feat: add P2 harness machine contracts

harness/tht/cli/preprocess_cmd.py

@@ -1,17 +1,18 @@
 """One-shot preprocessing commands."""
+# ruff: noqa: BLE001
 
 from __future__ import annotations
 
+import hashlib
 import json
 import re
-import hashlib
 from pathlib import Path
 
 import typer
 
 from tht.cli.config_cmd import CONFIG_OPT
-from tht.config import workspace_id_from_path
-
+from tht.cli.schema_cmd import _load_config_or_exit
+from tht.config import workspace_id_for_config
 
 preprocess_app = typer.Typer(help="Materialize versioned preprocessing artifacts")
 
@@ -22,7 +23,8 @@ def run_dwh_from_config(
     from tht.cli.lsh_cmd import build_lsh_artifacts
     from tht.cli.schema_cmd import _load_config_or_exit, refresh_catalog
     from tht.jobs.dwh_pipeline import (
-        DwhPreprocessPipeline, config_dwh_binding,
+        DwhPreprocessPipeline,
+        config_dwh_binding,
     )
 
     cfg = _load_config_or_exit(config)
@@ -64,7 +66,6 @@ def _parse_dwh_steps(value: str) -> tuple[str, ...]:
 
 def run_from_config(config: Path, *, dry_run: bool = False, resume: str | None = None):
     from tht.adapters.factory import build_evidence_sources, build_vector_store
-    from tht.cli.schema_cmd import _load_config_or_exit
     from tht.cli.vector_cmd import make_embedder
     from tht.corpus.chunk import ChunkPolicy
     from tht.corpus.pipeline import CorpusPipeline
@@ -87,7 +88,7 @@ def run_from_config(config: Path, *, dry_run: bool = False, resume: str | None =
         return "sha256:" + hashlib.sha256(value.encode()).hexdigest()
 
     return pipeline.run_as_job(
-        workspace_id=workspace_id_from_path(config),
+        workspace_id=workspace_id_for_config(cfg, config),
         workspace_root=corpus_root.parent,
         config_fingerprint=fingerprint(cfg.model_dump_json()),
         input_fingerprint=fingerprint(config.resolve().as_posix()),
@@ -116,7 +117,7 @@ def gc_from_config(config: Path, *, dry_run: bool = False):
         pipeline_version="evidence-v1",
         retain_published_generations=cfg.vector.retain_published_generations,
     )
-    pipeline.workspace_id = workspace_id_from_path(config)
+    pipeline.workspace_id = workspace_id_for_config(cfg, config)
     return pipeline.gc(workspace_root=corpus_root.parent, dry_run=dry_run)