1 |
;;; nnweb.el --- retrieving articles via web search engines |
;;; nnweb.el --- retrieving articles via web search engines |
2 |
;; Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001 |
;; Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003 |
3 |
;; Free Software Foundation, Inc. |
;; Free Software Foundation, Inc. |
4 |
|
|
5 |
;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org> |
;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org> |
24 |
|
|
25 |
;;; Commentary: |
;;; Commentary: |
26 |
|
|
27 |
;; Note: You need to have `url' and `w3' installed for this |
;; Note: You need to have `w3' installed for some functions to work. |
|
;; backend to work. |
|
28 |
|
|
29 |
;;; Code: |
;;; Code: |
30 |
|
|
36 |
(require 'gnus) |
(require 'gnus) |
37 |
(require 'nnmail) |
(require 'nnmail) |
38 |
(require 'mm-util) |
(require 'mm-util) |
39 |
(eval-when-compile |
(require 'mm-url) |
40 |
|
(eval-and-compile |
41 |
(ignore-errors |
(ignore-errors |
42 |
(require 'w3) |
(require 'url))) |
43 |
(require 'url) |
(autoload 'w3-parse-buffer "w3-parse") |
|
(require 'w3-forms))) |
|
|
|
|
|
;; Report failure to find w3 at load time if appropriate. |
|
|
(unless noninteractive |
|
|
(eval '(progn |
|
|
(require 'w3) |
|
|
(require 'url) |
|
|
(require 'w3-forms)))) |
|
44 |
|
|
45 |
(nnoo-declare nnweb) |
(nnoo-declare nnweb) |
46 |
|
|
49 |
|
|
50 |
(defvoo nnweb-type 'google |
(defvoo nnweb-type 'google |
51 |
"What search engine type is being used. |
"What search engine type is being used. |
52 |
Valid types include `google', `dejanews', `dejanewsold', `reference', |
Valid types include `google', `dejanews', and `gmane'.") |
|
and `altavista'.") |
|
53 |
|
|
54 |
(defvar nnweb-type-definition |
(defvar nnweb-type-definition |
55 |
'( |
'((google |
|
(google |
|
|
;;(article . nnweb-google-wash-article) |
|
|
;;(id . "http://groups.google.com/groups?as_umsgid=%s") |
|
56 |
(article . ignore) |
(article . ignore) |
57 |
(id . "http://groups.google.com/groups?selm=%s&output=gplain") |
(id . "http://groups.google.com/groups?selm=%s&output=gplain") |
|
;;(reference . nnweb-google-reference) |
|
58 |
(reference . identity) |
(reference . identity) |
59 |
(map . nnweb-google-create-mapping) |
(map . nnweb-google-create-mapping) |
60 |
(search . nnweb-google-search) |
(search . nnweb-google-search) |
61 |
(address . "http://groups.google.com/groups") |
(address . "http://groups.google.com/groups") |
62 |
(identifier . nnweb-google-identity)) |
(identifier . nnweb-google-identity)) |
63 |
(dejanews ;; alias of google |
(dejanews ;; alias of google |
|
;;(article . nnweb-google-wash-article) |
|
|
;;(id . "http://groups.google.com/groups?as_umsgid=%s") |
|
64 |
(article . ignore) |
(article . ignore) |
65 |
(id . "http://groups.google.com/groups?selm=%s&output=gplain") |
(id . "http://groups.google.com/groups?selm=%s&output=gplain") |
|
;;(reference . nnweb-google-reference) |
|
66 |
(reference . identity) |
(reference . identity) |
67 |
(map . nnweb-google-create-mapping) |
(map . nnweb-google-create-mapping) |
68 |
(search . nnweb-google-search) |
(search . nnweb-google-search) |
69 |
(address . "http://groups.google.com/groups") |
(address . "http://groups.google.com/groups") |
70 |
(identifier . nnweb-google-identity)) |
(identifier . nnweb-google-identity)) |
71 |
;;; (dejanews |
(gmane |
72 |
;;; (article . ignore) |
(article . nnweb-gmane-wash-article) |
73 |
;;; (id . "http://search.dejanews.com/msgid.xp?MID=%s&fmt=text") |
(id . "http://gmane.org/view.php?group=%s") |
74 |
;;; (map . nnweb-dejanews-create-mapping) |
(reference . identity) |
75 |
;;; (search . nnweb-dejanews-search) |
(map . nnweb-gmane-create-mapping) |
