Repository navigation

Native lookup_by_key raises without an index, as the JVM build does #213

Workflow file for this run

.github/workflows/test-python-examples.yml at 68ca636

	name: Test Python Examples

	on:
	push:
	branches: [main]
	paths-ignore:
	# Changes that cannot alter what these jobs build or run. README.md and
	# tests/README.md are deliberately NOT here: the wheel build copies README.md
	# (scripts/Dockerfile.build, pyproject `readme`) and test_no_monorepo_wording.py
	# reads both. docs/** is safe to skip while tests/conftest.py collect-ignores
	# test_docs_examples.py (it needs docs/index.md); if that Markdown site comes
	# back, take docs/** out of this list. Schedule, dispatch and workflow_call
	# ignore these filters and always run everything.
	- "docs/**"
	- "*/.adoc"
	- "CLAUDE.md"
	- "examples/README.md"
	- "LICENSE*"
	# the native build has its own workflows; nothing under native/ is in the JVM wheel
	- "native/**"
	- ".github/workflows/test-native.yml"
	- ".github/workflows/release-native.yml"
	pull_request:
	# ready_for_review: draft PRs skip the platform matrices (see the job conditions),
	# so marking one ready is what starts them.
	types: [opened, synchronize, reopened, ready_for_review]
	paths-ignore:
	# Same list as under push.
	- "docs/**"
	- "*/.adoc"
	- "CLAUDE.md"
	- "examples/README.md"
	- "LICENSE*"
	# the native build has its own workflows; nothing under native/ is in the JVM wheel
	- "native/**"
	- ".github/workflows/test-native.yml"
	- ".github/workflows/release-native.yml"

	# A new upstream snapshot is published as arcadedata/arcadedb:latest without
	# any change here, so build and test against it every night as well. The tag
	# moves (ARCADEDB_IMAGE_TAG defaults to latest), so a night with no commit
	# here still tests a new engine; that is why there is no skip-if-main-unchanged
	# guard. One nightly per workflow, staggered an hour apart.
	schedule:
	- cron: "0 4 * * *"

	# Allow being called by other workflows (release-jvm.yml)
	workflow_call:
	inputs:
	arcadedb-tag:
	description: "arcadedata/arcadedb image tag to take the jars from (e.g. 26.10.2); default latest"
	required: false
	type: string
	default: "latest"
	examples:
	description: "Glob pattern(s) for examples to run (space-separated, relative to examples)."
	required: false
	type: string
	default: "0[1-9]_.py 1[0-9]_.py 2[0-9]_*.py"
	build-version:
	description: "Override package version (PEP 440) for build.sh"
	required: false
	type: string

	# Allow manual trigger
	workflow_dispatch:
	inputs:
	arcadedb-tag:
	description: "arcadedata/arcadedb image tag to take the jars from (e.g. 26.10.2); default latest"
	required: false
	default: "latest"
	examples:
	description: "Glob pattern(s) for examples to run (space-separated, relative to examples)."
	required: false
	default: "0[1-9]_.py 1[0-9]_.py 2[0-9]_*.py"

	# SUPERSEDE, do not race (same rule in test-jvm.yml).
	#
	# A force-push to a PR branch left the OLD run going alongside the new one: on
	# #6889 both 118b347ce and 75293c7a2 ran at once, and the stale one reported a
	# failure that the newer commit had already fixed. Anyone reading the PR sees a
	# red check against work that is done. The stale run also holds macOS/Windows
	# runners the organisation shares with ArcadeData/arcadedb.
	#
	# cancel-in-progress is scoped to pull_request ONLY. On a push to a long-lived
	# branch a superseded run is often the one you still want -- it is the record
	# for that commit -- so those are left to finish, and they get a group of their
	# own (run_id) so a queued push or nightly run is not replaced by a newer one
	# either. The group names the workflow literally: github.workflow is the
	# CALLER's name in a workflow_call, which would put this workflow and its
	# sibling in one group.
	concurrency:
	group: test-python-examples-${{ github.event.pull_request.number \|\| github.run_id }}
	cancel-in-progress: ${{ github.event_name == 'pull_request' }}

	env:
	# The image download-jars takes the jars from. Unset (push, pull_request, schedule)
	# means latest; a release build passes its tag.
	ARCADEDB_IMAGE_TAG: ${{ inputs.arcadedb-tag \|\| 'latest' }}
	EXAMPLES: ${{ inputs.examples \|\| '0[1-9]_.py 1[0-9]_.py 2[0-9]_*.py' }}

	# Note: Example 21 runs in CI at reduced scale; its arguments and 15-minute
	# limit are set in the per-example case in the "Run all examples" step below.

