Behdad Esfahbod [Thu, 25 Feb 2016 05:40:09 +0000 (14:40 +0900)]
Speed up buffer variable allocation sanity check
This makes defining HB_NDEBUG much less relevant, to the
point of irrelevance. Sorry about all the fuss in previous
release!
Behdad Esfahbod [Thu, 25 Feb 2016 04:56:47 +0000 (13:56 +0900)]
Add HB_NDEBUG
API changes:
- If NDEBUG is defined, define HB_NDEBUG
- Disable costlier sanity checks if HB_NDEBUG is defined.
In 1.2.3 introduced some code to disable costly sanity checks if
NDEBUG is defined. NDEBUG, however, disables all assert()s as
well. With HB_NDEBUG, one can disable costlier checks but keep
assert()s.
I'll probably add a way to define HB_NDEBUG automatically in
release tarballs. But for now, production systems that do NOT
define NDEBUG, are encouraged to define HB_NDEBUG for our build.
Behdad Esfahbod [Thu, 25 Feb 2016 03:26:26 +0000 (12:26 +0900)]
1.2.3
Behdad Esfahbod [Thu, 25 Feb 2016 03:23:02 +0000 (12:23 +0900)]
Disable internal buffer variable bookkeeping in NDEBUG builds
Saves some sweet time and binary size!
Behdad Esfahbod [Thu, 25 Feb 2016 02:31:03 +0000 (11:31 +0900)]
Add blacklist signature for Times New Roman (Bold) Italic on OS X
Behdad Esfahbod [Thu, 25 Feb 2016 02:20:05 +0000 (11:20 +0900)]
[test] Pass closure to hb_font_set_glyph_func()
Exercises fix in
c335fd7986fe360ab8e1c032c9b988d0d30511eb
Behdad Esfahbod [Thu, 25 Feb 2016 02:11:15 +0000 (11:11 +0900)]
Blacklist GDEF table of timesi.ttf and timesbi.ttf on Win 7
See discussion:
https://lists.freedesktop.org/archives/harfbuzz/2016-February/005489.html
Behdad Esfahbod [Thu, 25 Feb 2016 00:16:05 +0000 (09:16 +0900)]
In trampoline implementation of get_glyph(), don't destroy user data twice!
Behdad Esfahbod [Wed, 24 Feb 2016 11:27:13 +0000 (20:27 +0900)]
[ot-font] Accelerate cmap format4 get_glyph
Behdad Esfahbod [Wed, 24 Feb 2016 10:52:36 +0000 (19:52 +0900)]
[ot-font] Towards accelerating get_glyph()
Behdad Esfahbod [Wed, 24 Feb 2016 10:32:43 +0000 (19:32 +0900)]
[ot-font] Remove level of indirection in get_glyph_variant
Behdad Esfahbod [Wed, 24 Feb 2016 10:10:06 +0000 (19:10 +0900)]
[TODO] Update
This file is rather obsolete. Still, give it a refresh.
Behdad Esfahbod [Wed, 24 Feb 2016 10:05:23 +0000 (19:05 +0900)]
Add get_nominal_glyph() and get_variation_glyph() instead of get_glyph()
New API:
- hb_font_get_nominal_glyph_func_t
- hb_font_get_variation_glyph_func_t
- hb_font_funcs_set_nominal_glyph_func()
- hb_font_funcs_set_variation_glyph_func()
- hb_font_get_nominal_glyph()
- hb_font_get_variation_glyph()
Deprecated API:
- hb_font_get_glyph_func_t
- hb_font_funcs_set_glyph_func()
Clients that implement their own font-funcs are encouraged to replace
their get_glyph() implementation with a get_nominal_glyph() and
get_variation_glyph() pair. The variation version can assume that
variation_selector argument is not zero.
Behdad Esfahbod [Wed, 24 Feb 2016 08:32:22 +0000 (17:32 +0900)]
1.2.2
Behdad Esfahbod [Wed, 24 Feb 2016 07:06:40 +0000 (16:06 +0900)]
Fix recent test
Not sure why the FT functions were returning advance 1024. This
caused failure on drone.io. Switch to hb-ot-font and disable
glyph names.