76 |
;;; (address . "http://www.deja.com/=dnc/qs.xp") |
(search . nnweb-gmane-search) |
77 |
;;; (identifier . nnweb-dejanews-identity)) |
(address . "http://gmane.org/") |
78 |
;;; (dejanewsold |
(identifier . nnweb-gmane-identity))) |
|
;;; (article . ignore) |
|
|
;;; (map . nnweb-dejanews-create-mapping) |
|
|
;;; (search . nnweb-dejanewsold-search) |
|
|
;;; (address . "http://www.deja.com/dnquery.xp") |
|
|
;;; (identifier . nnweb-dejanews-identity)) |
|
|
(reference |
|
|
(article . nnweb-reference-wash-article) |
|
|
(map . nnweb-reference-create-mapping) |
|
|
(search . nnweb-reference-search) |
|
|
(address . "http://www.reference.com/cgi-bin/pn/go") |
|
|
(identifier . identity)) |
|
|
(altavista |
|
|
(article . nnweb-altavista-wash-article) |
|
|
(map . nnweb-altavista-create-mapping) |
|
|
(search . nnweb-altavista-search) |
|
|
(address . "http://www.altavista.digital.com/cgi-bin/query") |
|
|
(id . "/cgi-bin/news?id@%s") |
|
|
(identifier . identity))) |
|
79 |
"Type-definition alist.") |
"Type-definition alist.") |
80 |
|
|
81 |
(defvoo nnweb-search nil |
(defvoo nnweb-search nil |
82 |
"Search string to feed to DejaNews.") |
"Search string to feed to Google.") |
83 |
|
|
84 |
(defvoo nnweb-max-hits 999 |
(defvoo nnweb-max-hits 999 |
85 |
"Maximum number of hits to display.") |
"Maximum number of hits to display.") |
163 |
(url (and header (mail-header-xref header)))) |
(url (and header (mail-header-xref header)))) |
164 |
(when (or (and url |
(when (or (and url |
165 |
(mm-with-unibyte-current-buffer |
(mm-with-unibyte-current-buffer |
166 |
(nnweb-fetch-url url))) |
(mm-url-insert url))) |
167 |
(and (stringp article) |
(and (stringp article) |
168 |
(nnweb-definition 'id t) |
(nnweb-definition 'id t) |
169 |
(let ((fetch (nnweb-definition 'id)) |
(let ((fetch (nnweb-definition 'id)) |
173 |
(when (and fetch art) |
(when (and fetch art) |
174 |
(setq url (format fetch art)) |
(setq url (format fetch art)) |
175 |
(mm-with-unibyte-current-buffer |
(mm-with-unibyte-current-buffer |
176 |
(nnweb-fetch-url url)) |
(mm-url-insert url)) |
177 |
(if (nnweb-definition 'reference t) |
(if (nnweb-definition 'reference t) |
178 |
(setq article |
(setq article |
179 |
(funcall (nnweb-definition |
(funcall (nnweb-definition |
203 |
(nnweb-possibly-change-server group server)) |
(nnweb-possibly-change-server group server)) |
204 |
|
|
205 |
(deffoo nnweb-asynchronous-p () |
(deffoo nnweb-asynchronous-p () |
206 |
t) |
nil) |
207 |
|
|
208 |
(deffoo nnweb-request-create-group (group &optional server args) |
(deffoo nnweb-request-create-group (group &optional server args) |
209 |
(nnweb-possibly-change-server nil server) |
(nnweb-possibly-change-server nil server) |
302 |
nnweb-type nnweb-search server)) |
nnweb-type nnweb-search server)) |
303 |
(current-buffer)))))) |
(current-buffer)))))) |
304 |
|
|
|
(defun nnweb-fetch-url (url) |
|
|
(let (buf) |
|
|
(save-excursion |
|
|
(if (not nnheader-callback-function) |
|
|
(progn |
|
|
(with-temp-buffer |
|
|
(mm-enable-multibyte) |
|
|
(let ((coding-system-for-read 'binary) |
|
|
(coding-system-for-write 'binary) |
|
|
(default-process-coding-system 'binary)) |
|
|
(nnweb-insert url)) |
|
|
(setq buf (buffer-string))) |
|
|
(erase-buffer) |
|
|
(insert buf) |
|
|
t) |
|
|
(nnweb-url-retrieve-asynch |
|
|
url 'nnweb-callback (current-buffer) nnheader-callback-function) |
|
|
t)))) |
|
|
|
|
|
(defun nnweb-callback (buffer callback) |
|
|
(when (gnus-buffer-live-p url-working-buffer) |
|
|
(save-excursion |
|
|
(set-buffer url-working-buffer) |
|
|
(funcall (nnweb-definition 'article)) |
|
|
(nnweb-decode-entities) |
|
|
(set-buffer buffer) |
|
|
(goto-char (point-max)) |
|
|
