Commit | Line | Data |
---|---|---|
23f87bed MB |
1 | ;;; spam-stat.el --- detecting spam based on statistics |
2 | ||
3 | ;; Copyright (C) 2002, 2003, 2004 Free Software Foundation, Inc. | |
4 | ||
5 | ;; Author: Alex Schroeder <alex@gnu.org> | |
6 | ;; Keywords: network | |
7 | ;; URL: http://www.emacswiki.org/cgi-bin/wiki.pl?SpamStat | |
8 | ||
9 | ;; This file is part of GNU Emacs. | |
10 | ||
11 | ;; This is free software; you can redistribute it and/or modify it | |
12 | ;; under the terms of the GNU General Public License as published by | |
13 | ;; the Free Software Foundation; either version 2, or (at your option) | |
14 | ;; any later version. | |
15 | ||
16 | ;; This is distributed in the hope that it will be useful, but WITHOUT | |
17 | ;; ANY WARRANTY; without even the implied warranty of MERCHANTABILITY | |
18 | ;; or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public | |
19 | ;; License for more details. | |
20 | ||
21 | ;; You should have received a copy of the GNU General Public License | |
22 | ;; along with GNU Emacs; see the file COPYING. If not, write to the | |
23 | ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
24 | ;; Boston, MA 02111-1307, USA. | |
25 | ||
26 | ;;; Commentary: | |
27 | ||
28 | ;; This implements spam analysis according to Paul Graham in "A Plan | |
29 | ;; for Spam". The basis for all this is a statistical distribution of | |
30 | ;; words for your spam and non-spam mails. We need this information | |
31 | ;; in a hash-table so that the analysis can use the information when | |
32 | ;; looking at your mails. Therefore, before you begin, you need tons | |
33 | ;; of mails (Graham uses 4000 non-spam and 4000 spam mails for his | |
34 | ;; experiments). | |
35 | ;; | |
36 | ;; The main interface to using spam-stat, are the following functions: | |
37 | ;; | |
38 | ;; `spam-stat-buffer-is-spam' -- called in a buffer, that buffer is | |
39 | ;; considered to be a new spam mail; use this for new mail that has | |
40 | ;; not been processed before | |
41 | ;; | |
42 | ;; `spam-stat-buffer-is-non-spam' -- called in a buffer, that buffer | |
43 | ;; is considered to be a new non-spam mail; use this for new mail that | |
44 | ;; has not been processed before | |
45 | ;; | |
46 | ;; `spam-stat-buffer-change-to-spam' -- called in a buffer, that | |
47 | ;; buffer is no longer considered to be normal mail but spam; use this | |
48 | ;; to change the status of a mail that has already been processed as | |
49 | ;; non-spam | |
50 | ;; | |
51 | ;; `spam-stat-buffer-change-to-non-spam' -- called in a buffer, that | |
52 | ;; buffer is no longer considered to be spam but normal mail; use this | |
53 | ;; to change the status of a mail that has already been processed as | |
54 | ;; spam | |
55 | ;; | |
56 | ;; `spam-stat-save' -- save the hash table to the file; the filename | |
57 | ;; used is stored in the variable `spam-stat-file' | |
58 | ;; | |
59 | ;; `spam-stat-load' -- load the hash table from a file; the filename | |
60 | ;; used is stored in the variable `spam-stat-file' | |
61 | ;; | |
62 | ;; `spam-stat-score-word' -- return the spam score for a word | |
63 | ;; | |
64 | ;; `spam-stat-score-buffer' -- return the spam score for a buffer | |
65 | ;; | |
66 | ;; `spam-stat-split-fancy' -- for fancy mail splitting; add | |
67 | ;; the rule (: spam-stat-split-fancy) to `nnmail-split-fancy' | |
68 | ;; | |
69 | ;; This requires the following in your ~/.gnus file: | |
70 | ;; | |
71 | ;; (require 'spam-stat) | |
72 | ;; (spam-stat-load) | |
73 | ||
74 | ;;; Testing: | |
75 | ||
76 | ;; Typical test will involve calls to the following functions: | |
77 | ;; | |
78 | ;; Reset: (spam-stat-reset) | |
79 | ;; Learn spam: (spam-stat-process-spam-directory "~/Mail/mail/spam") | |
80 | ;; Learn non-spam: (spam-stat-process-non-spam-directory "~/Mail/mail/misc") | |
81 | ;; Save table: (spam-stat-save) | |
82 | ;; File size: (nth 7 (file-attributes spam-stat-file)) | |
83 | ;; Number of words: (hash-table-count spam-stat) | |
84 | ;; Test spam: (spam-stat-test-directory "~/Mail/mail/spam") | |
85 | ;; Test non-spam: (spam-stat-test-directory "~/Mail/mail/misc") | |
86 | ;; Reduce table size: (spam-stat-reduce-size) | |
87 | ;; Save table: (spam-stat-save) | |
88 | ;; File size: (nth 7 (file-attributes spam-stat-file)) | |
89 | ;; Number of words: (hash-table-count spam-stat) | |
90 | ;; Test spam: (spam-stat-test-directory "~/Mail/mail/spam") | |
91 | ;; Test non-spam: (spam-stat-test-directory "~/Mail/mail/misc") | |
92 | ||
93 | ;;; Dictionary Creation: | |
94 | ||
95 | ;; Typically, you will filter away mailing lists etc. using specific | |
96 | ;; rules in `nnmail-split-fancy'. Somewhere among these rules, you | |
97 | ;; will filter spam. Here is how you would create your dictionary: | |
98 | ||
99 | ;; Reset: (spam-stat-reset) | |
100 | ;; Learn spam: (spam-stat-process-spam-directory "~/Mail/mail/spam") | |
101 | ;; Learn non-spam: (spam-stat-process-non-spam-directory "~/Mail/mail/misc") | |
102 | ;; Repeat for any other non-spam group you need... | |
103 | ;; Reduce table size: (spam-stat-reduce-size) | |
104 | ;; Save table: (spam-stat-save) | |
105 | ||
106 | ;;; Todo: | |
107 | ||
108 | ;; Speed it up. Integrate with Gnus such that it uses spam and expiry | |
109 | ;; marks to call the appropriate functions when leaving the summary | |
110 | ;; buffer and saves the hash table when leaving Gnus. More testing: | |
111 | ;; More mails, disabling SpamAssassin, double checking algorithm, find | |
112 | ;; improved algorithm. | |
113 | ||
114 | ;;; Thanks: | |
115 | ||
116 | ;; Ted Zlatanov <tzz@lifelogs.com> | |
117 | ;; Jesper Harder <harder@myrealbox.com> | |
118 | ;; Dan Schmidt <dfan@dfan.org> | |
119 | ;; Lasse Rasinen <lrasinen@iki.fi> | |
120 | ;; Milan Zamazal <pdm@zamazal.org> | |
121 | ||
122 | \f | |
123 | ||
124 | ;;; Code: | |
125 | ||
126 | (defgroup spam-stat nil | |
127 | "Statistical spam detection for Emacs. | |
128 | Use the functions to build a dictionary of words and their statistical | |
129 | distribution in spam and non-spam mails. Then use a function to determine | |
130 | whether a buffer contains spam or not." | |
e2642250 | 131 | :version "21.4" |
23f87bed MB |
132 | :group 'gnus) |
133 | ||
134 | (defcustom spam-stat-file "~/.spam-stat.el" | |
135 | "File used to save and load the dictionary. | |
136 | See `spam-stat-to-hash-table' for the format of the file." | |
137 | :type 'file | |
138 | :group 'spam-stat) | |
139 | ||
140 | (defcustom spam-stat-install-hooks t | |
141 | "Whether spam-stat should install its hooks in Gnus. | |
142 | This is set to nil if you use spam-stat through spam.el." | |
143 | :type 'boolean | |
144 | :group 'spam-stat) | |
145 | ||
146 | (defcustom spam-stat-unknown-word-score 0.2 | |
147 | "The score to use for unknown words. | |
148 | Also used for words that don't appear often enough." | |
149 | :type 'number | |
150 | :group 'spam-stat) | |
151 | ||
152 | (defcustom spam-stat-max-word-length 15 | |
153 | "Only words shorter than this will be considered." | |
154 | :type 'integer | |
155 | :group 'spam-stat) | |
156 | ||
157 | (defcustom spam-stat-max-buffer-length 10240 | |
158 | "Only the beginning of buffers will be analyzed. | |
