refactor: (GH-apache#85) Changes in benchmark after review-3

Signed-off-by: Saurabh Kumar Singh <[email protected]>

Author: singh1203

.github/workflows/benchmark.yml

@@ -19,6 +19,8 @@ name: Benchmarks
 on:
   push:
     branches: [main]
+  pull_request:
+    branches: [main]
     paths:
       - ".github/workflows/benchmark.yml"
       - "ci/scripts/bench_adapt.py"
@@ -35,6 +37,10 @@ jobs:
         arch: ['amd64']
     steps:
       - uses: actions/checkout@v4
+      - name: Install dependencies
+        run: |
+          apt-get update
+          apt-get install -y ca-certificates curl
       - name: Set up Python
         uses: actions/setup-python@v4
         with:
@@ -45,9 +51,10 @@ jobs:
           go-version: ${{ matrix.go }}
           cache: true
           cache-dependency-path: go.sum
+          check-latest: false
       - name: Run Benchmarks
-        run: |
-          ci/scripts/bench.sh --json
+        if: github.event_name != 'push'
+        run: bash ci/scripts/bench.sh . --json
       - name: Upload results
         if: github.event_name == 'push' && github.repository == 'apache/arrow-go' && github.ref_name == 'main'
         env:

ci/scripts/bench.sh

@@ -23,9 +23,25 @@
 
 set -ex
 
-source_dir=${1}
+# Validate input arguments
+if [ -z "$1" ]; then
+  echo "Error: Missing source directory argument"
+  exit 1
+fi
+
+source_dir="$1"
+
+PARQUET_TEST_DATA="${source_dir}/cpp/submodules/parquet-testing/data"
+
+# If parquet_test_data doesn't exist, download it from Apache Arrow GitHub
+if [ ! -d "$PARQUET_TEST_DATA" ]; then
+  echo "parquet_test_data not found. Fetching from Apache Arrow repository..."
+  mkdir -p "$PARQUET_TEST_DATA"
+  curl -L https://github.com/apache/parquet-testing/tree/master/data
+fi
+
+export PARQUET_TEST_DATA
 
-export PARQUET_TEST_DATA=${1}/cpp/submodules/parquet-testing/data
 pushd "${source_dir}"
 
 # lots of benchmarks, they can take a while