[Libreoffice-commits] core.git: Changes to 'feature/nativealpha'

[Tomaž Vajngerl]

New branch 'feature/nativealpha' available with the following commits:
commit 36cc464460c5049b21b7b56c6898aa2b2d893afb
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Sun Jun 12 16:55:33 2016 +0900

    Support native alpha in outdev / virdev + support in headless
    
    This adds support for native alpha surface (RGBA) to OutputDevice
    and VirtualDevice. Mainly this removes restrictions that set the
    alpha value to opaque (1.0f). All the functions still use a
    separate "alpha" device for now.
    
    As the headless backend already uses a 32-bit surface there was
    not a big change necessary to make it work.
    
    Change-Id: Ic2f2319a006764395d70b64c473749cd6d47f812

commit cc355197ce4bc195f5af20136b2d6b79376846e1
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Sun Jun 12 16:53:17 2016 +0900

    Add testcases for 32-bit bitmap
    
    Change-Id: I531b8ba7cd691a526392a85da9cb8456a1f16177

commit 6bdebe7168e7d3ca34d2f616bedc71f2462734be
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Sun Jun 12 16:50:39 2016 +0900

    general support for 32-bit Bitmaps + support in "gen" backend
    
    Change-Id: Iad57414a21dfa4e84e471d288923a1417a3535a3

commit 836a8bdc5b86622a3ff4451e291c821bb87dd38c
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Sun Jun 12 16:47:25 2016 +0900

    bitmap: support 32bit conversion
    
    Change-Id: Ia97cb00a7f311088a8ab2c52c069a570f8c269b4

commit 13cdb063c6495fff32e540793998918c63b248f9
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Sun Jun 12 16:38:02 2016 +0900

    ColorData - for RGB macro use TRGB with transparency set to 0
    
    Change-Id: Ia363e05768ac76a7cf4b4da6cbb57174b2d64682

commit ca0d9f07a12983167510078e39fe232c35b86ade
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Tue Jun 7 18:33:34 2016 +0900

    vcl: add "color channel bitcount" for an bitmap buffer (Linux)
    
    In addition to "bitcount" on bitmap buffer, which represents the
    number of bits one pixel takes, we also need the bitcount that the
    color channels take of one pixel take.
    
    For example RGB can be contined in 32 bits (because that the
    underlaying buffer supports) In this case GetBitCount returns
    32bit so we can traverse the buffer correctly. However, we also
    need to know the bitcount of the actual color channels - in this
    case this would be 24 bits.
    
    Change-Id: I0a64928336540dc66d60659c46d64f369e786c3a

commit 8d3cf63a809c20e9bbcc04240f17672b505aa880
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Sun Jun 5 18:58:58 2016 +0900

    Alpha channel in BitmapColor - change bIndex to alpha
    
    We what to store the alpha channel in BitmapColor. To achieve this
    we can repurpose bIndex attribute for alpha. Generally we don't
    need bIndex at all as we can infer if we use index colors or not
    from the context.
    
    Change-Id: I18fe748beeca59e2869368a1c3c2ee9d2062b41e