add nbtest job to the GitHub Actions CI

.github/workflows/tests.yml

@@ -0,0 +1,25 @@
+name: notebook-tests
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+jobs:
+  tests:
+    runs-on: ubuntu-latest
+    services:
+      elasticsearch:
+        image: docker.elastic.co/elasticsearch/elasticsearch:8.12.0
+        env:
+          discovery.type: single-node
+          xpack.security.enabled: false
+          xpack.security.http.ssl.enabled: false
+          xpack.license.self_generated.type: trial
+        ports:
+          - 9200:9200
+    steps:
+      - uses: actions/checkout@v4
+      - run: sleep 30
+      - run: PYTHONPATH=. MOCK_ES=1 make test

bin/mocks.py

@@ -0,0 +1,18 @@
+# This module is provided to nbtest when running under GitHub Actions CI to
+# mock the Elasticsearch client to always connect on localhost, regardless of
+# arguments provided
+
+import elasticsearch
+
+orig_es_init = elasticsearch.Elasticsearch.__init__
+
+
+def patched_es_init(self, *args, **kwargs):
+    if 'cloud_id' in kwargs:
+        assert kwargs['cloud_id'] == 'foo'
+    if 'api_key' in kwargs:
+        assert kwargs['api_key'] == 'bar'
+    return orig_es_init(self, 'http://localhost:9200')
+
+
+elasticsearch.Elasticsearch.__init__ = patched_es_init

bin/nbtest

@@ -4,5 +4,16 @@ SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
 if [[ ! -f $SCRIPT_DIR/../.venv/bin/nbtest ]]; then
     make nbtest
 fi
+
+NBTEST="nbtest"
+if [[ "$MOCK_ES" != "" ]]; then
+    export ELASTIC_CLOUD_ID=foo
+    export ELASTIC_API_KEY=bar
+    NBTEST="$NBTEST --mocks bin.mocks"
+
+    # ensure elasticsearch is installed so that it can be mocked
+    $SCRIPT_DIR/../.venv/bin/pip install elasticsearch
+fi
+
 source $SCRIPT_DIR/../.venv/bin/activate
-$SCRIPT_DIR/../.venv/bin/nbtest $*
+$SCRIPT_DIR/../.venv/bin/$NBTEST $*

notebooks/integrations/hugging-face/.nbtest.yml

@@ -0,0 +1,2 @@
+masks:
+- 'Score: [0-9]+\.[0-9][0-9]*'