159 | This variable says how many characters this will be." | |
160 | :type 'integer | |
161 | :group 'spam-stat) | |
162 | ||
163 | (defcustom spam-stat-split-fancy-spam-group "mail.spam" | |
164 | "Name of the group where spam should be stored, if | |
165 | `spam-stat-split-fancy' is used in fancy splitting rules. Has no | |
166 | effect when spam-stat is invoked through spam.el." | |
167 | :type 'string | |
168 | :group 'spam-stat) | |
169 | ||
170 | (defcustom spam-stat-split-fancy-spam-threshhold 0.9 | |
171 | "Spam score threshhold in spam-stat-split-fancy." | |
172 | :type 'number | |
173 | :group 'spam-stat) | |
174 | ||
175 | (defvar spam-stat-syntax-table | |
176 | (let ((table (copy-syntax-table text-mode-syntax-table))) | |
177 | (modify-syntax-entry ?- "w" table) | |
178 | (modify-syntax-entry ?_ "w" table) | |
179 | (modify-syntax-entry ?. "w" table) | |
180 | (modify-syntax-entry ?! "w" table) | |
181 | (modify-syntax-entry ?? "w" table) | |
182 | (modify-syntax-entry ?+ "w" table) | |
183 | table) | |
184 | "Syntax table used when processing mails for statistical analysis. | |
185 | The important part is which characters are word constituents.") | |
186 | ||
187 | (defvar spam-stat-dirty nil | |
188 | "Whether the spam-stat database needs saving.") | |
189 | ||
190 | (defvar spam-stat-buffer nil | |
191 | "Buffer to use for scoring while splitting. | |
192 | This is set by hooking into Gnus.") | |
193 | ||
194 | (defvar spam-stat-buffer-name " *spam stat buffer*" | |
195 | "Name of the `spam-stat-buffer'.") | |
196 | ||
197 | ;; Functions missing in Emacs 20 | |
198 | ||
199 | (when (memq nil (mapcar 'fboundp | |
200 | '(gethash hash-table-count make-hash-table | |
201 | mapc puthash))) | |
202 | (require 'cl) | |
203 | (unless (fboundp 'puthash) | |
204 | ;; alias puthash is missing from Emacs 20 cl-extra.el | |
205 | (defalias 'puthash 'cl-puthash))) | |
206 | ||
207 | (eval-when-compile | |
208 | (unless (fboundp 'with-syntax-table) | |
209 | ;; Imported from Emacs 21.2 | |
210 | (defmacro with-syntax-table (table &rest body) "\ | |
211 | Evaluate BODY with syntax table of current buffer set to a copy of TABLE. | |
212 | The syntax table of the current buffer is saved, BODY is evaluated, and the | |
213 | saved table is restored, even in case of an abnormal exit. | |
214 | Value is what BODY returns." | |
215 | (let ((old-table (make-symbol "table")) | |
216 | (old-buffer (make-symbol "buffer"))) | |
217 | `(let ((,old-table (syntax-table)) | |
218 | (,old-buffer (current-buffer))) | |
219 | (unwind-protect | |
220 | (progn | |
221 | (set-syntax-table (copy-syntax-table ,table)) | |
222 | ,@body) | |
223 | (save-current-buffer | |
224 | (set-buffer ,old-buffer) | |
225 | (set-syntax-table ,old-table)))))))) | |
226 | ||
227 | ;; Hooking into Gnus | |
228 | ||
229 | (defun spam-stat-store-current-buffer () | |
230 | "Store a copy of the current buffer in `spam-stat-buffer'." | |
231 | (save-excursion | |
232 | (let ((str (buffer-string))) | |
233 | (set-buffer (get-buffer-create spam-stat-buffer-name)) | |
234 | (erase-buffer) | |
235 | (insert str) | |
236 | (setq spam-stat-buffer (current-buffer))))) | |
237 | ||
238 | (defun spam-stat-store-gnus-article-buffer () | |
239 | "Store a copy of the current article in `spam-stat-buffer'. | |
240 | This uses `gnus-article-buffer'." | |
241 | (save-excursion | |
242 | (set-buffer gnus-original-article-buffer) | |
243 | (spam-stat-store-current-buffer))) | |
244 | ||
245 | ;; Data -- not using defstruct in order to save space and time | |
246 | ||
247 | (defvar spam-stat (make-hash-table :test 'equal) | |
