Describe the requested feature:
Create a fuzzer that generates three files: foo.proto, encode_foo.c and
decode_foo.c. These should be structured similar to the alltypes test cases.
The files should contain a wide variety of .proto constructs, from empty
messages to messages with hundreds of fields. The fuzzer should only create
valid .proto files. Any compilation or test error should indicate a bug in
either the fuzzer, nanopb or protoc itself.
The generated files should have a mix of callback/pointer/static fields in the
generated messages, as well as the various nanopb options.
In what situation would the feature be useful:
Many issues are only reproducible with a specific kind of message structure.
The AllTypes test case tests all the field types, but not the combinations of
them. Issues like #142 and #138 could potentially be detected with a fuzzer.
Original issue reported on code.google.com by Petteri.Aimonen on 27 Jan 2015 at 3:39
Original issue reported on code.google.com by
Petteri.Aimonenon 27 Jan 2015 at 3:39