Add permissively licensed OPEN ICM Polish publications pilot

artifacts/open_icm_pl_ontology_manifest.json

@@ -0,0 +1,132 @@
+{
+  "actors": [
+    {
+      "id": "actor:piotrsty",
+      "type": "Contributor"
+    },
+    {
+      "id": "actor:icm-uw",
+      "type": "Organization"
+    },
+    {
+      "id": "actor:codex",
+      "type": "Agent"
+    }
+  ],
+  "claims": [
+    {
+      "falsification_condition": "The pinned Parquet, decisions or overlap report do not reproduce these counts.",
+      "id": "claim:open-icm-target-retention",
+      "statement": "The proposed DynaWord slice contains 85 records after excluding 10 exact title/DOI overlaps.",
+      "supported_by": [
+        "evidence:open-icm-target-stats:49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c",
+        "evidence:open-icm-target-overlap:c05fccedde6965de8dcaf62fd1233fe3fa77cdd105c426f9f6d5474684303a61"
+      ]
+    },
+    {
+      "falsification_condition": "Target-wide text deduplication and controlled ablations establish those properties.",
+      "id": "claim:open-icm-training-value-untested",
+      "statement": "Net text novelty and training benefit remain untested hypotheses.",
+      "supported_by": [
+        "evidence:open-icm-target-overlap:c05fccedde6965de8dcaf62fd1233fe3fa77cdd105c426f9f6d5474684303a61"
+      ]
+    }
+  ],
+  "evidence": [
+    {
+      "artifact": "data/open_icm_pl/open_icm_pl.stats.json",
+      "content_address": "49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c",
+      "id": "evidence:open-icm-target-stats:49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c",
+      "observation_type": "target_slice_counts",
+      "produced_by": "run:open-icm-dynaword-integration:49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c"
+    },
+    {
+      "artifact": "data/open_icm_pl/open_icm_pl.overlap_audit.json",
+      "content_address": "c05fccedde6965de8dcaf62fd1233fe3fa77cdd105c426f9f6d5474684303a61",
+      "id": "evidence:open-icm-target-overlap:c05fccedde6965de8dcaf62fd1233fe3fa77cdd105c426f9f6d5474684303a61",
+      "observation_type": "metadata_overlap_audit",
+      "produced_by": "run:open-icm-dynaword-integration:49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c"
+    }
+  ],
+  "objects": [
+    {
+      "id": "object:dataset:open-icm-pl-pilot",
+      "type": "Dataset"
+    },
+    {
+      "id": "object:dataset:open-icm-pl-dynaword-slice",
+      "type": "Dataset"
+    }
+  ],
+  "pending": [
+    "cross-source text deduplication",
+    "benchmark contamination check",
+    "review of quoted third-party text",
+    "controlled training ablation"
+  ],
+  "protocols": [
+    {
+      "id": "protocol:open-icm-dynaword-integration-v1",
+      "procedure": "exclude source-QA failures and exact normalized-title or DOI overlaps with pinned biblioteka_nauki; preserve attribution and pending text-level gates"
+    }
+  ],
+  "relations": [
+    {
+      "predicate": "DERIVED_FROM",
+      "source": "version:dataset:open-icm-pl-dynaword:11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31",
+      "target": "hf:dataset:PiotrSty/open-icm-pl-publications@c4a20022e6947f55f82c3354bb5c86609ba681b8"
+    },
+    {
+      "predicate": "GENERATED_BY",
+      "source": "version:dataset:open-icm-pl-dynaword:11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31",
+      "target": "run:open-icm-dynaword-integration:49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c"
+    },
+    {
+      "predicate": "VALIDATED_AGAINST",
+      "source": "version:dataset:open-icm-pl-dynaword:11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31",