(insert-buffer-substring url-working-buffer)) |
|
|
(funcall callback t) |
|
|
(gnus-kill-buffer url-working-buffer))) |
|
|
|
|
|
(defun nnweb-url-retrieve-asynch (url callback &rest data) |
|
|
(let ((url-request-method "GET") |
|
|
(old-asynch url-be-asynchronous) |
|
|
(url-request-data nil) |
|
|
(url-request-extra-headers nil) |
|
|
(url-working-buffer (generate-new-buffer-name " *nnweb*"))) |
|
|
(setq-default url-be-asynchronous t) |
|
|
(save-excursion |
|
|
(set-buffer (get-buffer-create url-working-buffer)) |
|
|
(setq url-current-callback-data data |
|
|
url-be-asynchronous t |
|
|
url-current-callback-func callback) |
|
|
(url-retrieve url nil)) |
|
|
(setq-default url-be-asynchronous old-asynch))) |
|
|
|
|
|
(if (fboundp 'url-retrieve-synchronously) |
|
|
(defun nnweb-url-retrieve-asynch (url callback &rest data) |
|
|
(url-retrieve url callback data))) |
|
|
|
|
|
;;; |
|
|
;;; DejaNews functions. |
|
|
;;; |
|
|
|
|
|
(defun nnweb-dejanews-create-mapping () |
|
|
"Perform the search and create a number-to-url alist." |
|
|
(save-excursion |
|
|
(set-buffer nnweb-buffer) |
|
|
(erase-buffer) |
|
|
(when (funcall (nnweb-definition 'search) nnweb-search) |
|
|
(let ((i 0) |
|
|
(more t) |
|
|
(case-fold-search t) |
|
|
(active (or (cadr (assoc nnweb-group nnweb-group-alist)) |
|
|
(cons 1 0))) |
|
|
subject date from |
|
|
map url parse a table group text) |
|
|
(while more |
|
|
;; Go through all the article hits on this page. |
|
|
(goto-char (point-min)) |
|
|
(setq parse (w3-parse-buffer (current-buffer)) |
|
|
table (nth 1 (nnweb-parse-find-all 'table parse))) |
|
|
(dolist (row (nth 2 (car (nth 2 table)))) |
|
|
(setq a (nnweb-parse-find 'a row) |
|
|
url (cdr (assq 'href (nth 1 a))) |
|
|
text (nreverse (nnweb-text row))) |
|
|
(when a |
|
|
(setq subject (nth 4 text) |
|
|
group (nth 2 text) |
|
|
date (nth 1 text) |
|
|
from (nth 0 text)) |
|
|
(if (string-match "\\([0-9]+\\)/\\([0-9]+\\)/\\([0-9]+\\)" date) |
|
|
(setq date (format "%s %s 00:00:00 %s" |
|
|
(car (rassq (string-to-number |
|
|
(match-string 2 date)) |
|
|
parse-time-months)) |
|
|
(match-string 3 date) |
|
|
(match-string 1 date))) |
|
|
(setq date "Jan 1 00:00:00 0000")) |
|
|
(incf i) |
|
|
(setq url (concat url "&fmt=text")) |
|
|
(when (string-match "&context=[^&]+" url) |
|
|
(setq url (replace-match "" t t url))) |
|
|
(unless (nnweb-get-hashtb url) |
|
|
(push |
|
|
(list |
|
|
(incf (cdr active)) |
|
|
(make-full-mail-header |
|
|
(cdr active) (concat subject " (" group ")") from date |
|
|
(concat "<" (nnweb-identifier url) "@dejanews>") |
|
|
nil 0 0 url)) |
|
|
map) |
|
|
(nnweb-set-hashtb (cadar map) (car map))))) |
|
|
;; See whether there is a "Get next 20 hits" button here. |
|
|
(goto-char (point-min)) |
|
|
(if (or (not (re-search-forward |
|
|
"HREF=\"\\([^\"]+\\)\"[<>b]+Next result" nil t)) |
|
|
(>= i nnweb-max-hits)) |
|
|
(setq more nil) |
|
|
;; Yup -- fetch it. |
|
|
(setq more (match-string 1)) |
|
|
(erase-buffer) |
|
|
(url-insert-file-contents more))) |
|
|
;; Return the articles in the right order. |
|
|
(setq nnweb-articles |
|
|
(sort (nconc nnweb-articles map) 'car-less-than-car)))))) |
|
|
|
|
|
(defun nnweb-dejanews-search (search) |
|
|
(nnweb-insert |
|
|
(concat |
|
|
(nnweb-definition 'address) |
|
|
"?" |
|
|
(nnweb-encode-www-form-urlencoded |
|
|
`(("ST" . "PS") |
|
|
("svcclass" . "dnyr") |
|
|
("QRY" . ,search) |
|
|
("defaultOp" . "AND") |
|
|
("DBS" . "1") |
|
|
("OP" . "dnquery.xp") |
|
|
("LNG" . "ALL") |
|
|
("maxhits" . "100") |
|
|
("threaded" . "0") |
|
|
("format" . "verbose2") |
|
|
("showsort" . "date") |
|
|
("agesign" . "1") |
|
|
("ageweight" . "1"))))) |
|
|
t) |
|
|
|
|
|
