GitLab

as suggested in review.
also updated getRRSIGWithRdataCount() to generate the data from write,
not from text as I found it took much longer time if I did it from text
(which is simpler).  with large number of RRSIGs it's noticeable and not
good for unittests to run frequently.

it was indeed not expected to be copyable.  making this explicit would help
prevent accidental misuse.
as suggested in review.

this will help prevent accidental direct misuse of these variables

(this makes me realize we needed to pass RRClass to destroy().  so updated
the interface).

and, since it was the only purpose of the 'size' param of the constructor,
it's now removed.
test cases are adjusted accordingly.

some parameters are currently hardcoded; we need a revised version
of the reader implementation.

- need an extra set of parentheses to use a substitution for a boolean var
  as a boolean expression
- specify a default param for a templated method in the declaration, not
  in the implementation

(I needed these fixes for #2097)

Testing rewinding it and accessing the stuff out of the usual order
(intermixing iteration of Rdata and the signatures).

The RdataReader tried to find the first stored RRSig by substracting
from the end of the data. This didn't work, partly because people can
pass more data than needed and partly because the tests failed for some
reason.

Now it finds them when it iterates through the normal data.

The tests don't pass, for some reason, but it seems there are less
failures than before. There would probably be some off-by-something
error somewhere.

Just confusion of names

The testing support of the encoder was moved to the tests, so no need
for the definitions in header file.

The Result, empty callbacks, etc.

Provide three more decoders of the data, using the RdataReader. This
does not link, as the reader has no bodies of the methods.

The foreach functions to manually iterate over the data are used in
tests only, so moved to the tests.

The specifications are used by the tests and encoding and will be used
by the decoding machinery too, probably, so we are making them
accessible and reside in a separate file.

Discovered the hard way - it can crash.

Small trick with just-in-time initialization of a pointer in benchmark
is used instead of copying it and relying on the fact that it can be
copied while it was not yet used.

This will help checking the various supported interfaces to decoding.

This is logically part of the next commit. But then, the git diff would
show the old file disappear and the new file appear with new content,
which would be hard to review, so it is split. This commit should
contain no changes in the content (hopefully), so can be skipped during
review.

When changing the names, this kind of got forgotten.

Conflicts:
	tests/tools/perfdhcp/Makefile.am

Including the expected usage scenarios.

Also, the interface was slightly refined - renaming of things, adding
support for the RRSig iteration.

The oposite of the isc::datasrc::memory::RDataEncoder.

Only the methods are drafted, no documentation yet and no member
variables listed.

The previous version was slightly misleading, the reader could
understand it the queries would be just slow as answered from the DB,
instead of not answered at all.