1 ;;; org-feed.el --- Add RSS feed items to Org files -*- lexical-binding: t; -*-
3 ;; Copyright (C) 2009-2017 Free Software Foundation, Inc.
5 ;; Author: Carsten Dominik <carsten at orgmode dot org>
6 ;; Keywords: outlines, hypermedia, calendar, wp
7 ;; Homepage: http://orgmode.org
9 ;; This file is part of GNU Emacs.
11 ;; GNU Emacs is free software: you can redistribute it and/or modify
12 ;; it under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation, either version 3 of the License, or
14 ;; (at your option) any later version.
16 ;; GNU Emacs is distributed in the hope that it will be useful,
17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 ;; GNU General Public License for more details.
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
23 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
27 ;; This module allows entries to be created and changed in an Org mode
28 ;; file triggered by items in an RSS feed. The basic functionality
29 ;; is geared toward simply adding new items found in a feed as
30 ;; outline nodes to an Org file. Using hooks, arbitrary actions can
31 ;; be triggered for new or changed items.
33 ;; Selecting feeds and target locations
34 ;; ------------------------------------
36 ;; This module is configured through a single variable, `org-feed-alist'.
37 ;; Here is an example, using a notes/tasks feed from reQall.com.
39 ;; (setq org-feed-alist
41 ;; "http://www.reqall.com/user/feeds/rss/a1b2c3....."
42 ;; "~/org/feeds.org" "ReQall Entries")
44 ;; With this setup, the command `M-x org-feed-update-all' will
45 ;; collect new entries in the feed at the given URL and create
46 ;; entries as subheadings under the "ReQall Entries" heading in the
47 ;; file "~/org/feeds.org". Each feed should normally have its own
48 ;; heading - however see the `:drawer' parameter.
50 ;; Besides these standard elements that need to be specified for each
51 ;; feed, keyword-value pairs can set additional options. For example,
52 ;; to de-select transitional entries with a title containing
54 ;; "reQall is typing what you said",
56 ;; you could use the `:filter' argument:
58 ;; (setq org-feed-alist
60 ;; "http://www.reqall.com/user/feeds/rss/a1b2c3....."
61 ;; "~/org/feeds.org" "ReQall Entries"
62 ;; :filter my-reqall-filter)))
64 ;; (defun my-reqall-filter (e)
65 ;; (if (string-match "reQall is typing what you said"
66 ;; (plist-get e :title))
70 ;; See the docstring for `org-feed-alist' for more details.
73 ;; Keeping track of previously added entries
74 ;; -----------------------------------------
76 ;; Since Org allows you to delete, archive, or move outline nodes,
77 ;; org-feed.el needs to keep track of which feed items have been handled
78 ;; before, so that they will not be handled again. For this, org-feed.el
79 ;; stores information in a special drawer, FEEDSTATUS, under the heading
80 ;; that received the input of the feed.
86 ;; org-feed.el is based on ideas by Brad Bozarth who implemented a
87 ;; similar mechanism using shell and awk scripts.
94 (declare-function url-retrieve-synchronously
"url"
95 (url &optional silent inhibit-cookies timeout
))
96 (declare-function xml-node-children
"xml" (node))
97 (declare-function xml-get-children
"xml" (node child-name
))
98 (declare-function xml-get-attribute
"xml" (node attribute
))
99 (declare-function xml-get-attribute-or-nil
"xml" (node attribute
))
100 (declare-function xml-substitute-special
"xml" (string))
102 (declare-function org-capture-escaped-%
"org-capture" ())
103 (declare-function org-capture-expand-embedded-elisp
"org-capture" (&optional mark
))
104 (declare-function org-capture-inside-embedded-elisp-p
"org-capture" ())
106 (defgroup org-feed nil
107 "Options concerning RSS feeds as inputs for Org files."
111 (defcustom org-feed-alist nil
112 "Alist specifying RSS feeds that should create inputs for Org.
