wolne_lektury: drop 63 docs that duplicate eltec_pol novels (6,141 -> 6,078 docs) (#67)

- wolne_lektury: drop 63 docs that duplicate eltec_pol novels (v2#72) (dd905112dfe019283014a7cc537e25c1cd3cc0e2)
- Merge origin/main (global_voices #66) into v026-wolne_lektury-dedup (6eab276752cb81998c6dfff6674ba434bad70cb6)
- Merge origin/main (1000_novels rescreen, PR 65) into v026-wolne_lektury-dedup (e8f6722a3fdfb0fcea8e5f120cee689752ceac2c)
- wolne_lektury: move the containment code to src/shingle_containment.py (0783e5eb1e2eda3aab5edd4f76c7f85910e7ff3d)

artifacts/source_findings.md

@@ -1480,3 +1480,43 @@ unresolved, so the release keeps 887 texts and drops 18 in copyright and 95 unre
 `src/clean_1000_novels.py` pins the screen CSV by SHA-256
 (`7fefc75f8fcbd07e5ec98842276765ffef9520b6acb4e24b4751c67cc3d2081c`, also in
 `1000_novels.license-evidence.json`).
+
+## wolne_lektury — eltec_pol duplicates dropped (2026-10-06)
+
+**Verdict** — kept, 63 documents dropped: 6,141 → 6,078 documents, 103,009,797 →
+90,654,820 tokens (input: the v0.2.5 Parquet, unchanged since v0.2.1). Each dropped
+document is a novel, or one tome of a novel, that `eltec_pol` also carries. 50 ELTeC
+novels are concerned, and `eltec_pol` stays whole (maintainer decision A,
+polish-dynaword-v2 #72). The filter is `src/clean_wolne_lektury.py` and the pair list is
+`data/wolne_lektury/wolne_lektury.eltec_pairs.tsv`.
+- Score: the share of a WL document's 5-word shingles found in the ELTeC novel that
+  shares the most with it. Tokenisation follows `near_dedup`, and a trailing WL footer is
+  cut first (no WL row has one). Containment is used, not Jaccard, because WL splits
+  novels into tomes: Chłopi has 4, Faraon, Nad Niemnem and Stara baśń have 3.
+- Threshold 0.9, in an empty stretch: no document scores between 0.7588 and 0.9636. All
+  63 documents above it pair with an ELTeC novel by the same author and title (Żuławski's
+  two WL documents are headed with the trilogy's title).
+- Hand-checked: all 3 documents between 0.3 and 0.9, the 3 between 0.9 and 0.99, and 10
+  sampled at ≥ 0.99 (seed 72). In 12 of the 13 drops checked, the unmatched shingles
+  form short scattered runs of at most 27 words (title head, spelling variants). The
+  exception is Dzieje grzechu (0.9879): its WL copy ends with a "Dodatek" of draft
+  chapters, about 1,800 words that ELTeC lacks. It is dropped with the novel.
+- Kept and listed as `keep_partial`: both WL tomes of Mniszek's Trędowata (0.7588 and
+  0.6473 against POL0051). The ELTeC text is shorter: about 121,000 words against WL's
+  162,000. 95% of the ELTeC shingles are in WL's tomes, and the WL text missing from
+  ELTeC comes in runs of up to 5,418 words. The shared part stays duplicated across the
+  two sources. Also kept: Kochanowski's Pieśń IX from Pieśni, księgi wtóre (0.4267),
+  which Łoziński's Oko proroka quotes.
+- Below 0.3 (not listed, not dropped): 6 documents at 0.10–0.22, all poems of at most
+  2,321 tokens by Mickiewicz, Kochanowski and Miciński; their overlap with the paired
+  novels was not inspected. Nothing scores between 0.2203 and 0.4267.
+
+**Reason** — 51 of the 100 ELTeC novels duplicate WL texts. ELTeC's CC BY 4.0 licence
+is the clean one (`rights_audit_2026-10/reports/literary.md`: eltec_pol CLEAN). WL
+licence labels are not touched here; #57 relabels them per row.
+
+References: polish-dynaword-v2 #72 (decision), #45 (method);
+`rights_audit_2026-10/literary/eltec_meta.tsv` (ELTeC file names and titles).
+
+**Lesson** — Jaccard near-dedup misses a novel that one source splits into tomes.
+Measure containment from the split side.

