Convert pepper font list to new resource system.

Convert pepper font list to new resource system.

This converts the GetFontFamilies call to the new resource system using a singleton resource.

BUG=

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=174241

[brettw, 2012-12-14 23:44:34]

This should be pretty straightforward. I changed the caching of the font list to
be in the singleton resource rather than a global static. This seems a bit
cleaner to me, although may be slower in some cases (if many instances in the
same process each request the font list, it will be slower).

[raymes, 2012-12-17 00:08:28]

looks good just some minor nits.

Are you trying to move stuff out of ppb_instance_proxy or was this just more
part of the font cleanup.

https://codereview.chromium.org/11578038/diff/2001/content/browser/renderer_h...
File content/browser/renderer_host/pepper/pepper_browser_font_singleton_host.cc
(right):

https://codereview.chromium.org/11578038/diff/2001/content/browser/renderer_h...
content/browser/renderer_host/pepper/pepper_browser_font_singleton_host.cc:93:
return PP_OK_COMPLETIONPENDING;
You can actually change the above few lines to:
context->reply_msg =
PpapiPluginMsg_BrowserFontSingleton_GetFontFamiliesReply(output);
return PP_OK;

https://codereview.chromium.org/11578038/diff/2001/content/browser/renderer_h...
content/browser/renderer_host/pepper/pepper_browser_font_singleton_host.cc:98:
PepperBrowserFontSingletonHost::PepperBrowserFontSingletonHost(
We seem to be having this pattern quite a lot where the host is empty except for
having a message filter in it. I think we can make the code cleaner by having a
generic "MessageFilterHost" which just takes a filter on construction.

This CL is fine though. I'll make the change separately.

https://codereview.chromium.org/11578038/diff/2001/content/browser/renderer_h...
File content/browser/renderer_host/pepper/pepper_browser_font_singleton_host.h
(right):

https://codereview.chromium.org/11578038/diff/2001/content/browser/renderer_h...
content/browser/renderer_host/pepper/pepper_browser_font_singleton_host.h:10:
#include "ppapi/host/resource_message_filter.h"
don't think this is is needed

https://codereview.chromium.org/11578038/diff/2001/content/browser/renderer_h...
content/browser/renderer_host/pepper/pepper_browser_font_singleton_host.h:14:
struct ReplyMessageContext;
not needed (I see that you've just copied this fromthe pepper_flash_file_host.h
which is also wrong. I'll fix it up separately).

https://codereview.chromium.org/11578038/diff/2001/ppapi/proxy/ppapi_messages.h
File ppapi/proxy/ppapi_messages.h (right):

https://codereview.chromium.org/11578038/diff/2001/ppapi/proxy/ppapi_messages...
ppapi/proxy/ppapi_messages.h:1597: // family list.
nit: maybe add a note that it is encoded as a null separated string.

https://codereview.chromium.org/11578038/diff/2001/webkit/plugins/ppapi/ppapi...
File webkit/plugins/ppapi/ppapi_plugin_instance.cc (right):

https://codereview.chromium.org/11578038/diff/2001/webkit/plugins/ppapi/ppapi...
webkit/plugins/ppapi/ppapi_plugin_instance.cc:2147: case
::ppapi::BROKER_SINGLETON_ID:
Forgot to add the in-process case. I guess it isn't needed in-process.

[brettw, 2012-12-17 22:53:42]

new snap up, PTAL

[raymes, 2012-12-17 22:57:42]

lgtm

thanks

[brettw, 2012-12-20 21:16:26]

TBR jschuh for ppapi_messages.h