Behdad Esfahbod [Wed, 24 Feb 2016 07:06:23 +0000 (16:06 +0900)]
[tests] Fix for multiple options in test runner scripts
Behdad Esfahbod [Wed, 24 Feb 2016 06:53:40 +0000 (15:53 +0900)]
Partially revert
86c68c7a2c971efe8e35b1f1bd99401dc8b688d2
That commit moved the advance adjustment for mark positioning to
be applied immediately, instead of doing late before. This breaks
if mark advances are zeroed late, like in Arabic. Also, easier to
hit it in RTL scripts since a single mark with non-zero advance is
enough to hit the bug, whereas in LTR, at least two marks are needed.
This reopens https://github.com/behdad/harfbuzz/issues/211
The cursive+mark interaction is broken again. To be fixed in a
different way.
Behdad Esfahbod [Wed, 24 Feb 2016 06:52:37 +0000 (15:52 +0900)]
Add test for mark positioning in rtl with non-zero mark advance
Apparently I broke this
86c68c7a2c971efe8e35b1f1bd99401dc8b688d2.
Fix coming.
Behdad Esfahbod [Wed, 24 Feb 2016 06:50:33 +0000 (15:50 +0900)]
[tests] Allow commenting out tests to be skipped
Behdad Esfahbod [Wed, 24 Feb 2016 03:30:57 +0000 (12:30 +0900)]
[Android.mk] Update with warning flags
Behdad Esfahbod [Tue, 23 Feb 2016 06:38:43 +0000 (15:38 +0900)]
1.2.1
Behdad Esfahbod [Tue, 23 Feb 2016 06:19:27 +0000 (15:19 +0900)]
Add note re only adding tests with Free Software fonts
Behdad Esfahbod [Tue, 23 Feb 2016 04:47:16 +0000 (13:47 +0900)]
[test] Add README about how to add shaping tests
Behdad Esfahbod [Tue, 23 Feb 2016 04:45:38 +0000 (13:45 +0900)]
Remove MANIFEST files
They are unused currently. We can add later if we hook them up
to anything useful.
Behdad Esfahbod [Mon, 22 Feb 2016 09:22:44 +0000 (18:22 +0900)]
In cluster-level=0, group ZWJ/ZWNJ with previous cluster
This better emulates Unicode grapheme clusters.
Note that Uniscribe does NOT do this, but should be harmless with most clients,
and improve fallback with clients that use HarfBuzz cluster as unit of fallback.
Fixes https://github.com/behdad/harfbuzz/issues/217
Behdad Esfahbod [Mon, 22 Feb 2016 07:00:59 +0000 (16:00 +0900)]
Minor
Behdad Esfahbod [Mon, 22 Feb 2016 06:56:29 +0000 (15:56 +0900)]
[coretext] Move CTFont construction to face_data
Behdad Esfahbod [Mon, 22 Feb 2016 06:50:12 +0000 (15:50 +0900)]
[coretext] Move code around
Behdad Esfahbod [Mon, 22 Feb 2016 06:42:53 +0000 (15:42 +0900)]
[coretext] Move code around
Behdad Esfahbod [Mon, 22 Feb 2016 06:37:18 +0000 (15:37 +0900)]
[build] git.mk fix for m4/gtk-doc.m4
Behdad Esfahbod [Mon, 22 Feb 2016 06:32:52 +0000 (15:32 +0900)]
[win32] Install git.mk
Behdad Esfahbod [Mon, 22 Feb 2016 02:44:45 +0000 (11:44 +0900)]
Make FixedVersion a template
Behdad Esfahbod [Mon, 22 Feb 2016 06:31:22 +0000 (15:31 +0900)]
[coretext] Fix leak!
Behdad Esfahbod [Mon, 22 Feb 2016 06:28:37 +0000 (15:28 +0900)]
[coretext] Move code around
Behdad Esfahbod [Mon, 22 Feb 2016 06:12:27 +0000 (15:12 +0900)]
[coretext] Ignore PPEM in font size selection
Behdad Esfahbod [Mon, 22 Feb 2016 06:07:20 +0000 (15:07 +0900)]
[coretext] Limit grapheme-cluster forming to cluster-level=0
Behdad Esfahbod [Mon, 22 Feb 2016 05:59:39 +0000 (14:59 +0900)]
[coretext] Fix shaping with varying font size
Fixes https://github.com/libass/libass/issues/212
Behdad Esfahbod [Fri, 19 Feb 2016 08:56:50 +0000 (15:56 +0700)]
1.2.0
Behdad Esfahbod [Fri, 19 Feb 2016 08:13:07 +0000 (15:13 +0700)]
[fuzzing] Add test for recent fix
Test from https://github.com/behdad/harfbuzz/issues/223
I forgot that we do run hb-fuzzer on tests in shaping/tests/fuzzed.tests.