data/wolne_lektury/wolne_lektury.eltec_pairs.tsv

@@ -0,0 +1,67 @@
+wl_id	eltec_id	score	action	wl_title
+wolne_lektury_2372617	eltec_pol_2455238	0.9995	drop	Wacław Berent / Ozimina
+wolne_lektury_2372618	eltec_pol_2455239	0.9997	drop	Wacław Berent / Próchno
+wolne_lektury_2373135	eltec_pol_2455240	0.9921	drop	Deotyma / Branki w jasyrze
+wolne_lektury_2373136	eltec_pol_2455241	0.9986	drop	Deotyma / Panienka z okienka
+wolne_lektury_2372152	eltec_pol_2455242	0.9998	drop	Deotyma / Zwierciadlana zagadka
+wolne_lektury_2374432	eltec_pol_2455246	0.9981	drop	Tadeusz Dołęga-Mostowicz / Kariera Nikodema Dyzmy
+wolne_lektury_2373182	eltec_pol_2455247	0.9999	drop	Pamiętnik pani Hanki / Tadeusz Dołęga-Mostowicz
+wolne_lektury_2372109	eltec_pol_2455248	0.9984	drop	Tadeusz Dołęga-Mostowicz / Znachor
+wolne_lektury_2374147	eltec_pol_2455249	0.9985	drop	Antonina Domańska / Historia żółtej ciżemki
+wolne_lektury_2374899	eltec_pol_2455250	0.9985	drop	Antonina Domańska / Krysia bezimienna
+wolne_lektury_2376177	eltec_pol_2455251	0.9981	drop	Antonina Domańska / Paziowie
+wolne_lektury_2373959	eltec_pol_2455258	0.9955	drop	Stefan Grabiński / Namiętność
+wolne_lektury_2376971	eltec_pol_2455259	0.9977	drop	Stefan Grabiński / Salamandra
+wolne_lektury_2373979	eltec_pol_2455260	0.9981	drop	Stefan Grabiński / Wyspa Itongo
+wolne_lektury_2374400	eltec_pol_2455262	0.9636	drop	Zygmunt Kaczkowski / Murdelio
+wolne_lektury_2374184	eltec_pol_2455267	0.9984	drop	Józef Ignacy Kraszewski / Hrabina Cosel
+wolne_lektury_2376438	eltec_pol_2455268	0.9980	drop	Józef Ignacy Kraszewski / Pogrobek
+wolne_lektury_2377370	eltec_pol_2455269	0.9984	drop	Józef Ignacy Kraszewski / Stara baśń
+wolne_lektury_2377371	eltec_pol_2455269	0.9983	drop	Józef Ignacy Kraszewski / Stara baśń
+wolne_lektury_2377372	eltec_pol_2455269	0.9983	drop	Józef Ignacy Kraszewski / Stara baśń
+wolne_lektury_2375201	eltec_pol_2455275	0.9974	drop	Władysław Łoziński / Oko proroka
+wolne_lektury_2376242	eltec_pol_2455275	0.4267	keep_partial	Jan Kochanowski / Pieśni, Księgi wtóre
+wolne_lektury_2375475	eltec_pol_2455281	0.7588	keep_partial	Tom pierwszy / I
+wolne_lektury_2375476	eltec_pol_2455281	0.6473	keep_partial	Helena Mniszek / Trędowata
+wolne_lektury_2375490	eltec_pol_2455283	0.9966	drop	Zuzanna Morawska / Na zgliszczach Zakonu
+wolne_lektury_2375491	eltec_pol_2455284	0.9966	drop	Zuzanna Morawska / Wilcze gniazdo
+wolne_lektury_2375933	eltec_pol_2455285	0.9999	drop	Eliza Orzeszkowa / Marta
+wolne_lektury_2375340	eltec_pol_2455286	0.9993	drop	Eliza Orzeszkowa / Meir Ezofowicz
+wolne_lektury_2375558	eltec_pol_2455287	0.9994	drop	Eliza Orzeszkowa / Nad Niemnem
+wolne_lektury_2375559	eltec_pol_2455287	0.9992	drop	Eliza Orzeszkowa / Nad Niemnem
+wolne_lektury_2375560	eltec_pol_2455287	0.9995	drop	Eliza Orzeszkowa / Nad Niemnem
+wolne_lektury_2373434	eltec_pol_2455291	0.9974	drop	Bolesław Prus / Emancypantki
+wolne_lektury_2373439	eltec_pol_2455292	0.9979	drop	Bolesław Prus / Faraon
