Katsumi Yamaoka | 1 Aug 2008 13:33
X-Face
Favicon
Gravatar

[emacs-w3m:10319] Re: Stack overflow in regexp matcher

>>>>> In [emacs-w3m : No.10313] 野宮賢 / NOMIYA Masaruさん wrote:

>> cf. http://www.jpl.org/elips/BUGS-ja.html

> これは拝見していました.
> ただ.

> 「これは byte-code(...) という関数が wl-message-buffer-display という
>   関数から呼ばれてエラーが起きているのですが、このように byte compile
>   された関数が何で、どこで定義されているかを知ることは事実上不可能です。」

> と書かれてありますので,小生には無理だと思っていました.

すみません、これは野宮さんの記事の後で読み返したらあんまり書いて
いなかったことに気がついて追加したものです。しかし、どうも言葉足
らずだったようですね。以下のように書き換えました:

,----
| ...という
| 関数から呼ばれてエラーが起きているのですが、このように byte compile
| された関数が現れても、それだけの情報では、それが何で、どこで定義され
| ているかを知ることは事実上不可能です。
| 
| そこで、...                                ...エラーの詳細を知るため
| の方法を書きましょう。二つ目のような場合でも、同様の方法でもっと人間
| が判読しやすい情報を得ることができます。
`----

> やってみましたところ,sb-itmedia.el の

(Continue reading)

Katsumi Yamaoka | 4 Aug 2008 02:21
X-Face
Favicon
Gravatar

[emacs-w3m:10320] Re: Stack overflow in regexp matcher

>>>>> In [emacs-w3m : No.10313] 野宮賢 / NOMIYA Masaruさん wrote:
>> やってみましたところ,sb-itmedia.el の

>> (when (and (re-search-forward "\
>> \\(<h[0-9]>[^<]+</h[0-9]>[^<]*\\(\\(?:<[^>h][^>]*>[^<]*\\)*\\)\\)?<!--BODY-->"
>> 				      nil t)

>> の正規表現が原因で,

> "</h[0-9]>" と "<!--BODY-->" の間の曖昧表現は、いかにも非常に広
> い範囲にマッチしてしまいそうですね。ご指摘ありがとうございます。
> いずれ直そうと思います。

対応しました。試してみてください。
--

-- 
山岡

Katsumi Yamaoka | 5 Aug 2008 02:46
X-Face
Favicon
Gravatar

[emacs-w3m:10321] Re: two w3m-view-header should be just like "= B"

>>>>> In [emacs-w3m : No.10287] jidanni <at> jidanni.org wrote:

> = runs the command w3m-view-header.
> = = puts one back where one was reading in w3m and lynx.
> = = puts one at the top of the page in emacs-w3m.

Improved.  So has been w3m-view-source.  Thanks for the suggestion.

0000 vk | 6 Aug 2008 18:49
Picon

[emacs-w3m:10322] problem with cvs version

Hello list,

I am using the latest emacs-w3m with emacs
23.0.60.1 (i686-pc-linux-gnu, GTK+ Version 2.12.0).

When I (require 'w3m), I get this backtrace (sorry for the long
lines, but it's clearer this way):

Debugger entered--Lisp error: (error "Cannot return from the debugger in an error")
* char-to-string(17220380)
* (puthash (car entity) (char-to-string (make-char ... ...)) table)
  (while --cl-dolist-temp-- (setq entity (car --cl-dolist-temp--)) (puthash (car entity) (char-to-string ...) table) (setq --cl-dolist-temp-- (cdr --cl-dolist-temp--)))
  (let ((--cl-dolist-temp-- ...) entity) (while --cl-dolist-temp-- (setq entity ...) (puthash ... ... table) (setq --cl-dolist-temp-- ...)) nil)
  (catch (quote --cl-block-nil--) (let (... entity) (while --cl-dolist-temp-- ... ... ...) nil))
  (cl-block-wrapper (catch (quote --cl-block-nil--) (let ... ... nil)))
  (block nil (let (... entity) (while --cl-dolist-temp-- ... ... ...) nil))
  (dolist (entity (quote ...)) (puthash (car entity) (char-to-string ...) table))
  (let ((table ...)) (dolist (entity ...) (puthash ... ... table)) (dolist (entity ...) (puthash ... ... table)) (dolist (entity ...) (puthash ... ... table)) (when (w3m-mule-unicode-p) (let ... ... ...)) table)
  (defconst w3m-entity-table (let (...) (dolist ... ...) (dolist ... ...) (dolist ... ...) (when ... ...) table) "Table of html character entities and values.")
  eval-buffer(#<buffer  *load*> nil "/home/vk00/elisp/emacs-w3m/w3m.el" nil t)  ; Reading at buffer position 80753
 
