From 26e92d087a888fe78430a404f44ff68a4cbaee06 Mon Sep 17 00:00:00 2001
From: Nathan Park <parknate@amazon.com>
Date: Fri, 18 Jul 2025 13:37:02 -0700
Subject: [PATCH 1/3] feature: s3 bucket path checking

---
 .githooks/pre-commit   | 31 ++++++++++++++++++++++++++
 s3_bucket_validator.py | 49 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 80 insertions(+)
 create mode 100644 .githooks/pre-commit
 create mode 100644 s3_bucket_validator.py

diff --git a/.githooks/pre-commit b/.githooks/pre-commit
new file mode 100644
index 0000000000..159b85a818
--- /dev/null
+++ b/.githooks/pre-commit
@@ -0,0 +1,31 @@
+#!/bin/sh
+# This pre-commit hook validates S3 bucket references in modified files
+
+set -e
+
+# Get list of staged files
+staged_files=$(git diff --cached --name-only --diff-filter=ACMR | grep -E '\.(py|ipynb|md|rst|yaml|yml|json)$' || true)
+
+if [ -z "$staged_files" ]; then
+    echo "No relevant files to check for S3 bucket references."
+    exit 0
+fi
+
+echo "Checking S3 bucket references in staged files..."
+has_invalid_buckets=0
+
+for file in $staged_files; do
+    echo "Validating S3 references in $file"
+    python s3_bucket_validator.py "$file"
+    if [ $? -ne 0 ]; then
+        has_invalid_buckets=1
+    fi
+done
+
+if [ $has_invalid_buckets -ne 0 ]; then
+    echo "ERROR: Invalid S3 bucket references found. Please fix them before committing."
+    exit 1
+fi
+
+echo "S3 bucket validation passed."
+exit 0
diff --git a/s3_bucket_validator.py b/s3_bucket_validator.py
new file mode 100644
index 0000000000..f73b8c3871
--- /dev/null
+++ b/s3_bucket_validator.py
@@ -0,0 +1,49 @@
+import re
+
+import boto3
+from botocore.exceptions import ClientError
+
+
+def is_bucket_accessible(bucket_name):
+    s3 = boto3.client('s3')
+    try:
+        s3.head_bucket(Bucket=bucket_name)
+        return True
+    except ClientError as e:
+        error_code = int(e.response['Error']['Code'])
+        if error_code == 403:
+            print(f"Bucket {bucket_name} exists, but you don't have permission to access it.")
+        elif error_code == 404:
+            print(f"Bucket {bucket_name} does not exist.")
+        else:
+            print(f"Error checking bucket {bucket_name}: {e}")
+        return False
+
+def validate_s3_references(file_path):
+    with open(file_path, 'r') as file:
+        content = file.read()
+
+    s3_pattern = re.compile(r's3:\/\/([a-zA-Z0-9._-]+)')
+    matches = s3_pattern.findall(content)
+
+    invalid_buckets = []
+    for bucket in matches:
+        if not is_bucket_accessible(bucket):
+            invalid_buckets.append(bucket)
+
+    return invalid_buckets
+
+if __name__ == "__main__":
+    import sys
+    if len(sys.argv) < 2:
+        print("Usage: python s3_bucket_validator.py <file_path>")
+        sys.exit(1)
+
+    file_path = sys.argv[1]
+    invalid_buckets = validate_s3_references(file_path)
+
+    if invalid_buckets:
+        print(f"Invalid or inaccessible S3 buckets found: {', '.join(invalid_buckets)}")
+        sys.exit(1)
+    else:
+        print("All referenced S3 buckets are valid and accessible.")
\ No newline at end of file

From ee438e4f049037caf57982fe2baa63191d8dfe5c Mon Sep 17 00:00:00 2001
From: Nathan Park <parknate@amazon.com>
Date: Fri, 18 Jul 2025 14:18:16 -0700
Subject: [PATCH 2/3] format: run black

---
 s3_bucket_validator.py | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/s3_bucket_validator.py b/s3_bucket_validator.py
index f73b8c3871..1dfa271c57 100644
--- a/s3_bucket_validator.py
+++ b/s3_bucket_validator.py
@@ -5,12 +5,12 @@
 
 
 def is_bucket_accessible(bucket_name):
-    s3 = boto3.client('s3')
+    s3 = boto3.client("s3")
     try:
         s3.head_bucket(Bucket=bucket_name)
         return True
     except ClientError as e:
-        error_code = int(e.response['Error']['Code'])
+        error_code = int(e.response["Error"]["Code"])
         if error_code == 403:
             print(f"Bucket {bucket_name} exists, but you don't have permission to access it.")
         elif error_code == 404:
@@ -19,11 +19,12 @@ def is_bucket_accessible(bucket_name):
             print(f"Error checking bucket {bucket_name}: {e}")
         return False
 
+
 def validate_s3_references(file_path):
-    with open(file_path, 'r') as file:
+    with open(file_path, "r") as file:
         content = file.read()
 
-    s3_pattern = re.compile(r's3:\/\/([a-zA-Z0-9._-]+)')
+    s3_pattern = re.compile(r"s3:\/\/([a-zA-Z0-9._-]+)")
     matches = s3_pattern.findall(content)
 
     invalid_buckets = []
@@ -33,8 +34,10 @@ def validate_s3_references(file_path):
 
     return invalid_buckets
 
+
 if __name__ == "__main__":
     import sys
+
     if len(sys.argv) < 2:
         print("Usage: python s3_bucket_validator.py <file_path>")
         sys.exit(1)
@@ -46,4 +49,4 @@ def validate_s3_references(file_path):
         print(f"Invalid or inaccessible S3 buckets found: {', '.join(invalid_buckets)}")
         sys.exit(1)
     else:
-        print("All referenced S3 buckets are valid and accessible.")
\ No newline at end of file
+        print("All referenced S3 buckets are valid and accessible.")

From 13696de059b1a8d81389e61cca7835d562a4246c Mon Sep 17 00:00:00 2001
From: Nathan Park <parknate@amazon.com>
Date: Fri, 18 Jul 2025 14:26:14 -0700
Subject: [PATCH 3/3] format: add future import

---
 s3_bucket_validator.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/s3_bucket_validator.py b/s3_bucket_validator.py
index 1dfa271c57..4d86294096 100644
--- a/s3_bucket_validator.py
+++ b/s3_bucket_validator.py
@@ -1,3 +1,5 @@
+from __future__ import absolute_import
+
 import re
 
 import boto3