+wolne_lektury_2373440	eltec_pol_2455292	0.9972	drop	Bolesław Prus / Faraon
+wolne_lektury_2373441	eltec_pol_2455292	0.9981	drop	Bolesław Prus / Faraon
+wolne_lektury_2375026	eltec_pol_2455293	0.9988	drop	Bolesław Prus / Lalka
+wolne_lektury_2375027	eltec_pol_2455293	0.9982	drop	Bolesław Prus / Lalka
+wolne_lektury_2372954	eltec_pol_2455294	0.9990	drop	Władysław Stanisław Reymont / Chłopi
+wolne_lektury_2372955	eltec_pol_2455294	0.9989	drop	Władysław Stanisław Reymont / Chłopi
+wolne_lektury_2372956	eltec_pol_2455294	0.9992	drop	Władysław Stanisław Reymont / Chłopi
+wolne_lektury_2376750	eltec_pol_2455294	0.9990	drop	Władysław Stanisław Reymont / Chłopi
+wolne_lektury_2374612	eltec_pol_2455295	0.9992	drop	Władysław Stanisław Reymont / Komediantka
+wolne_lektury_2372093	eltec_pol_2455296	0.9985	drop	Władysław St. Reymont / Ziemia obiecana, tom II
+wolne_lektury_2372094	eltec_pol_2455296	0.9990	drop	Władysław St. Reymont / Ziemia obiecana, tom I
+wolne_lektury_2376762	eltec_pol_2455297	0.9977	drop	Maria Rodziewiczówna / Lato
+wolne_lektury_2376763	eltec_pol_2455298	0.9985	drop	Maria Rodziewiczówna / Między ustami a
+wolne_lektury_2376764	eltec_pol_2455299	0.9984	drop	Maria Rodziewiczówna / Straszny dziadunio
+wolne_lektury_2375871	eltec_pol_2455300	0.9983	drop	Henryk Sienkiewicz / Ogniem i mieczem
+wolne_lektury_2375872	eltec_pol_2455300	0.9981	drop	Henryk Sienkiewicz / Ogniem i mieczem
+wolne_lektury_2376691	eltec_pol_2455301	0.9970	drop	Henryk Sienkiewicz / Quo vadis
+wolne_lektury_2376766	eltec_pol_2455302	0.9983	drop	Henryk Sienkiewicz / Rodzina Połanieckich
+wolne_lektury_2377398	eltec_pol_2455303	0.9998	drop	Andrzej Strug / Dzieje jednego pocisku
+wolne_lektury_2377720	eltec_pol_2455307	0.9899	drop	Aleksander Świętochowski / Tragikomedya prawdy
+wolne_lektury_2371802	eltec_pol_2455316	0.9991	drop	Stanisław Ignacy Witkiewicz / Nienasycenie
+wolne_lektury_2371803	eltec_pol_2455316	0.9988	drop	Stanisław Ignacy Witkiewicz / Nienasycenie
+wolne_lektury_2371806	eltec_pol_2455317	0.9993	drop	Stanisław Ignacy Witkiewicz / Pożegnanie jesieni
+wolne_lektury_2372038	eltec_pol_2455318	0.9999	drop	Gabriela Zapolska / Kaśka Kariatyda
+wolne_lektury_2372039	eltec_pol_2455319	0.9966	drop	Gabiela Zapolska / Kobieta bez skazy
+wolne_lektury_2377058	eltec_pol_2455320	0.9981	drop	Gabriela Zapolska / Sezonowa miłość
+wolne_lektury_2373356	eltec_pol_2455321	0.9879	drop	Stefan Żeromski / Dzieje grzechu
+wolne_lektury_2376559	eltec_pol_2455322	0.9995	drop	Stefan Żeromski / Przedwiośnie
+wolne_lektury_2377473	eltec_pol_2455323	0.9985	drop	Stefan Żeromski / Syzyfowe prace
+wolne_lektury_2372105	eltec_pol_2455326	0.9989	drop	Narcyza Żmichowska / Poganka
+wolne_lektury_2371605	eltec_pol_2455327	0.9988	drop	Jerzy Żuławski / Trylogia księżycowa
+wolne_lektury_2372130	eltec_pol_2455328	0.9986	drop	Jerzy Żuławski / Powrót
+wolne_lektury_2371607	eltec_pol_2455329	0.9981	drop	Jerzy Żuławski / Trylogia księżycowa