	permissions:
	contents: read

	jobs:
	# First job: Download ArcadeDB JARs (platform-agnostic)
	download-jars:
	name: Download ArcadeDB JARs
	runs-on: ubuntu-24.04
	# Draft PRs skip the whole workflow; marking the PR ready starts it.
	if: github.event_name != 'pull_request' \|\| github.event.pull_request.draft == false
	outputs:
	arcadedb-tag: ${{ steps.jars.outputs.arcadedb-tag }}
	matrix: ${{ steps.matrix.outputs.matrix }}
	steps:
	- name: Checkout code
	uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

	# The platform x Python legs of this run (scripts/select_matrix.py): a small matrix on
	# pull requests, the full one on every other event. The organisation shares one runner
	# cap with ArcadeData/arcadedb, so a pull request does not run every combination.
	# Done first, so a failure here costs no Docker pull.
	- name: Select the test matrix
	id: matrix
	shell: bash
	env:
	EVENT_NAME: ${{ github.event_name }}
	run: \|
	matrix="$(python3 scripts/select_matrix.py examples --event "$EVENT_NAME")"
	echo "$matrix"
	echo "matrix=$matrix" >> "$GITHUB_OUTPUT"

	- name: Set up Docker Buildx
	uses: docker/setup-buildx-action@f87e5991a6d7451dcb8d9637bfbc97413f497069 # v4.4.1

	- name: Download JARs from ArcadeDB Docker image
	id: jars
	shell: bash
	run: \|
	# Download JARs from the official Docker image: ARCADEDB_IMAGE_TAG,
	# `latest` (upstream's current snapshot) unless the caller pinned a tag. Every build job, Linux included, builds
	# from these jars (the artifact below), not from its own pull: upstream
	# re-pushes the snapshot tag on every deploy, so a pull per job could put
	# two different engine builds under one version in a single run.
	echo "📦 Downloading JARs from arcadedata/arcadedb:${ARCADEDB_IMAGE_TAG}..."

	# Create output directory
	mkdir -p src/arcadedb/jars

	# Create a temporary container and copy JARs from it
	CONTAINER_ID=$(docker create "arcadedata/arcadedb:${ARCADEDB_IMAGE_TAG}")
	docker cp $CONTAINER_ID:/home/arcadedb/lib/. src/arcadedb/jars/
	docker rm $CONTAINER_ID

	echo "✅ Downloaded $(ls -1 src/arcadedb/jars/*.jar \| wc -l) JARs"

	ARCADEDB_TAG=$(python3 scripts/extract_version.py src/arcadedb/jars --format=docker)
	echo "📌 ArcadeDB version: $ARCADEDB_TAG"
	# The build jobs pull arcadedata/arcadedb:<jar version> as their base image,
	# so a pinned tag must name that version exactly (26.10.2, not an alias).
	if [ "$ARCADEDB_IMAGE_TAG" != "latest" ] && [ "$ARCADEDB_IMAGE_TAG" != "$ARCADEDB_TAG" ]; then
	echo "❌ arcadedb-tag '$ARCADEDB_IMAGE_TAG' holds engine $ARCADEDB_TAG; pass the exact version tag"
	exit 1
	fi
	echo "arcadedb-tag=$ARCADEDB_TAG" >> "$GITHUB_OUTPUT"

	# Unfiltered on purpose: build-native.sh applies scripts/jar_exclusions.txt
	# to whatever lands in the jars directory, so the list lives in one place.
	- name: Upload JARs as artifact
	uses: actions/upload-artifact@cf430e030ddbb5b0abf93d22962f4752f3646cd9 # v7.0.2
	with:
	name: arcadedb-jars-examples
	path: src/arcadedb/jars/*.jar
	retention-days: 1

	# Second job: Test examples on each platform
	test-examples:
	name: Test Python Examples (${{ matrix.platform }}, Python ${{ matrix.python-version }})
	runs-on: ${{ matrix.runs-on }}
	needs: download-jars
	if: github.event_name != 'pull_request' \|\| github.event.pull_request.draft == false
	env:
	BUILD_VERSION: ${{ inputs.build-version }}
	# Pin the Hugging Face cache inside the workspace so it is cacheable with
	# one path on every platform (the default differs per OS). The examples
	# embed with sentence-transformers, and the full matrix is 16 jobs, so
	# uncached that is 16 independent model downloads per release and 16
	# chances for a transient hub outage to fail a release. It has happened.
	HF_HOME: ${{ github.workspace }}/.hf-cache
	strategy:
	fail-fast: false
	# Chosen by the "Select the test matrix" step of download-jars from the event that
	# started the run (scripts/select_matrix.py holds the platform and Python lists):
	# small on pull requests, the full 4 platforms x 4 Pythons on every other event.
	# Every leg carries platform, runs-on and python-version, so the job names stay
	# "Test Python Examples (<platform>, Python <version>)".
	matrix: ${{ fromJSON(needs.download-jars.outputs.matrix) }}

	steps:
	- name: Checkout code
	uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

