This improves commit bbd5ebbe8a. Several parameters did not need to go
inside draw_layer(). In particular this function should not take care of
filling return values on errors. It should just return a success boolean
with a GError which is processed on the main function. This also allows
proper shortcut to end the loop earlier on the first error.
Also the layer index parameter does not need to be a pointer since it
never requires updates.
Finally declare variables in smaller scopes for cleaner code.
(cherry picked from commit 5b2d89f794)
Added a recursive loop to draw layer children in layer groups. Text
layers in layer groups are now exported correctly.
(cherry picked from commit bbd5ebbe8a)
Nothing said what was going to be the order of the page, except by
testing. Now there will be an explicit text, which will be automatically
updated when checking the "reverse order" box.
(cherry picked from commit afe1de950f)
... only shows the color (and nothing else) when "Convert bitmaps to
vector graphics where possible" was set.
This is because gimp_drawable_histogram() only checks selected pixels.
So let's make sure we work on a duplicate of the image so that we can
safely remove the selection before processing the export.
Current comment was implying that PDF had some kind of fixed PPI, or
that a 72 PPI was hardcoded in cairo. This is not at all what this is
about. Cairo simply has a concept of user-space coordinates, and from
what I gather, it seems that this is set as "point", the typographical
unit of measure (not as a screen point). In such context, cairo_scale()
is used to update the transformation matrix (which will transform from
user to device units, i.e. pixels) so that we can draw with pixels. In
other words, both the user and device units will be pixels in subsequent
calls.
It turns out that 1 inch == 72 points. This is not a PPI/DPI at all and
the comment was completely misleading.
No need to loop through layers, which may be wrong anyway with
layers-as-pages. Also we were using the layer sizes without even
checking their offsets yet filling from (0, 0). This code was flawed in
many aspects.
Cairo doc is pretty clear about cairo_mask_surface():
> A drawing operator that paints the current source using the alpha
> channel of @surface as a mask
Therefore when creating a mask surface, it must be in one of
CAIRO_FORMAT_* with an alpha channel and the mask data must be copied
into this alpha channel. I chose CAIRO_FORMAT_A8 (a format with alpha
only) which I map to "Y u8" babl format in GEGL so that
gegl_buffer_copy() copies the contents of Y into A.
The `if` test was inverted. When the option is checked, we don't want to
add the GIMP_EXPORT_CAN_HANDLE_LAYER_MASKS capability so that
gimp_export_image() applies the mask.
This is a first step to fix bug 783528, but it's not finished. Right
now, when the option is unchecked, the mask is simply ignored. We want
to export a PDF mask instead.
And revert file-pdf-save to its previous signature before commit
d03de52. For compatibility reason, we can't change the signature of a
procedure available in previous stable releases.
Also revert file-pdf-save-multi for the same reason. I don't create a
file-pdf-save-multi2 though even though layers-as-pages could still be
meaningful there (creating a multi-page PDF from several multi-layers
images) but that's likely a special case which can be dealt with later.
Obviously the interactive export will use file-pdf-save2.
Otherwise default values are not taken into account in interactive mode.
In particular, we want ignore-hidden to be checked (TRUE) by default,
because this is the most obvious behavior (WYSIWYG).
Try to sort all GIMP_ICON_* defines into FDO categories like in
https://specifications.freedesktop.org/icon-naming-spec/latest/ar01s04.html
Add defines for all icons we override, rename some icons to their FDO
standard names, and mark the ones we duplicate with a comment so we
don't forget to rename those to standard names in 3.0.
PDF files have no transparency, still, transparent areas of
exported images would always display white, regardless of the
current background color.
(The misbehavior was found while checking BUG #752819,
but that is a cairo issue and is not fixed by this)
Set the resolution of the PangoFontMap from which the
PangoContext is created, set cairo font options only
on the PangoContext, explicitely set wrap mode and
do not force a layout height.
Most of the code was taken from the print plugin, and it includes iterating
over pixel regions (instead of fetching them at once), correct support for RGBA
(which was buggy previously), etc.
This commit removes the usage of A8 cairo surfaces for GRAY layers, simply
because it seems not to work. This should be investigated and fixed in order
to optimize the file size (by saving gray images as gray).
Also should be checked is why we treat the cairo bpp as 4 when calculating the
pixel locations in RGB cairo buffers. The print plug-in does this and it works,
and when I tried 3 it didn't work. May be because of word alignment...
This commit refactors the way the arguments are being processed into a much
cleaner code. It also makes the main save procedure the default save handler
instead of the removed defaults procedure - this means the order of its
arguments was changed accordingly.