data/wolne_lektury/wolne_lektury.md

@@ -19,16 +19,19 @@ Pulled from SpeakLeash's public redistribution (`speakleash-ds-pub`, key `wolne_
 ## Statistics
 | documents | characters | tokens (tiktoken proxy) |
 |---:|---:|---:|
-| 6,141 | 261,892,000 | 103,009,797 |
+| 6,078 | 230,764,171 | 90,654,820 |
 
 
 ## Per-document license metadata
 | license | documents |
 |---|---:|
-| `CC-BY-SA-4.0 / Wolna Sztuka 1.3` | 6,141 |
+| `CC-BY-SA-4.0 / Wolna Sztuka 1.3` | 6,078 |
 
 Author metadata present for **0** documents. Empty values mean the upstream record did not expose a machine-readable author field.
 
+## Legal scope note
+For v0.2.6, 63 documents whose text is an eltec_pol novel are dropped (share of a document's 5-word shingles found in one ELTeC novel >= 0.9; src/clean_wolne_lektury.py, polish-dynaword-v2#72). eltec_pol keeps those 50 novels whole under CC BY 4.0. Three documents only partly in an ELTeC novel are kept: both tomes of Mniszek's Trędowata (the ELTeC text is shorter) and Kochanowski's Pieśń IX, quoted in Łoziński's Oko proroka. Every pair is listed in wolne_lektury.eltec_pairs.tsv.
+
 Statistics were recomputed directly from the released parquet file.
 
 ## Filters applied (build_dynaword.py)
@@ -37,7 +40,7 @@ Minimal, per Dynaword guidelines (heavy filtering left to downstream use):
 - drop non-Polish (diacritic ratio): **0**
 - exact cross-source dedup (sha1): **0**
 - OCR alpha-ratio < 0.70 (OCR sources only): **0**
-- read 6,141 → kept 6,141
+- read 6,141 → kept 6,078
 
 Token counts are a fast tiktoken (cl100k) proxy (~1% off Llama-3); the canonical
 Llama-3 count is computed at release.

data/wolne_lektury/wolne_lektury.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5833c25db296f2479b3c09745ebedd550b2425dd6ebaf021f0095e684a0221a1
-size 128562077
+oid sha256:f308d49235551241d6c47c5e88138ba52e6235422ae272ddbc135c3b8696ae0e
+size 113055766

data/wolne_lektury/wolne_lektury.stats.json

@@ -1,16 +1,19 @@
 {
   "read": 6141,
-  "kept": 6141,
+  "kept": 6078,
   "drop_short": 0,
   "drop_lang": 0,
   "drop_dup": 0,
   "drop_ocr": 0,
-  "chars": 261892000,
-  "tokens": 103009797,
+  "chars": 230764171,
+  "tokens": 90654820,
   "licenses": {
-    "CC-BY-SA-4.0 / Wolna Sztuka 1.3": 6141
+    "CC-BY-SA-4.0 / Wolna Sztuka 1.3": 6078
   },
   "authors_with_value": 0,
   "license": "CC-BY-SA-4.0 / Wolna Sztuka 1.3",
-  "stats_recomputed_from_parquet": true
+  "stats_recomputed_from_parquet": true,
+  "drop_eltec_dup": 63,
+  "input_sha256": "5833c25db296f2479b3c09745ebedd550b2425dd6ebaf021f0095e684a0221a1",
+  "eltec_sha256": "ba40061741dff11165149f6c2cf3d5526e1b6c759e7ce1e03ccc164d58838325"
 }