	- name: Download JARs artifact
	uses: actions/download-artifact@9000827ccba6bdab643e8b6fd33ac0654aef8333 # v8.0.2
	with:
	name: arcadedb-jars-examples
	path: src/arcadedb/jars

	- name: Set up Java (for native builds on macOS/Windows)
	if: matrix.platform != 'linux/amd64' && matrix.platform != 'linux/arm64'
	uses: actions/setup-java@b843ebf9b913146b38da2eaa7aeddf76074ae1ad # v6.1.0
	with:
	distribution: 'corretto'
	java-version: '25'

	- name: Set up Docker Buildx (Linux only)
	if: matrix.platform == 'linux/amd64' \|\| matrix.platform == 'linux/arm64'
	uses: docker/setup-buildx-action@f87e5991a6d7451dcb8d9637bfbc97413f497069 # v4.4.1

	- name: Set up Python
	uses: actions/setup-python@6d4fd22b20465ccc0f7354476588e8df7c9ca0e0 # v7.1.0
	with:
	python-version: ${{ matrix.python-version }}

	- name: Setup UV package manager
	shell: bash
	run: \|
	# Retry transient network failures (a connection reset failed a whole examples job on #8897); the
	# script is downloaded first so a retried transfer is never piped half-written into sh.
	installer="$(mktemp)"
	curl -LsSf --retry 5 --retry-all-errors --retry-delay 5 -o "$installer" https://astral.sh/uv/install.sh
	sh "$installer"
	echo "$HOME/.local/bin" >> "$GITHUB_PATH"
	export PATH="$HOME/.local/bin:$PATH"
	uv --version

	- name: Install Python build dependencies
	shell: bash
	run: \|
	uv pip install --system build wheel setuptools

	# Windows currently enabled, no symlink needed
	# - name: Create python3 symlink (Windows only)
	# if: matrix.platform == 'windows/amd64' \|\| matrix.platform == 'windows/arm64'
	# shell: bash
	# run: \|
	# PYTHON_DIR=$(dirname "$(which python)")
	# ln -s "$PYTHON_DIR/python.exe" "$PYTHON_DIR/python3.exe" \|\| true
	# python3 --version

	- name: Build arcadedb (${{ matrix.platform }})
	shell: bash
	env:
	# The engine version download-jars read from the ARCADEDB_IMAGE_TAG jars;
	# the Docker build takes its base image by this tag, the jars from the artifact
	ARCADEDB_TAG: ${{ needs.download-jars.outputs.arcadedb-tag }}
	run: \|
	echo "🔨 Building arcadedb for ${{ matrix.platform }} with Python ${{ matrix.python-version }}..."
	# The jars download-jars fetched, on every platform, so all wheels of this
	# run carry the same engine build (and build.sh checks the integration jar)
	./scripts/build.sh ${{ matrix.platform }} ${{ matrix.python-version }} src/arcadedb/jars

	# Note: Java is NOT required - arcadedb has bundled JRE!

	- name: Install ArcadeDB Python bindings
	shell: bash
	run: \|
	uv pip install --system dist/arcadedb_jvm-*.whl

	- name: Install example dependencies
	shell: bash
	run: \|
	# Install dependencies needed by examples
	# PyTorch is required by sentence-transformers, even on macOS
	# For Python 3.13+, wheels might only be on PyTorch official index
	if [[ "$OSTYPE" == "darwin"* ]]; then
	# macOS: Install CPU-only PyTorch (supports MPS acceleration)
	uv pip install --system torch --index-url https://download.pytorch.org/whl/cpu
	# numpy is not capped: the old macOS `numpy<2.0` was for Intel Macs, where PyTorch stops at 2.2 (built against
	# NumPy 1.x). That leg (darwin/amd64) is disabled; re-enabling it needs the cap back for that platform only.
	# On Apple Silicon the cap breaks pyarrow, which requires NumPy 2 (example 27).
	uv pip install --system numpy requests sentence-transformers pandas polars pyarrow
	else
	# Linux & Windows: Install CPU-only PyTorch to save space (avoid CUDA)
	uv pip install --system torch --index-url https://download.pytorch.org/whl/cpu
	uv pip install --system numpy requests sentence-transformers pandas polars pyarrow
	fi

