Use the same mechanism to process int/float literals

Use the same mechanism to process int/float literals

This also fixes the float overflow errno leaking bug.

BUG=
R=alokp@chromium.org

Committed: https://code.google.com/p/angleproject/source/detail?r=f382bff

[Zhenyao Mo, 2013-09-09 19:16:36]

Alok, please take a look.

Tested on my linux/windows with webgl conformance tests.

[Alok Priyadarshi, 2013-09-09 20:00:41]

lgtm with minor comments.

https://codereview.appspot.com/13368050/diff/1/src/compiler/util.cpp
File src/compiler/util.cpp (right):

https://codereview.appspot.com/13368050/diff/1/src/compiler/util.cpp#newcode7
src/compiler/util.cpp:7: #include "util.h"
this should be compiler/util.h

https://codereview.appspot.com/13368050/diff/1/src/compiler/util.cpp#newcode11
src/compiler/util.cpp:11: #include "preprocessor/numeric_lex.h"
same here

[Zhenyao Mo, 2013-09-09 20:20:19]

https://codereview.appspot.com/13368050/diff/1/src/compiler/util.cpp
File src/compiler/util.cpp (right):

https://codereview.appspot.com/13368050/diff/1/src/compiler/util.cpp#newcode7
src/compiler/util.cpp:7: #include "util.h"
On 2013/09/09 20:00:42, Alok Priyadarshi wrote:
> this should be compiler/util.h

Done.

https://codereview.appspot.com/13368050/diff/1/src/compiler/util.cpp#newcode11
src/compiler/util.cpp:11: #include "preprocessor/numeric_lex.h"
On 2013/09/09 20:00:42, Alok Priyadarshi wrote:
> same here

Done.

[Zhenyao Mo, 2013-09-09 20:20:29]

Committed patchset #2 manually as rf382bff (presubmit successful).