+      "target": "hf:dataset:SlayerLab/polish-dynaword@fe44303ea2db290c7a195264226477f724973357"
+    }
+  ],
+  "runs": [
+    {
+      "actor": "actor:codex",
+      "finished_at": "2026-09-10T16:47:55.647670+00:00",
+      "id": "run:open-icm-dynaword-integration:49bfaa6003187e9089ef804aa5ca1bd196bf75925443fab1f352fe0cef19ea0c",
+      "protocol": "protocol:open-icm-dynaword-integration-v1",
+      "started_at": "2026-09-10T16:47:55.647670+00:00",
+      "stats": {
+        "added": "2026-09-10",
+        "author_coverage": 1.0,
+        "characters": 6219019,
+        "kept": 85,
+        "license_counts": {
+          "CC-BY-3.0-PL": 32,
+          "CC-BY-4.0": 7,
+          "CC-BY-SA-3.0": 1,
+          "CC-BY-SA-3.0-PL": 38,
+          "CC-BY-SA-4.0": 7
+        },
+        "sample_count": 12,
+        "source_pilot_acquired": 100,
+        "source_pilot_kept": 95,
+        "target_overlap_excluded": 10,
+        "target_revision": "fe44303ea2db290c7a195264226477f724973357",
+        "tokens": 2358174
+      },
+      "success": true
+    }
+  ],
+  "schema": "slayer-research-ontology-profile-v1",
+  "versions": [
+    {
+      "content_address": "c4a20022e6947f55f82c3354bb5c86609ba681b8",
+      "id": "hf:dataset:PiotrSty/open-icm-pl-publications@c4a20022e6947f55f82c3354bb5c86609ba681b8",
+      "object": "object:dataset:open-icm-pl-pilot"
+    },
+    {
+      "content_address": "11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31",
+      "id": "version:dataset:open-icm-pl-dynaword:11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31",
+      "object": "object:dataset:open-icm-pl-dynaword-slice"
+    }
+  ]
+}

artifacts/source_findings.md

@@ -971,3 +971,40 @@ Quarterly peer-reviewed journal published by the **Biuro Analiz Sejmowych (BAS)*
   `biblioteka_nauki` still needs cross-source dedup because its current upstream
   catalogue exposes 105 of these articles even though the released local shard
   does not contain them.
+
+---
+
+## open_icm_pl — OPEN ICM open-access publications (pilot)
+
+**Investigated** — 2026-09-10 (review round 2026-09-28).
+
+**Verdict** — `added` (proposed with `release: None`). 85 publications / 2.36M
+`cl100k_base` tokens — a deterministic 100-record pilot of the OPEN ICM
+repository (ICM UW). Small on purpose: yields must not be extrapolated to the
+6,183 eligible items.
+
+**Enumeration** — official DSpace API (`/server/api/core/items`), fail-closed
+license/language gate; repository-generated TEXT bitstreams (MD5-verified)
+fetched after the origin's interactive CAPTCHA. 100 acquired -> 95 after QA ->
+85 after removing 10 exact title/DOI matches against `biblioteka_nauki`.
+
+**License evidence** — verified AT THE ORIGIN: per-item `dc.rights` +
+`dc.rights.uri` in the portal's item metadata (example: "Uznanie autorstwa 3.0
+Polska" -> creativecommons.org/licenses/by/3.0/pl/ on
+<https://open.icm.edu.pl/items/0078ca93-45a5-4bef-912d-297c069ee6ab>, also
+machine-readable at `/server/api/core/items/0078ca93-...`); 77 of 85 items
+additionally ship a per-item `license.txt` bitstream (checksum-pinned).
+Allowlist: CC BY / CC BY-SA only (3.0-PL / 3.0 / 4.0); NC and "dozwolony
+użytek" rejected.
+
+**Blockers** — none to merge with `release: None`. Before release: NERGAL pass -
+the residual spot check FOUND one glyph-garbled contact the patterns missed
+(`JKolbusz @wsiz.rzeszow.pl` in `open_icm_pl_6ec903ca-...`); extraction noise
+defeats pattern matching, so the model-based scrubber is the real gate. Also
+cross-source text dedup (the 10 exclusions were metadata-level) and review of
+quoted third-party passages.