	# Restored before the dataset step, which is what triggers the model
	# download. Keyed on the model name so bumping the model invalidates it.
	- name: Cache Hugging Face model
	uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
	with:
	path: ${{ github.workspace }}/.hf-cache
	key: hf-all-MiniLM-L6-v2-${{ runner.os }}-${{ runner.arch }}
	restore-keys: \|
	hf-all-MiniLM-L6-v2-${{ runner.os }}-

	# Pre-warm with bounded retries, separately from the dataset step, so a
	# transient hub failure is retried here rather than surfacing mid-example
	# as an OSError that reads like a bindings bug. Retrying is appropriate
	# for this step specifically: it exercises third-party I/O, not our code.
	# Deliberately NOT fail-fast: if the hub is truly down the dataset step
	# below still runs and fails with its own message, so this cannot turn a
	# real breakage into a silent skip.
	- name: Pre-warm embedding model
	shell: bash
	continue-on-error: true
	run: \|
	uv pip install --system sentence-transformers >/dev/null 2>&1 \|\| true
	for attempt in 1 2 3; do
	if python3 -c "from sentence_transformers import SentenceTransformer; SentenceTransformer('all-MiniLM-L6-v2'); print('model ready')"; then
	echo "✅ model cached on attempt $attempt"
	exit 0
	fi
	echo "⚠️ model fetch failed (attempt $attempt/3)"
	[ "$attempt" -lt 3 ] && sleep $((attempt * 20))
	done
	echo "::warning::could not pre-warm the embedding model after 3 attempts; the dataset step will retry"

	- name: Download datasets
	shell: bash
	run: \|
	uv pip install --system tqdm py7zr lxml
	cd examples

	# Determine which examples are selected
	examples=$(ls $EXAMPLES 2>/dev/null \|\| true)

	# EXIT 75 MEANS "THE DATASET HOST IS DOWN", NOT "THE BINDINGS BROKE".
	#
	# On 2026-08-29 the TLS certificate for files.grouplens.org expired and
	# every job here went red on BOTH this fork and upstream's main -- 11
	# and 12 jobs, every OS, Python 3.10 through 3.14 -- at this step,
	# before a single example ran. Re-running does not help: a re-run
	# happens now, and the certificate is still expired.
	#
	# The cost is not the lost coverage, it is that upstream's main showed
	# a failing run titled with a Python bindings PR. A third party letting
	# a certificate lapse must not be indistinguishable from a regression
	# we introduced.
	#
	# So: skip the examples that need the dataset, keep running the rest,
	# and say plainly why. Any other non-zero exit is still a hard failure.
	SKIPPED_DATASETS=""
	fetch() { # fetch <label> <download_data.py args...>
	local label="$1"; shift
	# rc MUST be captured in the else branch. After `if cmd; then ...; fi`
	# with no else, the IF STATEMENT returns 0, so a `$?` read after `fi`
	# is the if's status and not the command's -- which silently turned
	# every failure into "exited 0" and would have made this skip
	# everything, including real ones.
	local rc=0
	if python3 download_data.py "$@"; then
	return 0
	else
	rc=$?
	fi
	if [ "$rc" -eq 75 ]; then
	echo "::warning title=Dataset unavailable::$label host unreachable; skipping examples that need it"
	SKIPPED_DATASETS="$SKIPPED_DATASETS $label"
	return 0
	fi
	echo "::error title=Dataset download failed::$label exited $rc"
	return "$rc"
	}

	if echo "$examples" \| grep -qE '(^\|[[:space:]])(04_csv_import_documents\|05_csv_import_graph\|06_vector_search_recommendations)\.py([[:space:]]\|$)'; then
	echo "📥 Downloading MovieLens Small dataset..."
	fetch movielens-small movielens-small
	fi

	if echo "$examples" \| grep -qE '(^\|[[:space:]])(07_stackoverflow_tables_oltp\|08_stackoverflow_tables_olap\|09_stackoverflow_graph_oltp\|10_stackoverflow_graph_olap\|13_stackoverflow_hybrid_queries)\.py([[:space:]]\|$)'; then
	echo "📥 Downloading Stack Overflow Tiny dataset (no vectors)..."
	fetch stackoverflow-tiny-novec stackoverflow-tiny --no-vectors
	fi

	if echo "$examples" \| grep -qE '(^\|[[:space:]])(11_vector_index_build\|12_vector_search)\.py([[:space:]]\|$)'; then
	echo "📥 Downloading Stack Overflow Tiny dataset (with vectors)..."
	fetch stackoverflow-tiny stackoverflow-tiny
	fi

