From 2e8541cf15b98bd4e92e6af2b6b126c93ac749e9 Mon Sep 17 00:00:00 2001
From: "Taylor R. Campbell" <net/mumble/campbell>
Date: Mon, 12 Jun 2006 20:46:28 +0000
Subject: [PATCH] Implement two kinds of filtering in Edwin's simple news
reader: filtering out messages that match criteria, such as spam; and
filtering headers of kept messages.
See the documentation strings for the new Edwin variables
NEWS-HEADER-FILTER and NEWS-KEPT-HEADERS. New procedures
NEWS-HEADER-SPLITTING-FILTER and NEWS-HEADER-REGEXP-FILTER are useful
for constructing the value of NEWS-HEADER-FILTER. For example,
I have this in my .edwin, so that any messages with xref headers
whose values contain `gmane.spam.detected' on any Gmane news servers
will be ignored:
(add-event-receiver! (ref-variable news-group-mode-hook)
(let ((gmane-spam-filter
(news-header-regexp-filter
'(("xref" . ".*gmane\\.spam\\.detected.*")))))
(lambda (buffer)
(if (string-prefix? "gmane." (buffer-name buffer))
(local-set-variable! news-header-filter
gmane-spam-filter
buffer)))))
---
v7/src/edwin/edwin.pkg | 10 +-
v7/src/edwin/snr.scm | 261 ++++++++++++++++++++++++++++++++---------
2 files changed, 213 insertions(+), 58 deletions(-)
diff --git a/v7/src/edwin/edwin.pkg b/v7/src/edwin/edwin.pkg
index 030e8559d..2709519ff 100644
--- a/v7/src/edwin/edwin.pkg
+++ b/v7/src/edwin/edwin.pkg
@@ -1,6 +1,6 @@
#| -*-Scheme-*-
-$Id: edwin.pkg,v 1.293 2006/02/06 18:46:44 cph Exp $
+$Id: edwin.pkg,v 1.294 2006/06/12 20:46:28 riastradh Exp $
Copyright 1989,1990,1991,1992,1993,1994 Massachusetts Institute of Technology
Copyright 1995,1996,1997,1998,1999,2000 Massachusetts Institute of Technology
@@ -1761,12 +1761,14 @@ USA.
edwin-command$news-move-to-summary
edwin-command$news-new-groups
edwin-command$news-next-article
+ edwin-command$news-next-article-in-thread
edwin-command$news-next-thread-article
edwin-command$news-next-unread-article
edwin-command$news-next-unread-article-in-thread
edwin-command$news-output-article
edwin-command$news-output-article-to-rmail-file
edwin-command$news-previous-article
+ edwin-command$news-previous-article-in-thread
edwin-command$news-previous-thread-article
edwin-command$news-previous-unread-article
edwin-command$news-previous-unread-article-in-thread
@@ -1815,9 +1817,11 @@ USA.
edwin-variable$news-group-show-author-name
edwin-variable$news-group-show-context-headers
edwin-variable$news-group-show-seen-headers
+ edwin-variable$news-header-filter
edwin-variable$news-hide-groups
edwin-variable$news-initially-collapse-threads
edwin-variable$news-join-threads-with-same-subject
+ edwin-variable$news-kept-headers
edwin-variable$news-refresh-group-when-selected
edwin-variable$news-server
edwin-variable$news-server-initial-refresh
@@ -1828,7 +1832,9 @@ USA.
edwin-variable$news-show-nonexistent-groups
edwin-variable$news-show-unsubscribed-groups
edwin-variable$news-sort-groups
- edwin-variable$news-split-threads-on-subject-changes))
+ edwin-variable$news-split-threads-on-subject-changes
+ news-header-splitting-filter
+ news-header-regexp-filter))
(define-package (edwin nntp)
(files "nntp")
diff --git a/v7/src/edwin/snr.scm b/v7/src/edwin/snr.scm
index 1826b82da..8a17fc490 100644
--- a/v7/src/edwin/snr.scm
+++ b/v7/src/edwin/snr.scm
@@ -1,9 +1,9 @@
#| -*-Scheme-*-
-$Id: snr.scm,v 1.65 2004/06/07 19:49:55 cph Exp $
+$Id: snr.scm,v 1.66 2006/06/12 20:46:28 riastradh Exp $
Copyright 1995,1996,1997,1998,1999,2000 Massachusetts Institute of Technology
-Copyright 2001,2003,2004 Massachusetts Institute of Technology
+Copyright 2001,2003,2004,2006 Massachusetts Institute of Technology
This file is part of MIT/GNU Scheme.
@@ -81,6 +81,15 @@ This variable can be set to #F to disable the timeout altogether.
[THIS VARIABLE CURRENTLY HAS NO EFFECT.]"
#f
(lambda (object) (or (not object) (exact-nonnegative-integer? object))))
+
+(define-variable news-header-filter
+ "Procedure for filtering news headers, or #F for no filter.
+Every header read by the news reader is filtered through this filter. If it
+ returns false, the header is ignored."
+ #f
+ (lambda (filter)
+ (or (not filter)
+ (procedure-of-arity? filter 1))))
;;; Variables for News-group-list buffers:
@@ -263,6 +272,24 @@ The default value of this variable is (SUBSCRIBED (HEADERS BODIES))."
(or (eq? 'HEADERS element)
(eq? 'BODIES element))))
(null? (cddr object)))))
+
+(define-variable news-kept-headers
+ "A list of regular expressions matching header fields to display.
+Header fields matching these regexps are shown in the given order, and
+ other header fields are hidden.
+This variable overrides RMAIL-IGNORED-HEADERS; to use RMAIL-IGNORED-HEADERS,
+ set NEWS-KEPT-HEADERS to #F."
+ '("date:" "from:" "newsgroups:" "subject:")
+ (lambda (obj)
+ (or (not obj)
+ (list-of-type? obj regular-expression?))))
+
+(define (regular-expression? obj)
+ (or (string? obj)
+ (compiled-regexp? obj)
+ ;++ This should be stricter about strings, and it should also
+ ;++ support REXPs.
+ ))
(define-command rnews
"Start a News reader.
@@ -2352,15 +2379,53 @@ Otherwise, the standard pruned header is shown."
(define (insert-news-header header buffer truncate?)
(let ((hend (mark-left-inserting-copy (buffer-start buffer)))
- (text (news-header:text header)))
- (if (and (not (string-null? text))
- (char=? #\newline (string-ref text 0)))
- (insert-substring text 1 (string-length text) hend)
- (insert-string text hend))
- (insert-newline hend)
- (if truncate? (delete-ignored-headers (buffer-start buffer) hend))
- (mark-temporary! hend))
- (buffer-put! buffer 'NEWS-ARTICLE-HEADER-TRUNCATED? truncate?))
+ (text (news-header:text header)))
+ (cond ((and truncate?
+ (ref-variable news-kept-headers))
+ => (lambda (regexps)
+ (insert-filtered-news-header text regexps hend buffer)
+ (insert-newline hend)))
+ (else
+ (if (and (not (string-null? text))
+ (char=? #\newline (string-ref text 0)))
+ (insert-substring text 1 (string-length text) hend)
+ (insert-string text hend))
+ (insert-newline hend)
+ (if truncate? (delete-ignored-headers (buffer-start buffer) hend))))
+ (mark-temporary! hend)
+ (buffer-put! buffer 'NEWS-ARTICLE-HEADER-TRUNCATED? truncate?)))
+
+(define (insert-filtered-news-header text regexps mark buffer)
+ (for-each (lambda (regexp)
+ (cond ((re-string-search-forward (string-append "^" regexp)
+ text
+ #t)
+ => (lambda (match)
+ (let ((start-index (re-match-start-index 0 match)))
+ (insert-substring text start-index
+ (find-header-end text
+ start-index)
+ mark))
+ (insert-newline mark)))))
+ regexps))
+
+(define (find-header-end text start-index)
+ (let* ((limit (string-length text))
+ (scan-line (lambda (start)
+ (cond ((substring-find-next-char text start limit
+ #\newline)
+ => fix:1+)
+ (else #f)))))
+ (let loop ((index (scan-line start-index)))
+ (cond ((or (not index) (fix:= index limit))
+ limit)
+ ((let ((char (string-ref text index)))
+ (or (char=? char #\space)
+ (char=? char #\tab)))
+ (loop (scan-line index)))
+ (else
+ ;; Lose the trailing newline.