248 | "Hash table used to store the statistics. | |
249 | Use `spam-stat-load' to load the file. | |
250 | Every word is used as a key in this table. The value is a vector. | |
251 | Use `spam-stat-ngood', `spam-stat-nbad', `spam-stat-good', | |
252 | `spam-stat-bad', and `spam-stat-score' to access this vector.") | |
253 | ||
254 | (defvar spam-stat-ngood 0 | |
255 | "The number of good mails in the dictionary.") | |
256 | ||
257 | (defvar spam-stat-nbad 0 | |
258 | "The number of bad mails in the dictionary.") | |
259 | ||
260 | (defsubst spam-stat-good (entry) | |
261 | "Return the number of times this word belongs to good mails." | |
262 | (aref entry 0)) | |
263 | ||
264 | (defsubst spam-stat-bad (entry) | |
265 | "Return the number of times this word belongs to bad mails." | |
266 | (aref entry 1)) | |
267 | ||
268 | (defsubst spam-stat-score (entry) | |
269 | "Set the score of this word." | |
270 | (if entry | |
271 | (aref entry 2) | |
272 | spam-stat-unknown-word-score)) | |
273 | ||
274 | (defsubst spam-stat-set-good (entry value) | |
275 | "Set the number of times this word belongs to good mails." | |
276 | (aset entry 0 value)) | |
277 | ||
278 | (defsubst spam-stat-set-bad (entry value) | |
279 | "Set the number of times this word belongs to bad mails." | |
280 | (aset entry 1 value)) | |
281 | ||
282 | (defsubst spam-stat-set-score (entry value) | |
283 | "Set the score of this word." | |
284 | (aset entry 2 value)) | |
285 | ||
286 | (defsubst spam-stat-make-entry (good bad) | |
287 | "Return a vector with the given properties." | |
288 | (let ((entry (vector good bad nil))) | |
289 | (spam-stat-set-score entry (spam-stat-compute-score entry)) | |
290 | entry)) | |
291 | ||
292 | ;; Computing | |
293 | ||
294 | (defun spam-stat-compute-score (entry) | |
295 | "Compute the score of this word. 1.0 means spam." | |
296 | ;; promote all numbers to floats for the divisions | |
297 | (let* ((g (* 2.0 (spam-stat-good entry))) | |
298 | (b (float (spam-stat-bad entry)))) | |
299 | (cond ((< (+ g b) 5) | |
300 | .2) | |
301 | ((= 0 spam-stat-ngood) | |
302 | .99) | |
303 | ((= 0 spam-stat-nbad) | |
304 | .01) | |
305 | (t | |
306 | (max .01 | |
307 | (min .99 (/ (/ b spam-stat-nbad) | |
308 | (+ (/ g spam-stat-ngood) | |
309 | (/ b spam-stat-nbad))))))))) | |
310 | ||
311 | ;; Parsing | |
312 | ||
313 | (defmacro with-spam-stat-max-buffer-size (&rest body) | |
314 | "Narrows the buffer down to the first 4k characters, then evaluates BODY." | |
315 | `(save-restriction | |
316 | (when (> (- (point-max) | |
317 | (point-min)) | |
318 | spam-stat-max-buffer-length) | |
319 | (narrow-to-region (point-min) | |
320 | (+ (point-min) spam-stat-max-buffer-length))) | |
321 | ,@body)) | |
322 | ||
323 | (defun spam-stat-buffer-words () | |
324 | "Return a hash table of words and number of occurences in the buffer." | |
325 | (with-spam-stat-max-buffer-size | |
326 | (with-syntax-table spam-stat-syntax-table | |
327 | (goto-char (point-min)) | |
328 | (let ((result (make-hash-table :test 'equal)) | |
329 | word count) | |
330 | (while (re-search-forward "\\w+" nil t) | |
331 | (setq word (match-string-no-properties 0) | |
332 | count (1+ (gethash word result 0))) | |
333 | (when (< (length word) spam-stat-max-word-length) | |
334 | (puthash word count result))) | |
335 | result)))) | |
336 | ||
337 | (defun spam-stat-buffer-is-spam () | |
338 | "Consider current buffer to be a new spam mail." | |
339 | (setq spam-stat-nbad (1+ spam-stat-nbad)) | |
340 | (maphash | |
341 | (lambda (word count) | |
342 | (let ((entry (gethash word spam-stat))) | |