(defun nnweb-dejanewsold-search (search) |
|
|
(nnweb-fetch-form |
|
|
(nnweb-definition 'address) |
|
|
`(("query" . ,search) |
|
|
("defaultOp" . "AND") |
|
|
("svcclass" . "dnold") |
|
|
("maxhits" . "100") |
|
|
("format" . "verbose2") |
|
|
("threaded" . "0") |
|
|
("showsort" . "date") |
|
|
("agesign" . "1") |
|
|
("ageweight" . "1"))) |
|
|
t) |
|
|
|
|
|
(defun nnweb-dejanews-identity (url) |
|
|
"Return an unique identifier based on URL." |
|
|
(if (string-match "AN=\\([0-9]+\\)" url) |
|
|
(match-string 1 url) |
|
|
url)) |
|
|
|
|
|
;;; |
|
|
;;; InReference |
|
|
;;; |
|
|
|
|
|
(defun nnweb-reference-create-mapping () |
|
|
"Perform the search and create a number-to-url alist." |
|
|
(save-excursion |
|
|
(set-buffer nnweb-buffer) |
|
|
(erase-buffer) |
|
|
(when (funcall (nnweb-definition 'search) nnweb-search) |
|
|
(let ((i 0) |
|
|
(more t) |
|
|
(case-fold-search t) |
|
|
(active (or (cadr (assoc nnweb-group nnweb-group-alist)) |
|
|
(cons 1 0))) |
|
|
Subject Score Date Newsgroups From Message-ID |
|
|
map url) |
|
|
(while more |
|
|
;; Go through all the article hits on this page. |
|
|
(goto-char (point-min)) |
|
|
(search-forward "</pre><hr>" nil t) |
|
|
(delete-region (point-min) (point)) |
|
|
(goto-char (point-min)) |
|
|
(while (re-search-forward "^ +[0-9]+\\." nil t) |
|
|
(narrow-to-region |
|
|
(point) |
|
|
(if (re-search-forward "^$" nil t) |
|
|
(match-beginning 0) |
|
|
(point-max))) |
|
|
(goto-char (point-min)) |
|
|
(when (looking-at ".*href=\"\\([^\"]+\\)\"") |
|
|
(setq url (match-string 1))) |
|
|
(nnweb-remove-markup) |
|
|
(goto-char (point-min)) |
|
|
(while (search-forward "\t" nil t) |
|
|
(replace-match " ")) |
|
|
(goto-char (point-min)) |
|
|
(while (re-search-forward "^\\([^:]+\\): \\(.*\\)$" nil t) |
|
|
(set (intern (match-string 1)) (match-string 2))) |
|
|
(widen) |
|
|
(search-forward "</pre>" nil t) |
|
|
(incf i) |
|
|
(unless (nnweb-get-hashtb url) |
|
|
(push |
|
|
(list |
|
|
(incf (cdr active)) |
|
|
(make-full-mail-header |
|
|
(cdr active) (concat "(" Newsgroups ") " Subject) From Date |
|
|
Message-ID |
|
|
nil 0 (string-to-int Score) url)) |
|
|
map) |
|
|
(nnweb-set-hashtb (cadar map) (car map)))) |
|
|
(setq more nil)) |
|
|
;; Return the articles in the right order. |
|
|
(setq nnweb-articles |
|
|
(sort (nconc nnweb-articles map) 'car-less-than-car)))))) |
|
|
|
|
|
(defun nnweb-reference-wash-article () |
|
|
(let ((case-fold-search t)) |
|
|
(goto-char (point-min)) |
|
|
(re-search-forward "^</center><hr>" nil t) |
|
|
(delete-region (point-min) (point)) |
|
|
(search-forward "<pre>" nil t) |
|
|
(forward-line -1) |
|
|
(let ((body (point-marker))) |
|
|
(search-forward "</pre>" nil t) |
|
|
(delete-region (point) (point-max)) |
|
|
(nnweb-remove-markup) |
|
|
(goto-char (point-min)) |
|
|
(while (looking-at " *$") |
|
|
(gnus-delete-line)) |
|
|
(narrow-to-region (point-min) body) |
|
|
(while (and (re-search-forward "^$" nil t) |
|
|
(not (eobp))) |
|
|
(gnus-delete-line)) |
|
|
(goto-char (point-min)) |
|
|
(while (looking-at "\\(^[^ ]+:\\) *") |
|
|
(replace-match "\\1 " t) |
|
|
(forward-line 1)) |
|
|
(goto-char (point-min)) |
|
|
(when (re-search-forward "^References:" nil t) |
|
|
(narrow-to-region |
|
|
(point) (if (re-search-forward "^$\\|^[^:]+:" nil t) |
|
|
(match-beginning 0) |
|
|
(point-max))) |
|
|
(goto-char (point-min)) |
|
|
(while (not (eobp)) |
|
|
(unless (looking-at "References") |
|
|
(insert "\t") |
|
|
(forward-line 1))) |
|
|
(goto-char (point-min)) |
|
|
(while (search-forward "," nil t) |
|
|
(replace-match " " t t))) |
|
|
(widen) |
|
|
(nnweb-decode-entities) |
|
|
(set-marker body nil)))) |
|
|
|
|
|
(defun nnweb-reference-search (search) |
|
|
