directory to suppress generation of .pyc files. According to the
Python docs there isn't a more elegant solution (currently).
* builds/toplevel.mk (dist): New target which builds .tar.gz,
.tar.bz2, and .zip files. Note that the version number is still
hard-coded.
(do-dist): Sub-target of `dist'.
(CONFIG_GUESS, CONFIG_SUB): New variables.
(.PHONY): Updated.
Watson <pwatson@redlinepy.com>.
* builds/unix/configure.ac: Renamed to...
* builds/unix/configure.raw: This which now serves (with appropriate
modifications) as a template for configure.ac.
* version.sed: New script.
* autogen.sh: Generate configure.ac from configure.raw, using
FREETYPE_MAJOR, FREETYPE_MINOR, and FREETYPE_PATCH from freetype.h.
src/base/ftmac.c, src/base/ftrfork.c, src/lzw/ftzopen.c,
src/raster/ftrend1.c, src/sfnt/ttpost.c, src/truetype/ttgxvar.c,
src/type42/t42parse.c, src/winfonts/winfnt.c: hardening the code
against out-of-bounds conditions when allocating arrays. This is
for the cases where FT_NEW_ARRAY and FT_RENEW_ARRAY are not used
already. Introducing the new FT_ALLOC_MULT and FT_REALLOC_MULT
macros.
include/freetype/internal/ftmemory.h, src/base/ftdbgmem.c,
src/base/ftutil.c: udpating the memory management functions and
macros to safely deal with array size buffer overflows, this
corresponds to attemps to allocate arrays that are too large. For
an example, consider the following code:
count = read_uint32_from_file();
array = malloc( sizeof(Item) * count );
for ( nn = 0; nn < count; nn++ )
array[nn] = read_item_from_file();
if 'count' is larger than FT_UINT_MAX/sizeof(Item), the multiplication
will overflow and the array allocated will be smaller than the data
read from the file. In this case, the heap will be trashed, and this
can be used as a denial-of-service, or make the engine crash later.
the FT_ARRAY_NEW and FT_ARRAY_RENEW macro now check that the new
count is no more than FT_INT_MAX/item_size, otherwise, a new error,
named 'FT_Err_Array_Too_Large' will be returned.
note that the memory debugger now works again when FT_DEBUG_MEMORY
is defined, and FT_STRICT_ALIASING has disappeared, the corresponding
code being now the default.
msn2@bidyut.com.
* include/freetype/config/ftstdlib.h (FT_CHAR_BIT, FT_FILE,
ft_fopen, ft_fclose, ft_fseek, ft_ftell, ft_fread, ft_smalloc,
ft_scalloc, ft_srealloc, ft_sfree, ft_labs): New wrapper macros for
C library functions. Update all users accordingly (and catch some
other places where the C library function was used instead of the
wrapper functions).
* src/base/ftsystem.c: Don't include stdio.h and stdlib.h.
* src/gzip/zutil.h [MSDOS && !(__TURBOC__ || __BORLANDC__)]: Don't
include malloc.h.
* src/cff/cffload.c (cff_charset_load): Set `charset->max_cid'.
* src/cff/cffgload.c (cff_slot_load): Change type of third parameter
to `FT_UInt'.
Check range of `glyph_index'.
* src/cff/cffgload.h: Updated.
include/freetype/cache/ftccmap.h,
include/freetype/cache/ftcglyph.h
include/freetype/cache/ftcimage.h
include/freetype/cache/ftcmanag.h
include/freetype/cache/ftcmru.h
include/freetype/cache/ftcsbits.h:
removing these header files from the public include directory.
* include/freetype/config/ftheader.h:
changing the definition of FT_CACHE_INTERNAL_XXX_H macros to
redirect to FT_CACHE_H instead
* src/cache/ftcbasic.c, src/cache/ftccache.c, src/cache/ftccache.h,
src/cache/ftccback.h, src/cache/ftccmap.c, src/cache/ftcglyph.c,
src/cache/ftcglyph.h, src/cache/ftcimage.c, src/cache/ftcimage.h,
src/cache/ftcmanag.c, src/cache/ftcmanag.h, src/cache/ftcmru.c,
src/cache/ftcmru.h, src/cache/ftcsbits.c, src/cache/ftcsbits.h:
modifications to prevent using the FT_CACHE_INTERNAL_XXX_H macros,
and grab the headers in 'src/cache' instead (see below).
src/cache/ftcsbits.h, src/cache/ftcmanag.h, src/cache/ftccmap.h,
src/cache/ftcmru.h: copying the cache's internal header files which
were located in 'include/freetype/cache' to the 'src/cache' directory
instead.
