arango-entity-resolution 3.3.0__tar.gz → 3.8.0__tar.gz

{arango_entity_resolution-3.3.0 → arango_entity_resolution-3.8.0}/.githooks/pre-commit

@@ -29,13 +29,20 @@ fi
 # Set Python path
 export PYTHONPATH="${PWD}/src:${PYTHONPATH}"
 
+# Prefer the project virtualenv (has dependencies); fall back to system python3
+if [ -x ".venv/bin/python" ]; then
+    PYTHON=".venv/bin/python"
+else
+    PYTHON="python3"
+fi
+
 # Run quick smoke tests
 echo "[1/4] Checking Python syntax..."
-python3 -m py_compile src/entity_resolution/utils/config.py || {
+"$PYTHON" -m py_compile src/entity_resolution/utils/config.py || {
     echo -e "${RED}[FAIL] Syntax error in config.py${NC}"
     exit 1
 }
-python3 -m py_compile src/entity_resolution/utils/constants.py || {
+"$PYTHON" -m py_compile src/entity_resolution/utils/constants.py || {
     echo -e "${RED}[FAIL] Syntax error in constants.py${NC}"
     exit 1
 }
@@ -56,8 +63,25 @@ echo -e "${GREEN}[OK] No hardcoded credentials found${NC}"
 
 echo ""
 echo "[3/4] Checking for emoji characters..."
-# Check Python files only (not JSON data)
-if grep -r "[\x{2713}\x{2714}\x{2715}\x{274C}\x{26A0}\x{1F389}\x{1F44D}]" src/ demo/*.py demo/scripts/*.py 2>/dev/null | grep -v "Binary"; then
+# Check Python source files only (portable: BSD grep does not support \x{...} classes)
+if "$PYTHON" - << 'EMOJICHECK'
+import pathlib, sys
+emoji = {"✓", "✔", "✕", "❌", "⚠", "\U0001F389", "\U0001F44D"}
+hits = []
+for path in pathlib.Path("src").rglob("*.py"):
+    try:
+        text = path.read_text(encoding="utf-8", errors="ignore")
+    except OSError:
+        continue
+    if any(ch in text for ch in emoji):
+        hits.append(str(path))
+if hits:
+    print("\n".join(hits))
+    sys.exit(1)
+EMOJICHECK
+then
+    : # clean
+else
     echo -e "${YELLOW}[WARNING] Found emoji characters (should use ASCII only)${NC}"
     echo "Consider replacing emojis with [OK], [ERROR], [WARNING] indicators"
     # Don't fail, just warn
@@ -66,7 +90,7 @@ echo -e "${GREEN}[OK] Code style check passed${NC}"
 
 echo ""
 echo "[4/4] Running critical module imports..."
-python3 << 'PYTEST'
+"$PYTHON" << 'PYTEST'
 import sys
 import os
 

{arango_entity_resolution-3.3.0 → arango_entity_resolution-3.8.0}/.githooks/pre-push

@@ -32,6 +32,13 @@ fi
 # Set Python path
 export PYTHONPATH="${PWD}/src:${PYTHONPATH:-}"
 
+# Prefer the project virtualenv (has dependencies); fall back to system python3
+if [ -x ".venv/bin/python" ]; then
+    PYTHON=".venv/bin/python"
+else
+    PYTHON="python3"
+fi
+
 # Check if we should skip tests (for emergency pushes)
 if [ -n "${SKIP_TESTS:-}" ]; then
     echo -e "${YELLOW}[WARNING] Skipping tests (SKIP_TESTS environment variable set)${NC}"
@@ -67,13 +74,13 @@ if command -v docker >/dev/null 2>&1 && command -v curl >/dev/null 2>&1; then
     fi
 else
     echo -e "${YELLOW}[WARNING] docker and/or curl not found; running unit tests only${NC}"
-    if ! run_test_section "Unit Tests (no Docker)" "python3 -m pytest -q -m \"not integration and not performance\" 2>&1 | tail -50"; then
+    if ! run_test_section "Unit Tests (no Docker)" "\"$PYTHON\" -m pytest -q -m \"not integration and not performance\" 2>&1 | tail -50"; then
         FAILED_TESTS+=("Unit Tests (no Docker)")
     fi
 fi
 
