cjm-transcript-decomp-core 0.0.1__tar.gz → 0.0.4__tar.gz

{cjm_transcript_decomp_core-0.0.1/cjm_transcript_decomp_core.egg-info → cjm_transcript_decomp_core-0.0.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cjm-transcript-decomp-core
-Version: 0.0.1
+Version: 0.0.4
 Summary: A frontend-agnostic core for the transcript decomposition workflow — composes isolated capability workers (forced alignment, VAD, graph storage) into a headless pipeline that decomposes transcription run manifests into a VAD-aligned context-graph spine with traceable provenance, and a CLI as its first driver.
 Author-email: "Christian J. Mills" <9126128+cj-mills@users.noreply.github.com>
 License: Apache-2.0
@@ -10,10 +10,10 @@ Classifier: Programming Language :: Python :: 3 :: Only
 Requires-Python: >=3.12
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: cjm_substrate>=0.0.51
-Requires-Dist: cjm_context_graph_primitives>=0.0.12
+Requires-Dist: cjm_substrate>=0.0.57
+Requires-Dist: cjm_context_graph_primitives>=0.0.15
 Requires-Dist: cjm_capability_primitives>=0.0.10
-Requires-Dist: cjm-context-graph-layer>=0.0.8
+Requires-Dist: cjm-context-graph-layer>=0.0.11
 Requires-Dist: cjm-transcript-graph-schema>=0.0.7
 Dynamic: license-file
 
@@ -26,15 +26,10 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 ## Modules
 
 - **`cjm_transcript_decomp_core.alignment`** — Pure forced-alignment logic (no capability calls): map FA words back to character spans in the original text, assign words to VAD chunks by timestamp, and build one text segment per VAD chunk. Extracted from the page-centric ForcedAlignmentService (Tier-1 logic).
-- **`cjm_transcript_decomp_core.cli`** — The CLI driver — the decomposition core's first (and currently only) frontend. Ships in-package as the cjm-transcript-decomp-core console script so the driver can never skew from the core. GUI presentation drivers come later and consume the same pipeline module; they never reimplement it (CLI-first / headless-core principle).
+- **`cjm_transcript_decomp_core.cli`** — The CLI driver — the decomposition core's first (and currently only) frontend.
 - **`cjm_transcript_decomp_core.graph`** — Graph-spine EXTENSION + skeptical-lens verification (stage 5, CR-18 revolution 2). Decomp no longer creates a Document: it RECOMPUTES the transcription-emitted root's deterministic node ids from the consumed manifest (no search), verifies the root exists, and attaches the fine Segment spine under the existing AudioSegment nodes — PART_OF to the owning rendition, STARTS_WITH per rendition (the coarse-seam jump anchor), source-wide NEXT. Each Segment carries the audio TimeSlice ref plus per-transcriber CharSlice refs into the Transcript nodes (the D4/P10 framing, finally expressible). Commit goes through the layer's idempotent extend_graph.
 - **`cjm_transcript_decomp_core.models`** — Lean data shapes for the transcript-decomposition pipeline: in-core mirrors of the forced-alignment / VAD / text DTOs (no FastHTML deps), run configuration, the committed graph-segment carrier, and the decomposition run manifest (proto-bundle).
 - **`cjm_transcript_decomp_core.pipeline`** — The headless decomposition pipeline (stage 5: decomp is an EXTENDER). Load a transcription run manifest, verify the transcription-emitted graph root exists (the graph begins at transcription), then per source per pipeline-segment run VAD + per-transcriber forced alignment, build one aligned segment per VAD chunk with per-transcriber text variants, and attach the fine spine under the existing AudioSegment nodes via the layer's idempotent extend_graph — with HITL approval seams between alignment, commit, and the next source.
-- **`tests.test_alignment`** — Tests for cjm_transcript_decomp_core.alignment — pure forced-alignment logic.
-- **`tests.test_graph`** — Tests for cjm_transcript_decomp_core.graph — fine-spine extension payload.
-- **`tests.test_models`** — Tests for cjm_transcript_decomp_core.models — decomposition data shapes.
-- **`tests_manual.measure_cold_fanout_overlap_e2e`** — Cold-run M×(VAD ∥ T×FA) fan-out overlap measurement (stage-5 closeout;
-- **`tests_manual.validate_stage7_volume_journal_e2e`** — Stage-7 stress part 6 — volume regression on the REAL corpus.
 
 ## API
 
@@ -50,7 +45,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `build_parser` _function_ — Build the CLI parser (subcommands: run).
 - `load_capabilities` _function_ — Discover manifests + load each requested capability (default instance).
 - `main` _function_ — CLI entry point (console script: `cjm-transcript-decomp-core`).
-- `run_command` _function_ — Execute the `run` subcommand: extend a transcription-emitted root with the fine spine.
+- `run_command` _function_ — Execute the `run` subcommand: extend transcription-run manifest(s) with the fine spine.
 
