feat: add compression metadata flatbuffer schema and tests

Add a flatbuffer schema for describing compressed models.
Flatbuffers with this schema are to be used as the value in a
.tflite model flatbuffer metadata field, and contain the extra
information necessary to describe a compressed model.

Include tests to ensure basic functionality and demonstrate
integration with C++, Python, and Bazel.

BUG=tensorflow#2636

tensorflow/lite/micro/compression/BUILD

@@ -0,0 +1,97 @@
+load(
+    "//tensorflow/lite/micro:build_def.bzl",
+    "tflm_cc_library",
+    "tflm_cc_test",
+)
+load(
+    "@flatbuffers//:build_defs.bzl",
+    "flatbuffer_cc_library",
+    "flatbuffer_py_library",
+)
+load("@rules_python//python:defs.bzl", "py_test")
+load("@tflm_pip_deps//:requirements.bzl", "requirement")
+
+package(
+    default_visibility = [
+        "//visibility:public",
+    ],
+)
+
+flatbuffer_cc_library(
+    # Generates the header-only library "metadata_generated.h", used to read
+    # the metadata flatbuffer.
+    name = "metadata_cc",
+    srcs = ["metadata.fbs"],
+)
+
+tflm_cc_library(
+    # The header-only library generated by flatc in ":metadata_cc" is saved to
+    # the source tree and comitted to git as "metadata_saved.h", which is used
+    # by code which builds via the Make build system, which has no means of
+    # generating the header on the fly. Code which builds via both bazel and
+    # Make should #include the saved header and use this target in its bazel
+    # BUILD deps. Code built exclusively via bazel would typically depend
+    # directly on ":metadata_cc", which would generate a header from the schema
+    # on the fly, during the build.
+    #
+    # When the schema definition "metadata.fbs" is changed, this saved header
+    # should be updated by running the script "./metadata_saved_update.sh",
+    # outside of bazel (because bazel cannot modify the source tree). The
+    # script regenerates the header from the schema and copies it to the source
+    # tree as "metadata_saved.h".
+    #
+    # Comitting the generated file risks inconsistency between the schema and
+    # the saved header, so consistency ensured by the unit test
+    # ":metadata_saved_test".
+    #
+    name = "metadata_saved",
+    hdrs = ["metadata_saved.h"],
+)
+
+sh_test(
+    # Ensures consistency bewteen the schema and the saved generated header.
+    # Fails if they mismatch, in which case, ./metadata_saved_update.sh should
+    # be run. See :metadata_saved above.
+    name = "metadata_saved_test",
+    size = "small",
+    srcs = ["metadata_saved_test.sh"],
+    args = [
+        "$(location metadata_saved.h)",
+        "$(location :metadata_cc_srcs)",
+    ],
+    data = [
+        "metadata_saved.h",
+        ":metadata_cc_srcs",
+    ],
+)
+
+tflm_cc_test(
+    name = "metadata_test_cc",
+    size = "small",
+    srcs = ["metadata_test.cc"],
+    deps = [
+        ":metadata_saved",
+        "//tensorflow/lite/micro:hexdump",
+        "//tensorflow/lite/micro/testing:micro_test",
+        "@flatbuffers//:runtime_cc",
+    ],
+)
+
+flatbuffer_py_library(
+    # Generates the Python module "metadata_py_generated", used to read the
+    # metadata flatbuffer.
+    name = "metadata_py",
+    srcs = ["metadata.fbs"],
+)
+
+py_test(
+    name = "metadata_test_py",
+    size = "small",
+    srcs = ["metadata_test.py"],
+    main = "metadata_test.py",
+    deps = [
+        "metadata_py",
+        "@flatbuffers//:runtime_py",
+        requirement("hexdump"),
+    ],
+)

tensorflow/lite/micro/compression/metadata.fbs

@@ -0,0 +1,56 @@
+// Copyright 2024 The TensorFlow Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+namespace tflite.micro.compression;
+
+table Metadata {
+    // Compression data root, to be used in a tflite.Model.metadata field with
+    // the key "COMPRESSION_METADATA".
+
+    schema_version:int = 1;
+        // ^ Incremented whenever there are backward-incompatible changes. Code
+        // should accept models with versions less than or equal the version
+        // for which the code is built. I.e., code should accept older models,
+        // but not necessarily newer ones.
+
+    subgraphs:[Subgraph];
+        // ^ Compression data indexed by subgraph index.
+}
+
+table Subgraph {
+    // Per-subgraph compression metadata.
+
+    lut_tensors:[LutTensor];
+        // ^ A list of tensors which are compressed using the
+        // (L)ook-(U)p-(T)able method. The indices of this vector are not
+        // significant.
+}
+
+table LutTensor {
+    // Look-Up-Table Tensor: a tensor representation where elements are
+    // compressed into indices into a table of values. The indices are unsigned
+    // integers, index_bitwidth-wide, in big-endian bit order, packed into the
+    // buffer identified by the corresponding tflite.Tensor's buffer field. The
+    // values are located in a newly-created buffer, encoded according to the
+    // tflite.Tensor.type. Tensors with multiple channels have distinct value
+    // tables for each channel, typically along their quantization axis,
+    // concatenated one after another. An element's index must be looked up in
+    // the value table corresponding to its channel.
+
+    tensor:int;            // index of the corresponding tflite.Tensor
+    value_buffer:uint;     // index of the buffer containing LUT values
+    index_bitwidth:uint8;  // bit-width of LUT indexes
+}
+
+root_type Metadata;