(url-insert-file-contents |
|
|
(concat |
|
|
(nnweb-definition 'address) |
|
|
"?" |
|
|
(nnweb-encode-www-form-urlencoded |
|
|
`(("search" . "advanced") |
|
|
("querytext" . ,search) |
|
|
("subj" . "") |
|
|
("name" . "") |
|
|
("login" . "") |
|
|
("host" . "") |
|
|
("organization" . "") |
|
|
("groups" . "") |
|
|
("keywords" . "") |
|
|
("choice" . "Search") |
|
|
("startmonth" . "Jul") |
|
|
("startday" . "25") |
|
|
("startyear" . "1996") |
|
|
("endmonth" . "Aug") |
|
|
("endday" . "24") |
|
|
("endyear" . "1996") |
|
|
("mode" . "Quick") |
|
|
("verbosity" . "Verbose") |
|
|
("ranking" . "Relevance") |
|
|
("first" . "1") |
|
|
("last" . "25") |
|
|
("score" . "50"))))) |
|
|
(setq buffer-file-name nil) |
|
|
t) |
|
|
|
|
|
;;; |
|
|
;;; Alta Vista |
|
|
;;; |
|
|
|
|
|
(defun nnweb-altavista-create-mapping () |
|
|
"Perform the search and create a number-to-url alist." |
|
|
(save-excursion |
|
|
(set-buffer nnweb-buffer) |
|
|
(erase-buffer) |
|
|
(let ((part 0)) |
|
|
(when (funcall (nnweb-definition 'search) nnweb-search part) |
|
|
(let ((i 0) |
|
|
(more t) |
|
|
(case-fold-search t) |
|
|
(active (or (cadr (assoc nnweb-group nnweb-group-alist)) |
|
|
(cons 1 0))) |
|
|
subject date from id group |
|
|
map url) |
|
|
(while more |
|
|
;; Go through all the article hits on this page. |
|
|
(goto-char (point-min)) |
|
|
(search-forward "<dt>" nil t) |
|
|
(delete-region (point-min) (match-beginning 0)) |
|
|
(goto-char (point-min)) |
|
|
(while (search-forward "<dt>" nil t) |
|
|
(replace-match "\n<blubb>")) |
|
|
(nnweb-decode-entities) |
|
|
(goto-char (point-min)) |
|
|
(while (re-search-forward "<blubb>.*href=\"\\([^\"]+\\)\"><strong>\\([^>]*\\)</strong></a><dd>\\([^-]+\\)- <b>\\([^<]+\\)<.*href=\"news:\\([^\"]+\\)\">.*\">\\(.+\\)</a><P>" |
|
|
nil t) |
|
|
(setq url (match-string 1) |
|
|
subject (match-string 2) |
|
|
date (match-string 3) |
|
|
group (match-string 4) |
|
|
id (concat "<" (match-string 5) ">") |
|
|
from (match-string 6)) |
|
|
(incf i) |
|
|
(unless (nnweb-get-hashtb url) |
|
|
(push |
|
|
(list |
|
|
(incf (cdr active)) |
|
|
(make-full-mail-header |
|
|
(cdr active) (concat "(" group ") " subject) from date |
|
|
id nil 0 0 url)) |
|
|
map) |
|
|
(nnweb-set-hashtb (cadar map) (car map)))) |
|
|
;; See if we want more. |
|
|
(when (or (not nnweb-articles) |
|
|
(>= i nnweb-max-hits) |
|
|
(not (funcall (nnweb-definition 'search) |
|
|
nnweb-search (incf part)))) |
|
|
(setq more nil))) |
|
|
;; Return the articles in the right order. |
|
|
(setq nnweb-articles |
|
|
(sort (nconc nnweb-articles map) 'car-less-than-car))))))) |
|
|
|
|
|
(defun nnweb-altavista-wash-article () |
|
|
(goto-char (point-min)) |
|
|
(let ((case-fold-search t)) |
|
|
(when (re-search-forward "^<strong>" nil t) |
|
|
(delete-region (point-min) (match-beginning 0))) |
|
|
(goto-char (point-min)) |
|
|
(while (looking-at "<strong>\\([^ ]+\\) +</strong> +\\(.*\\)$") |
|
|
(replace-match "\\1: \\2" t) |
|
|
(forward-line 1)) |
|
|
(when (re-search-backward "^References:" nil t) |
|
|
(narrow-to-region (point) (progn (forward-line 1) (point))) |
|
|
(goto-char (point-min)) |
|
|
(while (re-search-forward "<A.*\\?id@\\([^\"]+\\)\">[0-9]+</A>" nil t) |
|
|
(replace-match "<\\1> " t))) |
|
|
(widen) |
|
|
(nnweb-remove-markup) |
|
|
(nnweb-decode-entities))) |
|
|
|
|
|
(defun nnweb-altavista-search (search &optional part) |
|
|
(url-insert-file-contents |
|
|
(concat |
|
|
(nnweb-definition 'address) |
|
|
"?" |
|
|
(nnweb-encode-www-form-urlencoded |
|
|
`(("pg" . "aq") |
|
|
("what" . "news") |
|
|
,@(when part `(("stq" . ,(int-to-string (* part 30))))) |
|
|
("fmt" . "d") |
|
|
("q" . ,search) |
|
|
("r" . "") |
|
|
("d0" . "") |