113 Each entry in this list specified an RSS feed tat should be queried
114 to create inbox items in Org. Each entry is a list with the following items:
116 name a custom name for this feed
118 file the target Org file where entries should be listed, when
119 nil the target becomes the current buffer (may be an
120 indirect buffer) each time the feed update is invoked
121 headline the headline under which entries should be listed
123 Additional arguments can be given using keyword-value pairs. Many of these
124 specify functions that receive one or a list of \"entries\" as their single
125 argument. An entry is a property list that describes a feed item. The
126 property list has properties for each field in the item, for example `:title'
127 for the `<title>' field and `:pubDate' for the publication date. In addition,
128 it contains the following properties:
130 `:item-full-text' the full text in the <item> tag
131 `:guid-permalink' t when the guid property is a permalink
133 Here are the keyword-value pair allows in `org-feed-alist'.
136 The name of the drawer for storing feed information. The default is
137 \"FEEDSTATUS\". Using different drawers for different feeds allows
138 several feeds to target the same inbox heading.
140 :filter filter-function
141 A function to select interesting entries in the feed. It gets a single
142 entry as parameter. It should return the entry if it is relevant, or
145 :template template-string
146 The default action on new items in the feed is to add them as children
147 under the headline for the feed. The template describes how the entry
148 should be formatted. If not given, it defaults to
149 `org-feed-default-template'.
151 :formatter formatter-function
152 Instead of relying on a template, you may specify a function to format
153 the outline node to be inserted as a child. This function gets passed
154 a property list describing a single feed item, and it should return a
155 string that is a properly formatted Org outline node of level 1.
157 :new-handler function
158 If adding new items as children to the outline is not what you want
159 to do with new items, define a handler function that is called with
160 a list of all new items in the feed, each one represented as a property
161 list. The handler should do what needs to be done, and org-feed will
162 mark all items given to this handler as \"handled\", i.e. they will not
163 be passed to this handler again in future readings of the feed.
164 When the handler is called, point will be at the feed headline.
166 :changed-handler function
167 This function gets passed a list of all entries that have been
168 handled before, but are now still in the feed and have *changed*
169 since last handled (as evidenced by a different sha1 hash).
170 When the handler is called, point will be at the feed headline.
173 This function gets passed a buffer, and should return a list
174 of entries, each being a property list containing the
175 `:guid' and `:item-full-text' keys. The default is
176 `org-feed-parse-rss-feed'; `org-feed-parse-atom-feed' is an
179 :parse-entry function
180 This function gets passed an entry as returned by the parse-feed
181 function, and should return the entry with interesting properties added.
182 The default is `org-feed-parse-rss-entry'; `org-feed-parse-atom-entry'
186 (list :value
("" "http://" "" "")
188 (string :tag
"Feed URL")
189 (file :tag
"File for inbox")
190 (string :tag
"Headline for inbox")
193 (list :inline t
:tag
"Filter"
195 (symbol :tag
"Filter Function"))
196 (list :inline t
:tag
"Template"
198 (string :tag
"Template"))
199 (list :inline t
:tag
"Formatter"
201 (symbol :tag
"Formatter Function"))
202 (list :inline t
:tag
"New items handler"
204 (symbol :tag
"Handler Function"))
205 (list :inline t
:tag
"Changed items"
206 (const :changed-handler
)
207 (symbol :tag
"Handler Function"))
208 (list :inline t
:tag
"Parse Feed"
210 (symbol :tag
"Parse Feed Function"))
211 (list :inline t
:tag
"Parse Entry"
213 (symbol :tag
"Parse Entry Function"))
216 (defcustom org-feed-drawer
"FEEDSTATUS"
217 "The name of the drawer for feed status information.
218 Each feed may also specify its own drawer name using the `:drawer'
219 parameter in `org-feed-alist'."
221 :type
'(string :tag
"Drawer Name"))
223 (defcustom org-feed-default-template
"\n* %h\n %U\n %description\n %a\n"
224 "Template for the Org node created from RSS feed items.
225 This is just the default, each feed can specify its own.
226 Any fields from the feed item can be interpolated into the template with
227 %name, for example %title, %description, %pubDate etc. In addition, the
228 following special escapes are valid as well:
230 %h The title, or the first line of the description
231 %t The date as a stamp, either from <pubDate> (if present), or
234 %u,%U Like %t,%T, but inactive time stamps
235 %a A link, from <guid> if that is a permalink, else from <link>
236 %(sexp) Evaluate elisp `(sexp)' and replace with the result, the simple
237 %-escapes above can be used as arguments, e.g. %(capitalize \\\"%h\\\")"
239 :type
'(string :tag
"Template"))
241 (defcustom org-feed-save-after-adding t
242 "Non-nil means save buffer after adding new feed items."