	# HAND THE SKIP LIST TO THE NEXT STEP.
	#
	# Each step is its own shell, so SKIPPED_DATASETS dies here. Without
	# this the download skips politely and the example loop then runs
	# 04_csv_import_documents.py --dataset movielens-small against data
	# that was never fetched -- the job still fails, just later and with a
	# worse error. Caught in review on #6889.
	echo "SKIPPED_DATASETS=$(echo $SKIPPED_DATASETS \| xargs)" >> "$GITHUB_ENV"
	if [ -n "$(echo $SKIPPED_DATASETS \| xargs)" ]; then
	echo "::notice title=Datasets unavailable::skipping examples that need:$SKIPPED_DATASETS"
	fi

	- name: Install timeout command (macOS only)
	if: matrix.platform == 'darwin/amd64' \|\| matrix.platform == 'darwin/arm64'
	shell: bash
	run: \|
	# macOS doesn't have timeout command by default, use coreutils
	brew install coreutils

	- name: Set UTF-8 encoding (Windows only)
	if: matrix.platform == 'windows/amd64' \|\| matrix.platform == 'windows/arm64'
	shell: bash
	run: \|
	echo "PYTHONIOENCODING=utf-8" >> $GITHUB_ENV
	echo "PYTHONUTF8=1" >> $GITHUB_ENV

	- name: Run all examples
	id: run_examples
	shell: bash
	run: \|
	cd examples

	echo "🚀 Running Python Examples..."
	echo ""

	# Initialize counters
	total=0
	passed=0
	failed=0
	skipped=0

	# Create results file
	results_file="example-results.txt"
	> $results_file

	# Select example files by pattern(s) in EXAMPLES (space-separated globs)
	examples=$(ls $EXAMPLES 2>/dev/null \| sort)

	if [ -z "$examples" ]; then
	echo "❌ No example files found!"
	exit 1
	fi

	# Detect timeout command (macOS uses gtimeout, Linux uses timeout)
	if command -v gtimeout &> /dev/null; then
	TIMEOUT_CMD="gtimeout"
	else
	TIMEOUT_CMD="timeout"
	fi

	# WHICH EXAMPLES NEED WHICH DATASET. Same grouping the download step
	# uses to decide what to fetch, kept in the same file so the two
	# cannot drift into disagreeing about what a dataset is for.
	needs_dataset() { # <example.py> -> prints the label, or nothing
	case "$1" in
	04_csv_import_documents.py\|05_csv_import_graph.py\|06_vector_search_recommendations.py)
	echo "movielens-small" ;;
	07_stackoverflow_tables_oltp.py\|08_stackoverflow_tables_olap.py\|09_stackoverflow_graph_oltp.py\|10_stackoverflow_graph_olap.py\|13_stackoverflow_hybrid_queries.py)
	echo "stackoverflow-tiny-novec" ;;
	11_vector_index_build.py\|12_vector_search.py)
	echo "stackoverflow-tiny" ;;
	esac
	}

	# Run each example
	for example in $examples; do
	total=$((total + 1))

	# Skip, do not fail, when the dataset host was unreachable. Counted
	# as skipped and named, so a reader can tell "not run" from "passed".
	need=$(needs_dataset "$(basename "$example")")
	if [ -n "$need" ] && echo " ${SKIPPED_DATASETS:-} " \| grep -q " $need "; then
	echo "SKIP $example - dataset '$need' unavailable (third-party host)"
	skipped=$((skipped + 1))
	continue
	fi

	# Set example-specific parameters and timeout
	case "$example" in
	"21_graph_analytical_view_sql.py")
	example_args="--base-cities 1200 --stale-growth-cities 300 --sync-growth-cities 200 --regions 12 --batch-size 300 --sample-limit 5"
	example_name="$example (reduced scale for CI)"
	timeout_duration=900 # 15 minutes
	example_jvm_args=""
	;;
	"22_numpy_bulk_io.py")
	example_args="--rows 50000 --points 100000 --vectors 5000"
	example_name="$example (reduced scale for CI)"
	timeout_duration=600 # 10 minutes
	example_jvm_args=""
	;;
	"04_csv_import_documents.py")
	example_args="--dataset movielens-small --export"
	example_name="$example (movielens-small dataset with export)"
	timeout_duration=900 # 15 minutes
	example_jvm_args="-Xmx8g -Xms8g"
	;;
	"05_csv_import_graph.py")
	example_args="--dataset movielens-small --method sql --import-jsonl ./exports/movielens_small_db.jsonl.tgz --export"
	example_name="$example (movielens-small dataset, embedded sql method, import/export)"
	timeout_duration=900 # 15 minutes
	example_jvm_args=""
	;;
	"06_vector_search_recommendations.py")
	example_args="--import-jsonl ./exports/movielens_graph_small_db.jsonl.tgz"
	example_name="$example (vector search, import from JSONL)"
	timeout_duration=900 # 15 minutes
	example_jvm_args=""
	;;
	"07_stackoverflow_tables_oltp.py")
	example_args="--dataset stackoverflow-tiny --db arcadedb_sql --threads 1 --transactions 1000 --batch-size 500 --mem-limit 2g --run-label ci07"
	example_name="$example (stackoverflow-tiny, arcadedb_sql, minimal oltp)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"08_stackoverflow_tables_olap.py")
	example_args="--dataset stackoverflow-tiny --db arcadedb_sql --threads 1 --batch-size 500 --query-runs 1 --query-order fixed --mem-limit 2g --run-label ci08"
	example_name="$example (stackoverflow-tiny, arcadedb_sql, minimal olap)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"09_stackoverflow_graph_oltp.py")
	example_args="--dataset stackoverflow-tiny --db arcadedb_cypher --threads 1 --transactions 1000 --batch-size 500 --mem-limit 2g --run-label ci09"
	example_name="$example (stackoverflow-tiny graph oltp, arcadedb cypher, minimal)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"10_stackoverflow_graph_olap.py")
	example_args="--dataset stackoverflow-tiny --db arcadedb_cypher --threads 1 --batch-size 500 --query-runs 1 --query-order fixed --mem-limit 2g --run-label ci10"
	example_name="$example (stackoverflow-tiny graph olap, arcadedb cypher, minimal)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"11_vector_index_build.py")
	example_args="--backend arcadedb_sql --dataset stackoverflow-tiny --threads 1 --mem-limit 2g --batch-size 500 --beam-width 100 --quantization NONE --run-label ci11_arcadedb_sql"
	example_name="$example (vector build, arcadedb_sql backend, minimal)"
	timeout_duration=1200
	example_jvm_args=""
	;;
	"12_vector_search.py")
	db_path=$(find ./my_test_databases -maxdepth 1 -type d -name 'backend=arcadedb_sql_dataset=stackoverflow-tiny_*run=ci11_arcadedb_sql' \| head -n 1)
	if [ -z "$db_path" ]; then
	echo "❌ Missing build output for 12_vector_search.py. Ensure 11_vector_index_build.py ran first." \| tee -a $results_file
	failed=$((failed + 1))
	echo ""
	continue
	fi
	example_args="--backend arcadedb_sql --dataset stackoverflow-tiny --db-path $db_path --ef-search-values 100 --k 10 --query-limit 100 --query-runs 1 --query-order fixed --threads 1 --mem-limit 2g --run-label ci12_arcadedb_sql"
	example_name="$example (vector search, arcadedb_sql backend, minimal)"
	timeout_duration=1200
	example_jvm_args=""
	;;
	"13_stackoverflow_hybrid_queries.py")
	example_args="--dataset stackoverflow-tiny --batch-size 500 --encode-batch-size 64 --top-k 5 --candidate-limit 100 --min-reputation 100 --infer-sample-limit 5000 --run-label ci13"
	example_name="$example (hybrid pipeline, tiny + minimal limits)"
	timeout_duration=1800
	example_jvm_args=""
	;;
	"14_lifecycle_timing.py")
	example_args="--runs 1 --table-records 2000 --graph-vertices 500 --vector-records 500 --vector-dimensions 32 --query-runs 10 --jvm-heap 4g"
	example_name="$example (lifecycle benchmark, minimal)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"15_import_database_vs_transactional_table_ingest.py")
	example_args="--rows-per-table 2000 --tables 2 --columns 6 --string-size 32 --batch-size 500 --async-parallel 1 --parallel 1 --heap-size 4g --work-dir ./my_test_databases/import_vs_txn_dummy_ci"
	example_name="$example (table ingest benchmark, minimal)"
	timeout_duration=1200
	example_jvm_args=""
	;;
	"16_import_database_vs_transactional_graph_ingest.py")
	example_args="--vertices 2000 --edges 4000 --vertex-int-props 3 --vertex-str-props 2 --edge-int-props 1 --edge-str-props 1 --string-size 32 --batch-size 500 --async-parallel 1 --parallel 1 --heap-size 4g --work-dir ./my_test_databases/import_vs_txn_graph_ci"
	example_name="$example (graph ingest benchmark, minimal)"
	timeout_duration=1200
	example_jvm_args=""
	;;
	"17_timeseries_end_to_end.py")
	example_args="--hours 2 --interval-minutes 10"
	example_name="$example (time-series end-to-end, minimal)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"18_geo_predicates_wkt.py")
	example_args=""
	example_name="$example (geo predicates with WKT points and polygons)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"19_hash_index_exact_match.py")
	example_args=""
	example_name="$example (hash index exact-match workflow)"
	timeout_duration=900
	example_jvm_args=""
	;;
	"20_graph_algorithms_route_planning.py")
	example_args=""
	example_name="$example (graph algorithms route planning workflow)"
	timeout_duration=900
	example_jvm_args=""
	;;
	*)
	example_args=""
	example_name="$example"
	timeout_duration=900 # 15 minutes default
	example_jvm_args=""
	;;
	esac

	log_file="${example%.py}.log"

	echo "----------------------------------------"
	echo "📝 Running: $example_name"
	echo "----------------------------------------"