src/clean_wolne_lektury.py

@@ -0,0 +1,129 @@
+#!/usr/bin/env python3
+"""Drop wolne_lektury documents whose text is an eltec_pol novel (slayerlabs/polish-dynaword-v2#72).
+
+Half of the 100 ELTeC-pol novels are also in Wolne Lektury. The maintainer chose (2026-10-06) to
+drop the Wolne Lektury copies and keep eltec_pol whole, because ELTeC's CC BY 4.0 is the clean
+licence. Wolne Lektury splits long novels into tomes, so the test is containment, not Jaccard:
+
+  score = |S(wl) & S(novel)| / |S(wl)|
+
+S() is the set of 5-word shingles over NFKC-casefolded \\w+ tokens (the tokenisation of the
+repo's near_dedup), and `novel` is the ELTeC novel sharing the most shingles with the WL doc;
+the shingling and scoring code is in src/shingle_containment.py.
+A trailing Wolne Lektury footer is cut from WL text before scoring (the #45 measurement's rule);
+the released text is not changed. Docs scoring >= DROP_AT are dropped. Docs scoring
+>= REPORT_AT but below DROP_AT are only partly in an ELTeC novel; they are kept and listed.
+Both groups are written to wolne_lektury.eltec_pairs.tsv.
+
+Both input Parquets are pinned by SHA-256. Restore the WL Parquet before a rerun:
+  git checkout v0.2.5 -- data/wolne_lektury/wolne_lektury.parquet
+  python3 src/clean_wolne_lektury.py
+"""
+from __future__ import annotations
+
+import argparse
+import csv
+import hashlib
+import json
+import os
+from collections import Counter
+from pathlib import Path
+
+import pyarrow as pa
+import pyarrow.compute as pc
+import pyarrow.parquet as pq
+
+from shingle_containment import best_containment, strip_wl_footer
+
+SOURCE = "wolne_lektury"
+ROOT = Path(__file__).resolve().parents[1]
+DATA = ROOT / "data" / SOURCE
+ELTEC = ROOT / "data" / "eltec_pol" / "eltec_pol.parquet"
+INPUT_SHA256 = "5833c25db296f2479b3c09745ebedd550b2425dd6ebaf021f0095e684a0221a1"  # v0.2.1-v0.2.5
+ELTEC_SHA256 = "ba40061741dff11165149f6c2cf3d5526e1b6c759e7ce1e03ccc164d58838325"  # v0.2.1-v0.2.5
+PAIRS = f"{SOURCE}.eltec_pairs.tsv"
+# Thresholds sit in empty stretches of the v0.2.5 score distribution: no doc scores between
+# 0.7588 and 0.9636 (above: 63 docs, the novels; below: two tomes of an abridged ELTeC
+# Trędowata), and none between 0.2203 and 0.4267 (below: poems quoted in a novel).
+DROP_AT = 0.9
+REPORT_AT = 0.3
+
+
+def wl_title(text: str) -> str:
+    """Author and title: the first two non-empty lines of a Wolne Lektury text."""
+    return " / ".join([line.strip() for line in text.splitlines() if line.strip()][:2])
+
+
+def check_pin(path: Path, expected: str, restore: str) -> None:
+    with path.open("rb") as stream:
+        if hashlib.file_digest(stream, "sha256").hexdigest() != expected:
+            raise SystemExit(f"{path} does not match the pinned SHA-256; {restore}")
+
+
+def build(input_path: Path, eltec_path: Path, out_dir: Path) -> dict:
+    check_pin(input_path, INPUT_SHA256,
+              f"restore it with `git checkout v0.2.5 -- data/{SOURCE}/{SOURCE}.parquet`")
+    check_pin(eltec_path, ELTEC_SHA256, "it is not the v0.2.5 eltec_pol Parquet")
+    table = pq.read_table(input_path)
+    eltec = pq.read_table(eltec_path, columns=["id", "text"])
+    texts = table.column("text").to_pylist()
+    eltec_ids = eltec.column("id").to_pylist()
+    scores = best_containment([strip_wl_footer(t) for t in texts], eltec.column("text").to_pylist())
+
+    keep, pairs = [], []
+    for wl_id, text, (novel, score) in zip(table.column("id").to_pylist(), texts, scores, strict=True):
+        keep.append(score < DROP_AT)
+        if score >= REPORT_AT:
+            pairs.append({"wl_id": wl_id, "eltec_id": eltec_ids[novel], "score": f"{score:.4f}",
+                          "action": "drop" if score >= DROP_AT else "keep_partial",
+                          "wl_title": wl_title(text)})
+    kept = table.filter(pa.array(keep))
+
+    stats_path = out_dir / f"{SOURCE}.stats.json"
+    stats = json.loads(stats_path.read_text(encoding="utf-8"))  # keeps build_dynaword's read/drop_* counts
+    stats.update(
+        kept=kept.num_rows,
+        drop_eltec_dup=table.num_rows - kept.num_rows,
+        chars=pc.sum(pc.utf8_length(kept.column("text"))).as_py(),
+        tokens=pc.sum(kept.column("token_count")).as_py(),
+        licenses=dict(sorted(Counter(kept.column("license").to_pylist()).items())),
+        authors_with_value=sum(map(bool, kept.column("author").to_pylist())),
+        input_sha256=INPUT_SHA256,
+        eltec_sha256=ELTEC_SHA256,
+    )
+    drops = ("drop_short", "drop_lang", "drop_dup", "drop_ocr", "drop_eltec_dup")
+    if stats["read"] - sum(stats[k] for k in drops) != stats["kept"]:
+        raise SystemExit("stats drop counts do not add up to the kept rows")
+
+    staged = {
+        out_dir / f"{SOURCE}.parquet.tmp": lambda p: pq.write_table(kept, p, compression="zstd"),
+        out_dir / f"{PAIRS}.tmp": lambda p: _write_pairs(p, pairs),
+        out_dir / f"{SOURCE}.stats.json.tmp": lambda p: p.write_text(
+            json.dumps(stats, indent=2, ensure_ascii=False) + "\n", encoding="utf-8"),
+    }
+    for path, write in staged.items():
+        write(path)
+    for path in staged:
+        os.replace(path, path.with_suffix(""))
+    return stats
+
+
+def _write_pairs(path: Path, pairs: list[dict]) -> None:
+    with path.open("w", encoding="utf-8", newline="") as handle:
+        writer = csv.DictWriter(handle, ["wl_id", "eltec_id", "score", "action", "wl_title"],
+                                delimiter="\t", lineterminator="\n")
+        writer.writeheader()
+        writer.writerows(sorted(pairs, key=lambda p: (p["eltec_id"], p["wl_id"])))
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description=__doc__.split("\n")[0])
+    parser.add_argument("--input", type=Path, default=DATA / f"{SOURCE}.parquet")
+    parser.add_argument("--eltec", type=Path, default=ELTEC)
+    parser.add_argument("--out", type=Path, default=DATA)
+    args = parser.parse_args()
+    print(json.dumps(build(args.input, args.eltec, args.out), indent=2, ensure_ascii=False))
+
+
+if __name__ == "__main__":
+    main()

