On Fri, Jan 8, 2010 at 11:54, Richard Quadling <[email protected]> wrote:
> 2010/1/8 Hannes Magnusson <[email protected]>:
>> On Fri, Jan 8, 2010 at 10:55, Richard Quadling <[email protected]> 
>> wrote:
>>> Hi.
>>>
>>> Something that I would like to update in PhD is when an error is
>>> reported (say VERBOSE_MISSING_ATTRIBUTES), to be able to include the
>>> xml:ID of the nearest ancestor with an xml:ID.
>>>
>>> So, ...
>>>
>>> [VERBOSE_MISSING_ATTRIBUTES] Missing alt attribute for
>>> en/features/figures/gc-algorithm.png
>>>
>>> becomes
>>>
>>> VERBOSE_MISSING_ATTRIBUTES] Missing alt attribute for
>>> en/features/figures/gc-algorithm.png for features.gc.collecting-cycles
>>>
>>> Is there an easy way to do this that doesn't cost require massive
>>> amount of searching?
>>
>> You can get the current chunk ID with little effort I think.
>> IIRC that will even work when building bightml, as chunks are still
>> registered, just not rendered as such.
>
> Any idea how?


$<format>->CURRENT_CHUNK


>> Speaking of errormessages, "VERBOSE_MISSING_ATTRIBUTES" doesn't really
>> have a great meaning.
>> We should map the constants to textual representation, so that error
>> message above becomes "Missing attributes: Missing alt attribute
>> for"..
>
> Those descriptions are in the errh() function in phpdotnet\phd\functions.php
>
> Is changing ...
>
>        // PhD warnings
>        VERBOSE_NOVERSION             => 'VERBOSE_NOVERSION',
>        VERBOSE_BROKEN_LINKS          => 'VERBOSE_BROKEN_LINKS',
>        VERBOSE_MISSING_ATTRIBUTES    => 'VERBOSE_MISSING_ATTRIBUTES',
>
> to ...
>
>        // PhD warnings
>        VERBOSE_NOVERSION             => 'No version information',
>        VERBOSE_BROKEN_LINKS          => 'Broken links',
>        VERBOSE_MISSING_ATTRIBUTES    => 'Missing attributes',
>
> enough for you?

That would be what I was thinking, yeah.

-Hannes

Reply via email to