Behdad Esfahbod [Fri, 19 Feb 2016 08:12:24 +0000 (15:12 +0700)]
[fuzzing] Make test runner actually work
Behdad Esfahbod [Fri, 19 Feb 2016 08:12:16 +0000 (15:12 +0700)]
[fuzzing] Add TODO item
Behdad Esfahbod [Fri, 19 Feb 2016 07:52:31 +0000 (14:52 +0700)]
Avoid buffer->move_to() in case of buffer error
Fixes https://github.com/behdad/harfbuzz/issues/223
Right now we cannot test this because it has to be tested using hb-fuzzer.
We should move all fuzzing tests from test/shaping/tests/fuzzed.tests to
test/fuzzing/ and have its own test runner. At that point, should add
test from this issue as well.
Behdad Esfahbod [Fri, 19 Feb 2016 07:43:39 +0000 (14:43 +0700)]
[fuzzing] Make "make hb-fuzzer" build lib dependency
Not going to do with util/, but is convenient here.
Behdad Esfahbod [Thu, 18 Feb 2016 12:31:51 +0000 (19:31 +0700)]
Remove pointless overflow check in pointer math
Fixes https://github.com/behdad/harfbuzz/issues/227
Behdad Esfahbod [Thu, 18 Feb 2016 10:06:25 +0000 (17:06 +0700)]
Emoji skin tone modifiers need to be treated as combining marks
Fixes https://github.com/behdad/harfbuzz/issues/169
Behdad Esfahbod [Tue, 16 Feb 2016 10:16:33 +0000 (17:16 +0700)]
[USE] Zero mark advances by GDEF early
This is what Microsoft's implementation does. Marks that need advance
need to add it back using 'dist' or other feature in GPOS. Update tests to
match.
Behdad Esfahbod [Tue, 16 Feb 2016 09:07:20 +0000 (16:07 +0700)]
[GPOS] Fix interaction of mark attachments and cursive chaining
Fixes https://github.com/behdad/harfbuzz/issues/211
What happens in that bug is that a mark is attached to base first,
then a second mark is cursive-chained to the first mark. This only
"works" because it's in the Indic shaper where mark advances are
not zeroed.
Before, we didn't allow cursive to run on marks at all. Fix that.
We also where updating mark major offsets at the end of GPOS, such
that changes in advance of base will not change the mark attachment
position. That was superior to the alternative (which is what Uniscribe
does BTW), but made it hard to apply cursive to the mark after it
was positioned. We could track major-direction offset changes and
apply that to cursive in the post process, but that's a much trickier
thing to do than the fix here, which is to immediately apply the
major-direction advance-width offsets... Ie.:
https://github.com/behdad/harfbuzz/issues/211#issuecomment-
183194739
If this breaks any fonts, the font should be fixed to do mark attachment
after all the advances are set up first (kerning, etc).
Finally, this, still doesn't make us match Uniscribe, for I explained
in that bug. Looks like Uniscribe applies minor-direction cursive
adjustment immediate as well. We don't, and we like it our way, at
least for now. Eg. the sequence in the test case does this:
- The first subscript attaches with mark-to-base, moving in x only,
- The second subscript attaches with cursive attachment to first subscript
moving in x only,
- A final context rule moves the first subscript up by 104 units.
The way we do, the final shift-up, also shifts up the second subscript
mark because it's cursively-attached. Uniscribe doesn't. We get:
[ttaorya=0+1307|casubscriptorya=0@-242,104+-231|casubscriptnarroworya=0@20,104+507]
while Uniscribe gets:
[ttaorya=0+1307|casubscriptorya=0@-242,104+-211|casubscriptnarroworya=0+487]
note the different y-offset of the last glyph. In our view, after cursive,
things move together, period.
Behdad Esfahbod [Fri, 12 Feb 2016 05:50:17 +0000 (12:50 +0700)]
Minor
Behdad Esfahbod [Thu, 11 Feb 2016 09:57:52 +0000 (16:57 +0700)]
[GPOS] Minor
No effect.
