134ce0560c
use expat for xml read
...
we were using libxml for xml load, use expat instead, we get it for free
with glib
2017-02-25 13:07:43 +00:00
dc2b567ee2
start looking at xml parse use in vips.c
2017-02-24 14:57:20 +00:00
420a918cbf
Merge branch 'master' into try-expat
2017-02-24 14:31:33 +00:00
dbbe8b77c1
vipsheader could crash on bad field names
...
we were not checking the return of vips_image_get_as_string()
2017-02-24 14:30:26 +00:00
9ba8d32083
add expat to package deps
2017-02-24 14:28:08 +00:00
6cbda61ec1
fix thumbnail crop on rotated images
...
crop after rotate so we don't need to rotate the crop box
thanks ElHeineken!
see https://github.com/jcupitt/libvips/issues/608
2017-02-23 20:26:59 +00:00
4bcbcbbf49
try to fix IM load of ico files
...
8-bit ico files were being reported at 32-bit, because we used the
->depth field to get pixel depth
try using GetImageChannelDepth() instead, who knows, it might work
see https://github.com/jcupitt/php-vips/issues/34
2017-02-23 20:05:35 +00:00
156f729d75
note new seq thing
2017-02-23 20:05:11 +00:00
12805808da
better fix for the reorder memleak
...
we could try reordering during image close, and after the arrays had
been freed
2017-02-23 17:35:52 +00:00
4bbd56fb0a
handle transparency in vips_region_shrink()
...
we weren't multiplying through alpha in vips_region_shrink(), causing
fringing on pyramids with transparent elements
thanks sickrandir!
see https://github.com/jcupitt/libvips/issues/599
2017-02-23 17:04:25 +00:00
f84825a634
fix a small memleak in reorder
...
if set_input was called more than once, it could leak the arrays
thanks @gczuba!
see https://github.com/jcupitt/libvips/issues/607
2017-02-23 07:50:35 +00:00
09774740a7
better cache sizing
2017-02-22 20:11:52 +00:00
4a9d10a0c8
Merge branch 'master' into remove-seq-stalling
2017-02-22 16:11:49 +00:00
abfabd98ce
tiny docs update
2017-02-22 13:31:27 +00:00
a155521baf
fix translation files
2017-02-22 13:28:17 +00:00
3564339362
notes
2017-02-22 11:41:41 +00:00
2a85d41570
Merge branch 'master' into remove-seq-stalling
2017-02-22 10:37:54 +00:00
09eafa4f9f
fix xoff/yoff in mosaic output
...
we didn't record xoff / yoff correctly for negative offsets
see https://github.com/jcupitt/libvips/issues/606
2017-02-22 10:27:01 +00:00
ec75e3a2cc
notes
2017-02-22 09:53:13 +00:00
cb83918a3f
remove readbehind
2017-02-21 17:14:33 +00:00
a4d3c2a754
more UNBUF fixes
...
try to keep compat better, remove the single-thread-first-tile idea
2017-02-21 16:53:48 +00:00
959f412380
remove stalling
2017-02-21 15:03:15 +00:00
c6eb9ee0f2
experiment with removal of seq stall
2017-02-21 09:44:21 +00:00
6691e07d72
notes on seq removal
2017-02-21 09:14:26 +00:00
f45f6ad52f
oop, forgot to comment out DEBUG
...
thanks Lovell
2017-02-21 08:24:46 +00:00
197a21ed0e
don't load 0 length profiles
...
some tiffs have 0 length icc profiles ... test profile length before we
try to load it
see https://github.com/jcupitt/libvips/issues/605
2017-02-20 22:36:35 +00:00
3f92cfc935
oops, don't copy header fields
...
in prev commit
2017-02-20 13:42:19 +00:00
abf6a250c2
fix wrapmany compat
...
it was not setting reorder hints, then failing on recomp
the nip2 display control bar was broken by this, for example
2017-02-20 13:32:12 +00:00
6d95cfd0c5
tweak docs
...
see https://github.com/jcupitt/libvips/issues/558
2017-02-20 11:10:04 +00:00
6fb2122a5f
notes
2017-02-19 17:51:36 +00:00
53f374434d
fix build on centos7
...
see https://github.com/jcupitt/libvips/issues/601
2017-02-19 10:20:37 +00:00
646960077d
fix cache hint for shrink
...
shrink had UNBUFFERED as a cache hint, but for non-int shrinks it used
affine, so you could get out of order reads
set regular SEQ as the cache hint, and switch to reduce for the final
resize in the non-int case
see https://github.com/jcupitt/libvips/issues/596
2017-02-09 09:10:03 +00:00
8834b55c47
oops
...
ahem missed a rename
2017-01-30 11:28:23 +00:00
21a5d0dfe4
speeling
2017-01-29 12:24:25 +00:00
e210411404
enable bigtiff for large, uncompressed writes
...
if compression is off and the image is >4gb, make sure bigtiff is
enabled
this won't always work: if the image is just under the 4gb limit, it
could be pushed over the edge by a lot of metadata. And of course the
libtiff vips is using could be missing bigtiff support
see https://github.com/jcupitt/libvips/issues/591
2017-01-29 12:20:57 +00:00
2b482fc2cf
seems to work!
2017-01-27 16:43:37 +00:00
9cc868a7e3
Merge branch 'master' into add-tiff-scanline-load
2017-01-27 11:36:28 +00:00
756e69cf96
make pngload from buffer set flags
...
and remove some dead code from jpegload
2017-01-27 11:15:48 +00:00
94434c2359
tiff buffer loader supports get_flags
...
the tiff buffer loader was not setting any flags, so seq was using a
temp file (for example)
see https://github.com/jcupitt/libvips/issues/590
2017-01-27 10:32:47 +00:00
ec093100fb
use fits_open_diskfile(), not fits_open_file()
...
so we can open any filename ... fits_open_file() parses fits extended
filename formats
see https://github.com/jcupitt/libvips/issues/584
2017-01-26 20:17:29 +00:00
386d44ef38
more dbg in foreign.c
2017-01-26 18:29:43 +00:00
a34cc2fdf3
notes
2017-01-26 12:07:14 +00:00
0108d1d605
experiment with scanline load for tiff
...
trying to get memuse down for images with huge rows-per-strip
2017-01-26 12:05:14 +00:00
276fc83fa1
more VIPS_INIT() docs
...
see https://github.com/jcupitt/libvips/issues/588
2017-01-26 08:46:27 +00:00
7136053451
sort max/min results by y and x coordinate
...
previously the list of maxima or minima found by max/min were in found
order, so essentially random
they now sort equal values by y then x coordinate, so the results are
consistent between runs, provided there are fewer maxima / minima than
are being tracked
2017-01-24 14:25:15 +00:00
fb544e3f25
add note about non-determinism in stats
...
added a note to the docs for max/min/stats about non-determinism in the
results if there are many maxima/minima
thanks Mark
2017-01-24 12:14:55 +00:00
61d9f03ef8
tiny fixes
2017-01-24 11:02:40 +00:00
c9ce8f70a1
fix an off-by-one error in hist_local
...
an odd window width caused an off-by-one error in histogram construction
2017-01-24 10:51:12 +00:00
36761bcfd7
make vips_gaussnoise() pixels reproducible
...
previously, pixel values were regenerated on every calculation, so they
changed on recomputation
pixel values are now generated from the pixel (x, y) coordinate plus a per-call
seed
thanks MvGulik, see
https://github.com/jcupitt/nip2/issues/60
https://github.com/jcupitt/libvips/issues/583
2017-01-24 09:36:25 +00:00
c963678549
more hist_local fixups
...
add tests, add restrict
2017-01-22 10:36:07 +00:00
John Cupitt