src/shingle_containment.py

@@ -0,0 +1,94 @@
+"""Shingle containment between documents and a set of reference documents.
+
+Shared by the cross-source dedup filters (src/clean_wolne_lektury.py, src/clean_1000_novels.py).
+S() is the set of K-word shingles over NFKC-casefolded \\w+ tokens, the tokenisation of the repo's
+near_dedup. Containment, not Jaccard: a document is scored by the share of its own shingles that
+a reference holds, because Wolne Lektury splits long novels into tomes.
+"""
+from __future__ import annotations
+
+import hashlib
+import re
+import unicodedata
+from functools import lru_cache
+from typing import NamedTuple
+
+import numpy as np
+
+K = 5
+WORD = re.compile(r"\w+")
+# Odd 64-bit multipliers that mix the K word hashes of a shingle in order.
+MULT = np.array([0x9E3779B97F4A7C15, 0xC2B2AE3D27D4EB4F, 0x165667B19E3779F9,
+                 0xD6E8FEB86659FD93, 0xFF51AFD7ED558CCD], dtype=np.uint64)
+
+
+def strip_wl_footer(text: str) -> str:
+    """Text before a trailing "-----" footer that names wolnelektury, else the text unchanged."""
+    j = text.rfind("\n-----")
+    if j >= 0 and len(text) - j < 5000 and "wolnelektury" in text[j:].lower():
+        return text[:j]
+    return text
+
+
+@lru_cache(maxsize=None)
+def _word_hash(word: str) -> int:
+    # blake2b, not hash(): str hashing is salted per process, and the scores must be reproducible.
+    return int.from_bytes(hashlib.blake2b(word.encode("utf-8"), digest_size=8).digest(), "little")
+
+
+def shingles(text: str) -> np.ndarray:
+    """Sorted unique uint64 hashes of the text's K-word shingles."""
+    words = WORD.findall(unicodedata.normalize("NFKC", text).casefold())
+    n = len(words) - K + 1
+    if n <= 0:
+        return np.empty(0, np.uint64)
+    h = np.fromiter((_word_hash(w) for w in words), dtype=np.uint64, count=len(words))
+    s = np.zeros(n, np.uint64)
+    for i in range(K):
+        s += h[i:i + n] * MULT[i]
+    s ^= s >> np.uint64(31)
+    return np.unique(s)
+
+
+class Match(NamedTuple):
+    """How much of one document's shingle set the references hold."""
+    shingles: int      # unique shingles in the document
+    covered: int       # of those, how many occur in at least one reference
+    refs: np.ndarray   # indices of the references sharing at least one shingle, ascending
+    shared: np.ndarray  # shingles shared with each of those references
+
+    @property
+    def cover(self) -> float:
+        """Share of the document's shingles found in any reference (the union of the references)."""
+        return self.covered / self.shingles if self.shingles else 0.0
+
+    def best(self) -> tuple[int, float]:
+        """(index of the reference holding most shingles, share it holds); (-1, 0.0) if none."""
+        if not len(self.refs):
+            return -1, 0.0
+        i = int(self.shared.argmax())  # ties go to the lowest reference index
+        return int(self.refs[i]), self.shared[i] / self.shingles
+
+
+def containment(docs: list[str], refs: list[str]) -> list[Match]:
+    """Match every doc against all refs at once."""
+    sets = [shingles(text) for text in refs] or [np.empty(0, np.uint64)]
+    hashes = np.concatenate(sets)
+    owner = np.concatenate([np.full(len(s), i, np.int32) for i, s in enumerate(sets)])
+    order = np.argsort(hashes, kind="stable")
+    hashes, owner = hashes[order], owner[order]
+    out = []
+    for text in docs:
+        a = shingles(text)
+        lo, hi = np.searchsorted(hashes, a, "left"), np.searchsorted(hashes, a, "right")
+        lens = hi - lo
+        # Every (doc shingle, ref) hit: each ref's set is unique, so a shingle counts once per ref.
+        idx = np.repeat(lo - np.cumsum(lens) + lens, lens) + np.arange(lens.sum())
+        ref_ids, shared = np.unique(owner[idx], return_counts=True)
+        out.append(Match(len(a), int((lens > 0).sum()), ref_ids, shared))
+    return out
+
+
+def best_containment(docs: list[str], refs: list[str]) -> list[tuple[int, float]]:
+    """For each doc: (index of the ref holding most of its shingles, share held), (-1, 0.0) if none."""
+    return [m.best() for m in containment(docs, refs)]

