From b7cdd5211c457c7e103ebb9dbd1ec937a0ceaee5 Mon Sep 17 00:00:00 2001
From: Jake Awe <jawe@nvidia.com>
Date: Thu, 1 Feb 2024 13:24:11 -0600
Subject: [PATCH] fix cugraph notebooks

---
 Dockerfile           |  2 ++
 context/notebooks.sh | 15 ++++++++++++++-
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/Dockerfile b/Dockerfile
index 61344057..e59923ed 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -24,6 +24,8 @@ RUN pip install --upgrade conda-merge rapids-dependency-file-generator
 COPY condarc /condarc
 COPY notebooks.sh /notebooks.sh
 
+RUN apt-get update && apt-get install -y rsync && rm -rf /var/lib/apt/lists/*
+
 RUN /notebooks.sh
 
 
diff --git a/context/notebooks.sh b/context/notebooks.sh
index c64b887c..9b6ede95 100755
--- a/context/notebooks.sh
+++ b/context/notebooks.sh
@@ -15,7 +15,20 @@ mkdir -p /notebooks /dependencies
 for REPO in "${NOTEBOOK_REPOS[@]}"; do
     echo "Cloning $REPO..."
     git clone -b "${RAPIDS_BRANCH}" --depth 1 --single-branch "https://github.com/rapidsai/$REPO" "$REPO"
-    cp -rL "$REPO"/notebooks /notebooks/"$REPO"
+
+    SOURCE="$REPO/notebooks"
+    DESTINATION="/notebooks/$REPO"
+    if [ "$REPO" = "cugraph" ]; then
+        echo "Special handling for $REPO..."
+        EXCLUDE_LIST=$(mktemp)
+        mkdir -p $DESTINATION
+        find "$SOURCE" -type f -name "SKIP_CI_TESTING" -printf "%h\n" | sed "s|^$SOURCE/||" > "$EXCLUDE_LIST"
+        rsync -avL --exclude-from="$EXCLUDE_LIST" "$SOURCE"/ "$DESTINATION"/
+        rm "$EXCLUDE_LIST"
+    else
+        cp -rL "$SOURCE" "$DESTINATION"
+    fi
+
     if [ -f "$REPO/dependencies.yaml" ] && yq -e '.files.test_notebooks' "$REPO/dependencies.yaml" >/dev/null; then
         echo "Running dfg on $REPO"
         rapids-dependency-file-generator \