Deprecate robotlocomotion_lcmtypes and bot2_core_lcmtypes

[jwnimmer-tri]

Port existing Drake code to use the new types.

Add helper header drake/systems/sensors/robotlocomotion_compat.h to alias the old names onto the new names, along with a deprecation warning.

Requires #14885 first. Towards #14362.

---

This change is 

[jwnimmer-tri]

\CC @RussTedrake on inbound change coming next week -- robotlocomotion.image_t and robotlocomotion.image_array_t will be renamed to drake.lcmt_image and drake.lcmt_image_array, alongside all of our usual Drake LCM message types. The manipulation_station uses these messages for its cameras.

The deprecation affordances here were enough to pass all Anzu CI tests, and I didn't immediately see anything in the underactuated or manipulation repositories that used these messages. But possibly you'd like to check this over.

@drake-jenkins-bot linux-bionic-unprovisioned-gcc-bazel-experimental-snopt-packaging please in case you need binaries to test on.

[ggould-tri]

with one small and impractical misgiving below.

Reviewed 32 of 32 files at r1.
Reviewable status: 1 unresolved discussion, needs platform reviewer assigned, needs at least two assigned reviewers, labeled "do not merge", commits need curation (https://drake.mit.edu/reviewable.html#curated-commits) (waiting on @jwnimmer-tri)

---

lcmtypes/BUILD.bazel, line 74 at r1 (raw file):

drake_lcm_cc_library(
    name = "header",
    lcm_package = "drake",

minor: This seems to me to violate the spirit of the namespace/directory rule ("The code in that namespace should usually be in a directory whose basename matches the namespace name") but not the letter of it ("... or in a subdirectory thereof" which would seem to argue that the bare drake namespace is always appropriate).

This is probably too much of a pain to deal with anyway. Author's discretion.

[jwnimmer-tri]

Reviewable status: needs platform reviewer assigned, needs at least two assigned reviewers, labeled "do not merge" (waiting on @jwnimmer-tri)

---

lcmtypes/BUILD.bazel, line 74 at r1 (raw file):

Previously, ggould-tri wrote…

minor: This seems to me to violate the spirit of the namespace/directory rule ("The code in that namespace should usually be in a directory whose basename matches the namespace name") but not the letter of it ("... or in a subdirectory thereof" which would seem to argue that the bare drake namespace is always appropriate).

This is probably too much of a pain to deal with anyway. Author's discretion.

OK Fixing the packaging and include paths is part of the #14362 scope already, but will happen later.

[jwnimmer-tri]

Tagging +@ggould-tri officially for feature review.

+@EricCousineau-TRI for platform review per schedule, please.

[ggould-tri]

Reviewable status: LGTM missing from assignee EricCousineau-TRI(platform) (waiting on @EricCousineau-TRI)

---

lcmtypes/BUILD.bazel, line 74 at r1 (raw file):

Previously, jwnimmer-tri (Jeremy Nimmer) wrote…

OK Fixing the packaging and include paths is part of the #14362 scope already, but will happen later.

👍

[EricCousineau-TRI]

with two minor nits.

Reviewed 26 of 32 files at r1.
Reviewable status: 2 unresolved discussions (waiting on @jwnimmer-tri)

---

systems/sensors/BUILD.bazel, line 120 at r2 (raw file):

# TODO(jwnimmer-tri) This header assists with the deprecation of the
# @lcmtypes_robotlocomotion repository; remove it on or after 2021-08-01.
drake_cc_library(

nit This is presently public (per default visibility).

Can you make this private?

---

systems/sensors/robotlocomotion_compat.h, line 20 at r2 (raw file):

namespace robotlocomotion {

using header_t DRAKE_DEPRECATED("2021-08-01",

nit From brief inspection, not sure if I see these aliases tested.

Consider adding minor unittest, e.g. static_assert(std::is_same_v<robotlocomotion::header_t, drake::lcmt_header>, ""), etc.?

[jwnimmer-tri]

Reviewable status: 1 unresolved discussion (waiting on @EricCousineau-TRI and @ggould-tri)

---

systems/sensors/BUILD.bazel, line 120 at r2 (raw file):

Previously, EricCousineau-TRI (Eric Cousineau) wrote…

nit This is presently public (per default visibility).

Can you make this private?

I think it should be public. Anyone who wants to ease their transition to the new spellings would want to #include this?

---

systems/sensors/robotlocomotion_compat.h, line 20 at r2 (raw file):

Previously, EricCousineau-TRI (Eric Cousineau) wrote…

nit From brief inspection, not sure if I see these aliases tested.

Consider adding minor unittest, e.g. static_assert(std::is_same_v<robotlocomotion::header_t, drake::lcmt_header>, ""), etc.?

Done.

That particular test is useless -- the compiler already catches those kind of errors. Instead, I forked one of the unit tests from prior to this PR, checking that it still passes (with the only change being that the #include "robotlocomotion/..." line has to be removed.

[EricCousineau-TRI]

Reviewed 1 of 32 files at r1, 2 of 2 files at r3.
Reviewable status: complete! all discussions resolved, LGTM from assignees ggould-tri(platform),EricCousineau-TRI(platform) (waiting on @jwnimmer-tri)

---

systems/sensors/BUILD.bazel, line 120 at r2 (raw file):

Previously, jwnimmer-tri (Jeremy Nimmer) wrote…

I think it should be public. Anyone who wants to ease their transition to the new spellings would want to #include this?

OK SGTM!

FWIW Would have been clearer w/ explicit visibility tag and comment but meh for now.

---

systems/sensors/robotlocomotion_compat.h, line 20 at r2 (raw file):

Previously, jwnimmer-tri (Jeremy Nimmer) wrote…

Done.

That particular test is useless -- the compiler already catches those kind of errors. Instead, I forked one of the unit tests from prior to this PR, checking that it still passes (with the only change being that the #include "robotlocomotion/..." line has to be removed.

OK Works for me! (not 100% on board with "useless", but I'm like 70% on board w/ it)

[jwnimmer-tri]

Reviewable status: complete! all discussions resolved, LGTM from assignees ggould-tri(platform),EricCousineau-TRI(platform) (waiting on @jwnimmer-tri)

---

systems/sensors/robotlocomotion_compat.h, line 20 at r2 (raw file):

Previously, EricCousineau-TRI (Eric Cousineau) wrote…

OK Works for me! (not 100% on board with "useless", but I'm like 70% on board w/ it)

I could not imagine an implementation error that such a test would find, but would otherwise go undetected.

[EricCousineau-TRI]

---

systems/sensors/robotlocomotion_compat.h, line 20 at r2 (raw file):

Previously, jwnimmer-tri (Jeremy Nimmer) wrote…

I could not imagine an implementation error that such a test would find, but would otherwise go undetected.

Could ensure that it's an alias to the right type (non-error typo).
True, should be caught via code review, but add'l active check.