 ### `cjm_transcript_decomp_core.graph`
 
@@ -76,6 +71,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `build_alignment_composition` _function_ — Build the whole-source M×(VAD ∥ T×FA) composition (D8 fan-in, stage-5 variants).
 - `collect_capability_info` _function_ — Record capability identity + data-DB pointers for the run manifest (provenance).
 - `confirm_seam` _function_ — HITL approval seam in its cheapest viable form (log + optional CLI prompt).
+- `decomp_replay_handlers` _function_ — The decomp core's replay vocabulary (DEC 426658f1, replay stays DOMAIN-OWNED).
 - `decompose_source` _function_ — Decompose one source into aligned fine segments with per-transcriber variants.
 - `fa_words_from_result` _function_ — Normalize a typed forced-alignment result into FA words (pure; stage 3).
 - `load_source_manifest` _function_ — Load + lightly validate a transcription-core run manifest.
@@ -83,43 +79,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `submit_and_wait` _function_ — Submit one capability job, wait for it, and return its result (raise on failure).
 - `vad_chunks_from_result` _function_ — Normalize a typed VAD result into segment-local VAD chunks.
 
-### `tests.test_alignment`
-
-- `test_assign_words_to_chunks_by_timestamp` _function_
-- `test_build_segments_per_chunk` _function_
-- `test_empty_chunk_yields_empty_segment_warning` _function_
-- `test_map_fa_words_to_text_spans` _function_
-
-### `tests.test_graph`
-
-- `test_edge_topology_spans_coarse_boundary` _function_
-- `test_extension_payload_deterministic_ids` _function_
-- `test_preprocessed_manifest_distinct_renditions_same_boundaries` _function_
-- `test_provenance_refs_and_text_from` _function_
-- `test_resolve_root_ids_recomputation` _function_
-
-### `tests.test_models`
-
-- `test_faword_from_wire_tolerates_extra_keys` _function_
-- `test_manifest_save_round_trip` _function_
-- `test_manifest_shape_and_run_id` _function_
-- `test_stage5_segment_shape` _function_
-- `test_vad_chunk_duration` _function_
-
-### `tests_manual.measure_cold_fanout_overlap_e2e`
-
-- `lane_intervals` _function_ — (start, end) UTC intervals for completed jobs of one instance.
-- `main` _function_
-- `merge` _function_ — Merge overlapping intervals so a side can't double-count itself.
-- `overlap_seconds` _function_ — Total seconds where any interval of A overlaps any interval of B.
-- `self_overlap_max` _function_ — Max simultaneously in-flight jobs within one lane + seconds at depth >= 2.
-
-### `tests_manual.validate_stage7_volume_journal_e2e`
-
-- `checkpoint_copy` _function_ — Checkpoint-then-copy (the stage-3 G3 backup discipline — never copy
-- `journal_rows` _function_
-- `main` _function_
-
 ## Dependencies
 
-**Depends on:** `cjm-substrate`
+**Depends on:** `cjm-capability-primitives`, `cjm-context-graph-layer`, `cjm-context-graph-primitives`, `cjm-substrate`, `cjm-transcript-graph-schema`
+**Used by:** `cjm-transcript-decomp-tui`

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4}/README.md

@@ -7,15 +7,10 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 ## Modules
 
 - **`cjm_transcript_decomp_core.alignment`** — Pure forced-alignment logic (no capability calls): map FA words back to character spans in the original text, assign words to VAD chunks by timestamp, and build one text segment per VAD chunk. Extracted from the page-centric ForcedAlignmentService (Tier-1 logic).
-- **`cjm_transcript_decomp_core.cli`** — The CLI driver — the decomposition core's first (and currently only) frontend. Ships in-package as the cjm-transcript-decomp-core console script so the driver can never skew from the core. GUI presentation drivers come later and consume the same pipeline module; they never reimplement it (CLI-first / headless-core principle).
+- **`cjm_transcript_decomp_core.cli`** — The CLI driver — the decomposition core's first (and currently only) frontend.
 - **`cjm_transcript_decomp_core.graph`** — Graph-spine EXTENSION + skeptical-lens verification (stage 5, CR-18 revolution 2). Decomp no longer creates a Document: it RECOMPUTES the transcription-emitted root's deterministic node ids from the consumed manifest (no search), verifies the root exists, and attaches the fine Segment spine under the existing AudioSegment nodes — PART_OF to the owning rendition, STARTS_WITH per rendition (the coarse-seam jump anchor), source-wide NEXT. Each Segment carries the audio TimeSlice ref plus per-transcriber CharSlice refs into the Transcript nodes (the D4/P10 framing, finally expressible). Commit goes through the layer's idempotent extend_graph.
 - **`cjm_transcript_decomp_core.models`** — Lean data shapes for the transcript-decomposition pipeline: in-core mirrors of the forced-alignment / VAD / text DTOs (no FastHTML deps), run configuration, the committed graph-segment carrier, and the decomposition run manifest (proto-bundle).
 - **`cjm_transcript_decomp_core.pipeline`** — The headless decomposition pipeline (stage 5: decomp is an EXTENDER). Load a transcription run manifest, verify the transcription-emitted graph root exists (the graph begins at transcription), then per source per pipeline-segment run VAD + per-transcriber forced alignment, build one aligned segment per VAD chunk with per-transcriber text variants, and attach the fine spine under the existing AudioSegment nodes via the layer's idempotent extend_graph — with HITL approval seams between alignment, commit, and the next source.
-- **`tests.test_alignment`** — Tests for cjm_transcript_decomp_core.alignment — pure forced-alignment logic.
-- **`tests.test_graph`** — Tests for cjm_transcript_decomp_core.graph — fine-spine extension payload.
-- **`tests.test_models`** — Tests for cjm_transcript_decomp_core.models — decomposition data shapes.
-- **`tests_manual.measure_cold_fanout_overlap_e2e`** — Cold-run M×(VAD ∥ T×FA) fan-out overlap measurement (stage-5 closeout;
-- **`tests_manual.validate_stage7_volume_journal_e2e`** — Stage-7 stress part 6 — volume regression on the REAL corpus.
 
