mirror of
https://github.com/snapotter-hq/SnapOtter.git
synced 2026-08-03 07:46:42 +02:00
The published arm64 set disagreed on huggingface_hub (four bundles baked 1.22.0), scipy, and scikit-image. All six affected bundles are rebuilt under the constrained closure and published at new -r2 paths so baked manifests keep working; no rebuilt bundle carries hub and the layered compatibility check passes. verify-bundle.sh drops its stale onnxruntime expectation for upscale-enhance. Completes #669.
296 lines
10 KiB
Bash
Executable File
296 lines
10 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
set -euo pipefail
|
|
|
|
# ──────────────────────────────────────────────────────────────────────────────
|
|
# verify-bundle.sh -- Verify an AI bundle tarball inside the Docker container
|
|
#
|
|
# Usage: verify-bundle.sh <bundleId> <arch>
|
|
#
|
|
# bundleId - One of: background-removal, face-detection, object-eraser-colorize,
|
|
# upscale-enhance, photo-restoration, transcription
|
|
# arch - Architecture variant: amd64-gpu or arm64-cpu
|
|
#
|
|
# Expects:
|
|
# /bundles/<bundleId>-<arch>.tar.gz (read-only mount)
|
|
# /bundles/<bundleId>-<arch>.tar.gz.sha256
|
|
# /fixtures/ (test fixtures, read-only mount)
|
|
#
|
|
# Runs with --entrypoint bash (no entrypoint bootstrap), so uses /opt/venv
|
|
# directly rather than /data/ai/venv.
|
|
#
|
|
# Exit codes: 0=pass, 1=integrity, 2=import, 3=smoke
|
|
# ──────────────────────────────────────────────────────────────────────────────
|
|
|
|
BUNDLE_ID="${1:?Usage: verify-bundle.sh <bundleId> <arch>}"
|
|
ARCH="${2:?Usage: verify-bundle.sh <bundleId> <arch>}"
|
|
|
|
ARCHIVE="/bundles/${BUNDLE_ID}-${ARCH}.tar.gz"
|
|
SHA_FILE="/bundles/${BUNDLE_ID}-${ARCH}.tar.gz.sha256"
|
|
STAGING="/tmp/verify-staging"
|
|
|
|
VENV="/opt/venv"
|
|
PYTHON="${VENV}/bin/python3"
|
|
SITE_PACKAGES="$("${PYTHON}" -c 'import sysconfig; print(sysconfig.get_paths()["purelib"])')"
|
|
|
|
export MODELS_PATH="${MODELS_PATH:-/tmp/verify-models}"
|
|
export U2NET_HOME="${MODELS_PATH}/rembg"
|
|
export PYTHONPATH="/app/packages/ai/python"
|
|
export SNAPOTTER_GPU=0
|
|
export CUDA_VISIBLE_DEVICES=""
|
|
|
|
# ── Helpers ──────────────────────────────────────────────────────────────────
|
|
|
|
log() {
|
|
echo "=== $* ==="
|
|
}
|
|
|
|
pass() {
|
|
echo -e "\033[32mPASS: $*\033[0m"
|
|
}
|
|
|
|
fail() {
|
|
local msg="$1"
|
|
local code="${2:-1}"
|
|
echo -e "\033[31mFAIL: ${msg}\033[0m" >&2
|
|
exit "${code}"
|
|
}
|
|
|
|
run_python() {
|
|
"${PYTHON}" "$@"
|
|
}
|
|
|
|
check_imports() {
|
|
local mods="$1"
|
|
for mod in ${mods}; do
|
|
if run_python -c "import ${mod}" 2>/dev/null; then
|
|
pass "import ${mod}"
|
|
else
|
|
fail "import ${mod} failed" 2
|
|
fi
|
|
done
|
|
}
|
|
|
|
# ── Phase 1: Integrity Checks ───────────────────────────────────────────────
|
|
|
|
log "Phase 1: Integrity Checks"
|
|
|
|
if [[ ! -f "${ARCHIVE}" ]]; then
|
|
fail "Tarball not found: ${ARCHIVE}"
|
|
fi
|
|
pass "Tarball exists"
|
|
|
|
if [[ ! -f "${SHA_FILE}" ]]; then
|
|
fail "SHA256 file not found: ${SHA_FILE}"
|
|
fi
|
|
pass "SHA256 file exists"
|
|
|
|
EXPECTED_SHA="$(cat "${SHA_FILE}" | awk '{print $1}')"
|
|
ACTUAL_SHA="$(sha256sum "${ARCHIVE}" | awk '{print $1}')"
|
|
|
|
if [[ "${EXPECTED_SHA}" != "${ACTUAL_SHA}" ]]; then
|
|
fail "SHA256 mismatch: expected=${EXPECTED_SHA} actual=${ACTUAL_SHA}"
|
|
fi
|
|
pass "SHA256 checksum matches"
|
|
|
|
rm -rf "${STAGING}"
|
|
mkdir -p "${STAGING}"
|
|
tar -xzf "${ARCHIVE}" -C "${STAGING}"
|
|