I tracked it down to line 2061 of w3m.el (cvs version 1.1377). In
the following

2060:      (puthash (car entity)
2061:           (char-to-string (make-char 'greek-iso8859-7 (cdr entity)))
2062:           table))

(cdr entity) returns 65, but the make-char is returning 17220380
which is causing the (char-to-string) to fail.

Please, can anyone help?

Thanks so much.

0000 vk | 6 Aug 2008 23:36
Picon

[emacs-w3m:10323] Re: problem with cvs version

On Wed, Aug 6, 2008 at 12:49 PM, 0000 vk <0000.vk <at> gmail.com> wrote:
Hello list,

I am using the latest emacs-w3m with emacs
23.0.60.1 (i686-pc-linux-gnu, GTK+ Version 2.12.0).

When I (require 'w3m), I get this backtrace (sorry for the long
lines, but it's clearer this way):

Debugger entered--Lisp error: (error "Cannot return from the debugger in an error")
* char-to-string(17220380)
* (puthash (car entity) (char-to-string (make-char ... ...)) table)
  (while --cl-dolist-temp-- (setq entity (car --cl-dolist-temp--)) (puthash (car entity) (char-to-string ...) table) (setq --cl-dolist-temp-- (cdr --cl-dolist-temp--)))
  (let ((--cl-dolist-temp-- ...) entity) (while --cl-dolist-temp-- (setq entity ...) (puthash ... ... table) (setq --cl-dolist-temp-- ...)) nil)
  (catch (quote --cl-block-nil--) (let (... entity) (while --cl-dolist-temp-- ... ... ...) nil))
  (cl-block-wrapper (catch (quote --cl-block-nil--) (let ... ... nil)))
  (block nil (let (... entity) (while --cl-dolist-temp-- ... ... ...) nil))
  (dolist (entity (quote ...)) (puthash (car entity) (char-to-string ...) table))
  (let ((table ...)) (dolist (entity ...) (puthash ... ... table)) (dolist (entity ...) (puthash ... ... table)) (dolist (entity ...) (puthash ... ... table)) (when (w3m-mule-unicode-p) (let ... ... ...)) table)
  (defconst w3m-entity-table (let (...) (dolist ... ...) (dolist ... ...) (dolist ... ...) (when ... ...) table) "Table of html character entities and values.")
  eval-buffer(#<buffer  *load*> nil "/home/vk00/elisp/emacs-w3m/w3m.el" nil t)  ; Reading at buffer position 80753
 
I tracked it down to line 2061 of w3m.el (cvs version 1.1377). In
the following

2060:      (puthash (car entity)
2061:           (char-to-string (make-char 'greek-iso8859-7 (cdr entity)))
2062:           table))

(cdr entity) returns 65, but the make-char is returning 17220380
which is causing the (char-to-string) to fail.

Please, can anyone help?

Thanks so much.


Never mind.  It works with the latest emacs cvs.
Thanks for reading.

-vk

 
Katsumi Yamaoka | 7 Aug 2008 00:57
X-Face
Favicon
Gravatar

[emacs-w3m:10324] Re: vs. split-window-vertically

>>>>> In [emacs-w3m : No.10295] jidanni <at> jidanni.org wrote:

> Use
> C-x 2 runs the command split-window-vertically
> then e.g.,
> B runs the command w3m-view-previous-page
> or
> H runs the command w3m-gohome
> etc.

> You'll see both windows' actions are in tandem. What use is that? They
> should act independently.

