The logic is a little clunky, but as before, let's get it right first,
then get it clean once we have a full battery of tests.

R=bsiegert
CC=golang-codereviews
https://codereview.appspot.com/50910043

R=bsiegert
CC=golang-codereviews, remyoudompheng
https://codereview.appspot.com/39440045

See this thread for background:
https://groups.google.com/forum/#!topic/golang-dev/xG7vPi21r8g

R=golang-codereviews, r
CC=golang-codereviews
https://codereview.appspot.com/36380046

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/40690045

rounding.

Pass all the hinting tests. Hooray! Removing the hintingBrokenAt field
will be a follow-up CL.

R=bsiegert, remyoudompheng, jeremyjackins
CC=golang-dev
https://codereview.appspot.com/36880043

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/36300043

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/36220043

The very subtle pp1x versus phantomPoints[0].X distinction is pretty
ugly code, but let's get all the tests passing first, so that we can
refactor with confidence.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/35850043

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/34790045

*before* hinting.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/35070044

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/34700043

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/33650044

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/33750043

opcodes. Calculate dot product with 32-bit math to match C Freetype's
rounding.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/32740043

2nd / 4th phantom point rounding.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/32040043

opcodes.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/29510043

Successfully hint all of DejaVu Sans Oblique, barring 4 exceptions
that are errors in the upstream C Freetype project. See
http://lists.nongnu.org/archive/html/freetype/2013-11/msg00004.html
for details.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/21580044

Also add API for vertical metrics.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/21330043

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/19090043

Also make phantom point delta-X adjustment.

Also make print-glyph-points require Freetype version >= 2.5, which is
the latest version, at the time of writing.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/18370043

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/16500043

The difference is quite large on glyph 509 (period character)
of Adobe Source Sans Pro Bold Italic.

R=golang-dev, nigeltao, bsiegert
CC=golang-dev
https://codereview.appspot.com/16820043

Committer: Nigel Tao <nigeltao@golang.org>

R=golang-dev, nigeltao, bsiegert
CC=golang-dev
https://codereview.appspot.com/16630044

Committer: Nigel Tao <nigeltao@golang.org>

These instructions are found in Adobe Source Sans Pro fonts.

R=golang-dev, nigeltao, bsiegert
CC=golang-dev
https://codereview.appspot.com/16440043

Committer: Nigel Tao <nigeltao@golang.org>

Fix MIRP opcodes, scaled CVT initialization and Hinter.move.

R=bsiegert
CC=golang-dev, remyoudompheng
https://codereview.appspot.com/14930050

Found in Adobe Source Sans Pro.

R=golang-dev, bsiegert, nigeltao
CC=golang-dev
https://codereview.appspot.com/15360043

Committer: Nigel Tao <nigeltao@golang.org>

I fail at copy/pasting from the main Go repo.

R=adg, remyoudompheng
CC=golang-dev
https://codereview.appspot.com/14960051

R=dsymonds, adg
CC=golang-dev
https://codereview.appspot.com/15050048

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14802046

Also increase the maximum recursion depth from 4 to 8, since some
x-deja-vu-sans-oblique glyphs are actually depth 4. It's an arbitrary
limit. The C Freetype code doesn't seem to limit recursion (below
whatever the font says it needs), but it seems safer to have a limit.

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14494062

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14425064

one more test case.

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14691043

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14641043

"go test" drops from 5.203s to 0.474s.

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14566047

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14604044

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14548046

points accordingly.

Yes, it's a mess, but let's pass the tests first, and then refactor.

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14419052

R=bsiegert, r
CC=golang-dev
https://codereview.appspot.com/14414044

testdata/, in anticipation of having other test data.

R=dsymonds
CC=bsiegert, golang-dev
https://codereview.appspot.com/14361044

We now match Freetype C exactly for hinting luxisr.ttf. Yay.

R=bsiegert
CC=golang-dev
https://codereview.appspot.com/14336044