Note that these files are not used by FreeType clients, all cache
public APIs have been already moved to include/freetype/ftcache.h,
and the FT_CACHE_INTERNAL_XXXX_H macros all resolve to it.
the move is to allow us to modify the internals without intereference
from rogue clients. Note that there are no known client that accesses
the cache internals at the moment.
error if a negative size is passed in parameters.
* src/cache/ftccache.c: make ftc_node_destroy FT_BASE_DEF, it needs to
be exported for rogue clients
* src/pshinter/pshglob.c: prevent problems with malformed fonts which
have an odd number of blue values (these are broken according to the
specs).
* src/cff/cffload.c, src/type1/t1load.c: modify the loaders to force
even-ness of 'num_blue_values'. Also change the CFF loader so that
invalid entries in index files are ignored.
Re-enable glyph metrics grid-fitting. It is now done in the base
layer.
(FT_Set_Char_Size, FT_Set_Pixel_Sizes): Make sure the width and height
are not too small or too large, just like we were doing in 2.1.10.
* src/autofit/afloader.c (af_loader_load_g): The vertical metrics are
not scaled.
src/base/ftobjs.c, src/base/ftutil.c, src/cff/cffobjs.c,
src/psaux/afmparse.c, src/sfnt/ttbdf.c, src/tools/apinames.c,
src/truetype/ttdriver.c:
solved compiler warnings as well as C++ compilation problems
ft_contour_enclosed, ft_outline_get_orientation): Commented out. We
have to wait until `FT_GlyphSlot_Own_Bitmap' is stabilized.
(FT_Outline_Embolden): Use `FT_Outline_Get_Orientation'.
documentation.
* include/freetype/ftsynth.h (FT_GlyphSlot_Own_Bitmap),
src/base/ftsynth.c (FT_GlyphSlot_Own_Bitmap): New function to make
sure a glyph slot owns its bitmap. It is also marked experimental and
due to change.
(FT_GlyphSlot_Embolden): Undo the last change. It turns out rendering
the outline confuses some applications.
`width' and `height' to `FT_Long'.
(enum FT_Size_Request_Type), src/base/ftobjs.c (FT_Request_Metrics):
New request type `FT_SIZE_REQUEST_TYPE_SCALES' to specify the scales
directly.
`FT_ORIENTATION_NONE'.
* src/base/ftoutln.c (FT_OUTLINE_GET_CONTOUR, ft_contour_has,
ft_contour_enclosed, ft_outline_get_orientation): Another version of
`FT_Outline_Get_Orientation'. This version differs from the public
one in that each part (contour not enclosed in another contour) of the
outline is checked for orientation.
(FT_Outline_Embolden): Use `ft_outline_get_orientation'.
* src/base/ftsynth.c (FT_GlyphSlot_Embolden): Render the outline and
use bitmap's embolden routine when the outline one failed.
include/internal/services/svtteng.h, src/base/ftobjs.c,
src/truetype/ttdriver.c:
adding a new API named FT_Get_TrueType_Engine_Type to determine
wether we have a patented, unpatented or unimplemented TrueType
bytecode interpreter.
the FT_Get_Module_Flags API was removed consequently.
src/sfnt/ttsbit.h, src/sfnt/ttsbit.c: moving the definition
of 'set_sbit_strike' and 'load_sbit_metrics' to their 2.1.8
location. This is used to support binary compatibility with
the X.Org server's libXfont library.
note that this disables memory optimizations for the embedded
bitmap loader.Sigh....
anymore, this is used to ensure that all FT_BASE internal
functions are available for dynamic linking
* include/freetype/ftcache.h, src/cache/ftcbasic.c,
src/cache/ftccmap.c: try to revive old functions of the
cache sub-system. We also try to support FTC_ImageCache_Lookup
and FTC_ImageCache_SBit with legacy signatures through a gross
hack (hope it works)
Werner Lemberg