shr: alt text truncation

shr: alt text truncation

[Katsumi Yamaoka]

Currently the width of an alt text is limited by 8, so I come
across ones like the following occasionally:

+--------+
|Amazon.c|
+--------+
[   image   ]

Is it really necessary?  Even so, can't the limit be extended to
shr-width at least in a table?  For instance:

--- shr.el~	2010-11-04 22:03:07 +0000
+++ shr.el	2010-11-05 00:51:26 +0000
@@ -557,3 +557,6 @@
-	  (let ((shr-state 'space))
-	    (if (> (string-width alt) 8)
-		(shr-insert (truncate-string-to-width alt 8))
+	  (let ((shr-state 'space)
+		(limit (if (zerop shr-table-depth) 8 shr-width)))
+	    (if (> (string-width alt) limit)
+		(shr-insert
+		 (concat (truncate-string-to-width alt (- limit 3))
+			 (make-string 3 ?.)))

Re: shr: alt text truncation

[Lars Magne Ingebrigtsen]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

> Is it really necessary?  Even so, can't the limit be extended to
> shr-width at least in a table?  For instance:

I added the limit because of tables, really.  :-)  I found that in my
test cases, the alt texts were typically the thing that took up most
space, because it's not uncommon to have obnoxiously long, but
nonsensical alt texts, and when rendering tables, this made many of them
somewhat unreadable.

So the limitation can be removed outside of tables, but not inside.  The
length (8) could probably be increased, though.  16?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: shr: alt text truncation

[Steinar Bang]

>>>>> Lars Magne Ingebrigtsen <larsi@gnus.org>:

> I added the limit because of tables, really.  :-) I found that in my
> test cases, the alt texts were typically the thing that took up most
> space, because it's not uncommon to have obnoxiously long, but
> nonsensical alt texts, and when rendering tables, this made many of
> them somewhat unreadable.

Couldn't you keep them out of the width calculation for the table
columns, but let them expand to the width determined by the other cells
in that column?

Or is that what you tried, that looked bad?

Re: shr: alt text truncation

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@dod.no> writes:

> Couldn't you keep them out of the width calculation for the table
> columns, but let them expand to the width determined by the other cells
> in that column?
>
> Or is that what you tried, that looked bad?

I don't think it makes all that much difference.  The usual use case is
that you have a table with only an image in a <td>.  You have to use
something for the width, and the alt texts are often just the url
itself, so it's awkward.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: shr: alt text truncation

[Steinar Bang]

>>>>> Lars Magne Ingebrigtsen <larsi@gnus.org>:

> ..., and the alt texts are often just the url itself, so it's awkward.

In the early days of the web, 1994 or thereabouts, I experimented with
putting ASCII art versions of the image in tha alt attribute.

I was quickly told to stop doing that.

Re: shr: alt text truncation

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@dod.no> writes:

> In the early days of the web, 1994 or thereabouts, I experimented with
> putting ASCII art versions of the image in tha alt attribute.

Heh heh.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: shr: alt text truncation

[Ted Zlatanov]

On Tue, 09 Nov 2010 09:39:20 +0100 Steinar Bang <sb@dod.no> wrote: 

>>>>>> Lars Magne Ingebrigtsen <larsi@gnus.org>:
>> ..., and the alt texts are often just the url itself, so it's awkward.

SB> In the early days of the web, 1994 or thereabouts, I experimented with
SB> putting ASCII art versions of the image in tha alt attribute.

SB> I was quickly told to stop doing that.

Why?  You can probably get very close these days if you allow Unicode
glyphs :)

Ted

Re: shr: alt text truncation

[Steinar Bang]

>>>>> Ted Zlatanov <tzz@lifelogs.com>:

> On Tue, 09 Nov 2010 09:39:20 +0100 Steinar Bang <sb@dod.no> wrote: 
>>>>>>> Lars Magne Ingebrigtsen <larsi@gnus.org>:
>>> ..., and the alt texts are often just the url itself, so it's awkward.

SB> In the early days of the web, 1994 or thereabouts, I experimented with
SB> putting ASCII art versions of the image in tha alt attribute.

SB> I was quickly told to stop doing that.

> Why?

Well the argument used by one person was that he used lynx to read the
web pages, because it was fast.  And my ASCII art got in the way of the
fast aquiring of information.

Imagine that!

Re: shr: alt text truncation

[Rupert Swarbrick]

[-- Attachment #1: Type: text/plain, Size: 808 bytes --]

Steinar Bang <sb@dod.no> writes:
>>>>>> Ted Zlatanov <tzz@lifelogs.com>:
>
>> On Tue, 09 Nov 2010 09:39:20 +0100 Steinar Bang <sb@dod.no> wrote: 
>>>>>>>> Lars Magne Ingebrigtsen <larsi@gnus.org>:
>>>> ..., and the alt texts are often just the url itself, so it's awkward.
>
> SB> In the early days of the web, 1994 or thereabouts, I experimented with
> SB> putting ASCII art versions of the image in tha alt attribute.
>
> SB> I was quickly told to stop doing that.
>
>> Why?
>
> Well the argument used by one person was that he used lynx to read the
> web pages, because it was fast.  And my ASCII art got in the way of the
> fast aquiring of information.
>
> Imagine that!

Nowadays, I guess that screen-readers would have trouble too. "hyphen
hyphen slash slash slash..."

Rupert

[-- Attachment #2: Type: application/pgp-signature, Size: 315 bytes --]