343 | (if entry | |
344 | (spam-stat-set-bad entry (+ count (spam-stat-bad entry))) | |
345 | (setq entry (spam-stat-make-entry 0 count))) | |
346 | (spam-stat-set-score entry (spam-stat-compute-score entry)) | |
347 | (puthash word entry spam-stat))) | |
348 | (spam-stat-buffer-words)) | |
349 | (setq spam-stat-dirty t)) | |
350 | ||
351 | (defun spam-stat-buffer-is-non-spam () | |
352 | "Consider current buffer to be a new non-spam mail." | |
353 | (setq spam-stat-ngood (1+ spam-stat-ngood)) | |
354 | (maphash | |
355 | (lambda (word count) | |
356 | (let ((entry (gethash word spam-stat))) | |
357 | (if entry | |
358 | (spam-stat-set-good entry (+ count (spam-stat-good entry))) | |
359 | (setq entry (spam-stat-make-entry count 0))) | |
360 | (spam-stat-set-score entry (spam-stat-compute-score entry)) | |
361 | (puthash word entry spam-stat))) | |
362 | (spam-stat-buffer-words)) | |
363 | (setq spam-stat-dirty t)) | |
364 | ||
365 | (defun spam-stat-buffer-change-to-spam () | |
366 | "Consider current buffer no longer normal mail but spam." | |
367 | (setq spam-stat-nbad (1+ spam-stat-nbad) | |
368 | spam-stat-ngood (1- spam-stat-ngood)) | |
369 | (maphash | |
370 | (lambda (word count) | |
371 | (let ((entry (gethash word spam-stat))) | |
372 | (if (not entry) | |
373 | (error "This buffer has unknown words in it.") | |
374 | (spam-stat-set-good entry (- (spam-stat-good entry) count)) | |
375 | (spam-stat-set-bad entry (+ (spam-stat-bad entry) count)) | |
376 | (spam-stat-set-score entry (spam-stat-compute-score entry)) | |
377 | (puthash word entry spam-stat)))) | |
378 | (spam-stat-buffer-words)) | |
379 | (setq spam-stat-dirty t)) | |
380 | ||
381 | (defun spam-stat-buffer-change-to-non-spam () | |
382 | "Consider current buffer no longer spam but normal mail." | |
383 | (setq spam-stat-nbad (1- spam-stat-nbad) | |
384 | spam-stat-ngood (1+ spam-stat-ngood)) | |
385 | (maphash | |
386 | (lambda (word count) | |
387 | (let ((entry (gethash word spam-stat))) | |
388 | (if (not entry) | |
389 | (error "This buffer has unknown words in it.") | |
390 | (spam-stat-set-good entry (+ (spam-stat-good entry) count)) | |
391 | (spam-stat-set-bad entry (- (spam-stat-bad entry) count)) | |
392 | (spam-stat-set-score entry (spam-stat-compute-score entry)) | |
393 | (puthash word entry spam-stat)))) | |
394 | (spam-stat-buffer-words)) | |
395 | (setq spam-stat-dirty t)) | |
396 | ||
397 | ;; Saving and Loading | |
398 | ||
399 | (defun spam-stat-save (&optional force) | |
400 | "Save the `spam-stat' hash table as lisp file. | |
401 | With a prefix argument save unconditionally." | |
402 | (interactive "P") | |
403 | (when (or force spam-stat-dirty) | |
404 | (with-temp-buffer | |
405 | (let ((standard-output (current-buffer)) | |
406 | (font-lock-maximum-size 0)) | |
407 | (insert "(setq spam-stat-ngood " | |
408 | (number-to-string spam-stat-ngood) | |
409 | " spam-stat-nbad " | |
410 | (number-to-string spam-stat-nbad) | |
411 | " spam-stat (spam-stat-to-hash-table '(") | |
412 | (maphash (lambda (word entry) | |
413 | (prin1 (list word | |
414 | (spam-stat-good entry) | |
415 | (spam-stat-bad entry)))) | |
416 | spam-stat) | |
417 | (insert ")))") | |
418 | (write-file spam-stat-file))) | |
419 | (setq spam-stat-dirty nil))) | |
420 | ||
421 | (defun spam-stat-load () | |
422 | "Read the `spam-stat' hash table from disk." | |
423 | ;; TODO: maybe we should warn the user if spam-stat-dirty is t? | |
424 | (load-file spam-stat-file) | |
425 | (setq spam-stat-dirty nil)) | |
426 | ||
427 | (defun spam-stat-to-hash-table (entries) | |
428 | "Turn list ENTRIES into a hash table and store as `spam-stat'. | |
