docs: record the sorter's default (non-provenance) mode

- docs/SORTING.md: the extraction section now shows the default sort command (no --provenance) and presents --provenance as an optional mode.
- AGENTS.md 7: note that sorting runs in the default mode and that mixing the two modes creates duplicate placements.
Docs-only change; no code touched.
This commit is contained in:
2026-10-11 12:55:30 +02:00
parent 6c6a3273bb
commit bc165093e2
2 changed files with 15 additions and 5 deletions
+5
View File
@@ -230,6 +230,11 @@ sorting pipeline — **change it first, then the code**. Summary:
candidates). Weak hits ≥ `WEAK_MIN = 0.50` are accepted but flagged candidates). Weak hits ≥ `WEAK_MIN = 0.50` are accepted but flagged
low-confidence. Tokenless normal adoption needs `+0.08` low-confidence. Tokenless normal adoption needs `+0.08`
(`TOKENLESS_BONUS`). (`TOKENLESS_BONUS`).
- Sorting runs in the sorter's **default (non-provenance) mode**: every
file is classified on its own name (unclassified files land directly in
`_unsorted/<file>`). `--provenance` (classify whole extraction dirs as
units) exists but the live library does **not** use it; mixing the two
modes creates duplicate placements.
- `DROPPED_SYSTEMS` (default `{Middle-Earth}`) filters catalog entries at load - `DROPPED_SYSTEMS` (default `{Middle-Earth}`) filters catalog entries at load
and counts keyword hits as unclassified; `--drop-system`/`--keep-system` and counts keyword hits as unclassified; `--drop-system`/`--keep-system`
adjust. The committed catalog snapshot keeps all systems. adjust. The committed catalog snapshot keeps all systems.
+10 -5
View File
@@ -119,12 +119,17 @@ plan/apply-only):
password archives, disk full) are moved to `<out>/.broken/` and the run password archives, disk full) are moved to `<out>/.broken/` and the run
continues; re-runs skip already-extracted destinations. continues; re-runs skip already-extracted destinations.
- `<out>/extract_manifest.jsonl` records `{archive, dest, unit}` per - `<out>/extract_manifest.jsonl` records `{archive, dest, unit}` per
extraction. The sorter's `--provenance` flag reads it: all files inside extraction. By **default the sorter classifies each file on its own
an extraction directory are classified **once under the archive's name**; the optional `--provenance <manifest>` mode instead classifies
name** (a loose `Backpack.stl` alone is noise; under each extraction directory **once under the archive's name** (a loose
`The Twin Lance/` it belongs to that unit) and placed together, keeping `Backpack.stl` alone is noise; under `The Twin Lance/` it belongs to
the archive's inner directory structure: that unit) and keeps the archive's inner directory structure. The live
library uses the default (non-provenance) mode:
# default — what Warhammer_Library_Live uses
dlphelper sorter <extracted_root> --library <lib> --apply
# optional provenance mode (whole extraction dirs as units)
dlphelper sorter <extracted_root> --library <lib> \ dlphelper sorter <extracted_root> --library <lib> \
--provenance <extracted_root>/extract_manifest.jsonl --apply --provenance <extracted_root>/extract_manifest.jsonl --apply