;;; The integration with Gnus is not yet complete. See various `FIXME'
;;; comments, below, for supplementary explanations or discussions.
+;;; Several TODO items are marked as such
+
;;; Code:
(require 'gnus-sum)
-;; FIXME! We should not require `dns' nor `message' until we actually
-;; need them. Best would be to declare needed functions as auto-loadable.
-(require 'dns)
-(require 'message)
+(require 'gnus-uu) ; because of key prefix issues
+(require 'gnus) ; for the definitions of group content classification and spam processors
-(autoload 'bbdb-records "bbdb-com")
+;; FIXME! We should not require `message' until we actually need
+;; them. Best would be to declare needed functions as auto-loadable.
+(require 'message)
;; Attempt to load BBDB macros
(eval-when-compile
(file-error (defalias 'bbdb-search 'ignore))))
;; autoload executable-find
-(autoload 'executable-find "executable")
+(eval-and-compile
+ ;; executable-find is not autoloaded in Emacs 20
+ (autoload 'executable-find "executable"))
;; autoload ifile-spam-filter
-(autoload 'ifile-spam-filter "ifile-gnus")
-
-;;; Main parameters.
-
-(defvar spam-use-blacklist t
- "True if the blacklist should be used.")
-
-(defvar spam-use-whitelist t
- "True if the whitelist should be used.")
-
-(defvar spam-use-blackholes nil
- ;; FIXME! Turned off for now. The DNS routines are said to be flaky.
- "True if blackholes should be used.")
-
-(defvar spam-use-bogofilter t
- "True if bogofilter should be used.")
-
-(defvar spam-use-bbdb t
- "True if BBDB should be used.")
+(eval-and-compile
+ (autoload 'ifile-spam-filter "ifile-gnus"))
-(defvar spam-use-bbdb t
- "True if ifile should be used.")
+;; autoload query-dig
+(eval-and-compile
+ (autoload 'query-dig "dig"))
-(defvar spam-split-group "spam"
- "Usual group name where spam should be split.")
+;; autoload query-dns
+(eval-and-compile
+ (autoload 'query-dns "dns"))
-(defvar spam-junk-mailgroups
- ;; FIXME! The mailgroup list evidently depends on other choices made by the
- ;; user, so the built-in default below is not likely to be appropriate.
- (cons spam-split-group '("mail.junk" "poste.pourriel"))
- "Mailgroups which are dedicated by splitting to receive various junk.
-All unmarked article in such group receive the spam mark on group entry.")
+;;; Main parameters.
-;; FIXME! For `spam-ham-marks' and `spam-spam-marks', I wonder if it would
-;; not be easier for the user to just accept a string of mark letters, instead
-;; of a list of Gnus variable names. In such case, the stunt of deferred
-;; evaluation would not be useful anymore. Lars?? :-)
+(defgroup spam nil
+ "Spam configuration.")
-;; FIXME! It is rather questionable to see `K', `X' and `Y' as indicating
-;; positive ham. It much depends on how and why people use kill files, score
-;; files, and the kill command. Maybe it would be better, by default, to not
-;; process a message neither as ham nor spam, that is, just ignore it for
-;; learning purposes, when we are not sure of how the user sees it.
-;; But `r' and `R' should undoubtedly be seen as ham.
+(defcustom spam-directory "~/News/spam/"
+ "Directory for spam whitelists and blacklists."
+ :type 'directory
+ :group 'spam)
-;; FIXME! Some might consider overkill to define a list of spam marks. On
-;; the other hand, who knows, some users might for example like that
-;; explicitly `E'xpired articles be processed as positive spam.
+(defcustom spam-whitelist (expand-file-name "whitelist" spam-directory)
+ "The location of the whitelist.
+The file format is one regular expression per line.
+The regular expression is matched against the address."
+ :type 'file
+ :group 'spam)
-(defvar spam-ham-marks
- (list gnus-del-mark gnus-read-mark gnus-killed-mark
- gnus-kill-file-mark gnus-low-score-mark)
+(defcustom spam-blacklist (expand-file-name "blacklist" spam-directory)
+ "The location of the blacklist.
+The file format is one regular expression per line.
+The regular expression is matched against the address."
