Add 'cbcs' encryption scheme support in Android media.

media/base/encryption_scheme.h

Index: media/base/encryption_scheme.h
diff --git a/media/base/encryption_scheme.h b/media/base/encryption_scheme.h
index 37bea6710d264390a5b74cd920e10951ff34eba7..901a2940c43f94b3d7235da41264828fab3c13ac 100644
--- a/media/base/encryption_scheme.h
+++ b/media/base/encryption_scheme.h
@@ -42,14 +42,14 @@ class MEDIA_EXPORT EncryptionScheme {
 
     bool Matches(const Pattern& other) const;
 
-    uint32_t encrypt_blocks() const { return encrypt_blocks_; }
-    uint32_t skip_blocks() const { return skip_blocks_; }
+    uint8_t encrypt_blocks() const { return encrypt_blocks_; }
+    uint8_t skip_blocks() const { return skip_blocks_; }
 
     bool IsInEffect() const;
 
    private:
-    uint32_t encrypt_blocks_ = 0;
-    uint32_t skip_blocks_ = 0;
+    uint8_t encrypt_blocks_ = 0;
+    uint8_t skip_blocks_ = 0;

[xhwang, 2016/10/20 23:52:07]

why do we make this change? Does it make sense to just use a simple type like
"int"?

Also, in general we don't use unsigned integers only because they can't be
negative:
https://google.github.io/styleguide/cppguide.html#Integer_Types

[dougsteed, 2016/10/21 17:15:05]

I am reverting this for this CL. The uint32_t originated from an error in my
reading of the spec. These numbers are allocated 4 bits (0-15) in the standard
whereas I had originally read that as 4 bytes. When I realized this error I
changed it to uint8_t (inconsistently). In any case I agree that there is
probably no reason not to just use int. But I think this probably for a separate
CL, so will leave it unchanged for now.

 
     // Allow copy and assignment.
   };