More context in slack message (#14)

.github/workflows/run.yml

@@ -63,14 +63,29 @@ jobs:
           job_id=$(echo $jobs | jq -r '.jobs[] | select(.runner_name=="${{ runner.name }}") | .id')
           echo "job_id=$job_id" >> $GITHUB_OUTPUT
 
+      - name: query google
+        run: ./scripts/download.sh
+        env:
+          URL: ${{ github.event.inputs.url }}
+          RANGE: ${{ github.event.inputs.range }}
+          ACCESS_TOKEN: ${{ steps.auth.outputs.access_token }}
+
       - name: Notify Slack on Start
         run: ./scripts/slack.sh
         env:
           SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL }}
-          MESSAGE: ":islandora-logo: :hammer_and_wrench: Workbench job started by ${{ github.actor }}. View details: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}/job/${{ steps.get-job-id.outputs.job_id }}#step:12:1"
+          MESSAGE: |
+            ${{ github.actor }} started workbench ingest for __TITLE__
+
+            Items being ingested: __LINE_COUNT__
+
+            Google Sheet: __URL__
+
+            Workbench execution log: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}/job/${{ steps.get-job-id.outputs.job_id }}#step:12:1
+          URL: ${{ github.event.inputs.url }}
 
-      - name: download and transform google sheet
-        run: ./scripts/download-transform.sh
+      - name: transform google sheet
+        run: ./scripts/transform.sh
         env:
           URL: ${{ github.event.inputs.url }}
           RANGE: ${{ github.event.inputs.range }}

scripts/download.sh

@@ -0,0 +1,55 @@
+#!/usr/bin/env bash
+
+set -eou pipefail
+
+regex='^(https:\/\/docs\.google\.com\/spreadsheets\/d\/[a-zA-Z0-9_-]+)'
+if [[ "$URL" =~ $regex ]]; then
+  URL="${BASH_REMATCH[1]}"
+else
+  echo "Invalid URL"
+  exit 1
+fi
+
+# extract sheet ID from https://docs.google.com/spreadsheets/d/foo/edit?gid=0#gid=0
+SHEET_ID=$(echo "$URL" | sed -n 's|.*/d/\(.*\)|\1|p')
+
+attempt=0
+max_attempts=5
+while [[ "$attempt" -lt "$max_attempts" ]]; do
+  GSHEET_STATUS=$(curl -s \
+    -H "Authorization: Bearer $ACCESS_TOKEN" \
+    -o gsheet.json \
+    -w '%{http_code}' \
+    "https://sheets.googleapis.com/v4/spreadsheets/$SHEET_ID/values/$RANGE")
+
+  TITLE_STATUS=$(curl -s \
+    -H "Authorization: Bearer $ACCESS_TOKEN" \
+    -o response.json \
+    -w '%{http_code}' \
+    "https://sheets.googleapis.com/v4/spreadsheets/$SHEET_ID?fields=properties(title)")
+
+  if [ "$GSHEET_STATUS" -lt 500 ] && [ "$TITLE_STATUS" -lt 500 ]; then
+    jq -r .properties.title response.json > title.txt
+    break
+  fi
+
+  delay=$((attempt * 30))
+  sleep "$delay"
+  attempt=$(( attempt + 1))
+done
+
+files=("gsheet.json" "title.txt")
+for file in "${files[@]}"; do
+  if [[ ! -f "$file" ]]; then
+    echo "$file does not exist."
+    exit 1
+  elif [[ ! -s "$file" ]]; then
+    echo "$file is empty."
+    exit 1
+  fi
+  contents=$(cat "$file")
+  if [ "$contents" = "null" ]; then
+    echo "$file is null."
+    exit 1
+  fi
+done

scripts/slack.sh

@@ -2,6 +2,24 @@
 
 set -eou pipefail
 
+if [ -v URL ]; then
+  regex='^(https:\/\/docs\.google\.com\/spreadsheets\/d\/[a-zA-Z0-9_-]+)'
+  if [[ "$URL" =~ $regex ]]; then
+    URL="${BASH_REMATCH[1]}"
+  else
+    echo "Invalid URL"
+    exit 1
+  fi
+
+  TITLE=$(cat title.txt)
+  MESSAGE="${MESSAGE//__TITLE__/$TITLE}"
+  MESSAGE="${MESSAGE//__URL__/$URL\/edit\?gid=0#gid=0}"
+
+  LINE_COUNT=$(jq '.values | length' gsheet.json)
+  LINE_COUNT=$(( LINE_COUNT - 1))
+  MESSAGE="${MESSAGE//__LINE_COUNT__/$LINE_COUNT}"
+fi
+
 escaped_message=$(echo "$MESSAGE" | jq -Rsa .)
 curl -s -o /dev/null -XPOST "$SLACK_WEBHOOK_URL" -d '{
   "text": '"$escaped_message"'

scripts/download-transform.sh → scripts/transform.sh

@@ -2,27 +2,15 @@
 
 set -eou pipefail
 
-regex='^(https:\/\/docs\.google\.com\/spreadsheets\/d\/[a-zA-Z0-9_-]+)'
-if [[ "$URL" =~ $regex ]]; then
-  URL="${BASH_REMATCH[1]}"
-else
-  echo "Invalid URL"
-  exit 1
-fi
-
-# extract sheet ID from https://docs.google.com/spreadsheets/d/foo/edit?gid=0#gid=0
-SHEET_ID=$(echo "$URL" | sed -n 's|.*/d/\(.*\)|\1|p')
-response=$(curl -s \
-    "https://sheets.googleapis.com/v4/spreadsheets/$SHEET_ID/values/$RANGE" \
-    -H "Authorization: Bearer $ACCESS_TOKEN")
+GSHEET=$(cat gsheet.json)
 
-if echo "$response" | jq -e .values >/dev/null; then
+if echo "$GSHEET" | jq -e .values >/dev/null; then
   # save as a CSV
-  echo "$response" | jq .values | jq -r '(map(keys) | add | unique) as $cols | map(. as $row | $cols | map($row[.])) as $rows | $cols, $rows[] | @csv' | tail -n +2 > source.csv
+  echo "$GSHEET" | jq .values | jq -r '(map(keys) | add | unique) as $cols | map(. as $row | $cols | map($row[.])) as $rows | $cols, $rows[] | @csv' | tail -n +2 > source.csv
   # and also as JSON in a format check my work expects
-  echo "$response" | jq -r '.values | map(map(tostring))' > csv.json
+  echo "$GSHEET" | jq -r '.values | map(map(tostring))' > csv.json
 else
-  echo "Failed to fetch data: $(echo "$response" | jq -r '.error.message')"
+  echo "Failed to fetch data: $(echo "$GSHEET" | jq -r '.error.message')"
   exit 1
 fi