+ (fix:-1+ index))))))
(define (delete-ignored-headers hstart hend)
(let ((regexp (ref-variable rmail-ignored-headers hstart)))
@@ -2405,6 +2470,8 @@ This mode's commands include:
\\[news-previous-article] read the previous article
\\[news-next-unread-article] read the next unread article
\\[news-previous-unread-article] read the previous unread article
+\\[news-next-article-in-thread] read the next article in this thread
+\\[news-previous-article-in-thread] read the previous article in this thiread
\\[news-next-unread-article-in-thread] read the next unread article in this thread
\\[news-previous-unread-article-in-thread] read the previous unread article in this thread
\\[news-next-thread-article] read the first article of the next thread
@@ -2435,6 +2502,7 @@ This mode's commands include:
(define-key 'news-article #\rubout '(news-article . #\m-v))
(define-key 'news-article #\c 'news-toggle-article-context)
(define-key 'news-article #\d 'news-next-article)
+(define-key 'news-article #\D 'news-next-article-in-thread)
(define-key 'news-article #\f 'news-forward-article)
(define-key 'news-article #\i 'news-ignore-article-thread)
(define-key 'news-article #\n 'news-next-unread-article)
@@ -2449,6 +2517,7 @@ This mode's commands include:
(define-key 'news-article #\R 'news-compose-followup-article)
(define-key 'news-article #\t 'news-toggle-article-header)
(define-key 'news-article #\u 'news-previous-article)
+(define-key 'news-article #\U 'news-previous-article-in-thread)
(define-command news-next-article
"Select a buffer containing the next article in the News group.
@@ -2490,6 +2559,28 @@ Kill the current buffer in either case."
(lambda (buffer header)
(news-group-buffer:previous-header buffer header
news-header:unread?)))))
+
+(define-command news-next-article-in-thread
+ "Select a buffer containing the next article in this thread.
+If there is no such article, return to the News-group buffer.
+Kill the current buffer in either case."
+ ()
+ (lambda ()
+ (news-article-header-motion-command
+ (lambda (buffer header)
+ buffer
+ (news-thread:next-header header news-header:real?)))))
+
+(define-command news-previous-article-in-thread
+ "Select a buffer containing the previous article in this thread.
+If there is no such article, return to the News-group buffer.
+Kill the current buffer in either case."
+ ()
+ (lambda ()
+ (news-article-header-motion-command
+ (lambda (buffer header)
+ buffer
+ (news-thread:previous-header header news-header:real?)))))
(define-command news-next-unread-article-in-thread
"Select a buffer containing the next unread article in this thread.
@@ -4049,53 +4140,28 @@ With prefix arg, replaces the file with the list information."
(define (news-group:get-headers group argument buffer)
(let ((connection (news-group:connection group))
- (all?
- (or (command-argument-multiplier-only? argument)
- (ref-variable news-group-show-seen-headers buffer)))
- (limit
- (and argument
- (not (command-argument-multiplier-only? argument))
- (command-argument-value argument))))
+ (all?
+ (or (command-argument-multiplier-only? argument)
+ (ref-variable news-group-show-seen-headers buffer)))
+ (limit
+ (and argument
+ (not (command-argument-multiplier-only? argument))
+ (command-argument-value argument))))
(if (and (command-argument-multiplier-only? argument)
- (nntp-connection:closed? connection))
- (nntp-connection:reopen connection))
+ (nntp-connection:closed? connection))
+ (nntp-connection:reopen connection))
(if (and (ref-variable news-refresh-group-when-selected
- (news-server-buffer buffer #f))
- (not (nntp-connection:closed? connection)))
- (news-group:update-ranges! group))
- (call-with-values
- (lambda ()
- (split-list
- (news-group:headers
- group
- (if all?
- (news-group:all-header-numbers group)
- (let ((ns (news-group:unread-header-numbers group)))
- (if limit
- (let ((lns (length ns)))
- (cond ((<= lns (abs limit)) ns)
- ((< limit 0) (list-head ns (- limit)))
- (else (list-tail ns (- (length ns) limit)))))
- ns)))
- (let ((table (news-group:get-ignored-subjects group #f)))
- (if table
- (let ((t (get-universal-time))
- (show-ignored? (not all?)))