Behdad Esfahbod [Thu, 11 Feb 2016 09:48:13 +0000 (16:48 +0700)]
[GPOS] Merge fixing of offsets for cursive and mark attachments
Part of fixing https://github.com/behdad/harfbuzz/issues/211
Behdad Esfahbod [Thu, 11 Feb 2016 09:34:28 +0000 (16:34 +0700)]
[GPOS] Divide position_finish() into two phases, for advances and offsets
Right now the position_finish_advances() is empty. To be used for
spacing attachments proposal later.
Behdad Esfahbod [Thu, 11 Feb 2016 09:27:41 +0000 (16:27 +0700)]
[ot] Minor shuffling code around
Behdad Esfahbod [Thu, 11 Feb 2016 08:28:55 +0000 (15:28 +0700)]
[GPOS] Add harmless recursion in fix_mark_attachment()
Will do nothing. Just useful for merging two functions.
Behdad Esfahbod [Thu, 11 Feb 2016 08:25:28 +0000 (15:25 +0700)]
[GPOS] Merge attach_chain() and cursive_chain()
Differentiate, using new attach_type().
Behdad Esfahbod [Thu, 11 Feb 2016 07:53:11 +0000 (14:53 +0700)]
[GPOS] Minor shuffling
Behdad Esfahbod [Thu, 11 Feb 2016 07:49:10 +0000 (14:49 +0700)]
[GPOS] Negate sign of attach_lookback(), and rename it to attach_chain()
No behavior change. Preparing to unify how cursive and mark attachments
work.
Behdad Esfahbod [Thu, 11 Feb 2016 05:15:38 +0000 (12:15 +0700)]
[tests] Add test for advance zeroing of an ASCII letter marked as mark in GDEF
Behdad Esfahbod [Thu, 11 Feb 2016 05:14:27 +0000 (12:14 +0700)]
Remove font-dependent ASCII-only perf hack
Is confusing. I already hit it myself. Remove. We can optimize
ASCII based on Unicode properties. But should not do based on
assumptions on the font.
Behdad Esfahbod [Wed, 10 Feb 2016 11:43:43 +0000 (18:43 +0700)]
Fix previous commit!
Behdad Esfahbod [Wed, 10 Feb 2016 11:39:59 +0000 (18:39 +0700)]
[tests] Fix fonts in
cc4a78bf22c4a735b84c89b04be3bb531f42135e
They had an invalid LookupFlag (32).
Behdad Esfahbod [Wed, 10 Feb 2016 11:29:54 +0000 (18:29 +0700)]
Remove now-unused mark zeroing BY_UNICODE
Behdad Esfahbod [Wed, 10 Feb 2016 11:24:08 +0000 (18:24 +0700)]
[tests] Add tests for Latin mark zeroing
Behdad Esfahbod [Wed, 10 Feb 2016 11:20:02 +0000 (18:20 +0700)]
[tests] Add tests for Thai mark zeroing
Behdad Esfahbod [Wed, 10 Feb 2016 11:10:15 +0000 (18:10 +0700)]
Change default mark advance zeroing behavior from Unicode to GDEF
See thread at:
https://lists.freedesktop.org/archives/harfbuzz/2016-February/005462.html
Behdad Esfahbod [Wed, 10 Feb 2016 07:11:43 +0000 (14:11 +0700)]
Minor
Behdad Esfahbod [Wed, 10 Feb 2016 04:11:36 +0000 (11:11 +0700)]
Merge pull request #164 from fanc999/msvc.build
Enable Build of HarfBuzz under Visual Studio using NMake Makefiles
Behdad Esfahbod [Wed, 10 Feb 2016 04:10:08 +0000 (11:10 +0700)]
[README] Add ABI Tracker link
Chun-wei Fan [Wed, 9 Dec 2015 06:57:15 +0000 (14:57 +0800)]
MSVC builds: Add support to build DirectWrite shaping backend
Add into the NMake Makefiles to build the DirectWrite shaping backend,
but as PR #134 mentions, this is considered to be in an experimental state,
so don't include this in the build by default for now. This is most probably
going to replace the Uniscribe backend eventually, since DirectWrite is meant
to be Uniscribe's replacement, and is needed for Windows Store apps if a
system shaping API is to be used.