 ## API
 
@@ -31,7 +26,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `build_parser` _function_ — Build the CLI parser (subcommands: run).
 - `load_capabilities` _function_ — Discover manifests + load each requested capability (default instance).
 - `main` _function_ — CLI entry point (console script: `cjm-transcript-decomp-core`).
-- `run_command` _function_ — Execute the `run` subcommand: extend a transcription-emitted root with the fine spine.
+- `run_command` _function_ — Execute the `run` subcommand: extend transcription-run manifest(s) with the fine spine.
 
 ### `cjm_transcript_decomp_core.graph`
 
@@ -57,6 +52,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `build_alignment_composition` _function_ — Build the whole-source M×(VAD ∥ T×FA) composition (D8 fan-in, stage-5 variants).
 - `collect_capability_info` _function_ — Record capability identity + data-DB pointers for the run manifest (provenance).
 - `confirm_seam` _function_ — HITL approval seam in its cheapest viable form (log + optional CLI prompt).
+- `decomp_replay_handlers` _function_ — The decomp core's replay vocabulary (DEC 426658f1, replay stays DOMAIN-OWNED).
 - `decompose_source` _function_ — Decompose one source into aligned fine segments with per-transcriber variants.
 - `fa_words_from_result` _function_ — Normalize a typed forced-alignment result into FA words (pure; stage 3).
 - `load_source_manifest` _function_ — Load + lightly validate a transcription-core run manifest.
@@ -64,43 +60,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `submit_and_wait` _function_ — Submit one capability job, wait for it, and return its result (raise on failure).
 - `vad_chunks_from_result` _function_ — Normalize a typed VAD result into segment-local VAD chunks.
 
-### `tests.test_alignment`
-
-- `test_assign_words_to_chunks_by_timestamp` _function_
-- `test_build_segments_per_chunk` _function_
-- `test_empty_chunk_yields_empty_segment_warning` _function_
-- `test_map_fa_words_to_text_spans` _function_
-
-### `tests.test_graph`
-
-- `test_edge_topology_spans_coarse_boundary` _function_
-- `test_extension_payload_deterministic_ids` _function_
-- `test_preprocessed_manifest_distinct_renditions_same_boundaries` _function_
-- `test_provenance_refs_and_text_from` _function_
-- `test_resolve_root_ids_recomputation` _function_
-
-### `tests.test_models`
-
-- `test_faword_from_wire_tolerates_extra_keys` _function_
-- `test_manifest_save_round_trip` _function_
-- `test_manifest_shape_and_run_id` _function_
-- `test_stage5_segment_shape` _function_
-- `test_vad_chunk_duration` _function_
-
-### `tests_manual.measure_cold_fanout_overlap_e2e`
-
-- `lane_intervals` _function_ — (start, end) UTC intervals for completed jobs of one instance.
-- `main` _function_
-- `merge` _function_ — Merge overlapping intervals so a side can't double-count itself.
-- `overlap_seconds` _function_ — Total seconds where any interval of A overlaps any interval of B.
-- `self_overlap_max` _function_ — Max simultaneously in-flight jobs within one lane + seconds at depth >= 2.
-
-### `tests_manual.validate_stage7_volume_journal_e2e`
-
-- `checkpoint_copy` _function_ — Checkpoint-then-copy (the stage-3 G3 backup discipline — never copy
-- `journal_rows` _function_
-- `main` _function_
-
 ## Dependencies
 
-**Depends on:** `cjm-substrate`
+**Depends on:** `cjm-capability-primitives`, `cjm-context-graph-layer`, `cjm-context-graph-primitives`, `cjm-substrate`, `cjm-transcript-graph-schema`
+**Used by:** `cjm-transcript-decomp-tui`

cjm_transcript_decomp_core-0.0.4/cjm_transcript_decomp_core/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.0.4"

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4}/cjm_transcript_decomp_core/cli.py

@@ -10,9 +10,11 @@ Prerequisite runtime (once, from the repo root):
     cjm-ctl --cjm-config cjm.yaml setup-runtime
     cjm-ctl --cjm-config cjm.yaml install-all --capabilities capabilities_test.yaml --force
 