+ :type 'file
+ :group 'spam)
+
+(defcustom spam-use-dig t
+ "Whether query-dig should be used instead of query-dns."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-use-blacklist t
+ "Whether the blacklist should be used by spam-split."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-use-whitelist nil
+ "Whether the whitelist should be used by spam-split."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-use-blackholes nil
+ "Whether blackholes should be used by spam-split."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-use-bogofilter nil
+ "Whether bogofilter should be used by spam-split."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-use-bbdb nil
+ "Whether BBDB should be used by spam-split."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-use-ifile nil
+ "Whether ifile should be used by spam-split."
+ :type 'boolean
+ :group 'spam)
+
+(defcustom spam-split-group "spam"
+ "Group name where incoming spam should be put by spam-split."
+ :type 'string
+ :group 'spam)
+
+;; FIXME! The mailgroup list evidently depends on other choices made by the
+;; user, so the built-in default below is not likely to be appropriate.
+(defcustom spam-junk-mailgroups (cons spam-split-group '("mail.junk" "poste.pourriel"))
+ "Mailgroups with spam contents.
+All unmarked article in such group receive the spam mark on group entry."
+ :type '(repeat (string :tag "Group"))
+ :group 'spam)
+
+(defcustom spam-blackhole-servers '("bl.spamcop.net" "relays.ordb.org" "dev.null.dk" "relays.visi.com")
+ "List of blackhole servers."
+ :type '(repeat (string :tag "Server"))
+ :group 'spam)
+
+(defcustom spam-ham-marks (list 'gnus-del-mark 'gnus-read-mark 'gnus-killed-mark 'gnus-kill-file-mark 'gnus-low-score-mark)
"Marks considered as being ham (positively not spam).
-Such articles will be transmitted to `bogofilter -n' on group exit.")
-
-(defvar spam-spam-marks
- (list gnus-spam-mark)
+Such articles will be processed as ham (non-spam) on group exit."
+ :type '(set
+ (variable-item gnus-del-mark)
+ (variable-item gnus-read-mark)
+ (variable-item gnus-killed-mark)
+ (variable-item gnus-kill-file-mark)
+ (variable-item gnus-low-score-mark))
+ :group 'spam)
+
+(defcustom spam-spam-marks (list 'gnus-spam-mark)
"Marks considered as being spam (positively spam).
-Such articles will be transmitted to `bogofilter -s' on group exit.")
-
-;; FIXME! Ideally, the remainder of this page should be fully integrated
-;; within `gnus-sum.el'.
+Such articles will be transmitted to `bogofilter -s' on group exit."
+ :type '(set
+ (variable-item gnus-spam-mark)
+ (variable-item gnus-killed-mark)
+ (variable-item gnus-kill-file-mark)
+ (variable-item gnus-low-score-mark))
+ :group 'spam)
+
+(defcustom spam-face 'gnus-splash-face
+ "Face for spam-marked articles"
+ :type 'face
+ :group 'spam)
+
+(defgroup spam-bogofilter nil
+ "Spam bogofilter configuration."
+ :group 'spam)
+
+(defcustom spam-bogofilter-output-buffer-name "*Bogofilter Output*"
+ "Name of buffer when displaying `bogofilter -v' output."
+ :type 'string
+ :group 'spam-bogofilter)
+
+(defcustom spam-bogofilter-initial-timeout 40
+ "Timeout in seconds for the initial reply from the `bogofilter' program."
+ :type 'integer
+ :group 'spam-bogofilter)
+
+(defcustom spam-bogofilter-subsequent-timeout 15
+ "Timeout in seconds for any subsequent reply from the `bogofilter' program."
+ :type 'integer
+ :group 'spam-bogofilter)
+
+(defcustom spam-bogofilter-path (executable-find "bogofilter")
+ "File path of the Bogofilter executable program."
+ :type '(choice (file :tag "Location of bogofilter")
+ (const :tag "Bogofilter is not installed"))
+ :group 'spam-bogofilter)
+
+;; FIXME! In the following regexp, we should explain which tool produces
+;; which kind of header. I do not even remember them all by now. X-Junk
+;; (and previously X-NoSpam) are produced by the `NoSpam' tool, which has
+;; never been published, so it might not be reasonable leaving it in the
+;; list.