246 (defcustom org-feed-retrieve-method
'url-retrieve-synchronously
247 "The method to be used to retrieve a feed URL.
248 This can be `curl' or `wget' to call these external programs, or it can be
249 an Emacs Lisp function that will return a buffer containing the content
250 of the file pointed to by the URL."
253 (const :tag
"Internally with url.el" url-retrieve-synchronously
)
254 (const :tag
"Externally with curl" curl
)
255 (const :tag
"Externally with wget" wget
)
256 (function :tag
"Function")))
258 (defcustom org-feed-before-adding-hook nil
259 "Hook that is run before adding new feed items to a file.
260 You might want to commit the file in its current state to version control,
265 (defcustom org-feed-after-adding-hook nil
266 "Hook that is run after new items have been added to a file.
267 Depending on `org-feed-save-after-adding', the buffer will already
272 (defvar org-feed-buffer
"*Org feed*"
273 "The buffer used to retrieve a feed.")
276 (defun org-feed-update-all ()
277 "Get inbox items from all feeds in `org-feed-alist'."
279 (let ((nfeeds (length org-feed-alist
))
280 (nnew (apply '+ (mapcar 'org-feed-update org-feed-alist
))))
281 (message "%s from %d %s"
282 (cond ((= nnew
0) "No new entries")
283 ((= nnew
1) "1 new entry")
284 (t (format "%d new entries" nnew
)))
286 (if (= nfeeds
1) "feed" "feeds"))))
289 (defun org-feed-update (feed &optional retrieve-only
)
290 "Get inbox items from FEED.
291 FEED can be a string with an association in `org-feed-alist', or
292 it can be a list structured like an entry in `org-feed-alist'."
293 (interactive (list (org-completing-read "Feed name: " org-feed-alist
)))
294 (if (stringp feed
) (setq feed
(assoc feed org-feed-alist
)))
296 (error "No such feed in `org-feed-alist"))
298 (let ((name (car feed
))
300 (file (or (nth 2 feed
) (buffer-file-name (or (buffer-base-buffer)
302 (headline (nth 3 feed
))
303 (filter (nth 1 (memq :filter feed
)))
304 (formatter (nth 1 (memq :formatter feed
)))
305 (new-handler (nth 1 (memq :new-handler feed
)))
306 (changed-handler (nth 1 (memq :changed-handler feed
)))
307 (template (or (nth 1 (memq :template feed
))
308 org-feed-default-template
))
309 (drawer (or (nth 1 (memq :drawer feed
))
311 (parse-feed (or (nth 1 (memq :parse-feed feed
))
312 'org-feed-parse-rss-feed
))
313 (parse-entry (or (nth 1 (memq :parse-entry feed
))
314 'org-feed-parse-rss-entry
))
315 feed-buffer inbox-pos new-formatted
316 entries old-status status new changed guid-alist guid olds
)
317 (setq feed-buffer
(org-feed-get-feed url
))
318 (unless (and feed-buffer
(bufferp (get-buffer feed-buffer
)))
319 (error "Cannot get feed %s" name
))
321 (throw 'exit feed-buffer
))
322 (setq entries
(funcall parse-feed feed-buffer
))
323 (ignore-errors (kill-buffer feed-buffer
))
325 (save-window-excursion
326 (setq inbox-pos
(org-feed-goto-inbox-internal file headline
))
327 (setq old-status
(org-feed-read-previous-status inbox-pos drawer
))
328 ;; Add the "handled" status to the appropriate entries
329 (setq entries
(mapcar (lambda (e)
331 (plist-put e
:handled
336 ;; Find out which entries are new and which are changed
338 (if (not (plist-get e
:handled
))
340 (setq olds
(nth 2 (assoc (plist-get e
:guid
) old-status
)))
343 (plist-get e
:item-full-text
))
347 ;; Parse the relevant entries fully
348 (setq new
(mapcar parse-entry new
)
349 changed
(mapcar parse-entry changed
))
353 (setq new
(delq nil
(mapcar filter new
))
