* pull from the new semconv repo

* update jinja2 template for the new structure of convention data

* stable semconv names will appear in OpenTelemetry::SemanticConventions
  while all semconv names (experimental, deprecated, stable, etc) appear
  in OpenTelemetry::SemanticCandidates
    - This maps to what other langs (Java, Python, JS) are putting into
      "Incubating" modules. We opted for "candidates" because
      OpenTelemetry::SemanticConventions::Incubating::... was getting
      too long and OpenTelemetry::SemanticIncubating didn't make much
      sense.

* generate a module per root semconv namespace to make navigating docs
  more focused and to decrease the number of constants loaded when
  using more precise requires in downstream code

* generates only attribute name constants and to an attributes subfolder
  to prepare for metric name constants to be generated to metrics
  subfolders, also for decreasing the scope of what gets loaded when
  downstream instrumentation requires these constants

Co-authored-by: Kayla Reopelle (she/her) <87386821+kaylareopelle@users.noreply.github.com>

Because converting to SCREAMING_SNAKE_CASE cannot distinguish
between ...

    screaming.snake_case

... and ...

    screaming.snake.case

So, like other languages, ignore the deprecated messaging.client_id in
favor of its replacement and let backends deal with the changed attr
name.

    277 runs, 500 assertions, 26 failures, 0 errors, 0 skips

Apparently there are some kinks to work out.

There's naming hijinks in the db.* namespace I would like to avoid in
this first effort to bring semconv up-to-recent-date.

Explain the auto-generated situation.

The v1.10 names whose comments were updated in this commit were renamed
or removed in version 1.11-1.25. Made note of their replacements or that
their use should be dropped.

Removed test for one-to-one matching of 1.10 constants to 1.25 "all"
constants. That will never pass without updating the source semconv YAML
to add back the missing names with dep notices. I believe our dep notes
will suffice and we won't auto-generate Resource or Trace again.

* Move to a tmp/ dir and ignore that.

* Update the git commands to be able to reuse one directory for
  the repo working copy, just update what's needed for a different
  tag between iterations. (Generally only an improvement for dev envs.)

* Use sh() for shell commands so that tasks fail if their commands fail.

* Appease Rubocop.

Co-authored-by: Kayla Reopelle <87386821+kaylareopelle@users.noreply.github.com>

Yard's default use of the rdoc markdown provider was choking on links
containing codeblocks, for example: [`ENV`](link/to/somewhere). These
style links are pretty popular in the text details provided with semconv
attributes.

Let's switch to kramdown which is a markdown parser and
renderer that:

* can handle inline code blocks inside a link anchor
* is pure Ruby (no C extensions) so works with JRuby

* out with Candidates, we're using Incubating like most other SIGs
* shorten the namespace to SemConv
* do not provide an all-in-one rollup require for all namespaces;
  downstream users must directly require the namespaces that they use
* attributes that are templates (prefixes, really, for the moment) are
  generated as lambdas with _LAMBDA appended to their name; users must
  call that lambda with a key for the template to return the attribute
  name string

With multiple template entries in weaver.yaml, code generation for
stable and incubating can be performed in one invocation of the docker
image. Rakefile targets trimmed down in light of that.

Each root namespace gets separate attributes and metrics files.

Include examples of possible values for attributes.

Include a usage example of the template attributes as lambdas.

Fix links in doc comments referencing stable constants.

Co-authored-by: Hannah Ramadan <hannahramadan@users.noreply.github.com>
Co-authored-by: Kayla Reopelle <kaylareopelle@users.noreply.github.com>