 # Test 4: Import Verification
-if ! run_test_section "Module Imports" "python3 << 'PYTEST'
+if ! run_test_section "Module Imports" "\"$PYTHON\" << 'PYTEST'
 import sys
 import warnings
 warnings.filterwarnings('ignore')
@@ -116,7 +123,7 @@ PYTEST"; then
 fi
 
 # Test 5: Code Quality (Quick Check)
-if ! run_test_section "Code Quality" "python3 -m py_compile src/entity_resolution/utils/config.py src/entity_resolution/utils/constants.py src/entity_resolution/core/entity_resolver.py 2>&1 | head -10"; then
+if ! run_test_section "Code Quality" "\"$PYTHON\" -m py_compile src/entity_resolution/utils/config.py src/entity_resolution/utils/constants.py src/entity_resolution/core/entity_resolver.py 2>&1 | head -10"; then
     FAILED_TESTS+=("Code Quality")
 fi
 

{arango_entity_resolution-3.3.0 → arango_entity_resolution-3.8.0}/.github/workflows/publish.yml

@@ -24,7 +24,7 @@ jobs:
       run: |
         python -m pip install --upgrade pip
         python -m pip install build twine
-        python -m pip install -e ".[test,ml,llm,mcp]"
+        python -m pip install -e ".[test,ml,llm,mcp,ui]"
         
     - name: Run unit tests
       env:

{arango_entity_resolution-3.3.0 → arango_entity_resolution-3.8.0}/.github/workflows/python-package.yml

@@ -15,17 +15,17 @@ jobs:
     steps:
     - uses: actions/checkout@v4
     
-    - name: Set up Python 3.10
+    - name: Set up Python 3.11
       uses: actions/setup-python@v5
       with:
-        python-version: "3.10"
+        python-version: "3.11"
         cache: 'pip'
         
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
         python -m pip install build
-        python -m pip install -e ".[test,ml,llm,mcp]"
+        python -m pip install -e ".[test,ml,llm,mcp,ui]"
         
