[Oyranos-devel] GSoC'12 Report - Weeks 3 + 4

[Joseph S. <jsi...@gm...>]

Hi Kai-Uwe and everyone.

This is an update of my progress from the last two weeks.  I did post on my
blog for those weeks [1], although I wanted to wait until I had reached a
few important milestones before posting an update on the list.

LibCmpx's Renderer module now has all of its major PDF spool components in
place - PDF OutputIntent check, default color space tagging, and
OutputIntent embedding.

The default color space tagging was one of the spots that I had difficulty
with back when this was the XCPD project, since Ghostscript did not seem to
have a direct way of doing this through its command line API.  However, I
have finally put in place an improvised solution: first call Ghostscript to
embed a default profile (sRGB.icc) and then, by hand, create a "tagged"
copy of the PDF so that each /DeviceRGB object is replaced with a reference
to the /ICCBased default profile.  In addition, the PDF xref tables are
then correctly modified after manually tagging each entry.

Just one bit of clarification on the print workflow requirement - are we
hoping to see that *each* DeviceXXX entry (where XXX could be 'RGB',
'CMYK', or 'Gray') points only to sRGB, or also to their corresponding
default profiles (sRGB, SWOP, and gray)?

Finally, libCmpx Selector can now parse ColorKeyWords-related settings from
the test dialog and is able to store them into oyConfig_s device objects.
LibCmpx does this by creating a new PPD file with only the selected options
from the dialog.  Using a device "properties" and
"default_context.ppdfile..." call, Oyranos reads the contents of the new
PPD file.

-Joe

[1] jsimon3.wordpress.com