354 changed
(delq nil
(mapcar filter new
))))
356 (when (not (or new changed
))
357 (message "No new items in feed %s" name
)
360 ;; Get alist based on guid, to look up entries
363 (mapcar (lambda (e) (list (plist-get e
:guid
) e
)) new
)
364 (mapcar (lambda (e) (list (plist-get e
:guid
) e
)) changed
)))
366 ;; Construct the new status
370 (setq guid
(plist-get e
:guid
))
372 ;; things count as handled if we handle them now,
373 ;; or if they were handled previously
374 (if (assoc guid guid-alist
) t
(plist-get e
:handled
))
375 ;; A hash, to detect changes
376 (sha1 (plist-get e
:item-full-text
))))
379 ;; Handle new items in the feed
383 (goto-char inbox-pos
)
384 (funcall new-handler new
))
385 ;; No custom handler, do the default adding
386 ;; Format the new entries into an alist with GUIDs in the car
389 (lambda (e) (org-feed-format-entry e template formatter
))
392 ;; Insert the new items
393 (org-feed-add-items inbox-pos new-formatted
))
395 ;; Handle changed items in the feed
396 (when (and changed-handler changed
)
397 (goto-char inbox-pos
)
398 (funcall changed-handler changed
))
400 ;; Write the new status
401 ;; We do this only now, in case something goes wrong above, so
402 ;; that would would end up with a status that does not reflect
403 ;; which items truely have been handled
404 (org-feed-write-status inbox-pos drawer status
)
406 ;; Normalize the visibility of the inbox tree
407 (goto-char inbox-pos
)
408 (outline-hide-subtree)
410 (org-cycle-hide-drawers 'children
)
412 ;; Hooks and messages
413 (when org-feed-save-after-adding
(save-buffer))
414 (message "Added %d new item%s from feed %s to file %s, heading %s"
415 (length new
) (if (> (length new
) 1) "s" "")
417 (file-name-nondirectory file
) headline
)
418 (run-hooks 'org-feed-after-adding-hook
)
422 (defun org-feed-goto-inbox (feed)
423 "Go to the inbox that captures the feed named FEED."
425 (list (if (= (length org-feed-alist
) 1)
427 (org-completing-read "Feed name: " org-feed-alist
))))
428 (if (stringp feed
) (setq feed
(assoc feed org-feed-alist
)))
430 (error "No such feed in `org-feed-alist"))
431 (org-feed-goto-inbox-internal (nth 2 feed
) (nth 3 feed
)))
434 (defun org-feed-show-raw-feed (feed)
435 "Show the raw feed buffer of a feed."
437 (list (if (= (length org-feed-alist
) 1)
439 (org-completing-read "Feed name: " org-feed-alist
))))
440 (if (stringp feed
) (setq feed
(assoc feed org-feed-alist
)))
442 (error "No such feed in `org-feed-alist"))
443 (pop-to-buffer-same-window
444 (org-feed-update feed
'retrieve-only
))
445 (goto-char (point-min)))
447 (defun org-feed-goto-inbox-internal (file heading
)
448 "Find or create HEADING in FILE.
449 Switch to that buffer, and return the position of that headline."
452 (goto-char (point-min))
453 (if (re-search-forward
454 (concat "^\\*+[ \t]+" heading
"[ \t]*\\(:.*?:[ \t]*\\)?$")
456 (goto-char (match-beginning 0))
457 (goto-char (point-max))
458 (insert "\n\n* " heading
"\n\n")
459 (org-back-to-heading t
))
462 (defun org-feed-read-previous-status (pos drawer
)
463 "Get the alist of old GUIDs from the entry at POS.
464 This will find DRAWER and extract the alist."
467 (let ((end (save-excursion (org-end-of-subtree t t
))))
468 (if (re-search-forward
469 (concat "^[ \t]*:" drawer
":[ \t]*\n\\([^\000]*?\\)\n[ \t]*:END:")
471 (read (match-string 1))
474 (defun org-feed-write-status (pos drawer status
)
475 "Write the feed STATUS to DRAWER in entry at POS."
