From 52da7cee1d21febb673897cbb1d04ed54361d13f Mon Sep 17 00:00:00 2001
From: James Bruten <james.bruten@metoffice.gov.uk>
Date: Thu, 18 Dec 2025 08:34:32 +0000
Subject: [PATCH 1/3] add specific excludes

---
 github_scripts/get_git_sources.py | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)

diff --git a/github_scripts/get_git_sources.py b/github_scripts/get_git_sources.py
index e9f94f2..97d0867 100644
--- a/github_scripts/get_git_sources.py
+++ b/github_scripts/get_git_sources.py
@@ -12,6 +12,7 @@
 from typing import Optional
 from pathlib import Path
 from shutil import rmtree
+import shlex
 
 
 def run_command(
@@ -24,7 +25,7 @@ def run_command(
     Outputs:
         - result object from subprocess.run
     """
-    command = command.split()
+    command = shlex.split(command)
     result = subprocess.run(
         command,
         capture_output=True,
@@ -117,8 +118,25 @@ def sync_repo(repo_source: str, repo_ref: str, loc: Path) -> None:
     # Create a clean clone location
     loc.mkdir(parents=True)
 
+    exclude_dirs = (
+        "applications/*/working",
+        "applications/*/bin",
+        "science/*/working",
+        "science/*/bin",
+        "interfaces/*/working",
+        "interfaces/*/bin",
+        "components/*/working",
+        "components/*/bin",
+        "infrastructure/*/working",
+        "infrastructure/*/bin",
+        "mesh_tools/*/working",
+        "mesh_tools/*/bin"
+    )
+
     # Trailing slash required for rsync
     command = f"rsync -av {repo_source}/ {loc}"
+    for item in exclude_dirs:
+        command = f"{command} --exclude '{item}'"
     run_command(command)
 
     # Fetch the main branch from origin

From 327c509378d7885974cb522a8a944aa8f5ca780f Mon Sep 17 00:00:00 2001
From: James Bruten <james.bruten@metoffice.gov.uk>
Date: Thu, 18 Dec 2025 08:48:02 +0000
Subject: [PATCH 2/3] add test directory too

---
 github_scripts/get_git_sources.py | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/github_scripts/get_git_sources.py b/github_scripts/get_git_sources.py
index 97d0867..571bc77 100644
--- a/github_scripts/get_git_sources.py
+++ b/github_scripts/get_git_sources.py
@@ -120,16 +120,22 @@ def sync_repo(repo_source: str, repo_ref: str, loc: Path) -> None:
 
     exclude_dirs = (
         "applications/*/working",
+        "applications/*/test",
         "applications/*/bin",
         "science/*/working",
+        "science/*/test",
         "science/*/bin",
         "interfaces/*/working",
+        "interfaces/*/test",
         "interfaces/*/bin",
         "components/*/working",
+        "components/*/test",
         "components/*/bin",
         "infrastructure/*/working",
+        "infrastructure/*/test",
         "infrastructure/*/bin",
         "mesh_tools/*/working",
+        "mesh_tools/*/test",
         "mesh_tools/*/bin"
     )
 

From a75ee3789e92d57cf414f6c4d333429525615f78 Mon Sep 17 00:00:00 2001
From: James Bruten <james.bruten@metoffice.gov.uk>
Date: Thu, 18 Dec 2025 09:56:35 +0000
Subject: [PATCH 3/3] run black

---
 github_scripts/get_git_sources.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/github_scripts/get_git_sources.py b/github_scripts/get_git_sources.py
index 571bc77..d5ec8d1 100644
--- a/github_scripts/get_git_sources.py
+++ b/github_scripts/get_git_sources.py
@@ -136,7 +136,7 @@ def sync_repo(repo_source: str, repo_ref: str, loc: Path) -> None:
         "infrastructure/*/bin",
         "mesh_tools/*/working",
         "mesh_tools/*/test",
-        "mesh_tools/*/bin"
+        "mesh_tools/*/bin",
     )
 
     # Trailing slash required for rsync