     - name: Build wheel and source distribution
       run: python -m build

arango_entity_resolution-3.8.0/.github/workflows/runtime-platform-matrix.yml

@@ -0,0 +1,279 @@
+name: Runtime Platform Matrix
+
+on:
+  pull_request:
+    branches: ["main"]
+    paths:
+      - ".github/workflows/runtime-platform-matrix.yml"
+      - "scripts/summarize_runtime_activation.py"
+      - "src/entity_resolution/cli.py"
+      - "src/entity_resolution/services/runtime_*.py"
+      - "tests/test_cli.py"
+      - "tests/test_runtime_*.py"
+      - "docs/development/RUNTIME_HEALTH_CI_RUNBOOK.md"
+      - "ci/runtime-quality/**"
+  push:
+    branches: ["main"]
+    paths:
+      - ".github/workflows/runtime-platform-matrix.yml"
+      - "scripts/summarize_runtime_activation.py"
+      - "src/entity_resolution/cli.py"
+      - "src/entity_resolution/services/runtime_*.py"
+      - "tests/test_cli.py"
+      - "tests/test_runtime_*.py"
+      - "docs/development/RUNTIME_HEALTH_CI_RUNBOOK.md"
+      - "ci/runtime-quality/**"
+  workflow_dispatch:
+    inputs:
+      enable_self_hosted:
+        description: "Enable Apple Silicon and Linux GPU self-hosted jobs"
+        required: false
+        default: false
+        type: boolean
+
+permissions:
+  contents: read
+
+concurrency:
+  group: runtime-platform-matrix-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  build-matrix:
+    name: Build test matrix
+    runs-on: ubuntu-latest
+    outputs:
+      matrix: ${{ steps.set.outputs.matrix }}
+    steps:
+      - name: Compute matrix include list
+        id: set
+        run: |
+          BASE='[{"name":"linux-cpu","runs_on":"[\"ubuntu-latest\"]","self_hosted":false,"platform_tag":"linux-cpu","run_quality_gate":true}]'
+          FULL='[{"name":"linux-cpu","runs_on":"[\"ubuntu-latest\"]","self_hosted":false,"platform_tag":"linux-cpu","run_quality_gate":true},{"name":"apple-silicon","runs_on":"[\"self-hosted\",\"macOS\",\"ARM64\"]","self_hosted":true,"platform_tag":"apple-silicon","run_quality_gate":false},{"name":"linux-gpu","runs_on":"[\"self-hosted\",\"linux\",\"x64\",\"gpu\",\"nvidia\"]","self_hosted":true,"platform_tag":"linux-gpu","run_quality_gate":false}]'
+
+          if [[ "${{ github.event_name }}" == "workflow_dispatch" && "${{ github.event.inputs.enable_self_hosted }}" == "true" ]]; then
+            echo "matrix={\"include\":$FULL}" >> "$GITHUB_OUTPUT"
+          else
+            echo "matrix={\"include\":$BASE}" >> "$GITHUB_OUTPUT"
+          fi
+
+  runtime-platform-tests:
+    name: ${{ matrix.name }}
+    needs: build-matrix
+    runs-on: ${{ fromJson(matrix.runs_on) }}
+    timeout-minutes: 30
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJson(needs.build-matrix.outputs.matrix) }}
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: "pip"
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install -e ".[test,ml,mcp]"
+
+      - name: Runtime CLI contract smoke checks
+        run: |
+          arango-er runtime-health-benchmark --help
+          arango-er runtime-health-gate --help
+          arango-er runtime-quality-policy-validate --help
+          arango-er runtime-activation-summary --help
+
+      - name: Runtime benchmark/gate unit checks
+        env:
+          USE_DEFAULT_PASSWORD: "true"
+        run: |
+          pytest -q tests/test_runtime_benchmark_service.py
+          pytest -q tests/test_cli.py -k "runtime_health_benchmark or runtime_health_gate"
+          pytest -q tests/test_runtime_quality_policy_service.py tests/test_runtime_quality_policy_ci.py
+
+      - name: Validate quality policy and artifact references
+        run: |
+          arango-er runtime-quality-policy-validate \
+            --policy-file ci/runtime-quality/quality_gate_policy.json
+
+      - name: Start ephemeral ArangoDB for quality gate
+        if: ${{ matrix.run_quality_gate }}
+        run: |
+          ARANGO_ROOT_PASSWORD="${ARANGO_ROOT_PASSWORD:-${ARANGO_PASSWORD:-$(python - <<'PY'
+          import secrets
+          print(secrets.token_urlsafe(24))
+          PY
+          )}}"
+          echo "ARANGO_ROOT_PASSWORD=$ARANGO_ROOT_PASSWORD" >> "$GITHUB_ENV"
+          echo "ARANGO_PASSWORD=$ARANGO_ROOT_PASSWORD" >> "$GITHUB_ENV"
+          docker run -d --name runtime-gate-arango -e ARANGO_ROOT_PASSWORD="$ARANGO_ROOT_PASSWORD" -p 8529:8529 arangodb:3.11
+          for i in {1..30}; do
+            if curl -fsS -u "root:${ARANGO_ROOT_PASSWORD}" "http://127.0.0.1:8529/_api/version" >/dev/null; then
+              echo "ArangoDB is ready"
+              exit 0
+            fi
+            sleep 2
+          done
+          echo "ArangoDB did not become ready in time" >&2
+          exit 1
+
+      - name: Resolve quality gate policy for lane
+        if: ${{ matrix.run_quality_gate }}
+        run: |
+          python - <<'PY'
+          import json
+          import os
+          from pathlib import Path
+
+          policy = json.loads(Path("ci/runtime-quality/quality_gate_policy.json").read_text(encoding="utf-8"))
+          profile = "${{ matrix.platform_tag }}"
+          cfg = policy["profiles"][profile]
+
+          env_path = Path(os.environ["GITHUB_ENV"])
+          with env_path.open("a", encoding="utf-8") as fh:
+              fh.write(f"QUALITY_LABEL={cfg['label']}\n")
