Merge pull request #18 from nalbam/main

chore: Update sync-notion and start-ingestion workflows

.github/workflows/5-sync-notion.yml

@@ -0,0 +1,68 @@
+name: sync-notion
+
+on:
+  push:
+    branches:
+      - main
+
+  schedule:
+    - cron: "0 20 * * 0-4"
+
+env:
+  ENABLE_NOTION_SYNC: ${{ vars.ENABLE_NOTION_SYNC }}
+
+  NOTION_PAGE_NAME: "nalbam"
+  NOTION_PAGE_ID: "7aace0412a82431996f61a29225a95ec"
+
+  AWS_DEST_PATH: ${{ vars.AWS_DEST_PATH }}
+
+  KNOWLEDGE_BASE_ID: ${{ vars.KNOWLEDGE_BASE_ID }}
+  DATA_SOURCE_ID: ${{ vars.DATA_SOURCE_ID }}
+
+jobs:
+  sync:
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout 🛎️
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Setup Python 3.9 🐍
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.9
+
+      - name: Setup Dependencies
+        run: pip install python-notion-exporter
+
+      - name: Run Notion Exporter
+        if: env.ENABLE_NOTION_SYNC == 'Yes'
+        env:
+          NOTION_TOKEN_V2: ${{ secrets.NOTION_TOKEN_V2 }}
+          NOTION_FILE_TOKEN: ${{ secrets.NOTION_FILE_TOKEN }}
+        run: |
+          python bin/notion_exporter.py
+
+      - name: Sync to AWS S3 Data Source
+        if: env.ENABLE_NOTION_SYNC == 'Yes'
+        env:
+          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+          AWS_REGION: "us-east-1"
+        run: |
+          aws s3 sync --delete \
+            build/${{ env.NOTION_PAGE_NAME }}/ \
+            ${{ env.AWS_DEST_PATH }}/${{ env.NOTION_PAGE_NAME }}/
+
+      - name: Sync to AWS Bedrock Knowledge Base
+        if: env.ENABLE_NOTION_SYNC == 'Yes'
+        env:
+          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+          AWS_REGION: "us-east-1"
+        run: |
+          aws bedrock-agent start-ingestion-job \
+            --knowledge-base-id ${{ env.KNOWLEDGE_BASE_ID }} \
+            --data-source-id ${{ env.DATA_SOURCE_ID }}

.github/workflows/6-start-ingestion.yml.stop

@@ -9,6 +9,8 @@ on:
     - cron: "0 21 * * 0-4"
 
 env:
+  ENABLE_NOTION_SYNC: ${{ vars.ENABLE_NOTION_SYNC }}
+
   KNOWLEDGE_BASE_ID: ${{ vars.KNOWLEDGE_BASE_ID }}
   DATA_SOURCE_ID: ${{ vars.DATA_SOURCE_ID }}
 
@@ -28,11 +30,12 @@ jobs:
           python-version: 3.9
 
       - name: Sync to AWS Bedrock Knowledge Base
-        run: |
-          aws bedrock-agent start-ingestion-job \
-            --knowledge-base-id ${{ env.KNOWLEDGE_BASE_ID }} \
-            --data-source-id ${{ env.DATA_SOURCE_ID }}
+        if: env.ENABLE_NOTION_SYNC == 'Yes'
         env:
           AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
           AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
           AWS_REGION: "us-east-1"
+        run: |
+          aws bedrock-agent start-ingestion-job \
+            --knowledge-base-id ${{ env.KNOWLEDGE_BASE_ID }} \
+            --data-source-id ${{ env.DATA_SOURCE_ID }}

bin/.env.example

@@ -1,5 +1,5 @@
-NOTION_TOKEN = "v02%3Auser_token_or_cookies"
+NOTION_TOKEN_V2 = "v02%3Auser_token_or_cookies"
 NOTION_FILE_TOKEN = "v02%3Afile_token"
 
-NOTION_PAGE_ID = "0c7c08203a9b4435a4ca07b6454151d7"
-NOTION_PAGE_NAME = "notion"
+NOTION_PAGE_NAME = "nalbam"
+NOTION_PAGE_ID = "7aace0412a82431996f61a29225a95ec"

bin/notion_exporter.py

@@ -3,16 +3,16 @@
 from python_notion_exporter import NotionExporter, ExportType, ViewExportType
 
 
-NOTION_TOKEN = os.getenv("NOTION_TOKEN")
+NOTION_TOKEN_V2 = os.getenv("NOTION_TOKEN_V2")
 NOTION_FILE_TOKEN = os.getenv("NOTION_FILE_TOKEN")
 
-NOTION_PAGE_ID = os.getenv("NOTION_PAGE_ID", "0c7c08203a9b4435a4ca07b6454151d7")
 NOTION_PAGE_NAME = os.getenv("NOTION_PAGE_NAME", "demo")
+NOTION_PAGE_ID = os.getenv("NOTION_PAGE_ID", "7aace0412a82431996f61a29225a95ec")
 
 
 if __name__ == "__main__":
     exporter = NotionExporter(
-        token_v2=NOTION_TOKEN,
+        token_v2=NOTION_TOKEN_V2,
         file_token=NOTION_FILE_TOKEN,
         pages={NOTION_PAGE_NAME: NOTION_PAGE_ID},
         export_directory="build",

serverless.yml

@@ -53,11 +53,19 @@ resources:
         TimeToLiveSpecification:
           AttributeName: expire_at
           Enabled: true
-    # S3Bucket:
-    #   Type: AWS::S3::Bucket
-    #   Properties:
-    #     BucketName:
-    #       Fn::Sub: ${self:provider.environment.BASE_NAME}-${AWS::AccountId}
+        Tags:
+          - Key: Project
+            Value: ${self:provider.environment.BASE_NAME}
+
+    S3Bucket:
+      Type: AWS::S3::Bucket
+      Properties:
+        BucketName:
+          Fn::Sub: ${self:provider.environment.BASE_NAME}-${AWS::AccountId}
+        Tags:
+          - Key: Project
+            Value: ${self:provider.environment.BASE_NAME}
+
     # KnowledgeBase:
     #   Type: AWS::Bedrock::KnowledgeBase
     #   Properties: