Commit graph

5312 commits

Author SHA1 Message Date
Behdad Esfahbod
e4753c82ff [src/README] s/Source.mk/Makefile.sources/ 2008-09-15 14:50:29 -04:00
Behdad Esfahbod
50f45d7768 [src/Makefile.am] Add comment referring to Makefile.sources 2008-09-15 14:50:07 -04:00
Chris Wilson
34f2506432 [surface] Trivial rejection of fill/stroke if clip extents is zero
Avoid tessellating the path if we know that the target extents is zero.
Besides the rare occurrence when everything is clipped out, a zero-sized
surface is often intended as a no-op surface for benchmarking.
2008-09-15 16:28:59 +01:00
Behdad Esfahbod
93325a3f08 [configure.ac.enable] Cleanup Makefile.*.config generation 2008-09-11 19:12:31 -04:00
Behdad Esfahbod
66d6cb642c [Makefile.am.common] Unify more common rules 2008-09-11 18:06:57 -04:00
Behdad Esfahbod
de5b8bce37 Makefile.am cleanup
Use a common build/Makefile.am.common file.
2008-09-11 15:49:09 -04:00
Adrian Johnson
1cf1b723c7 Return correct status 2008-09-11 18:36:55 +09:30
Adrian Johnson
15dd80c894 Don't overwrite previous status 2008-09-11 18:26:04 +09:30
Behdad Esfahbod
d1560bd687 Mark private symbol so 2008-09-11 04:23:05 -04:00
Behdad Esfahbod
4a01b1d9c9 Build -uninstalled.pc files 2008-09-11 00:38:30 -04:00
Adrian Johnson
d878924cf8 PS: execute type3 glyphs in their own dictionary
The PS emulation of the PDF text operators stores some data on the
user dict between calls to operators. When using the PDF text
operators in type 3 glyphs, the PDF operators were overwriting the
same data as the PDF operators in the page. Fix this by creating a new
dictionary to execute each type 3 glyph in.
2008-09-11 08:46:55 +09:30
Adrian Johnson
b56075ee51 Enable show_glyphs inside of in user-font glyphs 2008-09-11 08:41:39 +09:30
Adrian Johnson
c3663324fd Add _cairo_type3_glyph_surface_analyze_glyph()
This function emits the glyph to a null stream with the side effect
that other glyphs referenced by this user-font glyph will be added to
the font subsets.
2008-09-11 08:38:39 +09:30
Adrian Johnson
b74533b5a5 Add null output stream 2008-09-11 08:37:30 +09:30
Adrian Johnson
6772efcdf6 Put user fonts in a separate subset 2008-09-11 08:37:02 +09:30
Carl Worth
e0b4f8dc8f Avoid missing enum warnings with pixman 0.11.9 2008-09-10 14:54:18 -07:00
Carl Worth
bdcdc8400e Rename _init_pattern_with_snapshot to _cairo_pattern_init_snapshot 2008-09-10 14:54:18 -07:00
Chris Wilson
bf62798b12 [cairo-compiler-private.h] Divert popen to _popen for MSVC
Microsoft uses _popen instead of popen - required to build the
boilerplate but not actually used on Windows (by default)!
2008-09-10 18:21:50 +01:00
Behdad Esfahbod
5e5a2b9da7 More fixes for pkgconfig files
Should all work now.
2008-09-10 12:41:57 -04:00
Behdad Esfahbod
9dd9289c4a Rename cairo-backend.pc.in to cairo-features.pc.in 2008-09-10 12:18:27 -04:00
Behdad Esfahbod
87dfceb9ec [Makefile.am] Rename INCLUDES to AM_CPPFLAGS
Fixes automake warnings
2008-09-08 15:13:07 -04:00
Chris Wilson
d478b3e1ef [cairoint.h] Update LAST_STATUS. 2008-09-07 18:31:29 +01:00
Chris Wilson
7960fc94dc [ps] Set used level to 3 only if we use native gradients.
Set the used level after performing the checks that the native Type2/3
gradients support the pattern.
2008-09-07 15:03:58 +01:00
Chris Wilson
7439c45039 [ps] Only use fallback images if the gradient has non-constant alpha.
If the gradient has constant alpha, then we can express it as a flattened
linear gradient. Otherwise, should the opacity vary across the gradient we
need to fallback.
2008-09-07 15:03:47 +01:00
Chris Wilson
a33f2f6962 Revert "[ps] A flattened gradient is not linear."
This reverts commit 85ed37da33.

Adrian Johnson pointed out that iff the gradient has constant alpha then
it can be flattened. Reverting this commit in order to restore the
capability to handle gradient flattening.
2008-09-07 15:03:21 +01:00
Chris Wilson
1a6f2b08a1 [ps] Support B PDF operator.
Add the procedure definition for fill-stroke path operators, B and B*.
2008-09-07 15:02:21 +01:00
Chris Wilson
ab3bb369d5 [ps] Add comments about design of output.
Add some snippets from Adrian Johnson about how and why the PS output is
factored in terms of PDF operators.
2008-09-07 15:02:15 +01:00
Chris Wilson
d45b3168db [ps] Use a color_t rather than open-code.
Simplify the code by tracking the current solid color with a
cairo_color_t.
2008-09-06 09:10:30 +01:00
Chris Wilson
85ed37da33 [ps] A flattened gradient is not linear.
We cannot express an alpha-gradient as a simple linear interpolation
between 2 flattened colors. So fallback.
2008-09-06 09:10:25 +01:00
Chris Wilson
c1c86afa87 [ps] Use floating point values when flattening color.
Use the full-precision color values when computing the alpha-blended
color.
2008-09-06 09:10:20 +01:00
Chris Wilson
e583fb061d [type1] Allocate the temporary subsetter on the stack.
Avoid a reasonably small, short-lived malloc by allocating the
cairo_type1_font_subset_t on the stack.
2008-09-06 08:14:43 +01:00
Behdad Esfahbod
7ad076442d Rename src/*.mk to src/Makefile.*
Namely:

	src/Sources.mk      -> src/Makefile.sources
	src/Config.mk       -> src/Makefile.am.config
	src/Config.mk.win32 -> src/Makefile.win32.config
2008-09-04 22:23:42 -04:00
Behdad Esfahbod
37f82a4d1b Get rid of CAIRO_LDADD in favor of CAIRO_LIBS 2008-09-04 16:13:05 -04:00
Behdad Esfahbod
44155f7e59 Cleanup configure.in macros 2008-09-04 16:08:41 -04:00
Behdad Esfahbod
097c4ac022 [Makefile.am] Use check_PROGRAMS 2008-09-03 20:52:17 -04:00
Behdad Esfahbod
b1ced45f06 [doc] Fix out-of-tree build 2008-09-03 20:52:12 -04:00
Vladimir Vukicevic
2cf4b3161c [quartz] move _cairo_quartz_create_cgimage to cairo-quartz-surface.c 2008-09-03 15:34:40 -07:00
Behdad Esfahbod
99e2e2d0bb [src/Makefile.am] Minor fix in MAINTAINERCLEANFILES 2008-09-03 12:06:35 -04:00
Behdad Esfahbod
2791d51f6e [src/README] Refer to BIBLIOGRAPHY 2008-09-03 11:41:41 -04:00
Chris Wilson
8992312df0 [scaled-fonts] Cleanup a couple of valgrind warnings.
Beware the propagation of uninitialised variables and forgotten error
codes.
2008-09-03 16:33:49 +01:00
Chris Wilson
3b7726e3d4 [src/Makefile.am] s/libcairo_la_CFLAGS/INCLUDES/
Update the custom rules after the overall to the build system.
2008-09-03 14:07:37 +01:00
Chris Wilson
c18cc5b441 [src/Makefile.am] make check compile fix
Minor typo preventing TESTS from building.
2008-09-03 10:20:45 +01:00
Behdad Esfahbod
cf212ddaca Also test linking when cross-compiling
We build the test.  Just don't run it.
2008-09-02 22:05:06 -04:00
Behdad Esfahbod
2823ebaa6e [src/check-link.c] Simple test to check we can link to the built cairo.so 2008-09-02 21:31:43 -04:00
Behdad Esfahbod
5926257770 Revamp the build system.
Quick summary of changes:

  - Move list of cairo source files out of src/Makefile.am and into
    src/Sources.mk,

  - Generate files src/Config.mk and src/Config.mk.win32 that choose
    the right set of source files and headers based on configured
    backends and features.  This drastically simplifies building
    using other build systems.  The src/Makefile.win32 file needs
    to be updated to reflect these changes.

  - Add README files to various directories,

  - Add toplevel HACKING file.
2008-09-02 20:24:08 -04:00
Chris Wilson
a5d33bcbb4 [xlib] Keep a pointer to the cairo_xlib_display_t on the surface.
Cleanup the code somewhat by passing cairo_xlib_display_t around
internally as opposed to a Display and then having to lookup the
corresponding cairo_xlib_display_t each time.

[To get a cairo_xlib_display_t from a Display is a list traversal under
mutex (though the element we're looking for is most likely at the start),
but to get the Display is just a lockless pointer dereference.]
2008-09-02 10:49:12 +01:00
Chris Wilson
939b836bfa [xlib] Use an embedded doubly-linked list for the close display hooks.
Kill the allocation and linear search of the close display list on remove,
by embedding a list node into the parent structure.

Original patch by Karl Tomlinson <karlt+@karlt.net>, Mozilla Corporation.
https://bugzilla.mozilla.org/show_bug.cgi?id=453199#c5
2008-09-02 10:08:04 +01:00
Chris Wilson
38f779afcb [user-font] Propagate status return.
An error return possible went unpropagated - fix that. Also whilst in the
vicinity avoid the non-constant initialisation of the local variable
before the antiquated compilers moan.
2008-09-02 09:17:33 +01:00
Chris Wilson
83f390cfc5 [ps] Initialise has_creation_date.
Slipped by check-valgrind, since the boilerplate always initialised it
when explicitly setting the date.
2008-09-01 18:24:15 +01:00
Chris Wilson
e955b7399e Check that lvalue of BITSWAP8() is a uint8_t.
The bit-swapping macro uses the full register for intermediate storage so
we need to be careful to only read the low byte when using the result.

[Only the use in ps-surface.c was incorrect, I just converted the other
unsigned chars to uint8_t for consistency.]
2008-09-01 15:36:34 +01:00