+              fh.write(f"QUALITY_CORPUS={cfg['quality_corpus']}\n")
+              fh.write(f"QUALITY_BASELINE={cfg['quality_baseline_metrics']}\n")
+              fh.write(f"QUALITY_MODEL_NAME={cfg['quality_model_name']}\n")
+              fh.write(f"QUALITY_DEVICE={cfg['quality_device']}\n")
+              fh.write(f"QUALITY_BATCH_SIZE={cfg['quality_batch_size']}\n")
+              fh.write(f"QUALITY_COSINE_DRIFT_MAX={cfg['quality_cosine_drift_max']}\n")
+              fh.write(f"QUALITY_TOPK_OVERLAP_MIN={cfg['quality_topk_overlap_min']}\n")
+          PY
+
+      - name: Run corpus-driven runtime health gate
+        if: ${{ matrix.run_quality_gate }}
+        run: |
+          mkdir -p artifacts/runtime artifacts/quality
+          arango-er runtime-health-gate \
+            -c config/er_config.example.yaml \
+            --host 127.0.0.1 \
+            --port 8529 \
+            --username root \
+            --password "${ARANGO_ROOT_PASSWORD}" \
+            --database _system \
+            --registry-file artifacts/runtime/runtime_registry_${{ matrix.platform_tag }}.json \
+            --label "${QUALITY_LABEL}" \
+            --bootstrap-baseline \
+            --quality-corpus "${QUALITY_CORPUS}" \
+            --quality-model-name "${QUALITY_MODEL_NAME}" \
+            --quality-device "${QUALITY_DEVICE}" \
+            --quality-batch-size "${QUALITY_BATCH_SIZE}" \
+            --quality-baseline-metrics "${QUALITY_BASELINE}" \
+            --quality-cosine-drift-max "${QUALITY_COSINE_DRIFT_MAX}" \
+            --quality-topk-overlap-min "${QUALITY_TOPK_OVERLAP_MIN}" \
+            --fail-on-regression > artifacts/quality/quality_gate_${{ matrix.platform_tag }}.json
+
+      - name: Stop ephemeral ArangoDB
+        if: ${{ always() && matrix.run_quality_gate }}
+        run: |
+          docker rm -f runtime-gate-arango || true
+
+      - name: Collect runtime environment evidence
+        if: always()
+        run: |
+          mkdir -p artifacts/runtime
+          python - <<'PY'
+          import json
+          import platform
+          from pathlib import Path
+
+          payload = {
+              "platform_tag": "${{ matrix.platform_tag }}",
+              "python_version": platform.python_version(),
+              "system": platform.system(),
+              "machine": platform.machine(),
+              "processor": platform.processor(),
+          }
+
+          try:
+              import torch
+              payload["torch"] = {
+                  "available": True,
+                  "version": getattr(torch, "__version__", None),
+                  "cuda_available": bool(torch.cuda.is_available()),
+                  "mps_available": bool(
+                      hasattr(torch.backends, "mps")
+                      and torch.backends.mps.is_available()
+                  ),
+              }
+          except Exception as exc:
+              payload["torch"] = {"available": False, "error": str(exc)}
+
+          try:
+              import onnxruntime as ort
+              payload["onnxruntime"] = {
+                  "available": True,
+                  "version": ort.__version__,
+                  "providers": ort.get_available_providers(),
+              }
+          except Exception as exc:
+              payload["onnxruntime"] = {"available": False, "error": str(exc)}
+
+          out = Path("artifacts/runtime") / f"runtime_env_${{ matrix.platform_tag }}.json"
+          out.write_text(json.dumps(payload, indent=2), encoding="utf-8")
+          print(f"Wrote {out}")
+          PY
+
+      - name: Upload runtime matrix logs
+        if: always()
+        uses: actions/upload-artifact@v4
+        with:
+          name: runtime-matrix-${{ matrix.platform_tag }}
+          if-no-files-found: ignore
+          retention-days: 7
+          path: |
+            .pytest_cache
+            artifacts/runtime/runtime_env_${{ matrix.platform_tag }}.json
+            artifacts/runtime/runtime_registry_${{ matrix.platform_tag }}.json
+            artifacts/quality/quality_gate_${{ matrix.platform_tag }}.json
+
+  summarize-activation-evidence:
+    name: summarize-activation-evidence
+    runs-on: ubuntu-latest
+    needs: runtime-platform-tests
+    if: always()
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: "pip"
+
+      - name: Install package
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install -e .
+
+      - name: Download runtime matrix artifacts
+        continue-on-error: true
+        uses: actions/download-artifact@v4
+        with:
+          pattern: runtime-matrix-*
+          merge-multiple: true
+          path: artifacts
+
+      - name: Build activation evidence summary
+        run: |
+          arango-er runtime-activation-summary \
+            --artifacts-root artifacts \
+            --output-json artifacts/runtime/activation_summary.json \
+            --output-md artifacts/runtime/activation_summary.md
+
+      - name: Upload activation evidence summary
+        uses: actions/upload-artifact@v4
+        with:
+          name: runtime-activation-evidence
+          if-no-files-found: ignore
+          retention-days: 7
+          path: |
+            artifacts/runtime/activation_summary.json
+            artifacts/runtime/activation_summary.md

