Make style building for 'font-size' less custom.

Make style building for 'font-size' less custom.

This change adds a FontDescription::Size structure which represents a
(specified) font size in transit. The conversion from CSSValue* to
font size (now FontDescription::Size) has been moved to
StyleBuilderConverter. This reduces the dependency on RenderStyle/CSSValue
in FontBuilder.

Keyword sizes are now resolved into actual float sizes in FontBuilder::
setSize. This happens only if the FontDescription::Size has a keyword size,
but has no float value size. The idea is that if both a keyword size
and float value size is provided from the outside, it means we inherit
the "entire" Size as-is, and no re-resolving of the keyword size is
necessary.

The m_fontSizehasViewportUnits flag has been removed. StyleBuilderConverter
is now marking the relevant style directly (more or less) during conversion
instead.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=182154

[andersr, 2014-08-15 12:40:42]

https://codereview.chromium.org/468793003/diff/1/Source/platform/fonts/FontDe...
File Source/platform/fonts/FontDescription.cpp (right):

https://codereview.chromium.org/468793003/diff/1/Source/platform/fonts/FontDe...
Source/platform/fonts/FontDescription.cpp:101: return Size(0, size.value * 1.2,
size.isAbsolute);
Note a keyword size of '0' here, although the keyword size (if any) is known,
for compatibility with how it worked before.

It looks like the spec says that a 'larger' size for 'medium' for example, is
'large'. Should this be something like this instead?

if (size.keyword && size.keyword < 7)
    return Size(size.keyword + 1, 0, false);
return Size(0, size.value * 1.2, size.isAbsolute);

[dglazkov, 2014-09-15 19:29:42]

LGTM. My apologies for taking this long! Please bug me more annoyingly, I am
very sorry.

On 2014/08/15 at 12:40:42, andersr wrote:
>
https://codereview.chromium.org/468793003/diff/1/Source/platform/fonts/FontDe...
> File Source/platform/fonts/FontDescription.cpp (right):
> 
>
https://codereview.chromium.org/468793003/diff/1/Source/platform/fonts/FontDe...
> Source/platform/fonts/FontDescription.cpp:101: return Size(0, size.value *
1.2, size.isAbsolute);
> Note a keyword size of '0' here, although the keyword size (if any) is known,
for compatibility with how it worked before.
> 
> It looks like the spec says that a 'larger' size for 'medium' for example, is
'large'. Should this be something like this instead?
> 
> if (size.keyword && size.keyword < 7)
>     return Size(size.keyword + 1, 0, false);
> return Size(0, size.value * 1.2, size.isAbsolute);

Sounds good, but let's do it as another change.

[andersr, 2014-09-17 10:43:50]

On 2014/09/15 at 19:29:42, dglazkov wrote:
> LGTM. My apologies for taking this long! Please bug me more annoyingly, I am
very sorry.

Don't worry, this wasn't very urgent.

[andersr, 2014-09-17 10:44:03]

The CQ bit was checked by andersr@opera.com

[commit-bot: I haz the power, 2014-09-17 10:44:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patchset/468793003/40001

[commit-bot: I haz the power, 2014-09-17 11:50:22]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-09-17 11:50:24]

Try jobs failed on following builders:
  linux_blink_dbg on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/linux_blink_dbg/builds/2...)
  linux_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/2...)
  win_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/27432)

[andersr, 2014-09-17 12:45:15]

The CQ bit was checked by andersr@opera.com

[commit-bot: I haz the power, 2014-09-17 12:45:45]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patchset/468793003/60001

[commit-bot: I haz the power, 2014-09-17 13:51:39]

Committed patchset #4 (id:60001) as 182154