feat(preprocess): resume evidence jobs by run id

harness/tht/cli/preprocess_cmd.py

@@ -3,6 +3,8 @@
 from __future__ import annotations
 
 import json
+import re
+import hashlib
 from pathlib import Path
 
 import typer
@@ -25,9 +27,6 @@ def run_from_config(config: Path, *, dry_run: bool = False, resume: str | None =
     if cfg.embeddings is None:
         raise RuntimeError("embeddings are not configured")
     corpus_root = cfg.paths.artifacts.parent / "corpus"
-    generation = None
-    if resume:
-        generation = resume if resume.startswith("gen:") else f"gen:{resume}"
     pipeline = CorpusPipeline(
         store=CorpusStore(corpus_root), sources=build_evidence_sources(cfg),
         embedder=make_embedder(cfg.embeddings),
@@ -36,7 +35,17 @@ def run_from_config(config: Path, *, dry_run: bool = False, resume: str | None =
         chunk_policy=ChunkPolicy(version="chunk-v1", max_chars=cfg.vector.max_chunk_chars),
         pipeline_version="evidence-v1",
     )
-    return pipeline.run(dry_run=dry_run, resume=generation)
+    def fingerprint(value: str) -> str:
+        return "sha256:" + hashlib.sha256(value.encode()).hexdigest()
+
+    return pipeline.run_as_job(
+        workspace_id=config.stem.lower().replace(".", "-").replace("_", "-"),
+        workspace_root=corpus_root.parent,
+        config_fingerprint=fingerprint(cfg.model_dump_json()),
+        input_fingerprint=fingerprint(config.resolve().as_posix()),
+        dry_run=dry_run,
+        resume_run_id=resume,
+    )
 
 
 @preprocess_app.command("evidence")
@@ -46,6 +55,13 @@ def evidence_cmd(
     resume: str | None = typer.Option(None, "--resume"),
     json_output: bool = typer.Option(False, "--json"),
 ) -> None:
+    if resume is not None and re.fullmatch(r"[0-9a-f]{32}", resume) is None:
+        payload = {"status": "failed", "error": "resume requires a preprocessing run id"}
+        if json_output:
+            typer.echo(json.dumps(payload, sort_keys=True))
+        else:
+            typer.secho("ERRORE: resume requires a preprocessing run id", fg=typer.colors.RED, err=True)
+        raise typer.Exit(code=2)
     try:
         result = run_from_config(config, dry_run=dry_run, resume=resume)
     except Exception:
@@ -60,6 +76,6 @@ def evidence_cmd(
         typer.echo(json.dumps(payload, ensure_ascii=False, sort_keys=True))
     else:
         typer.echo(
-            f"OK: generation={payload['generation']} changed={len(payload['changed'])} "
+            f"OK: run={payload['run_id']} generation={payload['generation']} changed={len(payload['changed'])} "
             f"unchanged={len(payload['unchanged'])} removed={len(payload['removed'])}"
         )