-Then decompose a transcription-core run manifest:
+Then decompose transcription-core run manifest(s) — a multi-manifest batch
+shares ONE loaded capability stack (models load once, per-manifest runs):
 
     cjm-transcript-decomp-core run path/to/transcription-run.json --yes
+    cjm-transcript-decomp-core run runs/run_a.json runs/run_b.json --yes
     # GPU runs: opt into CR-7 GPU subtree attribution
     cjm-transcript-decomp-core run run.json --yes --sysmon-capability cjm-capability-monitor-nvidia
 """
@@ -21,11 +23,13 @@ import argparse
 import asyncio
 import getpass
 import logging
+import os
 from pathlib import Path
 from typing import Any, Dict, List, Optional
 
 from cjm_substrate.core.manager import CapabilityManager
 from cjm_substrate.core.queue import JobQueue
+from cjm_substrate.core.workspace import resolve_workspace
 from cjm_transcript_decomp_core.models import DecompConfig
 from cjm_transcript_decomp_core.pipeline import load_source_manifest, run_decomp
 
@@ -41,7 +45,8 @@ def build_parser() -> argparse.ArgumentParser:  # Configured CLI parser
     sub = parser.add_subparsers(dest="command", required=True)
 
     run = sub.add_parser("run", help="Extend a transcription-emitted graph root with the fine spine")
-    run.add_argument("manifest", help="Transcription-core run manifest JSON (the source to decompose)")
+    run.add_argument("manifests", nargs="+",
+                     help="Transcription-core run manifest JSON(s) — a batch shares ONE loaded capability stack")
     run.add_argument("--manifests-dir", default=".cjm/manifests", help="Capability manifests directory")
     run.add_argument("--vad-capability", default="cjm-capability-silero-vad", help="VAD capability name")
     run.add_argument("--fa-capability", default="cjm-capability-qwen3-forced-aligner", help="Forced-alignment capability name")
@@ -56,7 +61,15 @@ def build_parser() -> argparse.ArgumentParser:  # Configured CLI parser
     run.add_argument("--language", default="English", help="Forced-alignment language")
     run.add_argument("--force", action="store_true", help="Bypass capability-side caches (VAD + FA)")
     run.add_argument("-y", "--yes", action="store_true", help="Auto-accept HITL seams (headless mode)")
-    run.add_argument("--output", default=None, help="Decomp-manifest output path (default: runs/<run_id>.json)")
+    run.add_argument("--output", default=None, help="Decomp-manifest output path (single-manifest runs only; default: <output-dir>/<run_id>.json)")
+    run.add_argument("--output-dir", default=None,
+                     help="Decomp-manifest output directory (callers with a shared runs dir "
+                          "pass it so manifests land beside their sources; default: the "
+                          "workspace's runs/ when one is active, else runs/ under the cwd)")
+    run.add_argument("--workspace", default=None,
+                     help="Workspace root (5daadfc4; default: CJM_WORKSPACE env, else upward walk "
+                          "from cwd). Supplies the --output-dir default and is exported so "
+                          "substrate config + capability workers resolve workspace-scoped paths")
     run.add_argument("--actor", default=None,
                      help="Journal attribution for who/what initiated this run (default: cli:<username>)")
     run.add_argument("-v", "--verbose", action="store_true", help="DEBUG-level logging")
@@ -85,11 +98,31 @@ def load_capabilities(
 
 async def run_command(
     args: argparse.Namespace,  # Parsed CLI arguments for the `run` subcommand
-) -> int:  # Process exit code (0 = all sources extended + verified)
-    """Execute the `run` subcommand: extend a transcription-emitted root with the fine spine."""
-    manifest_path = str(Path(args.manifest).resolve())
-    if not Path(manifest_path).exists():
-        raise SystemExit(f"source manifest not found: {manifest_path}")
+) -> int:  # Process exit code (0 = every manifest fully extended + verified)
+    """Execute the `run` subcommand: extend transcription-run manifest(s) with the fine spine.
+
+    Batch shape (decomp-TUI demand, work item 0ff6bf0f): N manifests ride ONE
+    capability stack — the models load once and every manifest reuses them,
+    where N CLI invocations would pay N model loads. Each manifest keeps its
+    own run_id, decomp manifest, and RUN_STARTED/RUN_FINISHED journal bracket,
+    so a batch member is indistinguishable from a solo run downstream. A failed
+    member records and the batch continues (unattended queueing must not sink
+    the rest); the exit code aggregates across members.
+    """
+    # 5daadfc4 workspace: resolve BEFORE any substrate config loads; export so
+    # the process tree (substrate config, capability workers) is workspace-scoped.
+    ws = resolve_workspace(explicit=getattr(args, "workspace", None))
+    if ws is not None:
+        os.environ["CJM_WORKSPACE"] = str(ws.root)
+    output_dir = (Path(args.output_dir) if args.output_dir
+                  else (ws.runs_dir if ws is not None else Path("runs")))
+    manifest_paths = [str(Path(m).resolve()) for m in args.manifests]
+    missing = [m for m in manifest_paths if not Path(m).exists()]
+    if missing:
+        raise SystemExit(f"source manifest(s) not found: {', '.join(missing)}")
+    if args.output and len(manifest_paths) > 1:
+        raise SystemExit("--output names ONE decomp manifest — omit it for batch runs "
+                         "(each lands at runs/<run_id>.json)")
 
     cfg = DecompConfig(
         vad_capability=args.vad_capability,
@@ -118,11 +151,28 @@ async def run_command(
 
