Re: lynx-dev [PATCH]

[Klaus Weide]

On Thu, 22 Jun 2000, Vlad Harchev wrote:

>  Here is a patch that allows user to trigger execution of commands declared
> via EXTERNAL setting. To allow this, one more field to the syntax of the value
> of EXTERNAL setting is added; if that field is TRUE, then activating (using 
> arrow-right for example) the link with URL matching the one specified by 
> 'prefix' field will be equal to pressing '.' on it.
>  Personally I use it for invoking 'links' browser for ftp links - it caches
> ftp connections so it doesn't reconnect when directory is changed. Here is
> a setting I use:
> EXTERNAL:ftp://:links %s:TRUE:TRUE
> 
>  Seems Eduardo wanted this functionality.

Some comments after looking through the patch:

What happens if Enter is pressed on a form submit field?  What
should happen?

 It seems the EXTERN action doesn't kick in in that case.  That's
 probably the most reasonable behavior, but maybe needs to be stated.

What happens if not Enter, but some other key is pressed, and
handle_LYK_ACTIVATE() is called?  What should happen?

 You seem to assume that handle_LYK_ACTIVATE() only gets called for
keys bound to ACTIVATE, but it's not nearly that simple.  Without
looking closely, and somewhat form memory, the code may be reached
for
  - 'd'ownload
  - HEAD
  - NOCACHE ('x')
  - LYK_SUBMIT (somethign that can get triggered by a mouse action,
    otherwise similar to LYK_ACTIVATE)
  - maybe something else.

For each of these,
 does it make more sense to keep the original behavior, rather than
 executing the EXTERN command?  (I think, in general, yes.  Not sure
 about LYK_SUBMIT.)
OTOH, if you do want to go into EXTERN execution, there are various
flags that probably need to be reset before returning (reloading,
LYforce_*, HEAD_request...).

   Klaus


; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden