be properly set. writeTo() may be invoked without a call to
getSerializedSize(), so the generated serialization methods would
write a length of 0 for non-empty packed fields. Just call
getSerializedSize() at the beginning of writeTo(): although this
means that we may compute the byte size needlessly when there
are no packed fields, in practice, getSerializedSize() will
already have been called - all of the writeTo() wrappers in
AbstractMessageLite invoke it.
Tested: new unittest case in WireFormatTest.java now passes
This should make it both easier to use and easier to
optimize, in exchange for a small amount of generality.
In practice, any remotely normal case is still very
natural.
The cost is that a upb_msg will now always have an overhead
of 2*sizeof(void*). This is comparable to proto2 overhead.
The benefit is that upb_msg is now self-describing, and
read-only algorithms can now operate on a upb_msg regardless
of the memory-management scheme.
Also, upb_array and upb_string now know inherently if they
own their associated memory, and upb_array has a generic
pointer for memory management purposes like upb_msg does.
There is significant refactoring here, as well as some more trivial
name changes. upb_msg has become upb_msgdef, to reflect the fact
that a upb_msg is not *itself* a message, it describes a message.
There are other renamings, such as upb_parse_state -> upb_stream_parser.
More significantly, the upb_msg class and parser have been refactored
to reflect my recent realization about how memory management should
work. upb_msg now has no memory management, and a memory mangement
scheme (that works beautifully with multiple language runtimes) will
be layered on top of it.
This iteration has the new, read-only upb_msg. upb_mm_msg (a
memory-managed message class) will come in the next change.
jasonh+personal@google.com
Joshua Haberman
Dan Egnor
Jon Skeet
kenton@google.com
temporal