     queue = JobQueue(deps=manager, sysmon_capability_name=args.sysmon_capability)
     await queue.start()
+    all_ok = True
     try:
         # CR-14 follow-up: actor attribution (operator identity by default;
         # agents/services pass --actor explicitly).
         actor = args.actor or f"cli:{getpass.getuser()}"
-        manifest = await run_decomp(manager, queue, cfg, manifest_path, actor=actor)
+        for mp in manifest_paths:
+            try:
+                manifest = await run_decomp(manager, queue, cfg, mp, actor=actor)
+            except Exception as e:  # One member's failure must not sink the batch
+                logger.error(f"decomp failed for {mp}: {e}")
+                print(f"FAILED {mp}: {e}")
+                all_ok = False
+                continue
+            out = (Path(args.output) if args.output
+                   else output_dir / f"{manifest.run_id}.json")
+            manifest.save(out, workspace=ws)
+            n_manifest_sources = len(load_source_manifest(mp).get("sources", []) or [])
+            n_sources = len(manifest.sources)
+            n_segs = sum(s.segment_count for s in manifest.sources)
+            print(f"decomp manifest: {out}")
+            print(f"sources extended: {n_sources}/{n_manifest_sources}  segment nodes: {n_segs}")
+            all_ok = all_ok and (n_sources == n_manifest_sources)
     finally:
         await queue.stop()
         for iid in reversed(load_order):  # Reverse load order; the monitor unloads last
@@ -130,15 +180,7 @@ async def run_command(
                 manager.unload_capability(iid)
             except Exception as e:  # Best-effort teardown; never mask the run's outcome
                 logger.warning(f"unload {iid} failed: {e}")
-
-    out = Path(args.output) if args.output else Path("runs") / f"{manifest.run_id}.json"
-    manifest.save(out)
-    n_manifest_sources = len(load_source_manifest(manifest_path).get("sources", []) or [])
-    n_sources = len(manifest.sources)
-    n_segs = sum(s.segment_count for s in manifest.sources)
-    print(f"decomp manifest: {out}")
-    print(f"sources extended: {n_sources}/{n_manifest_sources}  segment nodes: {n_segs}")
-    return 0 if n_sources == n_manifest_sources else 1
+    return 0 if all_ok else 1
 
 
 def main(

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4}/cjm_transcript_decomp_core/models.py

@@ -7,6 +7,8 @@ from dataclasses import asdict, dataclass, field
 from pathlib import Path
 from typing import Any, Dict, List, Optional, Union
 
+from cjm_substrate.core.workspace import relativize_recorded
+
 
 @dataclass
 class FAWord:
@@ -173,11 +175,16 @@ class DecompManifest:
     def save(
         self,
         path: Union[str, Path],  # Destination JSON file (parent dirs created)
+        workspace=None,  # Active Workspace; owned paths record as ${WS}/<rel> (5daadfc4 rung f)
     ) -> Path:  # The written path
-        """Write the manifest as pretty-printed JSON."""
+        """Write the manifest as pretty-printed JSON.
+
+        With `workspace`, recorded paths under its root take the ${WS}/ token
+        form (relativize_recorded), so the manifest relocates with the
+        workspace; readers resolve via resolve_recorded_tree at load."""
         out = Path(path)
         out.parent.mkdir(parents=True, exist_ok=True)
-        out.write_text(json.dumps(self.to_dict(), indent=2))
+        out.write_text(json.dumps(relativize_recorded(self.to_dict(), workspace), indent=2))
         return out
 
 

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4}/cjm_transcript_decomp_core/pipeline.py

@@ -10,12 +10,14 @@ from typing import Any, Dict, List, Optional, Tuple
 from cjm_capability_primitives.forced_alignment import ForcedAlignResult
 from cjm_capability_primitives.vad import VADResult
 from cjm_context_graph_layer.declare import Derivation, derivation_to_graph
-from cjm_context_graph_layer.ops import extend_graph, graph_task
+from cjm_context_graph_layer.journal import journal_extend, sidecar_journal_path, wires_handlers
+from cjm_context_graph_layer.ops import graph_task
 from cjm_substrate.core.empirical_store import compute_config_hash
 from cjm_substrate.core.journal_store import JournalEvent, SubstrateEventType
 from cjm_substrate.core.manager import CapabilityManager
 from cjm_substrate.core.ports import Composition, CompositionNode, NodeState
 from cjm_substrate.core.queue import JobQueue, JobStatus
+from cjm_substrate.core.workspace import resolve_recorded_tree
 from cjm_transcript_decomp_core.alignment import (assign_words_to_chunks,
                                                   build_segments_from_alignment,
                                                   map_fa_words_to_text, tier1_alignment_checks)
@@ -58,8 +60,11 @@ def load_source_manifest(
 