+
+**Lesson** — (1) pattern scrubbers fail on PDF-extraction glyph noise - always
+spot-check noisy sources, a "0 residual" pattern count is not evidence; (2)
+CAPTCHA-gated origins are reproducible only via pinned checksums, not re-fetch -
+say so in the datasheet instead of pretending a before/after can be regenerated.

data/open_icm_pl/NOTICE.md

@@ -0,0 +1,7 @@
+# Attribution and license notice
+
+Source: OPEN ICM, https://open.icm.edu.pl/home
+
+Each record retains its item URL, authorship, repository license label, normalized license identifier, bitstream metadata and license evidence where exposed by OPEN ICM. The applicable CC BY or CC BY-SA terms and attribution obligations apply per record; consult `artifacts/attribution.jsonl`.
+
+Preparation: Piotr Styla with OpenAI Codex. Changes: repository TEXT selection, Unicode and whitespace normalization, page-number-only removal, line-wrap repair, limited email and labelled-phone redaction, language/quality filtering and within-source deduplication. No endorsement by OPEN ICM or credited authors is implied.

data/open_icm_pl/open_icm_pl.attribution.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3019903082ad4b4eed4955a19b44e81dddf247341be0cae9467380d6532c7590
+size 172553

data/open_icm_pl/open_icm_pl.checksums.json

@@ -0,0 +1,11 @@
+{
+  "NOTICE.md": "e8b706e3aeddc659c8a20b3354ed78c75f5ff5d3801dfed13f100b5253097927",
+  "open_icm_pl.attribution.jsonl": "3019903082ad4b4eed4955a19b44e81dddf247341be0cae9467380d6532c7590",
+  "open_icm_pl.decisions.jsonl": "8f1d90bb052398bcd18eae457218f8b5b9d68ce8dd114168b1fff305570a4eda",
+  "open_icm_pl.md": "1f94660f48755dc3350c7bb467d23d1baebfb0fb36ebdbf925e0d0cb9b205ff0",
+  "open_icm_pl.overlap_audit.json": "95d29efa6a0038913bcb91a4bdc6ba25aa85d081eff5f85fedc9c63e2f50e967",
+  "open_icm_pl.parquet": "11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31",
+  "open_icm_pl.qa.json": "0f9bf4531e33d0ec141b50ad08756ac90c663eac0786691ac80bb0f6f265dcce",
+  "open_icm_pl.sample.jsonl": "90559234a0e0debca6671470bb434a37fe5f7e6962dd12f1ecdb232a234844df",
+  "open_icm_pl.stats.json": "4c62ecf5979a1936bff91657d14ef60c27de5e3d71afd77419b2f855b7f57520"
+}

data/open_icm_pl/open_icm_pl.decisions.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f1d90bb052398bcd18eae457218f8b5b9d68ce8dd114168b1fff305570a4eda
+size 34473

data/open_icm_pl/open_icm_pl.md

@@ -0,0 +1,84 @@
+# OPEN ICM Polish publications pilot
+
+This target-specific slice is derived from the deterministic 100-record OPEN ICM pilot.
+
+- Acquired: 100; retained after source QA: 95
+- Exact title/DOI overlaps with `biblioteka_nauki` excluded: 10
+- Proposed records: 85
+- Proposed tokens: 2,358,174 (`cl100k_base` proxy)
+- Characters: 6,219,019; author coverage: 100.0%
+- License: per-record allowlisted CC BY or CC BY-SA variant
+- Immutable source snapshot: https://huggingface.co/datasets/PiotrSty/open-icm-pl-publications/tree/c4a20022e6947f55f82c3354bb5c86609ba681b8
+
+Per-record attribution, item URLs, license labels, normalized identifiers,
+bitstream metadata and checksums are preserved. The DynaWord proposal excludes
+all exact normalized-title or DOI matches found against 42,071 pinned
+`biblioteka_nauki` rows. Cross-source text deduplication and benchmark checks
+remain integration gates. Repository-generated TEXT can retain PDF extraction
+errors, omit figures and flatten tables. Item-level licensing does not establish
+the status of every quoted third-party passage.
