)))
-
(defun atom-to-rss-modify-link (entry)
(let* ((link (assoc 'link entry))
(link-addr (xml-get-attribute-or-nil link 'href)))
"Return an XML node representing the author. AUTHOR can be:
- nil, in which case `user-full-name' and `user-mail-address' are
used;
-- a single string, the full name of the author;
+- a single string, the full name of the author; no email address
+ will be included;
- a list with two elements, the full name and the email address
of the author;
- something else, assumed to be a complete `atomPersonConstruct'."
(dolist (child (xml-node-children node))
(when (listp child) (atom-xhtml-convert-links child base))))
+(defun atom-generate-id (link creation-date)
+ "Generate a string suitable for use as an atom:id element. This
+implements Mark Pilgrom's tag: URI method, using the
+CREATION-DATE of the entry, and the domain part of LINK."
+ (format "tag:%s,%s:/%s"
+ (url-host (url-generic-parse-url link))
+ (format-time-string "%Y-%m-%d" creation-date)
+ (format-time-string "%Y%m%d%H%M%S" creation-date)))
+
+\f
+;;; Functions that should probably not be there
+
(defun url-canonalize (address base)
"Make ADRESS an absolute URL, taking it in the BASE context."
;; I feel such a function should exist in `url-parse'. Did I miss it?
(file-name-directory (url-filename url-base))))
(url-recreate-url url-base))))
-(defun atom-generate-id (link creation-date)
- "Generate a string suitable for use as an atom:id element. This
-implements Mark Pilgrom's tag: URI method, using the
-CREATION-DATE of the entry, and the domain part of LINK."
- (format "tag:%s,%s:/%s"
- (url-host (url-generic-parse-url link))
- (format-time-string "%Y-%m-%d" creation-date)
- (format-time-string "%Y%m%d%H%M%S" creation-date)))
+(defun xml-node-as-text (node)
+ "Return a string representing NODEn, an XML structure."
+ (with-temp-buffer
+ (xml-print (xml-node-children node))
+ (buffer-string)))
(provide 'atom)
;;; atom.el ends here