- (lambda (header)
- (and (news-header:ignore?! header table t)
- (begin
- (set-news-header:status! header #\I)
- (article-number-seen! group
- (news-header:number header))
- show-ignored?))))
- (lambda (header) header #f))))
- news-header?))
- (lambda (headers invalid)
- (for-each (lambda (entry)
- (if (not (eq? (car entry) 'UNREACHABLE-ARTICLE))
- (article-number-seen! group (cdr entry))))
- invalid)
- headers))))
+ (news-server-buffer buffer #f))
+ (not (nntp-connection:closed? connection)))
+ (news-group:update-ranges! group))
+ (receive (headers invalid)
+ (split-list (news-group:headers* group all? limit buffer)
+ news-header?)
+ (for-each (lambda (entry)
+ (if (not (eq? (car entry) 'UNREACHABLE-ARTICLE))
+ (article-number-seen! group (cdr entry))))
+ invalid)
+ headers)))
(define (news-group:get-unread-headers group buffer)
(news-group:update-ranges! group)
@@ -4106,6 +4172,89 @@ With prefix arg, replaces the file with the list information."
(news-group:get-headers group #f buffer))
(news-group:close-database group))
+(define (news-group:headers* group all? limit context)
+ (news-group:headers
+ group
+ (if all?
+ (news-group:all-header-numbers group)
+ (let ((ns (news-group:unread-header-numbers group)))
+ (if limit
+ (let ((lns (length ns)))
+ (cond ((<= lns (abs limit)) ns)
+ ((< limit 0) (list-head ns (- limit)))
+ (else (list-tail ns (- (length ns) limit)))))
+ ns)))
+ (let ((ignore-header?
+ (let ((filter (ref-variable news-header-filter context)))
+ (or (and filter
+ (lambda (header)
+ (not (filter header))))
+ (lambda (header) header #f))))
+ (table
+ (news-group:get-ignored-subjects group #f)))
+ (if table
+ (let ((t (get-universal-time))
+ (show-ignored? (not all?)))
+ (lambda (header)
+ (and (ignore-header? header)
+ (news-header:ignore?! header table t)
+ (begin
+ (set-news-header:status! header #\I)
+ (article-number-seen! group
+ (news-header:number header))
+ show-ignored?))))
+ ignore-header?))))
+
+;;;; Header Filter Combinators
+
+(define (news-header-splitting-filter unit-filter)
+ (lambda (header)
+ (let* ((text (news-header:text header))
+ (limit (string-length text))
+ (start (if (and (fix:> limit 1)
+ (char=? (string-ref text 0) #\newline))
+ 1
+ 0)))
+ (let loop ((start start) (index start))
+ (cond ((substring-find-next-char text index limit #\newline)
+ => (lambda (line-end)
+ (let ((next-line-start (fix:1+ line-end)))
+ (if (fix:= next-line-start limit)
+ (unit-filter text start line-end)
+ (let ((char (string-ref text next-line-start)))
+ (if (or (char=? char #\space)
+ (char=? char #\tab))
+ (loop start next-line-start)
+ (and (unit-filter text start line-end)
+ (loop next-line-start
+ next-line-start))))))))
+ ((fix:= start limit)
+ #t)
+ (else
+ (unit-filter text start limit)))))))
+
+(define (news-header-regexp-filter specifiers)
+ (news-header-splitting-filter
+ (let ((table (alist->string-table
+ (map (lambda (specifier)
+ (cons (car specifier)
+ (re-compile-pattern (cdr specifier)
+ #f))) ; Don't case-fold.
+ specifiers)
+ #t))) ; Case-insensitive
+ (lambda (text start end)
+ (cond ((substring-find-next-char text start end #\:)
+ => (lambda (colon-index)
+ (cond ((string-table-get table
+ (substring text start colon-index))
+ => (lambda (regexp)
+ (not (re-substring-match regexp text
+ ;; Skip colon & space.
+ (fix:+ colon-index 2)
+ end))))
+ (else #t))))
+ (else #t))))))
+
(define (article-number-seen! group number)
(set-news-group:ranges-deleted!
group
--
2.25.1