+
+## Origin and license at the origin
+
+- Origin portal: **OPEN ICM** (Interdisciplinary Centre for Modelling, University
+  of Warsaw) - <https://open.icm.edu.pl/home>. Items are discovered through the
+  official DSpace API (`/server/api/core/items`), each record pins its
+  `landing_url`, e.g.
+  <https://open.icm.edu.pl/items/0078ca93-45a5-4bef-912d-297c069ee6ab>.
+- License statement **on the portal** (not the mirror): every item carries
+  `dc.rights` + `dc.rights.uri` in its origin metadata - e.g. the item above
+  declares `dc.rights: "Uznanie autorstwa 3.0 Polska"` with
+  `dc.rights.uri: http://creativecommons.org/licenses/by/3.0/pl/`, readable
+  without JavaScript at
+  <https://open.icm.edu.pl/server/api/core/items/0078ca93-45a5-4bef-912d-297c069ee6ab>
+  and rendered on the landing page. 77 of 85 records additionally ship a
+  per-item `license.txt` bitstream, pinned by checksum in
+  `open_icm_pl.attribution.jsonl` (`license_evidence`).
+- Allowlisted labels only (fail-closed gate in the fetch script):
+  CC-BY-3.0-PL (32), CC-BY-SA-3.0-PL (38), CC-BY-SA-3.0 (1), CC-BY-4.0 (7),
+  CC-BY-SA-4.0 (7). `Dozwolony użytek` and NC variants are rejected.
+
+## What the texts are
+
+- **Genre**: Polish open-access academic publications - journal articles
+  (economics, psychology, pedagogy, law), monographs and edited volumes,
+  conference texts. One record per publication.
+- **Register**: formal academic written Polish with abstracts, bibliographies
+  and byline blocks (authors, affiliations, ORCID, contact e-mails). Text is
+  repository-generated from PDFs (`is_ocr`-like extraction): math/font glyph
+  noise ("❬❜❡♥✐❛✳❛❧..."), hex-encoded UTF-16 fragments and flattened tables
+  survive in places.
+- **Time span**: `issued` 1998 - 2024-09 (the deterministic 100-record pilot;
+  6,183 items are eligible in total - yields must not be extrapolated).
+
+## Scrubber and residual-PII spot check
+
+Scrubber that ran: **pattern redaction in `src/fetch_open_icm_pl.py`** (e-mail ->
+`[REDACTED:EMAIL]`, labelled phone -> `[REDACTED:PHONE]`, ID/IBAN patterns;
+pipeline counts over the shard: 112 e-mail and 4 labelled-phone redactions).
+This is **not** [SlayerLab/NERGAL](https://huggingface.co/SlayerLab/NERGAL); the
+shard must be re-run through NERGAL before it is admitted to a release. Merge
+with `release: None` is fine meanwhile.
+
+Residual-PII spot check (60 of 85 records, seed 20260928): 85 `[REDACTED:EMAIL]`
+placeholders confirm the redaction works on clean contacts; **one real residual
+contact was found** - a glyph-garbled work e-mail in
+`open_icm_pl_6ec903ca-db89-4a17-acc8-850420c468ac` ("JKolbusz @wsiz.rzeszow.pl",
+stray space before `@` in a line mangled by broken font encoding) that the
+pattern missed. Benign non-hits: 161 digit runs (DOIs, URL slugs), 5 26-digit
+runs (hex-encoded UTF-16 extraction garbage), 1 IPv4-like hit (a CiteSeerX DOI
+URL fragment), 28 ORCIDs and 25 institutional byline addresses. Residual risk is
+MODERATE - extraction noise breaks pattern matching (the missed contact is
+proof), which is precisely why NERGAL (model-based) is the release gate.
