From f47c763b3872a40eb0323f6958a55d235957ffbe Mon Sep 17 00:00:00 2001
From: Lee Rhodes <leerho@users.noreply.github.com>
Date: Sun, 4 Oct 2026 17:43:27 -0700
Subject: [PATCH] ci: download TCK snapshots from the serialization_test_data
 layout

The TCK moved its snapshots from serialization/<lang>/snapshots to
serialization_test_data/<lang>_generated_files (apache/datasketches-tck#14),
the same layout this repository uses. Match the new archive paths, strip
one fewer path component, and pin the TCK at e748b0a, its current main.

The serde-compat workflow's Java entry passed `java` to the download
script, which only accepts cpp and go. The check_java_files profile
generates its own snapshots before checking them, so skip the download
for that entry instead.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---
 .github/workflows/serde-compat.yml        | 2 ++
 tools/download_serialization_test_data.sh | 6 +++---
 2 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/serde-compat.yml b/.github/workflows/serde-compat.yml
index d3d85b369..a15b88526 100644
--- a/.github/workflows/serde-compat.yml
+++ b/.github/workflows/serde-compat.yml
@@ -30,7 +30,9 @@ jobs:
           java-version: '25'
           distribution: 'temurin'
 
+      # The Java profile generates its own snapshots before checking them.
       - name: Download ${{ matrix.name }} snapshots
+        if: matrix.language != 'java'
         run: ./tools/download_serialization_test_data.sh ${{ matrix.language }}
 
       - name: Run Java tests against ${{ matrix.name }} snapshots
diff --git a/tools/download_serialization_test_data.sh b/tools/download_serialization_test_data.sh
index 800e6206e..caa8f4239 100755
--- a/tools/download_serialization_test_data.sh
+++ b/tools/download_serialization_test_data.sh
@@ -20,7 +20,7 @@
 set -euo pipefail
 
 # Pin the archive so compatibility tests always use an immutable snapshot set.
-readonly TCK_REVISION="d363b12d293b395d90abb42677f9ea63178dbc0d"
+readonly TCK_REVISION="e748b0a038a17c6d5c771afc4efd9ee35a5a7ec5"
 readonly TCK_ARCHIVE_URL="https://api.github.com/repos/apache/datasketches-tck/tarball/${TCK_REVISION}"
 readonly SCRIPT_DIRECTORY="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 readonly REPOSITORY_ROOT="$(cd "${SCRIPT_DIRECTORY}/.." && pwd)"
@@ -103,7 +103,7 @@ for language in "${languages[@]}"; do
 
   while IFS= read -r member; do
     case "${member}" in
-      */serialization/"${language}"/snapshots/*.sk)
+      */serialization_test_data/"${language}"_generated_files/*.sk)
         name="${member##*/}"
         for existing_name in "${names[@]-}"; do
           if [[ "${name}" == "${existing_name}" ]]; then
@@ -126,7 +126,7 @@ for language in "${languages[@]}"; do
   tar \
     -xzf "${archive_path}" \
     -C "${staging_directory}" \
-    --strip-components=4 \
+    --strip-components=3 \
     "${members[@]}"
   for name in "${names[@]}"; do
     if [[ ! -f "${staging_directory}/${name}" ]]; then