[PYTHON-4160] Add LlamaIndex (#9)

* PYTHON-4160 Skeleton

* [PYTHON-4160] Add required env vars in run.sh

* CHanged display name of LlamaIndex build variant

* Typo in MONGO_URI

* Removed .idea directory mistakenly added.

Author: caseyclements

.evergreen/config.yml

@@ -94,6 +94,11 @@ tasks:
       - func: "fetch repo"
       - func: "execute tests"
 
+  - name: test-llama-index
+    commands:
+      - func: "fetch repo"
+      - func: "execute tests"
+
 buildvariants:
   - name: test-semantic-kernel-python-rhel
     display_name: Semantic-Kernel RHEL Python
@@ -142,3 +147,16 @@ buildvariants:
       - rhel87-small
     tasks:
       - name: test-chatgpt-retrieval-plugin
+
+  - name: test-llama-index-rhel
+    display_name: LlamaIndex RHEL
+    expansions:
+      DIR: llama_index
+      REPO_NAME: llama_index
+      # TODO - Update CLONE_URL once pull-request is merged
+      CLONE_URL: -b feature/mongodb-datastore --single-branch https://github.com/caseyclements/llama_index.git
+      DATABASE: llama_index_test_db
+    run_on:
+      - rhel87-small
+    tasks:
+      - name: test-llama-index

llama_index/indexConfig.json

@@ -0,0 +1,14 @@
+{
+  "fields": [
+    {
+      "numDimensions": 1536,
+      "path": "embedding",
+      "similarity": "cosine",
+      "type": "vector"
+    }
+  ],
+  "name": "vector_index",
+  "type": "vectorSearch",
+  "database": "llama_index_test_db",
+  "collectionName": "llama_index_test_vectorstore"
+}

llama_index/run.sh

@@ -0,0 +1,30 @@
+#!/bin/sh
+
+set -x
+
+. $workdir/src/.evergreen/utils.sh
+
+PYTHON_BINARY=$(find_python3)
+$PYTHON_BINARY -c "import sys; print(f'Python version found: {sys.version_info}')"
+
+# cd to the MongoDB integration. It has its own project
+cd llama-index-integrations/vector_stores/llama-index-vector-stores-mongodb
+
+# Install Poetry into base python
+$PYTHON_BINARY -m pip install -U pip poetry
+# Create a package specific poetry environment
+$PYTHON_BINARY -m poetry env use $PYTHON_BINARY
+# Activate the poetry env, which itself does not include poetry
+source $($PYTHON_BINARY -m poetry env info --path)/bin/activate
+# Recreate the poetry lock file
+$PYTHON_BINARY -m poetry lock --no-update
+# Install from pyproject.toml into package specific environment
+$PYTHON_BINARY -m poetry install --with dev
+
+# Run tests. Sensitive variables in Evergreen come from Evergreen project: ai-ml-pipeline-testing/
+OPENAI_API_KEY=$openai_api_key \
+MONGO_URI=$llama_index_mongodb_uri \
+MONGODB_DATABASE="llama_index_test_db" \
+MONGODB_COLLECTION="llama_index_test_vectorstore" \
+MONGODB_INDEX="vector_index" \
+$PYTHON_BINARY -m poetry run pytest -v tests