478 (let ((end (save-excursion (org-end-of-subtree t t
))))
479 (if (re-search-forward (concat "^[ \t]*:" drawer
":[ \t]*\n")
482 (goto-char (match-end 0))
483 (delete-region (point)
485 (and (re-search-forward "^[ \t]*:END:" nil t
)
486 (match-beginning 0)))))
487 (outline-next-heading)
488 (insert " :" drawer
":\n :END:\n")
489 (beginning-of-line 0))
490 (insert (pp-to-string status
)))))
492 (defun org-feed-add-items (pos entries
)
493 "Add the formatted items to the headline as POS."
497 (unless (looking-at org-complex-heading-regexp
)
498 (error "Wrong position"))
499 (setq level
(org-get-valid-level (length (match-string 1)) 1))
500 (org-end-of-subtree t t
)
501 (skip-chars-backward " \t\n")
502 (beginning-of-line 2)
504 (while (setq entry
(pop entries
))
505 (org-paste-subtree level entry
'yank
))
506 (org-mark-ring-push pos
))))
508 (defun org-feed-format-entry (entry template formatter
)
509 "Format ENTRY so that it can be inserted into an Org file.
510 ENTRY is a property list. This function adds a `:formatted-for-org' property
511 and returns the full property list.
512 If that property is already present, nothing changes."
513 (require 'org-capture
)
514 (if formatter
(funcall formatter entry
)
516 (org-split-string (or (plist-get entry
:description
) "???")
518 (time (or (if (plist-get entry
:pubDate
)
519 (org-read-date t t
(plist-get entry
:pubDate
)))
521 (v-h (or (plist-get entry
:title
) (car dlines
) "???"))
522 (v-t (format-time-string (org-time-stamp-format nil nil
) time
))
523 (v-T (format-time-string (org-time-stamp-format t nil
) time
))
524 (v-u (format-time-string (org-time-stamp-format nil t
) time
))
525 (v-U (format-time-string (org-time-stamp-format t t
) time
))
526 (v-a (let ((tmp (or (and (plist-get entry
:guid-permalink
)
527 (plist-get entry
:guid
))
528 (plist-get entry
:link
))))
529 (if tmp
(format "[[%s]]\n" tmp
) ""))))
532 (goto-char (point-min))
534 ;; Mark %() embedded elisp for later evaluation.
535 (org-capture-expand-embedded-elisp 'mark
)
537 ;; Simple %-escapes. `org-capture-escaped-%' may modify
538 ;; buffer and cripple match-data. Use markers instead.
539 (while (re-search-forward "%\\([a-zA-Z]+\\)" nil t
)
540 (let ((key (match-string 1))
541 (beg (copy-marker (match-beginning 0)))
542 (end (copy-marker (match-end 0))))
543 (unless (org-capture-escaped-%
)
544 (delete-region beg end
)
556 (let ((v (plist-get entry
(intern (concat ":" name
)))))
561 (concat "^\\([ \t]*\\)%" name
"[ \t]*$"))
562 (org-feed-make-indented-block
563 v
(org-get-indentation))
567 ;; Escape string delimiters within embedded lisp.
568 (if (org-capture-inside-embedded-elisp-p)
569 (replace-regexp-in-string "\"" "\\\\\"" replacement
)
572 ;; %() embedded elisp
573 (org-capture-expand-embedded-elisp)
575 (decode-coding-string
576 (buffer-string) (detect-coding-region (point-min) (point-max) t
))))))
578 (defun org-feed-make-indented-block (s n
)
579 "Add indentation of N spaces to a multiline string S."
580 (if (not (string-match "\n" s
))
583 (org-split-string s
"\n")
584 (concat "\n" (make-string n ?\
)))))
586 (defun org-feed-skip-http-headers (buffer)
587 "Remove HTTP headers from BUFFER, and return it.
588 Assumes headers are indeed present!"
589 (with-current-buffer buffer
591 (goto-char (point-min))
592 (search-forward "\n\n")
593 (delete-region (point-min) (point))
596 (defun org-feed-get-feed (url)
597 "Get the RSS feed file at URL and return the buffer."