arango_entity_resolution-3.8.0/.github/workflows/ui-contract.yml

@@ -0,0 +1,66 @@
+name: UI Contract
+
+# Regenerates the OpenAPI schema and the frontend's TypeScript types, then
+# fails if the committed artifacts are stale. This prevents the frontend
+# (ui/src/api) and the FastAPI backend (src/entity_resolution/ui) from drifting
+# out of contract.
+
+on:
+  push:
+    branches: [ "main" ]
+  pull_request:
+    branches: [ "main" ]
+
+permissions:
+  contents: read
+
+jobs:
+  contract:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Python 3.11
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: 'pip'
+
+      - name: Install Python package (UI extra)
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install -e ".[ui]"
+
+      - name: Set up Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: "20"
+          cache: 'npm'
+          cache-dependency-path: ui/package-lock.json
+
+      - name: Install UI dependencies
+        working-directory: ui
+        run: npm ci
+
+      - name: Regenerate OpenAPI schema
+        run: python scripts/export_openapi.py
+
+      - name: Regenerate TypeScript types
+        working-directory: ui
+        run: npm run gen:types
+
+      - name: Fail if generated contract artifacts are stale
+        run: |
+          if ! git diff --exit-code -- ui/openapi.json ui/src/api/schema.ts; then
+            echo "::error::OpenAPI schema or generated TypeScript types are out of date."
+            echo "Run 'python scripts/export_openapi.py && (cd ui && npm run gen:types)' and commit the result."
+            exit 1
+          fi
+
+      - name: Typecheck UI
+        working-directory: ui
+        run: npm run typecheck
+
+      - name: Unit test UI (Vitest)
+        working-directory: ui
+        run: npm run test

{arango_entity_resolution-3.3.0 → arango_entity_resolution-3.8.0}/.gitignore

@@ -74,6 +74,8 @@ npm-debug.log*
 yarn-debug.log*
 yarn-error.log*
 package-lock.json
+# The UI lockfile IS committed so CI `npm ci` and the contract gate are reproducible.
+!ui/package-lock.json
 yarn.lock
 
 # Data files
@@ -128,4 +130,20 @@ docs/customer/**
 # Legacy code (archived implementations)
 # Foxx services and related code moved to legacy/ for historical reference
 # Not actively maintained or tested in v2.0+
-legacy/
+legacy/
+
+# UI frontend build artifacts (rebuilt from ui/ source)
+src/entity_resolution/ui/static/
+
+# Runtime artifacts (models, benchmarks, caches)
+artifacts/
+# Coverage and analysis artifacts (never track)
+coverage.json
+COVERAGE_SUMMARY.txt
+detailed_coverage_analysis_*.json
+
+# Local Claude Code harness config (per-user, not project-shared)
+.claude/
+.cursor/
+.prd-drift-queue/
+CLAUDE.md