Because those windows display the identical buffer object.  How
about using `M-x clone-buffer' instead of `C-x 2' ?
An operation having some resemblance is `C-c C-t C-x 2 C-p'.

Katsumi Yamaoka | 7 Aug 2008 01:10
X-Face
Favicon
Gravatar

[emacs-w3m:10325] Re: vs. split-window-vertically

>>>>> In [emacs-w3m : No.10324] Katsumi Yamaoka wrote:
> An operation having some resemblance is `C-c C-t C-x 2 C-p'.

Sorry, it had had to be `C-c C-t C-x 2 C-c C-p'.

Katsumi Yamaoka | 7 Aug 2008 02:41
X-Face
Favicon
Gravatar

[emacs-w3m:10326] Re: faces w3m-bold and w3m-current-anchor are hard to read, if Chinese

>>>>> In [emacs-w3m : No.10314] jidanni <at> jidanni.org wrote:

> Help, the faces w3m-bold and w3m-current-anchor are hard to read, if
> Chinese.

> ^U what-cursor-position shows
>      -arphic-ar pl kaitim big5-bold-r-normal--16-160-72-72-c-160-big5-0

> Don't you think the Chinese in the line
>     圖 Images: 後 back, 前 front (黃峻崢,斯托曼,積丹尼
> on http://jidanni.org/comp/index.html
> is hard to read? Lots more on e.g., http://www.z-chang.net/qa2.html
> What can I do about it, short of
> (shell-command-on-region (region-beginning) (region-end) "cat").

First of all, I don't know at all how Emacs choose fonts for a
certain face.  It may be derived from fonts specified in a fontset,
or not.  And a font chosen differs according to Emacs versions.
But for that matter, Emacs 23 looks like doing fairly right thing.
Your page is displayed in my Emacs 23 as follows.  Though it may
not be your taste, it's not so bad for me.

`C-u C-x =' says that the fonts used are:

-misc-fixed-medium-r-normal--16-120-100-100-c-160-jisx0208.1983-0
-misc-fixed-bold-r-normal--16-120-100-100-c-80-iso8859-1

Note that the font for the Chinese characters is not bold (why
the Japanese font is used will be due to my language environment).
It seems that Emacs 23 has a capability to render it in bold.
In Emacs 22, those Chinese characters are displayed in normal
weight in the page.

You are using Emacs 22, aren't you?  If so, Emacs 23 that will
be released in the near future may bring a boon to you.

Regards,
Katsumi Yamaoka | 7 Aug 2008 13:01
X-Face
Favicon
Gravatar

[emacs-w3m:10327] update gnus-summary-w3m-safe-toggle-inline-images

Hi,

The command `gnus-summary-w3m-safe-toggle-inline-images' is an
example mentioned in the emacs-w3m Info: (info "(emacs-w3m)Gnus")
It is the Gnus summary command that toggles the images display
in html articles.  Today I realized that it doesn't work properly
if the article is broken into pages by `^L's.  If you use it,
especially for shimbun articles, please update it to:

Attachment: application/emacs-lisp, 688 bytes

The one in the Info has already been fixed in CVS.

Regards,

In Japanese:

コマンド `gnus-summary-w3m-safe-toggle-inline-images' は emacs-w3m
の Info に書いてある例です: (info "(emacs-w3m-ja)Gnus")
それは html 記事の画像表示を ON/OFF 切り換えするための、Gnus の
概略バッファーで使うコマンドです。今日わかったのですが、記事が
`^L' で改ページされていると、それがまともに動作しません。もしそ
れを、特に shimbun で使っているならば、上記のものに更新してくだ
さい。Info のものはすでに CVS で直っています。
--

-- 
山岡
jidanni | 7 Aug 2008 18:56
Favicon
Gravatar

[emacs-w3m:10328] Re: faces w3m-bold and w3m-current-anchor are hard to read, if Chinese

KY> You are using Emacs 22, aren't you?  If so, Emacs 23 that will
KY> be released in the near future may bring a boon to you.
OK, I will shut up (閉嘴) until then. Thanks.


Gmane