+
+## Normalization and reproducibility note
+
+`src/fetch_open_icm_pl.py` downloads the repository-generated TEXT bitstreams
+(MD5-verified against declared checksums) and applies Unicode/whitespace
+normalization, PII pattern redaction, language/quality filtering and exact/near
+deduplication. Raw->shipped before/after excerpts cannot be regenerated
+anonymously at review time: the OPEN ICM bitstream endpoints serve an
+interactive CAPTCHA page to anonymous clients (the acquisition recorded in
+`attribution.jsonl` went through it). The raw bitstreams are pinned by
+checksum per record; the extraction-artifact residue is visible in the corpus
+samples.

data/open_icm_pl/open_icm_pl.overlap_audit.json

@@ -0,0 +1,819 @@
+{
+  "candidate_records": 100,
+  "method": "DOI or normalized title substring; fallback SequenceMatcher >= 0.90 over attribution titles",
+  "observed_at": "2026-09-10T08:08:03.451328+00:00",
+  "records_with_exact_match": 10,
+  "records_with_fuzzy_match": 0,
+  "results": [
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "0078ca93-45a5-4bef-912d-297c069ee6ab",
+      "title": "Przedstawienia rodziny w komiksie – uwikłania ideologiczne"
+    },
+    {
+      "doi": [
+        "10.34616/wwr.2019.2.373.383"
+      ],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Grażyna Lubowicka | 2019 | Żądanie Antygony. Normy i nienormatywny model rodziny Judith Butler",
+          "id": "biblioteka_nauki_articles_2127396"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "02ed3b1f-9b49-4d37-aef3-0fb5ea3644a7",
+      "title": "Żądanie Antygony. Normy i nienormatywny model rodziny Judith Butler"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "033e43b6-ce8f-471d-a9e9-ddae41889117",
+      "title": "Wybrane modele doskonałości jako przykład kształtowania jakości w placówkach medycznych"
+    },
+    {
+      "doi": [
+        "10.6084/m9.figshare.980699"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "047046a5-1eda-4f87-86a8-6322a0b1c060",
+      "title": "Przestrzeń i tożsamość. Refleksje z antropologii przestrzeni"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "08486a8e-0ea6-4334-a018-32f33f4055ec",
+      "title": "Wywiad z Andrzejem Klimczukiem: Razem kreatywniej"
+    },
+    {
+      "doi": [
+        "10.23734/wwr20152.083.102"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "0afdc44e-d37d-4180-a084-3f05bdcd7925",
+      "title": "Aktualność Jana Pawła II koncepcji wychowania w rodzinie"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "0dccc307-5583-4ea7-9261-9300edab5fc6",
+      "title": "Dzień dobry! Ale co to znaczy dobry dzień?"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "0e96ee0c-a323-443f-8a97-9ba0c7a097ce",
+      "title": "Współczesne koncepcje filozofii i etyki"
+    },
+    {
+      "doi": [
+        "10.5604/01.3001.0015.7110"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "114d4650-7aab-4262-b142-97c5a4db09b8",
+      "title": "Moment poniesienia kosztów inwestycji w Polskiej Strefie Inwestycji"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "14f37950-1cda-420c-82e3-bba1c95c913a",
+      "title": "Nietrwałość wzorców osobowych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "168531fb-d5ad-4946-b065-e039614c095c",
+      "title": "Przezwyciężanie kryzysu w procesie integracji europejskiej poprzez zmiany regulacji dotyczących udziału obywateli UE w podejmowaniu i realizacji decyzji"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "1a7a6fd5-de21-4bce-a186-a6c896ab7a1e",
+      "title": "Znaczenie wartości moralnych w działalności gospodarczej przedsiębiorcy"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "1f4656fd-bcb6-4577-b8b4-82aff31e3873",
+      "title": "Procesy poznawcze a uzależnienie od grania hazardowego – błędy myślenia nałogowych hazardzistów"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "232f5743-6863-48e3-a87a-12a2b6a5137c",