pass "Tarball extracted to ${STAGING}"
|
|
|
|
if [[ ! -f "${STAGING}/bundle.json" ]]; then
|
|
fail "bundle.json not found in archive"
|
|
fi
|
|
|
|
# Validate required fields in bundle.json
|
|
run_python -c "
|
|
import json, sys
|
|
with open('${STAGING}/bundle.json') as f:
|
|
b = json.load(f)
|
|
for field in ('bundleId', 'arch', 'version', 'models'):
|
|
if field not in b:
|
|
print(f'Missing required field: {field}', file=sys.stderr)
|
|
sys.exit(1)
|
|
print(f' bundleId={b[\"bundleId\"]} arch={b[\"arch\"]} version={b[\"version\"]} models={len(b[\"models\"])}')
|
|
" || fail "bundle.json missing required fields"
|
|
pass "bundle.json valid"
|
|
|
|
# ── Phase 2: Installation ────────────────────────────────────────────────────
|
|
|
|
log "Phase 2: Installation"
|
|
|
|
if [[ -d "${STAGING}/site-packages" ]]; then
|
|
cp -a "${STAGING}/site-packages/." "${SITE_PACKAGES}/"
|
|
pass "site-packages merged into ${SITE_PACKAGES}"
|
|
else
|
|
echo " No site-packages/ in bundle, skipping"
|
|
fi
|
|
|
|
mkdir -p "${MODELS_PATH}"
|
|
if [[ -d "${STAGING}/models" ]]; then
|
|
cp -a "${STAGING}/models/." "${MODELS_PATH}/"
|
|
pass "Models copied to ${MODELS_PATH}"
|
|
else
|
|
echo " No models/ in bundle, skipping"
|
|
fi
|
|
|
|
if [[ -d "${STAGING}/fixups" ]]; then
|
|
WHEELS=("${STAGING}/fixups"/*.whl)
|
|
if [[ -f "${WHEELS[0]}" ]]; then
|
|
run_python -m pip install --no-index --no-cache-dir "${STAGING}/fixups"/*.whl 2>/dev/null
|
|
pass "Fixup wheels installed"
|
|
else
|
|
echo " No .whl files in fixups/, skipping"
|
|
fi
|
|
else
|
|
echo " No fixups/ directory, skipping"
|
|
fi
|
|
|
|
rm -rf "${STAGING}"
|
|
pass "Staging cleaned up"
|
|
|
|
df -h
|
|
|
|
# ── Phase 3: Import Checks ──────────────────────────────────────────────────
|
|
|
|
log "Phase 3: Import Checks"
|
|
|
|
case "${BUNDLE_ID}" in
|
|
background-removal)
|
|
check_imports "rembg onnxruntime"
|
|
;;
|
|
face-detection)
|
|
check_imports "mediapipe"
|
|
;;
|
|
object-eraser-colorize)
|
|
check_imports "onnxruntime cv2"
|
|
;;
|
|
upscale-enhance)
|
|
# onnxruntime was only ever a transitive of an older dep tree here; no
|
|
# upscale-enhance script imports it and gpu.py degrades cleanly without it.
|
|
check_imports "torch realesrgan"
|
|
;;
|
|
photo-restoration)
|
|
check_imports "torch onnxruntime mediapipe"
|
|
;;
|
|
ocr)
|
|
fail "OCR v3 runtimes must use verify-ocr-runtime.sh" 2
|
|
;;
|
|
transcription)
|
|
check_imports "faster_whisper"
|
|
;;
|
|
*)
|
|
fail "Unknown bundle: ${BUNDLE_ID}" 2
|
|
;;
|
|
esac
|
|
|
|
pass "All imports OK for ${BUNDLE_ID}"
|
|
|
|
# ── Phase 4: Functional Smoke Tests ─────────────────────────────────────────
|
|
|
|
log "Phase 4: Functional Smoke Tests"
|
|
|
|
AI_SCRIPTS="/app/packages/ai/python"
|
|
|
|
smoke_background_removal() {
|
|
local input="/fixtures/image/valid/test-200x150.png"
|
|
[[ -f "$input" ]] || fail "Fixture missing: $input" 3
|
|
local output="/tmp/verify-smoke-rembg.png"
|
|
timeout 300 "${PYTHON}" "${AI_SCRIPTS}/remove_bg.py" "${input}" "${output}" 2>/dev/null || true
|
|
[[ -f "${output}" && -s "${output}" ]] || fail "remove_bg output missing or empty" 3
|
|
# Verify it's a valid PNG (check magic bytes)
|
|
run_python -c "
|
|
import sys
|
|
with open('${output}', 'rb') as f:
|
|
magic = f.read(8)
|
|
if magic[:4] != b'\\x89PNG':
|
|
sys.exit(1)
|
|
" || fail "remove_bg output is not a valid PNG" 3
|
|
pass "remove_bg produced valid PNG output"
|
|
}
|
|
|
|
smoke_face_detection() {
|
|
local input="/fixtures/image/valid/sample-photo.jpg"