     Consumed as untyped JSON — the format/version tags are the only interchange
     contract (no shared schema type across cores; manifest-as-interchange, CR-20).
+    ${WS}/ recorded paths (5daadfc4 rung f) resolve to absolute here, anchored
+    at the manifest's own location (active workspace as fallback), so all
+    downstream consumers keep seeing absolute paths.
     """
-    data = json.loads(Path(path).read_text())
+    data = resolve_recorded_tree(json.loads(Path(path).read_text()), Path(path))
     fmt = data.get("format", "")
     if "transcription-core" not in fmt:
         logger.warning(f"unexpected source manifest format: {fmt!r} (continuing)")
@@ -380,6 +385,10 @@ async def run_decomp(
         capabilities=collect_capability_info(manager, [cfg.vad_capability, cfg.fa_capability, cfg.graph_capability]),
     )
     vad_config_hash = str((manifest.capabilities.get(cfg.vad_capability) or {}).get("config_hash") or "")
+    # Pipeline writes append through to the graph db's SIDECAR journal (DEC ccbab9f5 /
+    # finding 4219da27): unjournaled ingestion would erode db-rebuildability.
+    graph_db_path = (manifest.capabilities.get(cfg.graph_capability) or {}).get("db_path")
+    graph_journal_path = sidecar_journal_path(str(graph_db_path)) if graph_db_path else None
 
     sources = src.get("sources", []) or []
     status = "completed"
@@ -419,7 +428,12 @@ async def run_decomp(
                 status = "aborted"
                 break
 
-            res = await extend_graph(queue, cfg.graph_capability, nodes, edges)
+            res = await journal_extend(queue, cfg.graph_capability, nodes, edges,
+                                       journal_path=graph_journal_path, verb="spine-extension",
+                                       actor="pipeline:cjm-transcript-decomp-core", run=run_id,
+                                       args={"source_id": ids["source"],
+                                             "segments": len(ids["segments"]),
+                                             "text_from": text_from})
             logger.info(f"[src {i}] extension: +{res.nodes_added} node(s) "
                         f"({res.nodes_verified} verified present), +{res.edges_added} edge(s) "
                         f"({res.edges_existing} existing)")
@@ -431,7 +445,10 @@ async def run_decomp(
                                 "text_from": text_from},
                 )
                 dn, de = derivation_to_graph(d)
-                await extend_graph(queue, cfg.graph_capability, [dn], de)
+                await journal_extend(queue, cfg.graph_capability, [dn], de,
+                                     journal_path=graph_journal_path, verb="derivation",
+                                     actor="pipeline:cjm-transcript-decomp-core", run=run_id,
+                                     args={"method": "alignment-fold/v1"})
 
             # Fine spine hangs under the run's renditions — verify scopes there.
             vr = await verify_source(queue, cfg.graph_capability, ids["source"], ids["renditions"])
@@ -473,3 +490,14 @@ async def run_decomp(
         "segments": sum(s.segment_count for s in manifest.sources),
     })
     return manifest
+
+
+def decomp_replay_handlers() -> Dict[str, Any]:  # verb -> async handler(queue, graph_id, op)
+    """The decomp core's replay vocabulary (DEC 426658f1, replay stays DOMAIN-OWNED).
+
+    Exported through the `cjm_context_graph_layer.replay` entry-point group and
+    unioned by `composed_replay_handlers`. Both verbs are `journal_extend` ops —
+    wire-carrying by construction — so they register the layer's shared
+    `apply_wires` (identity-comparable across cores: transcription also emits
+    `derivation`, and the shared handler keeps that collision legal)."""
+    return wires_handlers("spine-extension", "derivation")

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4/cjm_transcript_decomp_core.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cjm-transcript-decomp-core
-Version: 0.0.1
+Version: 0.0.4
 Summary: A frontend-agnostic core for the transcript decomposition workflow — composes isolated capability workers (forced alignment, VAD, graph storage) into a headless pipeline that decomposes transcription run manifests into a VAD-aligned context-graph spine with traceable provenance, and a CLI as its first driver.
 Author-email: "Christian J. Mills" <9126128+cj-mills@users.noreply.github.com>
 License: Apache-2.0
@@ -10,10 +10,10 @@ Classifier: Programming Language :: Python :: 3 :: Only
 Requires-Python: >=3.12
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: cjm_substrate>=0.0.51
-Requires-Dist: cjm_context_graph_primitives>=0.0.12
+Requires-Dist: cjm_substrate>=0.0.57
+Requires-Dist: cjm_context_graph_primitives>=0.0.15
 Requires-Dist: cjm_capability_primitives>=0.0.10
-Requires-Dist: cjm-context-graph-layer>=0.0.8
+Requires-Dist: cjm-context-graph-layer>=0.0.11
 Requires-Dist: cjm-transcript-graph-schema>=0.0.7
 Dynamic: license-file
 
@@ -26,15 +26,10 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 ## Modules
 
