Issue #4: temporary lint and test workflow

ai-cfia · Mar 26, 2024 · d9149ab · d9149ab
1 parent 44672ba
commit d9149ab
Show file tree

Hide file tree

Showing 4 changed files with 65 additions and 9 deletions.
diff --git a/.env.template b/.env.template
@@ -42,3 +42,9 @@ DB_NAME=
 # wish to query. This will depend on your database schema 
 # TABLE_NAME=
 # COLUMN=
+
+# ailab-llama-search parameters. Adapt to your needs.
+# Ask your cloud admin for the secrets.
+LLAMAINDEX_DB_EMBED_MODEL_PARAMS='{"model": "text-embedding-ada-002", "deployment_name": "ada", "api_key": "<azure_openai_api_key>", "api_version": "2023-07-01-preview", "azure_endpoint": "<azure_openai_endpoint>"}'
+LLAMAINDEX_DB_VECTOR_STORE_PARAMS='{"database": "llamaindex_db_legacy", "host": "<postgres_host>", "password": "<postgres_password>", "port": "5432", "user": "<postgres_user>", "embed_dim": 1536}'
+LLAMAINDEX_DB_TRANS_PATHS={"id": "node/metadata/id", "chunk_id": "node/metadata/chunk_id", "url": "node/metadata/url", "title": "node/metadata/title", "subtitle": "node/metadata/subtitle", "tokens_count": "node/metadata/tokens_count", "last_updated": "node/metadata/last_updated", "score": "node/metadata/score", "llama_id": "node/id_", "llama_score": "score", "content": "node/text"}
diff --git a/.github/workflows/github-workflows.yaml b/.github/workflows/github-workflows.yaml
@@ -8,7 +8,57 @@ on:
       - synchronize
 
 jobs:
-  #   lint-test: to be determined
+  # temporary
+  list-packages:
+    runs-on: ubuntu-latest
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
+    steps:
+      - uses: actions/checkout@v3
+      - id: set-matrix
+        run: |
+          echo "Finding package directories..."
+          PACKAGES=$(find . -maxdepth 1 -type d -exec test -e '{}/requirements.txt' \; -print | sed 's|^\./||' | jq -R . | jq -cs .)
+          PACKAGES=$(find . -maxdepth 1 -type d ! -path . -exec test -e '{}/requirements.txt' \; -print | sed 's|^\./||' | jq -R . | jq -cs .)
+          echo "matrix=${PACKAGES}" >> $GITHUB_OUTPUT
+  # temporary
+  lint-test:
+    needs: list-packages
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        package: ${{ fromJson(needs.list-packages.outputs.matrix) }}
+    steps:
+      - uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.11.x"
+          cache: "pip"
+
+      - name: Install dependencies for ${{ matrix.package }}
+        run: |
+          if [ -f ${{ matrix.package }}/requirements.txt ]; then
+            python -m pip install --upgrade pip
+            pip install -r ${{ matrix.package }}/requirements.txt
+          fi
+          pip install ruff
+
+      - name: Lint with ruff (selected rules) in ${{ matrix.package }}
+        run: |
+          ruff check --output-format=github --select=E9,F63,F7,F82 --target-version=py311 ${{ matrix.package }}
+          ruff check --output-format=github --target-version=py311 ${{ matrix.package }}
+
+      - name: Test with unittest in ${{ matrix.package }}
+        run: |
+          cd ${{ matrix.package }}
+          python -m unittest discover -s tests
+        env:
+          LLAMAINDEX_DB_EMBED_MODEL_PARAMS: ${{ secrets.LLAMAINDEX_DB_EMBED_MODEL_PARAMS }}
+          LLAMAINDEX_DB_VECTOR_STORE_PARAMS: ${{ secrets.LLAMAINDEX_DB_VECTOR_STORE_PARAMS }}
+          LLAMAINDEX_DB_TRANS_PATHS: ${{ secrets.LLAMAINDEX_DB_TRANS_PATHS }}
 
   markdown-check:
     uses: ai-cfia/github-workflows/.github/workflows/workflow-markdown-check.yml@main

diff --git a/ailab-llama-search/requirements.txt b/ailab-llama-search/requirements.txt
@@ -3,3 +3,4 @@ llama-index-vector-stores-postgres
 llama-index-embeddings-azure-openai
 llama-index-storage-kvstore-postgres
 python-dotenv
+dpath
diff --git a/ailab-llama-search/tests/test_integration.py b/ailab-llama-search/tests/test_integration.py
@@ -1,6 +1,5 @@
 import json
 import os
-import time
 import unittest
 
 from ailab_llama_search import create_index_object, search
@@ -11,22 +10,22 @@ class AilabLlamaSearchIntegrationTests(unittest.TestCase):
 
     def setUp(self):
         load_dotenv()
-        self.embed_model_params = json.loads(os.getenv("EMBED_MODEL_PARAMS"))
-        self.vector_store_params = json.loads(os.getenv("VECTOR_STORE_PARAMS"))
-        self.trans_paths = json.loads(os.getenv("TRANS_PATHS"))
+        self.embed_model_params = json.loads(
+            os.getenv("LLAMAINDEX_DB_EMBED_MODEL_PARAMS")
+        )
+        self.vector_store_params = json.loads(
+            os.getenv("LLAMAINDEX_DB_VECTOR_STORE_PARAMS")
+        )
+        self.trans_paths = json.loads(os.getenv("LLAMAINDEX_DB_TRANS_PATHS"))
         self.search_params = {"similarity_top_k": 5}
         self.index = create_index_object(
             self.embed_model_params, self.vector_store_params
         )
 
     def test_search(self):
         query = "steps and considerations of the sampling procedures for food safety"
-        start_time = time.time()
         results = search(query, self.index, self.search_params, self.trans_paths)
-        end_time = time.time()
-        duration = (end_time - start_time) * 1000
         n = self.search_params["similarity_top_k"]
-        self.assertLess(duration, 2000)
         self.assertEqual(len(results), n)
         for result in results:
             for key in self.trans_paths.keys():