Fix trace reuse and packaged model assets

tests/test_ali_trace_pipeline.py

@@ -1,5 +1,8 @@
 import json
+import shutil
+import subprocess
 import tempfile
+import tomllib
 import unittest
 from datetime import datetime, timezone
 from pathlib import Path
@@ -655,6 +658,33 @@ class AliTracePipelineTest(unittest.TestCase):
 
             self.assertEqual(Path(captured["dir"]), output_dir)
 
+    def test_format_and_sort_trace_supports_zstd_input_during_time_inference(self):
+        if shutil.which("zstd") is None or shutil.which("zstdcat") is None:
+            self.skipTest("zstd/zstdcat are required for .jsonl.zst formatter smoke test")
+        with tempfile.TemporaryDirectory() as temp_dir:
+            root = Path(temp_dir)
+            raw_path = root / "0417-1500-1530.jsonl"
+            zst_path = root / "0417-1500-1530.jsonl.zst"
+            output_path = root / "formatted" / "trace-raw.jsonl"
+            raw_path.write_text(
+                json.dumps(
+                    make_raw_row(
+                        "req-zst",
+                        utc_ms("2026-04-17 15:00:01.000"),
+                        time_text="2026-04-17 15:00:01.000",
+                    )
+                )
+                + "\n",
+                encoding="utf-8",
+            )
+            subprocess.run(["zstd", "-q", str(raw_path), "-o", str(zst_path)], check=True)
+
+            stats = format_and_sort_trace(input_dir=zst_path, output_path=output_path, chunk_bytes=256)
+
+            self.assertEqual(stats["row_count"], 1)
+            formatted_rows = [json.loads(line) for line in output_path.read_text(encoding="utf-8").splitlines()]
+            self.assertEqual([row["meta"]["request_id"] for row in formatted_rows], ["req-zst"])
+
     def test_export_release_ready_trace_defaults_temp_dir_to_output_parent(self):
         with tempfile.TemporaryDirectory() as temp_dir:
             root = Path(temp_dir)
@@ -855,3 +885,57 @@ class AliTracePipelineTest(unittest.TestCase):
                 (analysis_dir / "details" / "details_summary.json").stat().st_mtime_ns,
                 details_summary_mtime_ns,
             )
+
+    def test_trace_analyzer_rebuilds_when_same_output_gets_different_input(self):
+        with tempfile.TemporaryDirectory() as temp_dir:
+            root = Path(temp_dir)
+            formatted_root = root / "outputs" / "formatted"
+            analysis_root = root / "outputs" / "analysis"
+
+            def build_trace(label: str, request_id: str, ready: str) -> Path:
+                raw_dir = root / label
+                raw_dir.mkdir()
+                raw_path = raw_dir / "0417-1500-1530.jsonl"
+                raw_path.write_text(
+                    json.dumps(make_raw_row(request_id, utc_ms(ready), user_id=f"user-{request_id}")) + "\n",
+                    encoding="utf-8",
+                )
+                self.assertEqual(formatter_main(["format", str(raw_dir), "--output-root", str(formatted_root)]), 0)
+                formatted_path = formatted_root / "041715-041715-raw.jsonl"
+                renamed_path = formatted_root / f"{label}-raw.jsonl"
+                formatted_path.replace(renamed_path)
+                self.assertEqual(formatter_main(["build-release", str(renamed_path), "--jobs", "1", "--block-size", "8"]), 0)
+                return renamed_path
+
+            first_path = build_trace("first", "req-first", "2026-04-17 15:00:01.000")
+            second_path = build_trace("second", "req-second", "2026-04-17 15:00:02.000")
+
+            common_args = [
+                "--output-root",
+                str(analysis_root),
+                "--dataset-name",
+                "same-dataset",
+                "--segment-mode",
+                "bytes",
+                "--block-size",
+                "8",
+            ]
+            self.assertEqual(analyzer_main(["analyze", str(first_path), *common_args]), 0)
+            analysis_dir = analysis_root / "same-dataset"
+            self.assertIn("req-first", (analysis_dir / "features.csv").read_text(encoding="utf-8"))
+
+            self.assertEqual(analyzer_main(["analyze", str(second_path), *common_args]), 0)
+
+            features_text = (analysis_dir / "features.csv").read_text(encoding="utf-8")
+            self.assertIn("req-second", features_text)
+            self.assertNotIn("req-first", features_text)
+            details_summary = json.loads((analysis_dir / "details" / "details_summary.json").read_text(encoding="utf-8"))
+            self.assertTrue(str(details_summary["release_path"]).endswith("second.jsonl"))
+
+    def test_pyproject_includes_trace_model_meta_package_data(self):
+        pyproject = tomllib.loads(Path("pyproject.toml").read_text(encoding="utf-8"))
+        package_data = pyproject["tool"]["setuptools"]["package-data"]
+
+        self.assertIn("trace_model_meta", package_data)
+        self.assertIn("**/*.json", package_data["trace_model_meta"])
+        self.assertIn("**/*.jinja", package_data["trace_model_meta"])