Linux: fix fake italics for font's without italic variants.

Linux: fix fake italics for font's without italic variants.

Before this patch we assumed that the style which we got from
fontconfig was the one that we asked for. Rather than do this we need
to query the resulting style and plumb that back into WebKit. Once
WebKit knows that there's a mismatch between the request and actual
styles it can trigger faking.

BUG=14810

[Evan Martin, 2009-06-23 03:17:22]

http://codereview.chromium.org/147005/diff/1/4
File skia/ext/SkFontHost_fontconfig_direct.cpp (right):

http://codereview.chromium.org/147005/diff/1/4#newcode71
Line 71: fcvalue.u.i = is_bold ? FC_WEIGHT_BOLD : FC_WEIGHT_NORMAL;
It seems strange to test is_bold here when you've just tested it to get into
this "if" statement's predicate.  Same patterns below.

http://codereview.chromium.org/147005/diff/1/4#newcode185
Line 185: *is_bold = resulting_bold >= FC_WEIGHT_BOLD;
This can be used uninitialized if FcPatternGetInteger fails above.  Same pattern
below with italic.

http://codereview.chromium.org/147005/diff/1/7
File skia/ext/SkFontHost_fontconfig_ipc.cpp (right):

http://codereview.chromium.org/147005/diff/1/7#newcode51
Line 51: request.WriteBool(*is_bold);
Why not just:  request.WriteBool(is_bold && *is_bold);

same below for italics

[agl, 2009-06-23 03:24:57]

http://codereview.chromium.org/147005/diff/1/4
File skia/ext/SkFontHost_fontconfig_direct.cpp (right):

http://codereview.chromium.org/147005/diff/1/4#newcode71
Line 71: fcvalue.u.i = is_bold ? FC_WEIGHT_BOLD : FC_WEIGHT_NORMAL;
On 2009/06/23 03:17:22, Evan Martin wrote:
> It seems strange to test is_bold here when you've just tested it to get into
> this "if" statement's predicate.  Same patterns below.

Good point.

http://codereview.chromium.org/147005/diff/1/4#newcode185
Line 185: *is_bold = resulting_bold >= FC_WEIGHT_BOLD;
On 2009/06/23 03:17:22, Evan Martin wrote:
> This can be used uninitialized if FcPatternGetInteger fails above.  Same
pattern
> below with italic.

Sorry, I don't see it.

http://codereview.chromium.org/147005/diff/1/7
File skia/ext/SkFontHost_fontconfig_ipc.cpp (right):

http://codereview.chromium.org/147005/diff/1/7#newcode51
Line 51: request.WriteBool(*is_bold);
On 2009/06/23 03:17:22, Evan Martin wrote:
> Why not just:  request.WriteBool(is_bold && *is_bold);
> 
> same below for italics

Good point.

[Evan Martin, 2009-06-23 03:32:39]

http://codereview.chromium.org/147005/diff/1/4
File skia/ext/SkFontHost_fontconfig_direct.cpp (right):

http://codereview.chromium.org/147005/diff/1/4#newcode185
Line 185: *is_bold = resulting_bold >= FC_WEIGHT_BOLD;
You're testing resulting_bold, but resulting_bold is only set if the test on
L177 passes.

[Evan Martin, 2009-06-23 03:33:29]

http://codereview.chromium.org/147005/diff/1/4
File skia/ext/SkFontHost_fontconfig_direct.cpp (right):

http://codereview.chromium.org/147005/diff/1/4#newcode185
Line 185: *is_bold = resulting_bold >= FC_WEIGHT_BOLD;
On 2009/06/23 03:32:39, Evan Martin wrote:
> You're testing resulting_bold, but resulting_bold is only set if the test on
> L177 passes.

Durrr, I see, it's the last arg to the function.


Stupid functions returning true on error.

[Evan Martin, 2009-06-23 03:33:54]

LGTM

though I'd love to see a layout test cover thsi