|
|
("d1" . ""))))) |
|
|
(setq buffer-file-name nil) |
|
|
t) |
|
|
|
|
305 |
;;; |
;;; |
306 |
;;; Deja bought by google.com |
;;; Deja bought by google.com |
307 |
;;; |
;;; |
320 |
(goto-char (point-min)) |
(goto-char (point-min)) |
321 |
(while (search-forward "<br>" nil t) |
(while (search-forward "<br>" nil t) |
322 |
(replace-match "\n")) |
(replace-match "\n")) |
323 |
(nnweb-remove-markup) |
(mm-url-remove-markup) |
324 |
(goto-char (point-min)) |
(goto-char (point-min)) |
325 |
(while (re-search-forward "^[ \t]*\n" nil t) |
(while (re-search-forward "^[ \t]*\n" nil t) |
326 |
(replace-match "")) |
(replace-match "")) |
330 |
(narrow-to-region (point) (point-max)) |
(narrow-to-region (point) (point-max)) |
331 |
(search-forward "</pre>" nil t) |
(search-forward "</pre>" nil t) |
332 |
(delete-region (point) (point-max)) |
(delete-region (point) (point-max)) |
333 |
(nnweb-remove-markup) |
(mm-url-remove-markup) |
334 |
(widen))) |
(widen))) |
335 |
|
|
336 |
(defun nnweb-google-parse-1 (&optional Message-ID) |
(defun nnweb-google-parse-1 (&optional Message-ID) |
352 |
"http://groups.google.com/groups?selm=%s&output=gplain" mid)) |
"http://groups.google.com/groups?selm=%s&output=gplain" mid)) |
353 |
(narrow-to-region (search-forward ">" nil t) |
(narrow-to-region (search-forward ">" nil t) |
354 |
(search-forward "</a>" nil t)) |
(search-forward "</a>" nil t)) |
355 |
(nnweb-remove-markup) |
(mm-url-remove-markup) |
356 |
(nnweb-decode-entities) |
(mm-url-decode-entities) |
357 |
(setq Subject (buffer-string)) |
(setq Subject (buffer-string)) |
358 |
(goto-char (point-max)) |
(goto-char (point-max)) |
359 |
(widen) |
(widen) |
360 |
(forward-line 1) |
(forward-line 2) |
361 |
(when (looking-at "<br><font[^>]+>") |
(when (looking-at "<br><font[^>]+>") |
362 |
(goto-char (match-end 0))) |
(goto-char (match-end 0))) |
363 |
(if (not (looking-at "<a[^>]+>")) |
(if (not (looking-at "<a[^>]+>")) |
364 |
(skip-chars-forward " \t") |
(skip-chars-forward " \t") |
365 |
(narrow-to-region (point) |
(narrow-to-region (point) |
366 |
(search-forward "</a>" nil t)) |
(search-forward "</a>" nil t)) |
367 |
(nnweb-remove-markup) |
(mm-url-remove-markup) |
368 |
(nnweb-decode-entities) |
(mm-url-decode-entities) |
369 |
(setq Newsgroups (buffer-string)) |
(setq Newsgroups (buffer-string)) |
370 |
(goto-char (point-max)) |
(goto-char (point-max)) |
371 |
(widen) |
(widen) |
372 |
(skip-chars-forward "- \t")) |
(skip-chars-forward "- \t")) |
373 |
(when (looking-at |
(when (looking-at |
374 |
"\\([0-9]+[/ ][A-Za-z]+[/ ][0-9]+\\)[ \t]*by[ \t]*\\([^<]*\\) - <a") |
"\\([0-9]+\\)[/ ]\\([A-Za-z]+\\)[/ ]\\([0-9]+\\)[ \t]*by[ \t]*\\([^<]*\\) - <a") |
375 |
(setq From (match-string 2) |
(setq From (match-string 4) |
376 |
Date (match-string 1))) |
Date (format "%s %s 00:00:00 %s" |
377 |
|
(match-string 2) (match-string 1) |
378 |
|
(match-string 3)))) |
379 |
(forward-line 1) |
(forward-line 1) |
380 |
(incf i) |
(incf i) |
381 |
(unless (nnweb-get-hashtb url) |
(unless (nnweb-get-hashtb url) |
398 |
(nconc nnweb-articles map)) |
(nconc nnweb-articles map)) |
399 |
(when (setq header (cadar map)) |
(when (setq header (cadar map)) |
400 |
(mm-with-unibyte-current-buffer |
(mm-with-unibyte-current-buffer |
401 |
(nnweb-fetch-url (mail-header-xref header))) |
(mm-url-insert (mail-header-xref header))) |
402 |
(caar map)))) |
(caar map)))) |
403 |
|
|
404 |
(defun nnweb-google-create-mapping () |
(defun nnweb-google-create-mapping () |
407 |
(set-buffer nnweb-buffer) |
(set-buffer nnweb-buffer) |
408 |
(erase-buffer) |
(erase-buffer) |
409 |
(when (funcall (nnweb-definition 'search) nnweb-search) |
(when (funcall (nnweb-definition 'search) nnweb-search) |