Chun-wei Fan [Wed, 3 Feb 2016 10:31:23 +0000 (18:31 +0800)]
build: Build the DirectWrite sources if requested
This adds to the autotools build system so that the (experimental)
DirectWrite support for HarfBuzz is built (and dist'ed).
Chun-wei Fan [Tue, 3 Nov 2015 11:00:42 +0000 (19:00 +0800)]
build: Support Visual Studio builds using NMake
This adds a set of NMake Makefiles that can be used to build HarfBuzz, from
the standard basic build building the minimal HarfBuzz DLL (consisting
of OpenType, fallback and Uniscribe support only), to a full fledged build
consisting of GLib and FreeType support, as well as building the utilities,
the test programs in src/ and test/api, and HarfBuzz-ICU and
HarfBuzz-GObject, and up to building the introspection files. This means a
flexible build mechanism is supported here, so anything that is supported
for a Windows build (code-wise), should all be supported by this build
system.
As in an earlier commit, the source listings are shared with the autotools
builds with the various Makefile.sources in src/, src/hb-ucdn and util/, and
this set of NMake Makefiles will transform these lists into the form they
want.
In the current form, all the test programs in test/api pass, and this has
been checked successfully with 'make -j8 distcheck'.
Chun-wei Fan [Tue, 3 Nov 2015 10:52:46 +0000 (18:52 +0800)]
MSVC builds: Add a pre-configured config.h(.win32) template
This adds a pre-configured config.h template that can be used for Visual
Studio builds, where autotools is not normally available. This has the
configs that are suitable for Visual Studio builds, as well as all the
features used for Windows builds enabled (HAVE_OT, HAVE_FALLBACK and
HAVE_UNISCRIBE).
Note that the optional features are not enabled here, they are enabled by
/D's (or -D's) in the NMake Makefiles as requested.
Chun-wei Fan [Tue, 3 Nov 2015 08:56:27 +0000 (16:56 +0800)]
build: Move source listings into separate Makefile snippets
This moves all the source listings in src/Makefile.am,
src/hb-ucdn/Makefile.am and util/Makefile.am into separate Makefile
snippets, so that they may be shared between different Makefile-based
build systems, such as NMake for Visual Studio.
Behdad Esfahbod [Tue, 2 Feb 2016 08:39:19 +0000 (16:39 +0800)]
Fix more hangs in case of buffer allocation errors
Hopefully
fixes https://github.com/behdad/harfbuzz/issues/214
Behdad Esfahbod [Tue, 2 Feb 2016 04:33:32 +0000 (12:33 +0800)]
Merge pull request #222 from n8willis/master
Add usermanual chapter on cluster levels
n8willis [Thu, 28 Jan 2016 18:21:32 +0000 (12:21 -0600)]
Update Makefile.am
n8willis [Thu, 28 Jan 2016 18:14:12 +0000 (12:14 -0600)]
Added initial usermanual chapter on cluster levels.
Behdad Esfahbod [Mon, 18 Jan 2016 11:47:18 +0000 (12:47 +0100)]
Merge pull request #216 from mhosken/master
Refactor graphite2 to make ovleraps into negative width spaces
Martin Hosken [Sat, 16 Jan 2016 23:24:00 +0000 (17:24 -0600)]
Refactor graphite2 to make ovleraps into negative width spaces
Behdad Esfahbod [Tue, 12 Jan 2016 16:17:21 +0000 (16:17 +0000)]
Fix compiler warning
Fixes https://github.com/behdad/harfbuzz/issues/212
Behdad Esfahbod [Tue, 12 Jan 2016 13:05:01 +0000 (13:05 +0000)]
Minor
Behdad Esfahbod [Mon, 11 Jan 2016 18:31:04 +0000 (18:31 +0000)]
[fuzzing] Fuzz input text
Very rudimentary right now, but will get kcc's bot going.
From
https://github.com/behdad/harfbuzz/issues/139#issuecomment-
163580783
Behdad Esfahbod [Mon, 11 Jan 2016 17:38:41 +0000 (17:38 +0000)]
[OT] Get list of lookup subtables once and loop through them
This speeds up shaping the Amiri font by over 15%.
This was primarily needed for my work on OpenType GX, since
we will be collecting only sublookups that are "active" for
current font instance; but it's a nice boost in general as
well.