+      "title": "Efektywność i sposoby jej pomiaru oraz rola państwa w jej kreowaniu"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "23f7e569-c7e0-4e49-a38e-d7896bacf996",
+      "title": "Zbieranie fragmentów, czyli literackie egzystencje Europy Środkowej"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "241e63dc-f2b5-4e65-a425-c0a142ae2f12",
+      "title": "Nowoczesny/ponowoczesny – Grotowski w poszukiwaniu esencji tożsamości Performera"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "28253160-f06b-45cf-84b0-22d9bf73773c",
+      "title": "Pielgrzymowanie i turystyka religijna na przykładzie ośrodków pielgrzymkowych w diecezji krakowskiej"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "29a47f77-0b5c-4a0a-9949-c50b789e8d5b",
+      "title": "Wyzwania dla jakości w kontekście wymagań normy spalinowej Euro6d – studium przypadku projektu badawczo-rozwojowego realizowanego przy wsparciu NCBR"
+    },
+    {
+      "doi": [
+        "10.23734/wwr20171.361.380"
+      ],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Beata Gromadzka | 2017 | Kryzys w rodzinie czy kryzys pisania o rodzinie? Na podstawie literatury dla młodzieży na przełomie XX i XXI wieku",
+          "id": "biblioteka_nauki_articles_2105687"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "2a574e9b-718c-4231-8078-5df5920fc4f3",
+      "title": "Kryzys w rodzinie czy kryzys pisania o rodzinie? Na podstawie literatury dla młodzieży na przełomie XX i XXI wieku"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "3cc80540-3717-4da7-b93b-dafe5713d3bf",
+      "title": "Człowiek zalogowany 2. Wirtualne społeczności"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "3e1de4ed-5bab-43b9-a0e4-50bb9984ce3e",
+      "title": "Coaching jako kluczowy instrument wspierający pracę menedżera"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "410c4772-f264-4874-9ab3-e9cca1d04f20",
+      "title": "Retoryka w służbie homiletyki"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "42111533-761c-4afd-b565-170ef8943c2c",
+      "title": "„Witajcie w nowej(?) bajce...” O promocji współczesnych bibliotek szkolnych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "43b4bf1c-ced7-4bbc-befd-87458955a228",
+      "title": "Kompetencja społeczna i inteligencja emocjonalna a zaangażowanie w Internet"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "44e7e9f0-254a-4f6d-bc36-4abcc23b8d1d",
+      "title": "Uwarunkowania rozwoju połączeń intermodalnych w obsłudze Terminalu Promowego Świnoujście"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "48a62029-a25d-4f60-bdd9-d3be82ab5cff",
+      "title": "Być w Górce Klasztornej i pamiętać Kalwarię"
+    },
+    {
+      "doi": [
+        "10.12657/czageo-"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "49250798-98c0-432f-bcdb-b78eebde7f43",
+      "title": "Samoocena poziomu stresu mieszkańców Trójmiasta, Kielc i Lwowa podczas pandemii COVID-19"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "4a657320-9b3f-4281-abee-1b3b7f0ce452",
+      "title": "Świadkowie cyberbullyingu – między obojętnością a reakcją"
+    },
+    {
+      "doi": [
+        "10.61905/wwr/175101"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "53a41688-3bd3-49f9-bb41-0b8cacfc24f8",
+      "title": "Macierzyństwo podczas studiów w opinii studentek pedagogiki – komunikat z badań"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "5841c253-2189-4702-898d-602318542ec1",
+      "title": "Koncepcja pomiaru efektywności nauczycieli szkół średnich w Polsce"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "596d6e52-f282-4235-adf3-d48fcbae050d",
+      "title": "Referat powinien być niczym pocałunek… Metodyczne przesłanie Profesora Floriana Plita wygłoszone do koleżanek i kolegów Wydziału Geografii i Studiów Regionalnych Uniwersytetu Warszawskiego na jednym z pożegnań"