	# Run the example with appropriate parameters
	if [ -n "$example_jvm_args" ]; then
	if ARCADEDB_JVM_ARGS="$example_jvm_args" JAVA_TOOL_OPTIONS="" _JAVA_OPTIONS="" $TIMEOUT_CMD $timeout_duration python "$example" $example_args > "$log_file" 2>&1; then
	exit_code=0
	else
	exit_code=$?
	fi
	else
	if ARCADEDB_JVM_ARGS="" JAVA_TOOL_OPTIONS="" _JAVA_OPTIONS="" $TIMEOUT_CMD $timeout_duration python "$example" $example_args > "$log_file" 2>&1; then
	exit_code=0
	else
	exit_code=$?
	fi
	fi
	# ANSWERS, NOT ONLY A CLEAN EXIT (humemai/arcadedb-embedded-python#12). Example 10 once returned 0 rows for
	# 5 of its 10 queries on every platform and passed. After a clean run, the
	# same arguments on the pure-Python reference backend must give the same
	# row count and hash for every query.
	if [ $exit_code -eq 0 ] && [ "$example" = "10_stackoverflow_graph_olap.py" ]; then
	ref_args="${example_args/--db arcadedb_cypher/--db python_memory}"
	ref_args="${ref_args/--run-label ci10/--run-label ci10ref}"
	ref_log="${example%.py}-reference.log"
	if ! $TIMEOUT_CMD $timeout_duration python "$example" $ref_args > "$ref_log" 2>&1; then
	echo "❌ reference run (python_memory) of $example failed; last 20 lines:"
	tail -n 20 "$ref_log"
	exit_code=1
	elif ! python scripts/compare_query_hashes.py "$log_file" "$ref_log"; then
	echo "❌ $example: answers differ from the python_memory reference (above)"
	exit_code=1
	fi
	fi
	if [ $exit_code -eq 0 ]; then
	echo "✅ PASSED: $example_name" \| tee -a $results_file
	passed=$((passed + 1))
	else
	if [ $exit_code -eq 124 ]; then
	echo "⏱️ TIMEOUT: $example_name (exceeded $((timeout_duration/60)) minutes)" \| tee -a $results_file
	failed=$((failed + 1))
	else
	echo "❌ FAILED: $example_name (exit code: $exit_code)" \| tee -a $results_file
	failed=$((failed + 1))
	fi
	# Show last 20 lines of error log
	echo "Last 20 lines of output:"
	tail -n 20 "$log_file"
	fi
	echo ""
	done

	# Print summary
	echo "========================================"
	echo "📊 EXAMPLE TEST SUMMARY"
	echo "========================================"
	echo "Total: $total"
	echo "Passed: $passed ✅"
	echo "Failed: $failed ❌"
	echo "Skipped: $skipped ⏭️"
	echo "========================================"
	echo ""

	echo "Examples pattern(s): $EXAMPLES"

	# Output to GitHub Actions
	echo "total=$total" >> $GITHUB_OUTPUT
	echo "passed=$passed" >> $GITHUB_OUTPUT
	echo "failed=$failed" >> $GITHUB_OUTPUT
	echo "skipped=$skipped" >> $GITHUB_OUTPUT

	# Show detailed results
	echo "Detailed Results:"
	cat $results_file

	# Save list of examples for summary (before the failure exit, so
	# the summary is populated even when some examples fail)
	echo "$examples" > examples-ran.txt

	# Exit with error if any failed
	if [ $failed -gt 0 ]; then
	echo "❌ Some examples failed!"
	exit 1
	else
	echo "✅ All examples passed!"
	fi