+(defcustom spam-bogofilter-spaminfo-header-regexp "^X-\\(jf\\|Junk\\|NoSpam\\|Spam\\|SB\\)[^:]*:"
+ "Regexp for spam markups in headers.
+Markup from spam recognisers, as well as `Xref', are to be removed from
+articles before they get registered by Bogofilter."
+ :type 'regexp
+ :group 'spam-bogofilter)
;;; Key bindings for spam control.
-;; FIXME! The justification for `M-d' is that this is what Paul Graham
-;; suggests in his original article, and what Eric Raymond's patch for Mutt
-;; uses. But more importantly, that binding was still free in Summary mode!
-
-;; FIXME! Lars has not blessed the following key bindings yet. It looks
-;; convenient that the score analysis command uses a sequence ending with the
-;; letter `t', so it nicely parallels `B t' or `V t'. `M-d' is a kind of
-;; "alternate" `d', it is also the sequence suggested in Paul Graham article,
-;; and also in Eric Raymond's patch for Mutt. `S x' might be the more
-;; official key binding for `M-d'.
-
(gnus-define-keys gnus-summary-mode-map
"St" spam-bogofilter-score
"Sx" gnus-summary-mark-as-spam
+ "Mst" spam-bogofilter-score
+ "Msx" gnus-summary-mark-as-spam
"\M-d" gnus-summary-mark-as-spam)
;;; How to highlight a spam summary line.
-;; FIXME! Of course, `gnus-splash-face' has another purpose. Maybe a
-;; special face should be created, named and used instead, for spam lines.
+;; TODO: How do we redo this every time spam-face is customized?
-(push '((eq mark gnus-spam-mark) . gnus-splash-face)
+(push '((eq mark gnus-spam-mark) . spam-face)
gnus-summary-highlight)
+;; convenience functions
+(defun spam-group-spam-contents-p (group)
+ (if (stringp group)
+ (or (member group spam-junk-mailgroups)
+ (memq 'gnus-group-spam-classification-spam (gnus-parameter-spam-contents group)))
+ nil))
+
+(defun spam-group-ham-contents-p (group)
+ (if (stringp group)
+ (memq 'gnus-group-spam-classification-ham (gnus-parameter-spam-contents group))
+ nil))
+
+
;;; Hooks dispatching. A bit raw for now.
(defun spam-summary-prepare ()
(spam-mark-junk-as-spam-routine))
(defun spam-summary-prepare-exit ()
- (spam-bogofilter-register-routine))
+ (spam-bogofilter-register-routine)
+ (when (spam-group-spam-contents-p gnus-newsgroup-name)
+ (
+ ;; TODO: the spam processors here
+ ;; TODO: the spam-processed articles will be moved here
+ ))
+ (when (spam-group-ham-contents-p gnus-newsgroup-name)
+ (
+ ;; TODO: the ham processors here
+ )))
(add-hook 'gnus-summary-prepare-hook 'spam-summary-prepare)
(add-hook 'gnus-summary-prepare-exit-hook 'spam-summary-prepare-exit)
(defun spam-mark-junk-as-spam-routine ()
- (when (member gnus-newsgroup-name spam-junk-mailgroups)
+ ;; check the global list of group names spam-junk-mailgroups and the group parameters
+ (when (spam-group-spam-contents-p gnus-newsgroup-name)
(let ((articles gnus-newsgroup-articles)
article)
(while articles
decision)
(while (and list-of-checks (not decision))
(let ((pair (pop list-of-checks)))
- (when (eval (car pair))
- (setq decision (apply (cdr pair))))))
+ (when (symbol-value (car pair))
+ (setq decision (funcall (cdr pair))))))
(if (eq decision t)
nil
decision)))
\f
;;;; Blackholes.
-(defvar spam-blackhole-servers '("bl.spamcop.net"
- "relays.ordb.org"
- "dev.null.dk"
- "relays.visi.com"
- "rbl.maps.vix.com")
- "List of blackhole servers.")
-
(defun spam-check-blackholes ()
- "Check the Receieved headers for blackholed relays."
+ "Check the Received headers for blackholed relays."
(let ((headers (message-fetch-field "received"))
ips matches)
(when headers
ips)))
(dolist (server spam-blackhole-servers)
(dolist (ip ips)
- (when (query-dns (concat ip "." server))
- (push (list ip server (query-dns (concat ip "." server) 'TXT))
- matches)))))
+ (let ((query-string (concat ip "." server)))
+ (if spam-use-dig
+ (let ((query-result (query-dig query-string)))
+ (when query-result
+ (message "spam detected with blackhole check of relay %s (dig query result '%s')" query-string query-result)
+ (push (list ip server query-result)
+ matches)))
+ ;; else, if not using dig.el
+ (when (query-dns query-string)
+ (push (list ip server (query-dns query-string 'TXT))
+ matches)))))))
(when matches
spam-split-group)))
\f
;;;; Blacklists and whitelists.
-(defvar spam-directory "~/News/spam/"
- "When spam files are kept.")
-
-(defvar spam-whitelist (expand-file-name "whitelist" spam-directory)
- "The location of the whitelist.
-The file format is one regular expression per line.
-The regular expression is matched against the address.")
-
-(defvar spam-blacklist (expand-file-name "blacklist" spam-directory)
- "The location of the blacklist.
-The file format is one regular expression per line.
-The regular expression is matched against the address.")
-
(defvar spam-whitelist-cache nil)
(defvar spam-blacklist-cache nil)
spam-list-of-checks))))
;;; check the ifile backend; return nil if the mail was NOT classified as spam
-(condition-case nil
- (progn
- (require 'ifile-gnus)
- ;;;
- (defun spam-check-ifile ()
- (let ((ifile-primary-spam-group spam-split-group))
- (ifile-spam-filter nil))))
- (file-error (setq spam-list-of-checks
- (delete (assoc 'spam-use-ifile spam-list-of-checks)
- spam-list-of-checks))))
+;;; TODO: we can't (require) ifile, because it will insinuate itself automatically
+(defun spam-check-ifile ()
+ (let ((ifile-primary-spam-group spam-split-group))
+ (ifile-spam-filter nil)))
(defun spam-check-blacklist ()
;; FIXME! Should it detect when file timestamps change?
cache nil)))
found))
-(defun spam-display-buffer-contents ()
- "Display contents of the current buffer in the echo area if possible,
-otherwise pop-up the buffer itself."
- (unless (zerop (buffer-size))
- (if (<= (count-lines (point-min) (point-max)) 1)
- (progn
- (goto-char (point-max))
- (when (bolp)
- (backward-char 1))
- (message "%s" (buffer-substring (point-min) (point))))
- (goto-char (point-min))
- (display-buffer (current-buffer)))))
-
\f
;;;; Training via Bogofilter. Last updated 2002-09-02.
;;; make install
;;;
;;; Here as well, you need to become super-user for the last step. Now,
-;;; initialises your word lists by doing, under your own identity:
+;;; initialize your word lists by doing, under your own identity:
;;;
;;; mkdir ~/.bogofilter
;;; touch ~/.bogofilter/badlist
;;; * ? bogofilter
;;; | formail -bfI "X-Spam-Status: Yes"
-(defvar spam-output-buffer-name "*Bogofilter Output*"
- "Name of buffer when displaying `bogofilter -v' output.")
-
-(defvar spam-spaminfo-header-regexp
- ;; FIXME! In the following regexp, we should explain which tool produces
- ;; which kind of header. I do not even remember them all by now. X-Junk
- ;; (and previously X-NoSpam) are produced by the `NoSpam' tool, which has
- ;; never been published, so it might not be reasonable leaving it in the
- ;; list.
- "^X-\\(jf\\|Junk\\|NoSpam\\|Spam\\|SB\\)[^:]*:"
- "Regexp for spam markups in headers.
-Markup from spam recognisers, as well as `Xref', are to be removed from
-articles before they get registered by Bogofilter.")
-
-(defvar spam-bogofilter-path (executable-find "bogofilter")
- "File path of the Bogofilter executable program.
-Force this variable to nil if you want to inhibit the functionality.")
-
(defun spam-check-bogofilter ()
;; Dynamic spam check. I do not know how to check the exit status,
;; so instead, read `bogofilter -v' output.
