Introduce ENABLE_LATIN_1 compile flag

Introduce ENABLE_LATIN_1 compile flag

Mostly a bunch of renaming when flag is disabled.

R=yangguo@chromium.org
BUG=

Committed: https://code.google.com/p/v8/source/detail?r=13340

[Yang, 2013-01-07 16:15:07]

Some comments

- A lot of the changes from kMaxAsciiCode to kMaxOneByteCode are not gated by
ENABLE_LATIN_1. Are we sure this doesn't break things?
- Many comments and variable names still refer to ASCII, I assume CL(s) are
coming to fix this?
- I'm not sure moving to Latin1 is a good idea for FilterASCII, I put an inline
comment for this issue. I suggest keeping the ASCII semantics, based on
kAsciiDataHintTag.

https://codereview.chromium.org/11759008/diff/5001/src/jsregexp.cc
File src/jsregexp.cc (right):

https://codereview.chromium.org/11759008/diff/5001/src/jsregexp.cc#newcode2010
src/jsregexp.cc:2010: if (*border - 1 > String::kMaxOneByteCharCode &&  // ASCII
case.
Change comment.

https://codereview.chromium.org/11759008/diff/5001/src/jsregexp.cc#newcode2201
src/jsregexp.cc:2201: bool ascii,
Changing the parameter name would make sense.

https://codereview.chromium.org/11759008/diff/5001/src/jsregexp.cc#newcode2512
src/jsregexp.cc:2512: bool QuickCheckDetails::Rationalize(bool asc) {
Ditto (parameter name).

https://codereview.chromium.org/11759008/diff/5001/src/jsregexp.cc#newcode2567
src/jsregexp.cc:2567: if (compiler->ascii()) {
Maybe onebyte() instead of ascii()?

https://codereview.chromium.org/11759008/diff/5001/src/jsregexp.cc#newcode2868
src/jsregexp.cc:2868: if (quarks[j] > String::kMaxOneByteCharCode) {
Does the comment still hold true for the Latin1 char set? Apparently ÿ (\u00ff)
is part of Latin1, but its uppercase  Ÿ is is \u0178...

Maybe we want to keep FilterASCII the way it is, and use it in combination with
kAsciiDataHintTag.

https://codereview.chromium.org/11759008/diff/5001/src/objects.h
File src/objects.h (right):

https://codereview.chromium.org/11759008/diff/5001/src/objects.h#newcode7069
src/objects.h:7069: bool IsAscii() { return state_ == ASCII; }
I assume this is planned to be renamed to IsOneByte?

https://codereview.chromium.org/11759008/diff/5001/src/runtime.cc
File src/runtime.cc (right):

https://codereview.chromium.org/11759008/diff/5001/src/runtime.cc#newcode5937
src/runtime.cc:5937: #ifndef ENABLE_LATIN_1
I assume that this part will be used for strings with kAsciiDataHintTag, in a
later CL?

[drcarney, 2013-01-08 15:06:31]

I think this takes care of the FilterASCII problem to start.

I'm planning to do a series of mass renames to do with the ascii/one byte naming
problem after there's some stability.

[Yang, 2013-01-09 10:30:42]

On 2013/01/08 15:06:31, dcarney wrote:
> I think this takes care of the FilterASCII problem to start.
> 
> I'm planning to do a series of mass renames to do with the ascii/one byte
naming
> problem after there's some stability.

LGTM.