410 |
(let ((more t)) |
(let ((more t) |
411 |
|
(i 0)) |
412 |
(while more |
(while more |
413 |
(setq nnweb-articles |
(setq nnweb-articles |
414 |
(nconc nnweb-articles (nnweb-google-parse-1))) |
(nconc nnweb-articles (nnweb-google-parse-1))) |
415 |
;; FIXME: There is more. |
;; Check if there are more articles to fetch |
416 |
(setq more nil)) |
(goto-char (point-min)) |
417 |
|
(incf i 100) |
418 |
|
(if (or (not (re-search-forward |
419 |
|
"<td nowrap><a href=\\([^>]+\\).*<span class=b>Next</span>" nil t)) |
420 |
|
(>= i nnweb-max-hits)) |
421 |
|
(setq more nil) |
422 |
|
;; Yup, there are more articles |
423 |
|
(setq more (concat "http://groups.google.com" (match-string 1))) |
424 |
|
(when more |
425 |
|
(erase-buffer) |
426 |
|
(mm-url-insert more)))) |
427 |
;; Return the articles in the right order. |
;; Return the articles in the right order. |
428 |
(setq nnweb-articles |
(setq nnweb-articles |
429 |
(sort nnweb-articles 'car-less-than-car)))))) |
(sort nnweb-articles 'car-less-than-car)))))) |
430 |
|
|
431 |
(defun nnweb-google-search (search) |
(defun nnweb-google-search (search) |
432 |
(nnweb-insert |
(mm-url-insert |
433 |
(concat |
(concat |
434 |
(nnweb-definition 'address) |
(nnweb-definition 'address) |
435 |
"?" |
"?" |
436 |
(nnweb-encode-www-form-urlencoded |
(mm-url-encode-www-form-urlencoded |
437 |
`(("q" . ,search) |
`(("q" . ,search) |
438 |
("num". "100") |
("num". "100") |
439 |
("hq" . "") |
("hq" . "") |
450 |
url)) |
url)) |
451 |
|
|
452 |
;;; |
;;; |
453 |
|
;;; gmane.org |
454 |
|
;;; |
455 |
|
(defun nnweb-gmane-create-mapping () |
456 |
|
"Perform the search and create a number-to-url alist." |
457 |
|
(save-excursion |
458 |
|
(set-buffer nnweb-buffer) |
459 |
|
(erase-buffer) |
460 |
|
(when (funcall (nnweb-definition 'search) nnweb-search) |
461 |
|
(let ((more t) |
462 |
|
(case-fold-search t) |
463 |
|
(active (or (cadr (assoc nnweb-group nnweb-group-alist)) |
464 |
|
(cons 1 0))) |
465 |
|
subject group url |
466 |
|
map) |
467 |
|
;; Remove stuff from the beginning of results |
468 |
|
(goto-char (point-min)) |
469 |
|
(search-forward "Search Results</h1><ul>" nil t) |
470 |
|
(delete-region (point-min) (point)) |
471 |
|
(goto-char (point-min)) |
472 |
|
;; Iterate over the actual hits |
473 |
|
(while (re-search-forward ".*href=\"\\([^\"]+\\)\">\\(.*\\)" nil t) |
474 |
|
(setq url (concat "http://gmane.org/" (match-string 1))) |
475 |
|
(setq subject (match-string 2)) |
476 |
|
(unless (nnweb-get-hashtb url) |
477 |
|
(push |
478 |
|
(list |
479 |
|
(incf (cdr active)) |
480 |
|
(make-full-mail-header |
481 |
|
(cdr active) (concat "(" group ") " subject) nil nil |
482 |
|
nil nil 0 0 url)) |
483 |
|
map) |
484 |
|
(nnweb-set-hashtb (cadar map) (car map)))) |
485 |
|
;; Return the articles in the right order. |
486 |
|
(setq nnweb-articles |
487 |
|
(sort (nconc nnweb-articles map) 'car-less-than-car)))))) |
488 |
|
|
489 |
|
(defun nnweb-gmane-wash-article () |
490 |
|
(let ((case-fold-search t)) |
491 |
|
(goto-char (point-min)) |
492 |
|
(re-search-forward "<!--X-Head-of-Message-->" nil t) |
493 |
|
(delete-region (point-min) (point)) |
494 |
|
(goto-char (point-min)) |
495 |
|
(while (looking-at "^<li><em>\\([^ ]+\\)</em>.*</li>") |
496 |
|
(replace-match "\\1\\2" t) |
497 |
|
(forward-line 1)) |
498 |
|
(mm-url-remove-markup))) |
499 |
|
|
500 |
|
(defun nnweb-gmane-search (search) |
501 |
|
(mm-url-insert |
502 |
|
(concat |
503 |
|
(nnweb-definition 'address) |
504 |
|
"?" |
505 |
|
(mm-url-encode-www-form-urlencoded |
506 |
|
`(("query" . ,search))))) |
507 |
|
(setq buffer-file-name nil) |
508 |
|
t) |
509 |
|
|
510 |
|
|
511 |
|
(defun nnweb-gmane-identity (url) |