src/sources.py

@@ -311,6 +311,13 @@ SOURCES = {
         "upstream": "https://wolnelektury.pl/api/books/",
         "domain": "literature",
         "created": "1500-01-01, 2024-12-31",
+        "legal_note": "For v0.2.6, 63 documents whose text is an eltec_pol novel are dropped "
+                      "(share of a document's 5-word shingles found in one ELTeC novel >= 0.9; "
+                      "src/clean_wolne_lektury.py, polish-dynaword-v2#72). eltec_pol keeps those "
+                      "50 novels whole under CC BY 4.0. Three documents only partly in an ELTeC "
+                      "novel are kept: both tomes of Mniszek's Trędowata (the ELTeC text is "
+                      "shorter) and Kochanowski's Pieśń IX, quoted in Łoziński's Oko proroka. "
+                      "Every pair is listed in wolne_lektury.eltec_pairs.tsv.",
         "is_ocr": False,
     },
     "wikinews": {

src/test_shingle_containment.py

@@ -0,0 +1,52 @@
+"""Unit tests for the shared cross-source containment code (src/shingle_containment.py)."""
+
+import sys
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent))
+
+from shingle_containment import best_containment, containment, shingles, strip_wl_footer
+
+NOVEL = " ".join(f"w{i}" for i in range(400))
+OTHER = " ".join(f"x{i}" for i in range(400))
+
+
+def test_shingles_are_reproducible_across_processes():
+    # Fixed values: a salted hash() would change them on every interpreter start.
+    assert shingles("Ala ma kota, a kot ma Alę.").tolist() == [
+        1095117514565903941, 11979537242062189157, 12971353317152582320]
+    assert shingles("ALA  ma\nkota, a kot ma alę").tolist() == shingles("Ala ma kota, a kot ma Alę.").tolist()
+    assert shingles("a b c d e").tolist() != shingles("e d c b a").tolist()
+    assert len(shingles("a b c d")) == 0
+
+
+def test_strip_wl_footer_cuts_only_a_trailing_wolne_lektury_footer():
+    footer = "\n-----\nTa lektura pochodzi ze strony wolnelektury.pl.\n"
+    assert strip_wl_footer("Tekst." + footer) == "Tekst."
+    assert strip_wl_footer("Tekst.\n-----\nKoniec.") == "Tekst.\n-----\nKoniec."
+    assert strip_wl_footer("Tekst." + footer + "x" * 5000) == "Tekst." + footer + "x" * 5000
+
+
+def test_best_containment_finds_tomes_and_partial_copies():
+    words = NOVEL.split()
+    tome = " ".join(words[100:300]).upper()
+    half = " ".join(words[:100] + OTHER.split()[:100])
+    unrelated = " ".join(f"y{i}" for i in range(50))
+    novels = [" ".join(OTHER.split()[:50]), NOVEL]
+    scores = best_containment([tome, half, unrelated, "za krótki", NOVEL], novels)
+    assert scores[0] == (1, 1.0)
+    assert scores[1][0] == 1 and 0.45 < scores[1][1] < 0.55  # novel 1 holds more than novel 0
+    assert scores[2] == scores[3] == (-1, 0.0)
+    assert scores[4] == (1, 1.0)
+
+
+def test_cover_is_the_union_of_the_references():
+    words = NOVEL.split()
+    tomes = [" ".join(words[:200]), " ".join(words[196:])]  # two tomes, no shingle in both
+    whole, empty = containment([NOVEL, ""], tomes)
+    assert whole.shingles == whole.covered == 396 and whole.cover == 1.0
+    assert whole.refs.tolist() == [0, 1] and whole.shared.tolist() == [196, 200]
+    assert whole.best() == (1, 200 / 396)  # no single tome holds half of it
+    assert (empty.shingles, empty.covered, empty.cover, empty.best()) == (0, 0, 0.0, (-1, 0.0))
+    [alone] = containment([NOVEL], [])
+    assert (alone.covered, alone.cover, alone.best()) == (0, 0.0, (-1, 0.0))

