Handle an extension printer supporting only PWG raster data

chrome/browser/ui/webui/print_preview/printer_handler.h

Index: chrome/browser/ui/webui/print_preview/printer_handler.h
diff --git a/chrome/browser/ui/webui/print_preview/printer_handler.h b/chrome/browser/ui/webui/print_preview/printer_handler.h
index 92a9bf81bc6e60b2f12d0420a2fbcb57cced3cec..44188bbce46dccb872476bb5bd258c73449c0d1f 100644
--- a/chrome/browser/ui/webui/print_preview/printer_handler.h
+++ b/chrome/browser/ui/webui/print_preview/printer_handler.h
@@ -21,6 +21,10 @@ namespace content {
 class BrowserContext;
 }
 
+namespace gfx {
+class Size;
+}
+
 // Wrapper around PrinterProviderAPI to be used by print preview.
 // It makes request lifetime management easier, and hides details of more
 // complex operations like printing from the print preview handler.
@@ -64,6 +68,7 @@ class PrinterHandler {
       const std::string& destination_id,
       const std::string& capability,
       const std::string& ticket_json,
+      const gfx::Size& size,

[Vitaly Buka (NO REVIEWS), 2015/02/17 19:43:54]

size is a property of document, ideally we need to get it from there.
it's even invalid with multi-sized pdfs. Can you drop some TODO about that?

maybe TODO(): get source size from print_data.

[tbarzic, 2015/02/18 02:55:49]

So, basically, the page size should be extracted from document data when
processing a page in PDFiumEngineExports::RenderPDFPageToBitmap (and then used
to calculate the bitmap size), right?

Added TODO, and also renamed it to |page_size|, to be less vague.

       const scoped_refptr<base::RefCountedMemory>& print_data,
       const PrintCallback& callback) = 0;
 };