512 |
|
"Return a unique identifier based on URL." |
513 |
|
(if (string-match "group=\\(.+\\)" url) |
514 |
|
(match-string 1 url) |
515 |
|
url)) |
516 |
|
|
517 |
|
;;; |
518 |
;;; General web/w3 interface utility functions |
;;; General web/w3 interface utility functions |
519 |
;;; |
;;; |
520 |
|
|
536 |
(mapcar 'nnweb-insert-html (nth 2 parse)) |
(mapcar 'nnweb-insert-html (nth 2 parse)) |
537 |
(insert "</" (symbol-name (car parse)) ">\n"))) |
(insert "</" (symbol-name (car parse)) ">\n"))) |
538 |
|
|
|
(defun nnweb-encode-www-form-urlencoded (pairs) |
|
|
"Return PAIRS encoded for forms." |
|
|
(mapconcat |
|
|
(function |
|
|
(lambda (data) |
|
|
(concat (w3-form-encode-xwfu (car data)) "=" |
|
|
(w3-form-encode-xwfu (cdr data))))) |
|
|
pairs "&")) |
|
|
|
|
|
(defun nnweb-fetch-form (url pairs) |
|
|
"Fetch a form from URL with PAIRS as the data using the POST method." |
|
|
(let ((url-request-data (nnweb-encode-www-form-urlencoded pairs)) |
|
|
(url-request-method "POST") |
|
|
(url-request-extra-headers |
|
|
'(("Content-type" . "application/x-www-form-urlencoded")))) |
|
|
(url-insert-file-contents url) |
|
|
(setq buffer-file-name nil)) |
|
|
t) |
|
|
|
|
|
(defun nnweb-decode-entities () |
|
|
"Decode all HTML entities." |
|
|
(goto-char (point-min)) |
|
|
(while (re-search-forward "&\\(#[0-9]+\\|[a-z]+\\);" nil t) |
|
|
(let ((elem (if (eq (aref (match-string 1) 0) ?\#) |
|
|
(let ((c |
|
|
(string-to-number (substring |
|
|
(match-string 1) 1)))) |
|
|
(if (mm-char-or-char-int-p c) c 32)) |
|
|
(or (cdr (assq (intern (match-string 1)) |
|
|
w3-html-entities)) |
|
|
?#)))) |
|
|
(unless (stringp elem) |
|
|
(setq elem (char-to-string elem))) |
|
|
(replace-match elem t t)))) |
|
|
|
|
|
(defun nnweb-decode-entities-string (string) |
|
|
(with-temp-buffer |
|
|
(insert string) |
|
|
(nnweb-decode-entities) |
|
|
(buffer-substring (point-min) (point-max)))) |
|
|
|
|
|
(defun nnweb-remove-markup () |
|
|
"Remove all HTML markup, leaving just plain text." |
|
|
(goto-char (point-min)) |
|
|
(while (search-forward "<!--" nil t) |
|
|
(delete-region (match-beginning 0) |
|
|
(or (search-forward "-->" nil t) |
|
|
(point-max)))) |
|
|
(goto-char (point-min)) |
|
|
(while (re-search-forward "<[^>]+>" nil t) |
|
|
(replace-match "" t t))) |
|
|
|
|
|
(defun nnweb-insert (url &optional follow-refresh) |
|
|
"Insert the contents from an URL in the current buffer. |
|
|
If FOLLOW-REFRESH is non-nil, redirect refresh url in META." |
|
|
(let ((name buffer-file-name)) |
|
|
(if follow-refresh |
|
|
(save-restriction |
|
|
(narrow-to-region (point) (point)) |
|
|
(url-insert-file-contents url) |
|
|
(goto-char (point-min)) |
|
|
(when (re-search-forward |
|
|
"<meta[ \t\r\n]*http-equiv=\"Refresh\"[^>]*URL=\\([^\"]+\\)\"" nil t) |
|
|
(let ((url (match-string 1))) |
|
|
(delete-region (point-min) (point-max)) |
|
|
(nnweb-insert url t)))) |
|
|
(url-insert-file-contents url)) |
|
|
(setq buffer-file-name name))) |
|
|
|
|
539 |
(defun nnweb-parse-find (type parse &optional maxdepth) |
(defun nnweb-parse-find (type parse &optional maxdepth) |
540 |
"Find the element of TYPE in PARSE." |
"Find the element of TYPE in PARSE." |
541 |
(catch 'found |
(catch 'found |
585 |
(listp (cdr element))) |
(listp (cdr element))) |
586 |
(nnweb-text-1 element))))) |
(nnweb-text-1 element))))) |
587 |
|
|
|
(defun nnweb-replace-in-string (string match newtext) |
|
|
(while (string-match match string) |
|
|
(setq string (replace-match newtext t t string))) |
|
|
string) |
|
|
|
|
588 |
(provide 'nnweb) |
(provide 'nnweb) |
589 |
|
|
590 |
;;; arch-tag: f59307eb-c90f-479f-b7d2-dbd8bf51b697 |
;;; arch-tag: f59307eb-c90f-479f-b7d2-dbd8bf51b697 |