+    },
+    {
+      "doi": [
+        "10.24131/3724.24ns01"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "5a22a282-110d-4110-a11c-f837adf530a1",
+      "title": "Naturalne kręgi życia i wychowania człowieka"
+    },
+    {
+      "doi": [
+        "10.12797/9788383680491"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "5fc6e6dd-112b-4c8f-9ced-f1ede5525338",
+      "title": "Polskie studia wywiadowcze. Stan i perspektywy badań"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "627960f5-2ef2-44a6-83cd-6784d1f8b601",
+      "title": "Fotocodzienność. Próba analizy fenomenu aplikacji Instagram"
+    },
+    {
+      "doi": [
+        "10.31261/ppk.2021.05.08"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "635e0996-cfbe-4d40-863b-3055e3c35ace",
+      "title": "Przyjęcie do wykonania kary orzeczonej przez sądy państwa obcego a prawo łaski"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "65145d4d-d75d-4423-afec-7883814858ab",
+      "title": "Zastosowanie polimorfizmu mikrosatelitarnego DNA w kontroli rodowodów koni"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Anna Kacperczyk | 2014 | Autoetnografia – technika, metoda, nowy paradygmat? O metodologicznym statusie autoetnografii",
+          "id": "biblioteka_nauki_articles_623235"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "65bb87ae-682b-4b89-a57b-38f2a770a237",
+      "title": "Autoetnografia – technika, metoda, nowy paradygmat? O metodologicznym statusie autoetnografii"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "6922536c-8233-468b-8a4a-f82bc1a4a1ec",
+      "title": "Przyrodnicze uwarunkowania rozwoju ekoturystyki w krajach Azji Centralnej powstałych po rozpadzie Związku Radzieckiego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "6983bc0a-2b5d-4df6-9ccf-2313d845a7da",
+      "title": "Ewolucja programu w zakresie konkurencji pływackich oraz udział polskich pływaków w nowożytnych igrzyskach olimpijskich"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Marceli Kosman | 2018 | Stanisław Sławomir Nicieja, Kresowa Atlantyda. Historia i mitologia miast kresowych, t. X: Złoczów, Zadwórze, Kozaki, Gańczary, Łanowce, Zasmyki",
+          "id": "biblioteka_nauki_articles_616075"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "6d1010ff-08f8-4e4d-b529-347de48f82fc",
+      "title": "Złoczów"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "6ec903ca-db89-4a17-acc8-850420c468ac",
+      "title": "Innowacyjna Gmina. Informatyka w jednostkach samorządu terytorialnego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "6efbf73b-1271-4dff-825d-af549937c63c",
+      "title": "Ograniczenia rozwoju socjologii krytycznej. Próba alternatywnego wyjaśnienia"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "6f1b2524-2c2e-43bb-b980-4dc0176a7483",
+      "title": "Kategoria zadowolenia jako kryterium oceny konferencji (KSOW, PO Ryby 2007-2013)"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "6fe4e956-84a4-43f5-b865-7515a92700e8",
+      "title": "Polityka zagraniczna Hiszpanii w latach 1833-1914"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "70b6406d-d826-4c12-a210-71389f224e14",
+      "title": "Archiwalia pruskich izb handlowych oraz izb  gospodarczych III Rzeszy"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "72135cf8-5c7c-40e4-ae0f-d6f09653d808",
+      "title": "Koncepcja pomiaru efektywności pracy nauczycieli szkół średnich w Polsce"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "727d10bb-6b67-4995-9abe-152679f6dadc",
+      "title": "Orzecznictwo organów dyscyplinarnych w wybranych zawodach zaufania publicznego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "77cdac24-9c59-4b05-ac1f-11d5db87d0af",
