From 02896c465ffe53decfb38a141ae881118b4a3ade Mon Sep 17 00:00:00 2001
From: "tiffany.cheng" <tiffany.cheng@jisc.ac.uk>
Date: Thu, 18 Jun 2026 13:45:18 +0100
Subject: [PATCH 1/2] Add: decompress csv function in utils.py

---
 mario/utils.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/mario/utils.py b/mario/utils.py
index 6d43f6f..d5ab141 100644
--- a/mario/utils.py
+++ b/mario/utils.py
@@ -29,3 +29,13 @@ def gzip_file(input_path, output_path=None):
     with open(input_path, "rb") as f_in, gzip.open(output_path, "wb") as f_out:
         shutil.copyfileobj(f_in, f_out)
     return output_path
+
+
+def decompress_gzip_file(input_path, output_path=None):
+    import gzip
+    import shutil
+    if output_path is None:
+        output_path = input_path.removesuffix(".gz")
+    with gzip.open(input_path, "rb") as f_in, open(output_path, "wb") as f_out:
+        shutil.copyfileobj(f_in, f_out)
+    return output_path
\ No newline at end of file

From e2f138494b37f7c76f310bf018835aae466bffe9 Mon Sep 17 00:00:00 2001
From: "tiffany.cheng" <tiffany.cheng@jisc.ac.uk>
Date: Fri, 19 Jun 2026 15:05:13 +0100
Subject: [PATCH 2/2] Add: some checking code according to code review

---
 mario/utils.py | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/mario/utils.py b/mario/utils.py
index d5ab141..4dc529c 100644
--- a/mario/utils.py
+++ b/mario/utils.py
@@ -1,5 +1,7 @@
 import os
 from datetime import datetime
+import logging
+logger = logging.getLogger(__name__)
 
 
 def append_current_date_to_file_name(file_name: str) -> str:
@@ -33,9 +35,23 @@ def gzip_file(input_path, output_path=None):
 
 def decompress_gzip_file(input_path, output_path=None):
     import gzip
+    import os
     import shutil
+    if not input_path.endswith(".gz"):
+        raise ValueError(f"Input path must end with '.gz': {input_path}")
+
     if output_path is None:
         output_path = input_path.removesuffix(".gz")
+
+    # Create output directory if not exists
+    output_dir = os.path.dirname(output_path)
+    if output_dir:
+        os.makedirs(output_dir, exist_ok=True)
+
+    if os.path.exists(output_path):
+        logger.warning(f"Output file already exists and will be overwritten: {output_path}")
+
     with gzip.open(input_path, "rb") as f_in, open(output_path, "wb") as f_out:
         shutil.copyfileobj(f_in, f_out)
+
     return output_path
\ No newline at end of file