(when (and spam-use-bogofilter spam-bogofilter-path)
(spam-bogofilter-articles nil "-v" (list (gnus-summary-article-number)))
(when (save-excursion
- (set-buffer spam-output-buffer-name)
+ (set-buffer spam-bogofilter-output-buffer-name)
(goto-char (point-min))
(re-search-forward "Spamicity: \\(0\\.9\\|1\\.0\\)" nil t))
spam-split-group)))
(interactive)
(when (and spam-use-bogofilter spam-bogofilter-path)
(spam-bogofilter-articles nil "-v" (list (gnus-summary-article-number)))
- (save-excursion
- (set-buffer spam-output-buffer-name)
- (spam-display-buffer-contents))))
+ (with-current-buffer spam-bogofilter-output-buffer-name
+ (unless (zerop (buffer-size))
+ (if (<= (count-lines (point-min) (point-max)) 1)
+ (progn
+ (goto-char (point-max))
+ (when (bolp)
+ (backward-char 1))
+ (message "%s" (buffer-substring (point-min) (point))))
+ (goto-char (point-min))
+ (display-buffer (current-buffer)))))))
(defun spam-bogofilter-register-routine ()
(when (and spam-use-bogofilter spam-bogofilter-path)
(let ((articles gnus-newsgroup-articles)
- article mark ham-articles spam-articles)
+ article mark ham-articles spam-articles spam-mark-values ham-mark-values)
+
+ ;; marks are stored as symbolic values, so we have to dereference them for memq to work
+ ;; we wouldn't have to do this if gnus-summary-article-mark returned a symbol.
+ (dolist (mark spam-ham-marks)
+ (push (symbol-value mark) ham-mark-values))
+
+ (dolist (mark spam-spam-marks)
+ (push (symbol-value mark) spam-mark-values))
+
(while articles
(setq article (pop articles)
mark (gnus-summary-article-mark article))
- (cond ((memq mark spam-spam-marks) (push article spam-articles))
+ (cond ((memq mark spam-mark-values) (push article spam-articles))
((memq article gnus-newsgroup-saved))
- ((memq mark spam-ham-marks) (push article ham-articles))))
+ ((memq mark ham-mark-values) (push article ham-articles))))
(when ham-articles
(spam-bogofilter-articles "ham" "-n" ham-articles))
(when spam-articles
(spam-bogofilter-articles "SPAM" "-s" spam-articles)))))
-(defvar spam-bogofilter-initial-timeout 40
- "Timeout in seconds for the initial reply from the `bogofilter' program.")
-
-(defvar spam-bogofilter-subsequent-timeout 15
- "Timeout in seconds for any subsequent reply from the `bogofilter' program.")
-
(defun spam-bogofilter-articles (type option articles)
- (let ((output-buffer (get-buffer-create spam-output-buffer-name))
+ (let ((output-buffer (get-buffer-create spam-bogofilter-output-buffer-name))
(article-copy (get-buffer-create " *Bogofilter Article Copy*"))
- (remove-regexp (concat spam-spaminfo-header-regexp "\\|Xref:"))
+ (remove-regexp (concat spam-bogofilter-spaminfo-header-regexp "\\|Xref:"))
(counter 0)
prefix process article)
(when type
(message "%s %d" prefix counter))
(setq article (pop articles))
(gnus-summary-goto-subject article)
- (gnus-summary-select-article)
+ (gnus-summary-show-article t)
(gnus-eval-in-buffer-window article-copy
(insert-buffer-substring gnus-original-article-buffer)
;; Remove spam classification redundant headers: they may induce
;; unwanted biases in later analysis.
- (goto-char (point-min))
- (while (not (or (eobp) (= (following-char) ?\n)))
- (if (looking-at remove-regexp)
- (delete-region (point)
- (save-excursion (forward-line 1) (point)))
- (forward-line 1)))
- (goto-char (point-min))
+ (message-remove-header remove-regexp t)
;; Bogofilter really wants From envelopes for counting articles.
;; Fake one at the beginning, make sure there will be no other.
+ (goto-char (point-min))
(if (looking-at "From ")
(forward-line 1)
(insert "From nobody " (current-time-string) "\n"))