src/test_wolne_lektury_contract.py

@@ -0,0 +1,64 @@
+"""Contract checks for wolne_lektury after the eltec_pol dedup (src/clean_wolne_lektury.py, v2#72)."""
+
+import csv
+import hashlib
+import json
+import sys
+from pathlib import Path
+
+import pyarrow as pa
+import pyarrow.compute as pc
+import pyarrow.parquet as pq
+import pytest
+
+sys.path.insert(0, str(Path(__file__).resolve().parent))
+
+from clean_wolne_lektury import (DATA, DROP_AT, ELTEC, ELTEC_SHA256, PAIRS, REPORT_AT, build,
+                                 wl_title)
+
+# Hand-checked: only part of each is in its ELTeC novel, so they stay (see the findings entry).
+PARTIAL_IDS = {"wolne_lektury_2375475", "wolne_lektury_2375476", "wolne_lektury_2376242"}
+
+
+def test_wl_title_is_the_first_two_nonempty_lines():
+    assert wl_title("Stefan Żeromski\n\nDzieje grzechu\n\nISBN 1\n\nText") == "Stefan Żeromski / Dzieje grzechu"
+
+
+def test_build_fails_closed_on_unpinned_inputs(tmp_path):
+    fake = tmp_path / "wolne_lektury.parquet"
+    pq.write_table(pa.table({"id": ["a"], "text": ["b"]}), fake)
+    with pytest.raises(SystemExit, match="pinned SHA-256"):
+        build(fake, ELTEC, tmp_path)
+    assert sorted(p.name for p in tmp_path.iterdir()) == ["wolne_lektury.parquet"]
+
+
+def test_release_has_no_eltec_duplicates_and_matches_stats():
+    stats = json.loads((DATA / "wolne_lektury.stats.json").read_text(encoding="utf-8"))
+    table = pq.read_table(DATA / "wolne_lektury.parquet")
+    ids = set(table.column("id").to_pylist())
+    with (DATA / PAIRS).open(encoding="utf-8", newline="") as handle:
+        pairs = list(csv.DictReader(handle, delimiter="\t"))
+    dropped = {p["wl_id"] for p in pairs if p["action"] == "drop"}
+    partial = {p["wl_id"] for p in pairs if p["action"] == "keep_partial"}
+
+    assert table.num_rows == len(ids) == stats["kept"] == 6_078
+    assert len(dropped) == stats["drop_eltec_dup"] == 63
+    assert partial == PARTIAL_IDS
+    assert not dropped & ids and partial <= ids
+    assert all(float(p["score"]) >= DROP_AT for p in pairs if p["action"] == "drop")
+    assert all(REPORT_AT <= float(p["score"]) < DROP_AT for p in pairs if p["action"] == "keep_partial")
+    assert all(p["eltec_id"].startswith("eltec_pol_") for p in pairs)
+    drops = ("drop_short", "drop_lang", "drop_dup", "drop_ocr", "drop_eltec_dup")
+    assert stats["read"] - sum(stats[k] for k in drops) == stats["kept"]
+    assert pc.sum(table.column("token_count")).as_py() == stats["tokens"]
+    assert pc.sum(pc.utf8_length(table.column("text"))).as_py() == stats["chars"]
+    assert stats["licenses"] == {"CC-BY-SA-4.0 / Wolna Sztuka 1.3": stats["kept"]}  # #57 relabels later
+    with ELTEC.open("rb") as stream:  # eltec_pol is kept whole
+        assert hashlib.file_digest(stream, "sha256").hexdigest() == ELTEC_SHA256
+
+
+def test_datasheet_counts_match_stats():
+    stats = json.loads((DATA / "wolne_lektury.stats.json").read_text(encoding="utf-8"))
+    card = (DATA / "wolne_lektury.md").read_text(encoding="utf-8")
+    assert f"| {stats['kept']:,} | {stats['chars']:,} | {stats['tokens']:,} |" in card
+    assert f"{stats['drop_eltec_dup']} " in card and "eltec_pol" in card