|
|
[[ -f "$input" ]] || fail "Fixture missing: $input" 3
|
|
local output="/tmp/verify-smoke-faces.png"
|
|
timeout 300 "${PYTHON}" "${AI_SCRIPTS}/detect_faces.py" "${input}" "${output}" 2>/dev/null || true
|
|
[[ -f "${output}" ]] || fail "detect_faces output missing" 3
|
|
pass "detect_faces produced output"
|
|
}
|
|
|
|
smoke_object_eraser_colorize() {
|
|
local input="/tmp/verify-smoke-gray.png"
|
|
local output="/tmp/verify-smoke-color.png"
|
|
# Generate a 64x64 grayscale test image
|
|
run_python -c "
|
|
from PIL import Image
|
|
img = Image.new('L', (64, 64), 128)
|
|
img.save('${input}')
|
|
" 2>/dev/null
|
|
timeout 300 "${PYTHON}" "${AI_SCRIPTS}/colorize.py" "${input}" "${output}" '{"model":"ddcolor"}' 2>/dev/null || true
|
|
[[ -f "${output}" && -s "${output}" ]] || fail "colorize output missing or empty" 3
|
|
pass "colorize produced output"
|
|
}
|
|
|
|
smoke_upscale_enhance() {
|
|
local input="/fixtures/image/valid/test-100x100.jpg"
|
|
[[ -f "$input" ]] || fail "Fixture missing: $input" 3
|
|
local output="/tmp/verify-smoke-upscale.png"
|
|
timeout 300 "${PYTHON}" "${AI_SCRIPTS}/upscale.py" "${input}" "${output}" '{"scale":2}' 2>/dev/null || true
|
|
[[ -f "${output}" && -s "${output}" ]] || fail "upscale output missing or empty" 3
|
|
run_python -c "
|
|
from PIL import Image
|
|
inp = Image.open('${input}')
|
|
out = Image.open('${output}')
|
|
if out.width <= inp.width or out.height <= inp.height:
|
|
print(f'Output {out.size} not larger than input {inp.size}')
|
|
exit(1)
|
|
print(f' Upscale: {inp.size} -> {out.size}')
|
|
" || fail "upscale output dimensions not larger than input" 3
|
|
pass "upscale produced larger output"
|
|
}
|
|
|
|
smoke_photo_restoration() {
|
|
local input="/fixtures/image/valid/test-100x100.jpg"
|
|
[[ -f "$input" ]] || fail "Fixture missing: $input" 3
|
|
local output="/tmp/verify-smoke-restore.png"
|
|
timeout 300 "${PYTHON}" "${AI_SCRIPTS}/restore.py" "${input}" "${output}" 2>/dev/null || true
|
|
[[ -f "${output}" && -s "${output}" ]] || fail "restore output missing or empty" 3
|
|
pass "restore produced output"
|
|
}
|
|
|
|
smoke_transcription() {
|
|
local input="/fixtures/audio/valid/speech-10s.wav"
|
|
[[ -f "$input" ]] || fail "Fixture missing: $input" 3
|
|
local result
|
|
result="$(timeout 300 "${PYTHON}" "${AI_SCRIPTS}/transcribe.py" "${input}" 2>/dev/null)" || result=""
|
|
# Check stdout JSON has success=true and non-empty segments
|
|
echo "${result}" | run_python -c "
|
|
import json, sys
|
|
data = json.load(sys.stdin)
|
|
if not data.get('success'):
|
|
print('Transcription did not return success=true', file=sys.stderr)
|
|
sys.exit(1)
|
|
segments = data.get('segments', [])
|
|
if not segments:
|
|
print('Transcription returned empty segments', file=sys.stderr)
|
|
sys.exit(1)
|
|
print(f' Transcription segments: {len(segments)}')
|
|
" || fail "Transcription smoke test assertion failed" 3
|
|
pass "transcribe returned valid result"
|
|
}
|
|
|
|
case "${BUNDLE_ID}" in
|
|
background-removal) smoke_background_removal ;;
|
|
face-detection) smoke_face_detection ;;
|
|
object-eraser-colorize) smoke_object_eraser_colorize ;;
|
|
upscale-enhance) smoke_upscale_enhance ;;
|
|
photo-restoration) smoke_photo_restoration ;;
|
|
ocr) fail "OCR v3 runtimes must use verify-ocr-runtime.sh" 2 ;;
|
|
transcription) smoke_transcription ;;
|
|
esac
|
|
|
|
# ── Done ─────────────────────────────────────────────────────────────────────
|
|
|
|
log "All phases passed for ${BUNDLE_ID} (${ARCH})"
|
|
exit 0
|