From 20fa8247c19cae38ad95e0350b8352245f8d9f03 Mon Sep 17 00:00:00 2001
From: "Evan L. Ray" <elray@umass.edu>
Date: Tue, 10 Dec 2024 15:55:14 -0500
Subject: [PATCH 1/3] first draft workflow for getting nhsn data into s3 bucket

---
 .github/workflows/snapshot-nhsn-data.yml | 43 ++++++++++++++++++++++++
 1 file changed, 43 insertions(+)
 create mode 100644 .github/workflows/snapshot-nhsn-data.yml

diff --git a/.github/workflows/snapshot-nhsn-data.yml b/.github/workflows/snapshot-nhsn-data.yml
new file mode 100644
index 0000000..f76961a
--- /dev/null
+++ b/.github/workflows/snapshot-nhsn-data.yml
@@ -0,0 +1,43 @@
+name: Snapshot NHSN data and upload to S3
+
+on:
+  schedule:
+    - cron: "45 12 * * 3" # every Wednesday at 5:45PM UTC == 12:45PM EST
+  workflow_dispatch:
+
+env:
+  # Reich lab AWS account number
+  AWS_ACCOUNT: 312560106906
+
+jobs:
+  snapshot-nhsn-data:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Set up R 📊
+        uses: r-lib/actions/setup-r@v2
+        with:
+          r-version: 4.4.1
+          install-r: true
+          use-public-rspm: true
+          extra-repositories: 'https://hubverse-org.r-universe.dev'
+
+      - name: install R packages
+        run: |
+          Rscript -e "install.packages('remotes')"
+          Rscript -e "remotes::install_github('Chicago/RSocrata@v1.7.11')"
+      
+      - name: Snapshot NHSN data
+        run: Rscript -e "nhsn_data <- RSocrata::read.socrata('https://data.cdc.gov/resource/mpgq-jmmr.csv'); 
+                         write.csv(nhsn_data, paste0('nhsn-', Sys.Date(), '.csv'))"
+      
+      - name: Install rclone
+        run: |
+          curl https://rclone.org/install.sh | sudo bash
+          rclone version
+      
+      - name: Copy to cloud storage
+        # copy the created file to S3
+        run: |
+          rclone copy ./data/ ":s3,provider=AWS,env_auth:infectious-disease-data/data-raw/influenza-nhsn" \
+            --checksum --verbose --stats-one-line --config=/dev/null
+        shell: bash

From 9557921d9ab16a929dbf8bba25c80e6408858ecf Mon Sep 17 00:00:00 2001
From: "Evan L. Ray" <elray@umass.edu>
Date: Tue, 10 Dec 2024 15:57:46 -0500
Subject: [PATCH 2/3] fix run time

---
 .github/workflows/snapshot-nhsn-data.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/snapshot-nhsn-data.yml b/.github/workflows/snapshot-nhsn-data.yml
index f76961a..b6e5d5c 100644
--- a/.github/workflows/snapshot-nhsn-data.yml
+++ b/.github/workflows/snapshot-nhsn-data.yml
@@ -2,7 +2,7 @@ name: Snapshot NHSN data and upload to S3
 
 on:
   schedule:
-    - cron: "45 12 * * 3" # every Wednesday at 5:45PM UTC == 12:45PM EST
+    - cron: "45 5 * * 3" # every Wednesday at 5:45PM UTC == 12:45PM EST
   workflow_dispatch:
 
 env:

From 16e94409658d6696291851b004007758ee0e4528 Mon Sep 17 00:00:00 2001
From: "Evan L. Ray" <elray@umass.edu>
Date: Tue, 10 Dec 2024 16:24:43 -0500
Subject: [PATCH 3/3] fixed the action

---
 .github/workflows/snapshot-nhsn-data.yml | 31 ++++++++++++++----------
 1 file changed, 18 insertions(+), 13 deletions(-)

diff --git a/.github/workflows/snapshot-nhsn-data.yml b/.github/workflows/snapshot-nhsn-data.yml
index b6e5d5c..0af7a03 100644
--- a/.github/workflows/snapshot-nhsn-data.yml
+++ b/.github/workflows/snapshot-nhsn-data.yml
@@ -26,18 +26,23 @@ jobs:
           Rscript -e "install.packages('remotes')"
           Rscript -e "remotes::install_github('Chicago/RSocrata@v1.7.11')"
       
+      - name: Get file name
+        run: echo "FILE_NAME=nhsn-$(date +'%Y-%m-%d').csv" >> $GITHUB_ENV
+
       - name: Snapshot NHSN data
         run: Rscript -e "nhsn_data <- RSocrata::read.socrata('https://data.cdc.gov/resource/mpgq-jmmr.csv'); 
-                         write.csv(nhsn_data, paste0('nhsn-', Sys.Date(), '.csv'))"
-      
-      - name: Install rclone
-        run: |
-          curl https://rclone.org/install.sh | sudo bash
-          rclone version
-      
-      - name: Copy to cloud storage
-        # copy the created file to S3
-        run: |
-          rclone copy ./data/ ":s3,provider=AWS,env_auth:infectious-disease-data/data-raw/influenza-nhsn" \
-            --checksum --verbose --stats-one-line --config=/dev/null
-        shell: bash
+                         write.csv(nhsn_data, file = $FILE_NAME, row.names = FALSE)"
+        env:
+          FILE_NAME: ${{ env.FILE_NAME }}
+
+    - name: Configure AWS credentials
+      # request credentials to assume the hub's AWS role via OpenID Connect
+      if: env.CLOUD_ENABLED == 'true'
+      uses: aws-actions/configure-aws-credentials@v4
+      with:
+        role-to-assume: arn:aws:iam::${{ env.AWS_ACCOUNT }}:role/iddata-github-action 
+        aws-region: us-east-1
+
+    - name: Copy files to cloud storage
+      run: |
+        aws s3 cp "./$FILE_NAME" "s3://infectious-disease-data/data-raw/influenza-nhsn" --dryrun