 - **`cjm_transcript_decomp_core.alignment`** — Pure forced-alignment logic (no capability calls): map FA words back to character spans in the original text, assign words to VAD chunks by timestamp, and build one text segment per VAD chunk. Extracted from the page-centric ForcedAlignmentService (Tier-1 logic).
-- **`cjm_transcript_decomp_core.cli`** — The CLI driver — the decomposition core's first (and currently only) frontend. Ships in-package as the cjm-transcript-decomp-core console script so the driver can never skew from the core. GUI presentation drivers come later and consume the same pipeline module; they never reimplement it (CLI-first / headless-core principle).
+- **`cjm_transcript_decomp_core.cli`** — The CLI driver — the decomposition core's first (and currently only) frontend.
 - **`cjm_transcript_decomp_core.graph`** — Graph-spine EXTENSION + skeptical-lens verification (stage 5, CR-18 revolution 2). Decomp no longer creates a Document: it RECOMPUTES the transcription-emitted root's deterministic node ids from the consumed manifest (no search), verifies the root exists, and attaches the fine Segment spine under the existing AudioSegment nodes — PART_OF to the owning rendition, STARTS_WITH per rendition (the coarse-seam jump anchor), source-wide NEXT. Each Segment carries the audio TimeSlice ref plus per-transcriber CharSlice refs into the Transcript nodes (the D4/P10 framing, finally expressible). Commit goes through the layer's idempotent extend_graph.
 - **`cjm_transcript_decomp_core.models`** — Lean data shapes for the transcript-decomposition pipeline: in-core mirrors of the forced-alignment / VAD / text DTOs (no FastHTML deps), run configuration, the committed graph-segment carrier, and the decomposition run manifest (proto-bundle).
 - **`cjm_transcript_decomp_core.pipeline`** — The headless decomposition pipeline (stage 5: decomp is an EXTENDER). Load a transcription run manifest, verify the transcription-emitted graph root exists (the graph begins at transcription), then per source per pipeline-segment run VAD + per-transcriber forced alignment, build one aligned segment per VAD chunk with per-transcriber text variants, and attach the fine spine under the existing AudioSegment nodes via the layer's idempotent extend_graph — with HITL approval seams between alignment, commit, and the next source.
-- **`tests.test_alignment`** — Tests for cjm_transcript_decomp_core.alignment — pure forced-alignment logic.
-- **`tests.test_graph`** — Tests for cjm_transcript_decomp_core.graph — fine-spine extension payload.
-- **`tests.test_models`** — Tests for cjm_transcript_decomp_core.models — decomposition data shapes.
-- **`tests_manual.measure_cold_fanout_overlap_e2e`** — Cold-run M×(VAD ∥ T×FA) fan-out overlap measurement (stage-5 closeout;
-- **`tests_manual.validate_stage7_volume_journal_e2e`** — Stage-7 stress part 6 — volume regression on the REAL corpus.
 
 ## API
 
@@ -50,7 +45,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `build_parser` _function_ — Build the CLI parser (subcommands: run).
 - `load_capabilities` _function_ — Discover manifests + load each requested capability (default instance).
 - `main` _function_ — CLI entry point (console script: `cjm-transcript-decomp-core`).
-- `run_command` _function_ — Execute the `run` subcommand: extend a transcription-emitted root with the fine spine.
+- `run_command` _function_ — Execute the `run` subcommand: extend transcription-run manifest(s) with the fine spine.
 
 ### `cjm_transcript_decomp_core.graph`
 
@@ -76,6 +71,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `build_alignment_composition` _function_ — Build the whole-source M×(VAD ∥ T×FA) composition (D8 fan-in, stage-5 variants).
 - `collect_capability_info` _function_ — Record capability identity + data-DB pointers for the run manifest (provenance).
 - `confirm_seam` _function_ — HITL approval seam in its cheapest viable form (log + optional CLI prompt).
+- `decomp_replay_handlers` _function_ — The decomp core's replay vocabulary (DEC 426658f1, replay stays DOMAIN-OWNED).
 - `decompose_source` _function_ — Decompose one source into aligned fine segments with per-transcriber variants.
 - `fa_words_from_result` _function_ — Normalize a typed forced-alignment result into FA words (pure; stage 3).
 - `load_source_manifest` _function_ — Load + lightly validate a transcription-core run manifest.
@@ -83,43 +79,7 @@ A frontend-agnostic core for the transcript decomposition workflow — composes
 - `submit_and_wait` _function_ — Submit one capability job, wait for it, and return its result (raise on failure).
 - `vad_chunks_from_result` _function_ — Normalize a typed VAD result into segment-local VAD chunks.
 
-### `tests.test_alignment`
-
-- `test_assign_words_to_chunks_by_timestamp` _function_
-- `test_build_segments_per_chunk` _function_
-- `test_empty_chunk_yields_empty_segment_warning` _function_
-- `test_map_fa_words_to_text_spans` _function_
-
-### `tests.test_graph`
-
-- `test_edge_topology_spans_coarse_boundary` _function_
-- `test_extension_payload_deterministic_ids` _function_
-- `test_preprocessed_manifest_distinct_renditions_same_boundaries` _function_
-- `test_provenance_refs_and_text_from` _function_
-- `test_resolve_root_ids_recomputation` _function_
-
-### `tests.test_models`
-
-- `test_faword_from_wire_tolerates_extra_keys` _function_
-- `test_manifest_save_round_trip` _function_
-- `test_manifest_shape_and_run_id` _function_
-- `test_stage5_segment_shape` _function_
-- `test_vad_chunk_duration` _function_
-
-### `tests_manual.measure_cold_fanout_overlap_e2e`
-
-- `lane_intervals` _function_ — (start, end) UTC intervals for completed jobs of one instance.
-- `main` _function_
-- `merge` _function_ — Merge overlapping intervals so a side can't double-count itself.
-- `overlap_seconds` _function_ — Total seconds where any interval of A overlaps any interval of B.
-- `self_overlap_max` _function_ — Max simultaneously in-flight jobs within one lane + seconds at depth >= 2.
-
-### `tests_manual.validate_stage7_volume_journal_e2e`
-
-- `checkpoint_copy` _function_ — Checkpoint-then-copy (the stage-3 G3 backup discipline — never copy
-- `journal_rows` _function_
-- `main` _function_
-
 ## Dependencies
 
