[poppler] Annot rendering improvements

[Leonard Rosenthol]

The details of appearance (re)generation are left to the implementor (for
better or worse).

Leonard

On 4/2/12 2:56 PM, "Carlos Garcia Campos" <carlosgc at gnome.org> wrote:

>Excerpts from Leonard Rosenthol's message of dom abr 01 21:47:44 +0200
>2012:
>> Yes, the raw object has a /Rect of [0 0 100 100].
>> 
>> HOWEVER, it doesn't have a prebuilt appearance.  As such, Acrobat/Reader
>> (or any other viewer) has to regenerate an appearance in order to render
>> it to the page.  Since your line has an L (the actual line coords) of
>>[0 0
>> 200 200], that causes Acrobat/Reader to generate a new bounding box that
>> incorporates the entire line as that is what one would expect.
>
>Yes, we don't do that. Is this part of the spec and we missed it, or
>is a workaround for buggy documents?
>
>> So no bug in Acrobat, but Poppler should probably do the same thing.
>
>Sure, we'll do that, thank you very much for the clarification.
>
>> Leonard
>> 
>> On 4/1/12 4:03 AM, "Fabio D'Urso" <fabiodurso at hotmail.it> wrote:
>> 
>> >
>> >> From: lrosenth at adobe.com
>> >> To: fabiodurso at hotmail.it; poppler at lists.freedesktop.org
>> >> Date: Sat, 31 Mar 2012 21:19:06 -0700
>> >> Subject: Re: [poppler] Annot rendering improvements
>> >>
>> >> Acrobat & Adobe Reader must certainly clip an annotations appearance
>>to
>> >> the bounding area of the annotation.
>> >>
>> >> If you have an example PDF where this is not the case, I would very
>>much
>> >> like to see it.
>> >
>> >6 0 obj 
>> ><<
>> >/Rect [ 0 0 100 100 ]
>> >/Subtype /Line
>> >/P 3 0 R
>> >/Type /Annot
>> >/M (D:20120401093948+01'00')
>> >/F 4
>> >/NM (36ab4956-9486-42c6-93e6-6be2baa56df3)
>> >/C [ 1 0 0 ]
>> >/BS <<
>> >/W 2
>> >/S /S
>> >>>
>> >/CA 1
>> >/Cap false
>> >/LE [ /None /None ]
>> >/L [ 0 0 200 200 ]
>> >/T (fabio)
>> >/Subj (Line)
>> >/CreationDate (D:20120401093948+01'00')
>> >/Popup 8 0 R
>> >>>
>> >
>> >If I understand the specs correctly, it should only show the lower half
>> >(like current
>> >poppler does), no?
>> >
>> >               
>> 
>-- 
>Carlos Garcia Campos
>PGP key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x523E6462