From 1d0f13d727392f6c6aa82cf5c9ee8a9da66ec79e Mon Sep 17 00:00:00 2001
From: Justin McCarthy <justinamccarthy@gmail.com>
Date: Tue, 12 Mar 2024 15:11:22 -0700
Subject: [PATCH] integration_tests: automate corpus tests

---
 .github/workflows/corpus.yml     | 25 +++++++++++++++++++++++++
 integration_tests/.gitignore     |  1 +
 integration_tests/corpus_test.go | 20 ++++++++++----------
 3 files changed, 36 insertions(+), 10 deletions(-)
 create mode 100644 .github/workflows/corpus.yml
 create mode 100644 integration_tests/.gitignore

diff --git a/.github/workflows/corpus.yml b/.github/workflows/corpus.yml
new file mode 100644
index 00000000..7c3f259e
--- /dev/null
+++ b/.github/workflows/corpus.yml
@@ -0,0 +1,25 @@
+name: Nightly Test Corpus
+on:
+  schedule:
+    - cron: "0 0 * * *"
+  push:
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Go
+        uses: actions/setup-go@v4
+        with:
+          go-version: "1.22"
+
+      - name: Download
+        run: curl -L -o integration_tests/corpus-tests.tar.gz https://raw.githubusercontent.com/cedar-policy/cedar-integration-tests/main/corpus-tests.tar.gz
+
+      - name: Extract
+        run: mkdir -p integration_tests/tmp && tar -xzf integration_tests/corpus-tests.tar.gz -C integration_tests/tmp
+
+      - name: Corpus Tests
+        run: go test -tags corpus ./integration_tests/...
diff --git a/integration_tests/.gitignore b/integration_tests/.gitignore
new file mode 100644
index 00000000..a9a5aecf
--- /dev/null
+++ b/integration_tests/.gitignore
@@ -0,0 +1 @@
+tmp
diff --git a/integration_tests/corpus_test.go b/integration_tests/corpus_test.go
index b7602707..4a024f2f 100644
--- a/integration_tests/corpus_test.go
+++ b/integration_tests/corpus_test.go
@@ -1,3 +1,5 @@
+//go:build corpus
+
 package integration_tests
 
 import (
@@ -12,13 +14,7 @@ import (
 	"golang.org/x/exp/slices"
 )
 
-// TODO: a way to download the corpus tests and test them
-// for now, since this folder is 235MB.
-//
-// Remove the x before the next line if you have the corpus locally.
-// And remove t.Skip() from the test itself.
-//
-// x go:embed tmp/*
+//go:embed tmp/**/*.json tmp/**/*.cedar
 var integrationFS embed.FS
 
 type corpusTest struct {
@@ -40,11 +36,9 @@ type corpusTest struct {
 
 func TestCorpus(t *testing.T) {
 	t.Parallel()
-	t.Skip()
 	var tests []string
 	for _, p := range []string{
-		"tmp/abac-B/*.json",
-		"tmp/abac-type-directed-B/*.json",
+		"tmp/corpus-tests/*.json",
 	} {
 		more, err := fs.Glob(integrationFS, p)
 		if err != nil || len(more) == 0 {
@@ -79,6 +73,12 @@ func TestCorpus(t *testing.T) {
 	prefix := func(v string) string {
 		return "tmp/" + v
 	}
+
+	// detect possible corpus data pipeline failure
+	if len(tests) < 4_000 {
+		t.Fatalf("corpus test count too low: %v", len(tests))
+	}
+
 	for _, tn := range tests {
 		tn := tn
 		t.Run(tn, func(t *testing.T) {