-**Depends on:** `cjm-substrate`
+**Depends on:** `cjm-capability-primitives`, `cjm-context-graph-layer`, `cjm-context-graph-primitives`, `cjm-substrate`, `cjm-transcript-graph-schema`
+**Used by:** `cjm-transcript-decomp-tui`

cjm_transcript_decomp_core-0.0.4/cjm_transcript_decomp_core.egg-info/entry_points.txt

@@ -0,0 +1,5 @@
+[cjm_context_graph_layer.replay]
+decomp = cjm_transcript_decomp_core.pipeline:decomp_replay_handlers
+
+[console_scripts]
+cjm-transcript-decomp-core = cjm_transcript_decomp_core.cli:main

cjm_transcript_decomp_core-0.0.4/cjm_transcript_decomp_core.egg-info/requires.txt

@@ -0,0 +1,5 @@
+cjm_substrate>=0.0.57
+cjm_context_graph_primitives>=0.0.15
+cjm_capability_primitives>=0.0.10
+cjm-context-graph-layer>=0.0.11
+cjm-transcript-graph-schema>=0.0.7

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4}/pyproject.toml

@@ -14,7 +14,7 @@ classifiers = [
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3 :: Only",
 ]
-dependencies = ['cjm_substrate>=0.0.51', 'cjm_context_graph_primitives>=0.0.12', 'cjm_capability_primitives>=0.0.10', 'cjm-context-graph-layer>=0.0.8', 'cjm-transcript-graph-schema>=0.0.7']
+dependencies = ['cjm_substrate>=0.0.57', 'cjm_context_graph_primitives>=0.0.15', 'cjm_capability_primitives>=0.0.10', 'cjm-context-graph-layer>=0.0.11', 'cjm-transcript-graph-schema>=0.0.7']
 
 [project.urls]
 Repository = "https://github.com/cj-mills/cjm-transcript-decomp-core"
@@ -22,6 +22,9 @@ Repository = "https://github.com/cj-mills/cjm-transcript-decomp-core"
 [project.scripts]
 cjm-transcript-decomp-core = "cjm_transcript_decomp_core.cli:main"
 
+[project.entry-points."cjm_context_graph_layer.replay"]
+decomp = "cjm_transcript_decomp_core.pipeline:decomp_replay_handlers"
+
 [tool.setuptools.dynamic]
 version = {attr = "cjm_transcript_decomp_core.__version__"}
 

{cjm_transcript_decomp_core-0.0.1 → cjm_transcript_decomp_core-0.0.4}/tests/test_cli.py

@@ -8,12 +8,16 @@ def test_run_defaults():
     p = build_parser()
     args = p.parse_args(["run", "m.json", "--yes", "--language", "English"])
     assert args.command == "run"
-    assert args.manifest == "m.json"
+    assert args.manifests == ["m.json"]  # single manifest = a batch of one
     assert args.yes is True
     assert args.fa_capability == "cjm-capability-qwen3-forced-aligner"
     assert args.graph_capability == "cjm-capability-graph-sqlite"
     assert args.text_from is None  # authority defaults to the manifest's sole transcriber
     assert args.sysmon_capability is None
+    # 5daadfc4: None sentinel — run_command resolves the workspace's runs/ when
+    # one is active, else the legacy cwd-relative runs/; TUIs pin their browsed dir
+    assert args.output_dir is None
+    assert args.workspace is None
 
 
 def test_text_from_and_sysmon_flags():
@@ -22,3 +26,12 @@ def test_text_from_and_sysmon_flags():
                          "--sysmon-capability", "cjm-capability-monitor-nvidia"])
     assert args.text_from == "cjm-capability-voxtral-hf"
     assert args.sysmon_capability == "cjm-capability-monitor-nvidia"
+
+
+def test_batch_manifests():
+    # Batch shape (work item 0ff6bf0f): N manifests, one invocation, one stack.
+    p = build_parser()
+    args = p.parse_args(["run", "a.json", "b.json", "c.json", "--yes",
+                         "--text-from", "cjm-capability-voxtral-hf"])
+    assert args.manifests == ["a.json", "b.json", "c.json"]
+    assert args.text_from == "cjm-capability-voxtral-hf"

cjm_transcript_decomp_core-0.0.1/cjm_transcript_decomp_core/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.0.1"

cjm_transcript_decomp_core-0.0.1/cjm_transcript_decomp_core.egg-info/entry_points.txt

@@ -1,2 +0,0 @@
-[console_scripts]
-cjm-transcript-decomp-core = cjm_transcript_decomp_core.cli:main

cjm_transcript_decomp_core-0.0.1/cjm_transcript_decomp_core.egg-info/requires.txt

@@ -1,5 +0,0 @@
-cjm_substrate>=0.0.51
-cjm_context_graph_primitives>=0.0.12
-cjm_capability_primitives>=0.0.10
-cjm-context-graph-layer>=0.0.8
-cjm-transcript-graph-schema>=0.0.7