	- name: Generate test summary
	if: always()
	shell: bash
	run: \|
	cd examples

	echo "## 🎮 Python Examples Test Results (${{ matrix.platform }})" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY

	total="${{ steps.run_examples.outputs.total \|\| '0' }}"
	passed="${{ steps.run_examples.outputs.passed \|\| '0' }}"
	failed="${{ steps.run_examples.outputs.failed \|\| '0' }}"

	if [ "${{ steps.run_examples.outcome }}" = "success" ]; then
	echo "✅ Status: ALL EXAMPLES PASSED ($passed/$total)" >> $GITHUB_STEP_SUMMARY
	else
	echo "❌ Status: SOME EXAMPLES FAILED ($passed/$total passed)" >> $GITHUB_STEP_SUMMARY
	fi

	echo "" >> $GITHUB_STEP_SUMMARY
	echo "### Summary" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	echo "\| Metric \| Count \|" >> $GITHUB_STEP_SUMMARY
	echo "\|--------\|------:\|" >> $GITHUB_STEP_SUMMARY
	echo "\| 📝 Total \| $total \|" >> $GITHUB_STEP_SUMMARY
	echo "\| ✅ Passed \| $passed \|" >> $GITHUB_STEP_SUMMARY
	echo "\| ❌ Failed \| $failed \|" >> $GITHUB_STEP_SUMMARY
	echo "\| ⏭️ Skipped \| ${{ steps.run_examples.outputs.skipped \|\| '0' }} \|" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY

	# Add detailed results if available
	if [ -f example-results.txt ]; then
	echo "### Detailed Results" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	echo "\`\`\`" >> $GITHUB_STEP_SUMMARY
	cat example-results.txt >> $GITHUB_STEP_SUMMARY
	echo "\`\`\`" >> $GITHUB_STEP_SUMMARY
	fi

	echo "" >> $GITHUB_STEP_SUMMARY
	echo "### Examples Tested" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	echo "Pattern(s): $EXAMPLES" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	if [ -f examples-ran.txt ]; then
	while IFS= read -r example; do
	[ -n "$example" ] && echo "- $example" >> $GITHUB_STEP_SUMMARY
	done < examples-ran.txt
	fi
	echo "" >> $GITHUB_STEP_SUMMARY

	- name: Upload example logs
	if: always()
	uses: actions/upload-artifact@cf430e030ddbb5b0abf93d22962f4752f3646cd9 # v7.0.2
	with:
	name: example-logs-${{ matrix.platform == 'linux/amd64' && 'linux-amd64' \|\| matrix.platform == 'linux/arm64' && 'linux-arm64' \|\| matrix.platform == 'darwin/amd64' && 'darwin-amd64' \|\| matrix.platform == 'darwin/arm64' && 'darwin-arm64' \|\| matrix.platform == 'windows/amd64' && 'windows-amd64' \|\| 'windows-arm64' }}-py${{ matrix.python-version }}
	path: \|
	examples/*.log
	examples/example-results.txt
	retention-days: 7

	- name: Upload example databases
	if: failure()
	uses: actions/upload-artifact@cf430e030ddbb5b0abf93d22962f4752f3646cd9 # v7.0.2
	with:
	name: example-databases-${{ matrix.platform == 'linux/amd64' && 'linux-amd64' \|\| matrix.platform == 'linux/arm64' && 'linux-arm64' \|\| matrix.platform == 'darwin/amd64' && 'darwin-amd64' \|\| matrix.platform == 'darwin/arm64' && 'darwin-arm64' \|\| matrix.platform == 'windows/amd64' && 'windows-amd64' \|\| 'windows-arm64' }}-py${{ matrix.python-version }}
	path: examples/my_test_databases/
	retention-days: 3

	# Summary job that checks all platforms
	test-examples-summary:
	name: Examples Test Summary
	needs: test-examples
	runs-on: ubuntu-latest
	# Skipped (not red) when the matrix was skipped for a draft PR.
	if: always() && (github.event_name != 'pull_request' \|\| github.event.pull_request.draft == false)
	steps:
	- name: Check test results
	shell: bash
	run: \|
	echo "## 🎯 Overall Examples Test Summary" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY

	if [ "${{ needs.test-examples.result }}" = "success" ]; then
	echo "✅ All platforms passed example testing!" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	echo "All examples ran successfully across all enabled platforms." >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	echo "Platforms tested: linux/amd64, linux/arm64, darwin/arm64, windows/amd64" >> $GITHUB_STEP_SUMMARY
	else
	echo "❌ Some platforms failed example testing" >> $GITHUB_STEP_SUMMARY
	echo "" >> $GITHUB_STEP_SUMMARY
	echo "Please check the individual platform jobs for details." >> $GITHUB_STEP_SUMMARY
	exit 1
	fi

Navigation Menu

Repository navigation

Native lookup_by_key raises without an index, as the JVM build does #213

Workflow file

Native lookup_by_key raises without an index, as the JVM build does #213

Uh oh!

Workflow file for this run