429 | Every element in ENTRIES has the form \(WORD GOOD BAD) where WORD is | |
430 | the word string, NGOOD is the number of good mails it has appeared in, | |
431 | NBAD is the number of bad mails it has appeared in, GOOD is the number | |
432 | of times it appeared in good mails, and BAD is the number of times it | |
433 | has appeared in bad mails." | |
434 | (let ((table (make-hash-table :test 'equal))) | |
435 | (mapc (lambda (l) | |
436 | (puthash (car l) | |
437 | (spam-stat-make-entry (nth 1 l) (nth 2 l)) | |
438 | table)) | |
439 | entries) | |
440 | table)) | |
441 | ||
442 | (defun spam-stat-reset () | |
443 | "Reset `spam-stat' to an empty hash-table. | |
444 | This deletes all the statistics." | |
445 | (interactive) | |
446 | (setq spam-stat (make-hash-table :test 'equal) | |
447 | spam-stat-ngood 0 | |
448 | spam-stat-nbad 0) | |
449 | (setq spam-stat-dirty t)) | |
450 | ||
451 | ;; Scoring buffers | |
452 | ||
453 | (defvar spam-stat-score-data nil | |
454 | "Raw data used in the last run of `spam-stat-score-buffer'.") | |
455 | ||
456 | (defsubst spam-stat-score-word (word) | |
457 | "Return score for WORD. | |
458 | The default score for unknown words is stored in | |
459 | `spam-stat-unknown-word-score'." | |
460 | (spam-stat-score (gethash word spam-stat))) | |
461 | ||
462 | (defun spam-stat-buffer-words-with-scores () | |
463 | "Process current buffer, return the 15 most conspicuous words. | |
464 | These are the words whose spam-stat differs the most from 0.5. | |
465 | The list returned contains elements of the form \(WORD SCORE DIFF), | |
466 | where DIFF is the difference between SCORE and 0.5." | |
467 | (with-spam-stat-max-buffer-size | |
468 | (with-syntax-table spam-stat-syntax-table | |
469 | (let (result word score) | |
470 | (maphash (lambda (word ignore) | |
471 | (setq score (spam-stat-score-word word) | |
472 | result (cons (list word score (abs (- score 0.5))) | |
473 | result))) | |
474 | (spam-stat-buffer-words)) | |
475 | (setq result (sort result (lambda (a b) (< (nth 2 b) (nth 2 a))))) | |
476 | (setcdr (nthcdr 14 result) nil) | |
477 | result)))) | |
478 | ||
479 | (defun spam-stat-score-buffer () | |
480 | "Return a score describing the spam-probability for this buffer." | |
481 | (setq spam-stat-score-data (spam-stat-buffer-words-with-scores)) | |
482 | (let* ((probs (mapcar (lambda (e) (cadr e)) spam-stat-score-data)) | |
483 | (prod (apply #'* probs))) | |
484 | (/ prod (+ prod (apply #'* (mapcar #'(lambda (x) (- 1 x)) | |
485 | probs)))))) | |
486 | ||
487 | (defun spam-stat-split-fancy () | |
488 | "Return the name of the spam group if the current mail is spam. | |
489 | Use this function on `nnmail-split-fancy'. If you are interested in | |
490 | the raw data used for the last run of `spam-stat-score-buffer', | |
491 | check the variable `spam-stat-score-data'." | |
492 | (condition-case var | |
493 | (progn | |
494 | (set-buffer spam-stat-buffer) | |
495 | (goto-char (point-min)) | |
496 | (when (> (spam-stat-score-buffer) spam-stat-split-fancy-spam-threshhold) | |
497 | (when (boundp 'nnmail-split-trace) | |
498 | (mapc (lambda (entry) | |
499 | (push entry nnmail-split-trace)) | |
500 | spam-stat-score-data)) | |
501 | spam-stat-split-fancy-spam-group)) | |
502 | (error (message "Error in spam-stat-split-fancy: %S" var) | |
503 | nil))) | |
504 | ||
505 | ;; Testing | |
506 | ||
507 | (defun spam-stat-process-directory (dir func) | |
508 | "Process all the regular files in directory DIR using function FUNC." | |
509 | (let* ((files (directory-files dir t "^[^.]")) | |