We might, in the future, collect subtables in the lookup_accel.
That would also allow us to do a per-subtbale set-digest, which
should speed things up some more, specially for ContextChainFormat3
lookups... Amiri, for example, contains one lookup with 53
subtables!
Behdad Esfahbod [Mon, 11 Jan 2016 13:16:35 +0000 (13:16 +0000)]
1.1.3
Behdad Esfahbod [Mon, 11 Jan 2016 12:58:45 +0000 (12:58 +0000)]
[docs] Minor fixes.
Behdad Esfahbod [Mon, 11 Jan 2016 12:28:37 +0000 (12:28 +0000)]
[coretext] Speed up CoreText font fallback
Fixes https://code.google.com/p/chromium/issues/detail?id=547912
Behdad Esfahbod [Thu, 7 Jan 2016 22:53:59 +0000 (22:53 +0000)]
Add Homebrew build instructions
Behdad Esfahbod [Thu, 7 Jan 2016 22:52:16 +0000 (22:52 +0000)]
Don't build hb-fuzzer in "make", only "make check"
Behdad Esfahbod [Thu, 7 Jan 2016 19:11:41 +0000 (19:11 +0000)]
Add 9 reserved entries to hb_font_extents_t
Might add italic-angle, underline/strikethrough-position/thickness
in the future. Do this before new struct goes into a release.
Behdad Esfahbod [Wed, 6 Jan 2016 12:29:50 +0000 (12:29 +0000)]
[uniscribe] Fix negative advance
This happens with at least one test font I have.
Behdad Esfahbod [Wed, 6 Jan 2016 12:20:47 +0000 (12:20 +0000)]
[test] Add test text for Kaithi
Behdad Esfahbod [Tue, 5 Jan 2016 13:26:20 +0000 (13:26 +0000)]
Increase sanitize edit count from 8 to 32
See previous commit.
Behdad Esfahbod [Tue, 5 Jan 2016 13:23:45 +0000 (13:23 +0000)]
Don't count fixing-up FeatureParams offset as error
The font Garamond Premier Pro Caption (and possibly many other
Adobe fonts), have many FeatureParamsSize tables with the old
wrong offset. We handle fixing those up, but they were still
contributing to edit_count, and when I reduced HB_SANITIZE_MAX_EDIT
from 100 to 8 in
14c2de321826c36037adde859ccca3e2011325a9, these
fonts were now getting GPOS dropped and hence kerning disabled.
Fix, by not counting edits made towareds offset fix-up. I'll
also increase edit count again, in the next commit.
Behdad Esfahbod [Sat, 2 Jan 2016 13:25:18 +0000 (13:25 +0000)]
Check magicNumber in head table during sanitize
Behdad Esfahbod [Sat, 2 Jan 2016 00:59:14 +0000 (00:59 +0000)]
Merge pull request #208 from khaledhosny/travis-trusty
[travis] Try building on Trusty
Khaled Hosny [Fri, 1 Jan 2016 20:36:19 +0000 (00:36 +0400)]
[travis] Try building on Trusty
Trusty has gtk-doc 1.20 which has better markdown support which should
fix some of the brokenness in the newly added documentation. The build
seems to be as fast as the current container-based build on Precise.
Behdad Esfahbod [Fri, 1 Jan 2016 19:06:43 +0000 (19:06 +0000)]
[travis] Only run coveralls if testing behdad/harfbuzz
Fixes failures in others' forks, like:
https://travis-ci.org/khaledhosny/harfbuzz/jobs/
99719710#L3912
Behdad Esfahbod [Fri, 1 Jan 2016 16:44:01 +0000 (16:44 +0000)]
Merge pull request #207 from khaledhosny/docs-4
[docs] More buffers documentation
Khaled Hosny [Fri, 1 Jan 2016 16:38:21 +0000 (20:38 +0400)]
[docs] More buffers documentation
Behdad Esfahbod [Wed, 30 Dec 2015 21:20:14 +0000 (22:20 +0100)]
Merge pull request #205 from khaledhosny/more-docs
More docs
Behdad Esfahbod [Wed, 30 Dec 2015 21:20:00 +0000 (22:20 +0100)]
Merge pull request #206 from nico/patch-3
Remove unused HB_SHAPER_DATA_ENSURE_DECLARE lines.