From b767a743501c7c68c4d60e5bff551e3ddf5c9439 Mon Sep 17 00:00:00 2001
From: Stuart Axelbrooke <stuart@axelbrooke.com>
Date: Wed, 2 Jul 2025 21:49:15 -0700
Subject: [PATCH] Allow docker packaging of podcast review example

---
 examples/podcast_reviews/BUILD.bazel  | 66 ++++++++++++++++++++-------
 examples/podcast_reviews/MODULE.bazel | 27 ++++++++++-
 2 files changed, 76 insertions(+), 17 deletions(-)

diff --git a/examples/podcast_reviews/BUILD.bazel b/examples/podcast_reviews/BUILD.bazel
index fdce0f0..169c5cd 100644
--- a/examples/podcast_reviews/BUILD.bazel
+++ b/examples/podcast_reviews/BUILD.bazel
@@ -1,7 +1,7 @@
-load("//:py_repl.bzl", "py_repl")
-load("@databuild//databuild:rules.bzl", "databuild_job", "databuild_graph")
-load("@rules_python//python:pip.bzl", "compile_pip_requirements")
+load("@databuild//databuild:rules.bzl", "databuild_graph", "databuild_job")
 load("@pypi//:requirements.bzl", "requirement")
+load("@rules_python//python:pip.bzl", "compile_pip_requirements")
+load("//:py_repl.bzl", "py_repl")
 
 compile_pip_requirements(
     name = "py_requirements",
@@ -9,6 +9,22 @@ compile_pip_requirements(
     requirements_txt = "requirements_lock.txt",
 )
 
+platform(
+    name = "linux_arm",
+    constraint_values = [
+        "@platforms//os:linux",
+        "@platforms//cpu:arm64",
+    ],
+)
+
+platform(
+    name = "linux_x86",
+    constraint_values = [
+        "@platforms//os:linux",
+        "@platforms//cpu:x86_64",
+    ],
+)
+
 # Podcast Reviews Graph
 databuild_graph(
     name = "podcast_reviews_graph",
@@ -39,7 +55,10 @@ databuild_job(
 
 py_binary(
     name = "extract_reviews_binary",
-    srcs = ["extract_reviews_job.py", "duckdb_utils.py"],
+    srcs = [
+        "duckdb_utils.py",
+        "extract_reviews_job.py",
+    ],
     main = "extract_reviews_job.py",
     deps = [
         requirement("duckdb"),
@@ -58,7 +77,10 @@ databuild_job(
 
 py_binary(
     name = "extract_podcasts_binary",
-    srcs = ["extract_podcasts_job.py", "duckdb_utils.py"],
+    srcs = [
+        "duckdb_utils.py",
+        "extract_podcasts_job.py",
+    ],
     main = "extract_podcasts_job.py",
     deps = [
         requirement("duckdb"),
@@ -77,7 +99,10 @@ databuild_job(
 
 py_binary(
     name = "categorize_reviews_binary",
-    srcs = ["categorize_reviews_job.py", "duckdb_utils.py"],
+    srcs = [
+        "categorize_reviews_job.py",
+        "duckdb_utils.py",
+    ],
     main = "categorize_reviews_job.py",
     deps = [
         requirement("duckdb"),
@@ -96,7 +121,10 @@ databuild_job(
 
 py_binary(
     name = "phrase_modeling_binary",
-    srcs = ["phrase_modeling_job.py", "duckdb_utils.py"],
+    srcs = [
+        "duckdb_utils.py",
+        "phrase_modeling_job.py",
+    ],
     main = "phrase_modeling_job.py",
     deps = [
         requirement("duckdb"),
@@ -115,7 +143,10 @@ databuild_job(
 
 py_binary(
     name = "phrase_stats_binary",
-    srcs = ["phrase_stats_job.py", "duckdb_utils.py"],
+    srcs = [
+        "duckdb_utils.py",
+        "phrase_stats_job.py",
+    ],
     main = "phrase_stats_job.py",
     deps = [
         requirement("duckdb"),
@@ -134,7 +165,10 @@ databuild_job(
 
 py_binary(
     name = "daily_summary_binary",
-    srcs = ["daily_summary_job.py", "duckdb_utils.py"],
+    srcs = [
+        "daily_summary_job.py",
+        "duckdb_utils.py",
+    ],
     main = "daily_summary_job.py",
     deps = [
         requirement("duckdb"),
@@ -160,15 +194,15 @@ py_binary(
 py_binary(
     name = "test_jobs",
     srcs = [
-        "test_jobs.py",
-        "extract_reviews_job.py",
-        "extract_podcasts_job.py", 
         "categorize_reviews_job.py",
+        "daily_summary_job.py",
+        "duckdb_utils.py",
+        "extract_podcasts_job.py",
+        "extract_reviews_job.py",
+        "job_lookup.py",
         "phrase_modeling_job.py",
         "phrase_stats_job.py",
-        "daily_summary_job.py",
-        "job_lookup.py",
-        "duckdb_utils.py",
+        "test_jobs.py",
     ],
     main = "test_jobs.py",
     deps = [
@@ -184,5 +218,5 @@ py_repl(
     deps = [
         requirement("duckdb"),
         requirement("pydantic"),
-    ]
+    ],
 )
diff --git a/examples/podcast_reviews/MODULE.bazel b/examples/podcast_reviews/MODULE.bazel
index 69ccd92..8766c25 100644
--- a/examples/podcast_reviews/MODULE.bazel
+++ b/examples/podcast_reviews/MODULE.bazel
@@ -23,5 +23,30 @@ pip.parse(
     python_version = "3.13",
     requirements_lock = "//:requirements_lock.txt",
 )
-
 use_repo(pip, "pypi")
+
+# Rules OCI - necessary for producing a docker container
+bazel_dep(name = "rules_oci", version = "2.2.6")
+# For testing, we also recommend https://registry.bazel.build/modules/container_structure_test
+
+oci = use_extension("@rules_oci//oci:extensions.bzl", "oci")
+
+# Declare external images you need to pull, for example:
+oci.pull(
+    name = "debian",
+    image = "docker.io/library/python",
+    platforms = [
+        "linux/arm64/v8",
+        "linux/amd64",
+    ],
+    # 'latest' is not reproducible, but it's convenient.
+    # During the build we print a WARNING message that includes recommended 'digest' and 'platforms'
+    # values which you can use here in place of 'tag' to pin for reproducibility.
+    tag = "3.12-bookworm",
+)
+
+# For each oci.pull call, repeat the "name" here to expose them as dependencies.
+use_repo(oci, "debian", "debian_linux_amd64", "debian_linux_arm64_v8")
+
+# Platforms for specifying linux/arm
+bazel_dep(name = "platforms", version = "0.0.11")