510 | (max (/ (length files) 100.0)) | |
511 | (count 0)) | |
512 | (with-temp-buffer | |
513 | (dolist (f files) | |
514 | (when (and (file-readable-p f) | |
515 | (file-regular-p f) | |
516 | (> (nth 7 (file-attributes f)) 0)) | |
517 | (setq count (1+ count)) | |
518 | (message "Reading %s: %.2f%%" dir (/ count max)) | |
519 | (insert-file-contents f) | |
520 | (funcall func) | |
521 | (erase-buffer)))))) | |
522 | ||
523 | (defun spam-stat-process-spam-directory (dir) | |
524 | "Process all the regular files in directory DIR as spam." | |
525 | (interactive "D") | |
526 | (spam-stat-process-directory dir 'spam-stat-buffer-is-spam)) | |
527 | ||
528 | (defun spam-stat-process-non-spam-directory (dir) | |
529 | "Process all the regular files in directory DIR as non-spam." | |
530 | (interactive "D") | |
531 | (spam-stat-process-directory dir 'spam-stat-buffer-is-non-spam)) | |
532 | ||
533 | (defun spam-stat-count () | |
534 | "Return size of `spam-stat'." | |
535 | (interactive) | |
536 | (hash-table-count spam-stat)) | |
537 | ||
538 | (defun spam-stat-test-directory (dir) | |
539 | "Test all the regular files in directory DIR for spam. | |
540 | If the result is 1.0, then all files are considered spam. | |
541 | If the result is 0.0, non of the files is considered spam. | |
542 | You can use this to determine error rates." | |
543 | (interactive "D") | |
544 | (let* ((files (directory-files dir t "^[^.]")) | |
545 | (total (length files)) | |
546 | (score 0.0); float | |
547 | (max (/ total 100.0)); float | |
548 | (count 0)) | |
549 | (with-temp-buffer | |
550 | (dolist (f files) | |
551 | (when (and (file-readable-p f) | |
552 | (file-regular-p f) | |
553 | (> (nth 7 (file-attributes f)) 0)) | |
554 | (setq count (1+ count)) | |
555 | (message "Reading %.2f%%, score %.2f%%" | |
556 | (/ count max) (/ score count)) | |
557 | (insert-file-contents f) | |
558 | (when (> (spam-stat-score-buffer) 0.9) | |
559 | (setq score (1+ score))) | |
560 | (erase-buffer)))) | |
561 | (message "Final score: %d / %d = %f" score total (/ score total)))) | |
562 | ||
563 | ;; Shrinking the dictionary | |
564 | ||
565 | (defun spam-stat-reduce-size (&optional count) | |
566 | "Reduce the size of `spam-stat'. | |
567 | This removes all words that occur less than COUNT from the dictionary. | |
568 | COUNT defaults to 5" | |
569 | (interactive) | |
570 | (setq count (or count 5)) | |
571 | (maphash (lambda (key entry) | |
572 | (when (< (+ (spam-stat-good entry) | |
573 | (spam-stat-bad entry)) | |
574 | count) | |
575 | (remhash key spam-stat))) | |
576 | spam-stat) | |
577 | (setq spam-stat-dirty t)) | |
578 | ||
579 | (defun spam-stat-install-hooks-function () | |
580 | "Install the spam-stat function hooks" | |
581 | (interactive) | |
582 | (add-hook 'nnmail-prepare-incoming-message-hook | |
583 | 'spam-stat-store-current-buffer) | |
584 | (add-hook 'gnus-select-article-hook | |
585 | 'spam-stat-store-gnus-article-buffer)) | |
586 | ||
587 | (when spam-stat-install-hooks | |
588 | (spam-stat-install-hooks-function)) | |
589 | ||
590 | (defun spam-stat-unload-hook () | |
591 | "Uninstall the spam-stat function hooks" | |
592 | (interactive) | |
593 | (remove-hook 'nnmail-prepare-incoming-message-hook | |
594 | 'spam-stat-store-current-buffer) | |
595 | (remove-hook 'gnus-select-article-hook | |
596 | 'spam-stat-store-gnus-article-buffer)) | |
597 | ||
e0281050 RS |
598 | (add-hook 'spam-stat-unload-hook 'spam-stat-unload-hook) |
599 | ||
23f87bed MB |
600 | (provide 'spam-stat) |
601 | ||
602 | ;;; arch-tag: ff1d2200-8ddb-42fb-bb7b-1b5e20448554 | |
603 | ;;; spam-stat.el ends here |