Hello,
I am basically following the approach described under Union Types in
the techniques section of the documentation. I am using extensions. I
have found that when deserialising protobuffers in Java it does not
automatically deserialise the extensions. I have to manually create a
registry (I am
Status: New
Owner: ken...@google.com
Labels: Type-Defect Priority-Medium
New issue 207 by kleptog: python: MergeFrom copies fields not actually
present
http://code.google.com/p/protobuf/issues/detail?id=207
What steps will reproduce the problem?
1. Create the following protocol in test.proto
Status: New
Owner: ken...@google.com
Labels: Type-Defect Priority-Medium
New issue 208 by kleptog: python: constructor does not type check scalar
values
http://code.google.com/p/protobuf/issues/detail?id=208
What steps will reproduce the problem?
1. Create a simple protocol file, such as (tes
Comment #1 on issue 208 by kleptog: python: constructor does not type check
scalar values
http://code.google.com/p/protobuf/issues/detail?id=208
Sorry, that last line should ofcourse be:
etattr(self, field_name, field_value)
--
You received this message because you are subscribed t
Comment #2 on issue 208 by kleptog: python: constructor does not type check
scalar values
http://code.google.com/p/protobuf/issues/detail?id=208
Attached is a patch which fixes the above.
While testing this locally I found that it revealed problems in the test
suite where fields that weren
Comment #1 on issue 207 by kleptog: python: MergeFrom copies fields not
actually present
http://code.google.com/p/protobuf/issues/detail?id=207
Attached is a patch which does the above and fixes the issue for me.
Attachments:
mergefrom.patch 480 bytes
--
You received this message b
The C++ implementation does have a singleton registry, which allows it to
magically parse extensions that are linked into the binary. This behavior
can be somewhat confusing though since it relies on what the linker decides
to do with generated code. The Java implementation uses an object oriented
On Mon, Jul 26, 2010 at 2:09 PM, yy wrote:
>
> DynamicMessage thisthing= DynamicMessage.parseFrom(
> myDescriptorProto.getDescriptorForType(), inputM);
> thisthing.writeTo(System.out);
>
This writes raw binary to System.out. Use
System.out.print(thisthi
Sorry, that's currently the only solution. This isn't easy in C++ or Python
either -- it's an unusual use case.
On Thu, Jul 22, 2010 at 7:48 PM, Chung wrote:
> Hi all,
>
> In the Python API, I'd like to insert a message in the middle of a
> repeated message field. The only way I see for now is
Right, depth-first post-order *is* a (reverse) topological sort, which is
what we're looking for here.
On Thu, Jul 22, 2010 at 8:42 AM, Jason Hsueh wrote:
> The "outbound edge" in the topological sort is from dependency to
> dependent. This provides the guarantee that a file dependency will appe
On Tue, Jul 20, 2010 at 11:10 PM, Johan Philips wrote:
>
>
> On 21 jul, 00:22, Kenton Varda wrote:
> > IMO you should just include a tag field which contains some unique
> > identifier, as described in the docs you cite.
>
> So I just shouldn't have used your extensions mechanism but rather
> add
Consider the case where you define a message Foo in foo.proto, and then you
extend it in bar.proto. When generating code for foo.proto, protoc doesn't
necessarily know anything about bar.proto, so how could it generate anything
based on the extensions defined there? Instead, extensions are genera
Updates:
Status: Fixed
Labels: FixedIn-2.3.1
Comment #2 on issue 207 by ken...@google.com: python: MergeFrom copies
fields not actually present
http://code.google.com/p/protobuf/issues/detail?id=207
Thanks for the very clear bug report and patch.
I ended up fixing this slight
On Mon, Jul 26, 2010 at 7:54 AM, David wrote:
> I don't want to send the whole enchilada on the wire - just the
> descriptor for the message being sent.
>
The "whole enchilada" is needed if your message refers to any other message
types. If not, you can just build a dummy FileDescriptorProto co
Comment #3 on issue 208 by ken...@google.com: python: constructor does not
type check scalar values
http://code.google.com/p/protobuf/issues/detail?id=208
Another great bug report, thanks.
Fixed in revision 336. Testing revealed that the same problem existed for
nested messages, since Mer
Updates:
Status: Fixed
Labels: FixedIn-2.3.1
Comment #4 on issue 208 by ken...@google.com: python: constructor does not
type check scalar values
http://code.google.com/p/protobuf/issues/detail?id=208
(No comment was entered for this change.)
--
You received this message becau
Comment #11 on issue 187 by ken...@google.com: Command-line argument to
override the "optimize_for" option
http://code.google.com/p/protobuf/issues/detail?id=187
Hi all,
Attached is a little protoc plugin written in Python which converts all the
inputs to LITE_RUNTIME -- including renaming
Comment #12 on issue 187 by ken...@google.com: Command-line argument to
override the "optimize_for" option
http://code.google.com/p/protobuf/issues/detail?id=187
s/code code/core code/
s/retains/remains/
*sigh*
--
You received this message because you are subscribed to the Google Groups "P
On 27 jul, 22:22, Kenton Varda wrote:
>
> I agree, but we don't have a good API for querying which extension is set.
> We might fix that someday by introducing language-level support for
> "unions".
Ok. So my struggle to find this API is more or less justified :-)
> > for each defined extensi
19 matches
Mail list logo