599 ((eq org-feed-retrieve-method
'url-retrieve-synchronously
)
600 (org-feed-skip-http-headers (url-retrieve-synchronously url
)))
601 ((eq org-feed-retrieve-method
'curl
)
602 (ignore-errors (kill-buffer org-feed-buffer
))
603 (call-process "curl" nil org-feed-buffer nil
"--silent" url
)
605 ((eq org-feed-retrieve-method
'wget
)
606 (ignore-errors (kill-buffer org-feed-buffer
))
607 (call-process "wget" nil org-feed-buffer nil
"-q" "-O" "-" url
)
609 ((functionp org-feed-retrieve-method
)
610 (funcall org-feed-retrieve-method url
))))
612 (defun org-feed-parse-rss-feed (buffer)
613 "Parse BUFFER for RSS feed entries.
614 Returns a list of entries, with each entry a property list,
615 containing the properties `:guid' and `:item-full-text'."
617 (let ((case-fold-search t
)
618 entries beg end item guid entry
)
619 (with-current-buffer buffer
621 (goto-char (point-min))
622 (while (re-search-forward "<item\\>.*?>" nil t
)
624 end
(and (re-search-forward "</item>" nil t
)
625 (match-beginning 0)))
626 (setq item
(buffer-substring beg end
)
627 guid
(if (string-match "<guid\\>.*?>\\(.*?\\)</guid>" item
)
628 (xml-substitute-special (match-string-no-properties 1 item
))))
629 (setq entry
(list :guid guid
:item-full-text item
))
633 (nreverse entries
))))
635 (defun org-feed-parse-rss-entry (entry)
636 "Parse the `:item-full-text' field for xml tags and create new properties."
639 (insert (plist-get entry
:item-full-text
))
640 (goto-char (point-min))
641 (while (re-search-forward "<\\([a-zA-Z]+\\>\\).*?>\\([^\000]*?\\)</\\1>"
643 (setq entry
(plist-put entry
644 (intern (concat ":" (match-string 1)))
645 (xml-substitute-special (match-string 2)))))
646 (goto-char (point-min))
647 (unless (re-search-forward "isPermaLink[ \t]*=[ \t]*\"false\"" nil t
)
648 (setq entry
(plist-put entry
:guid-permalink t
))))
651 (defun org-feed-parse-atom-feed (buffer)
652 "Parse BUFFER for Atom feed entries.
653 Returns a list of entries, with each entry a property list,
654 containing the properties `:guid' and `:item-full-text'.
656 The `:item-full-text' property actually contains the sexp
657 formatted as a string, not the original XML data."
659 (with-current-buffer buffer
661 (let ((feed (car (xml-parse-region (point-min) (point-max)))))
665 :guid
(car (xml-node-children (car (xml-get-children entry
'id
))))
666 :item-full-text
(prin1-to-string entry
)))
667 (xml-get-children feed
'entry
)))))
669 (defun org-feed-parse-atom-entry (entry)
670 "Parse the `:item-full-text' as a sexp and create new properties."
671 (let ((xml (car (read-from-string (plist-get entry
:item-full-text
)))))
672 ;; Get first <link href='foo'/>.
673 (setq entry
(plist-put entry
:link
675 (car (xml-get-children xml
'link
))
677 ;; Add <title/> as :title.
678 (setq entry
(plist-put entry
:title
679 (xml-substitute-special
680 (car (xml-node-children
681 (car (xml-get-children xml
'title
)))))))
682 (let* ((content (car (xml-get-children xml
'content
)))
683 (type (xml-get-attribute-or-nil content
'type
)))
686 ((string= type
"text")
687 ;; We like plain text.
688 (setq entry
(plist-put entry
:description
689 (xml-substitute-special
690 (car (xml-node-children content
))))))
691 ((string= type
"html")
692 ;; TODO: convert HTML to Org markup.
693 (setq entry
(plist-put entry
:description
694 (xml-substitute-special
695 (car (xml-node-children content
))))))
696 ((string= type
"xhtml")
697 ;; TODO: convert XHTML to Org markup.
698 (setq entry
(plist-put entry
:description
700 (xml-node-children content
)))))
702 (setq entry
(plist-put entry
:description
704 "Unknown `%s' content." type
)))))))
710 ;; generated-autoload-file: "org-loaddefs.el"
713 ;;; org-feed.el ends here