+      "title": "Bez obrazu – nieistnienie: o traumatyzującej obawie przed byciem pominiętym przez obraz"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "7b9ddf4a-dca9-47e0-91c6-3fe5951494d9",
+      "title": "Geobotaniczne wskaźniki środowiska wysokogórskiego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "7be89182-597e-4cdc-ad6c-684c1947c7f3",
+      "title": "Szczegółowe poziome osnowy geodezyjne - przeszłość, stan aktualny i przyszłość w zakresie ich zakładania i funkcjonowania"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "7d5c4210-d8c7-4f61-8c0e-d1ea4cbbc8f7",
+      "title": "Pacynka kontra Faceopera. Zjawisko kont alternatywnych w mediach społecznościowych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "85a251c7-d59f-4de2-9ba7-d7e68ed396bf",
+      "title": "Temat śmierci w twórczości teatralnej i literackiej dla dzieci w początkach XXI wieku"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "8a1cca15-e1b9-4d23-b67c-6f48f649c616",
+      "title": "Wykaz publikacji w języku polskim poświęconych zarządzaniu i organizacji bibliotek (w wyborze)"
+    },
+    {
+      "doi": [
+        "10.26348/znwwsi.30.27"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "8c8e1344-c7b4-4247-962a-8fcd7c635f14",
+      "title": "Nauka programowania z ChatGPT: skrót do sukcesu czy droga na manowce?"
+    },
+    {
+      "doi": [
+        "10.61905/wwr/176527"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "8d289c40-5a70-4d24-8a4f-9c0453065268",
+      "title": "Komponenty resilience rodzin z defaworyzowanych sąsiedztw w obiektywie badań longitudinalnych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "8da1dee3-e9be-437b-8916-e9a8d3ac92b2",
+      "title": "Uzasadnienia konieczności postępowania zgodnie z zasadami moralnymi w działaniach marketingowych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Robert T. Tomczak | 2015 | Wychowanie i wykształcenie króla Michała Korybuta Wiśniowieckiego (1640–1660)",
+          "id": "biblioteka_nauki_articles_631708"
+        },
+        {
+          "attribution": "Ilona Czamańska | 2015 | Suplement do suplementu. Odpowiedź na recenzję Zbigniewa Anusika recenzję książki Wiśniowieccy. Monografia rodu",
+          "id": "biblioteka_nauki_articles_632035"
+        },
+        {
+          "attribution": "Robert T. Tomczak | 2015 | Wychowanie i wykształcenie króla Michała Korybuta Wiśniowieckiego (1640–1660)",
+          "id": "biblioteka_nauki_articles_632236"
+        },
+        {
+          "attribution": "Ilona Czamańska | 2015 | Suplement do suplementu. Odpowiedź na recenzję Zbigniewa Anusika recenzję książki Wiśniowieccy. Monografia rodu",
+          "id": "biblioteka_nauki_articles_632241"
+        },
+        {
+          "attribution": "Anna Penkała-Jastrzębska | 2021 | Janusz Antoni Wiśniowiecki, Ilias Polski, wyd. Przemysław P. Romaniuk, Jacek Burdowicz- -Nowicki, (Warszawa: Wydawnictwo Neriton, 2018), ISBN 978-83-660118-01-3",
+          "id": "biblioteka_nauki_articles_1834618"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "91a67fe4-4510-44dd-9a55-75bfb5f79392",
+      "title": "Wiśniowiec"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "92596827-a7ac-4757-839a-cb6d5671ab67",
+      "title": "Polityka handlowa Unii Europejskiej na podstawie implementacji założeń ruchu Sprawiedliwego handlu"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "93db5006-5cd3-4829-bc88-9b89650c5f1a",
+      "title": "Embrion ludzki - pytanie o status"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "95f04437-6495-4fb9-80c7-83a6bec4b6fe",
+      "title": "Postawy prokreacyjne młodych Polaków"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "9787af4b-b748-4921-9dd1-a5a7213bd6c9",
+      "title": "Wspólnota Izraela domem Boga w Starym Testamencie"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "99759031-988b-49d3-95f0-2f063d41102e",
+      "title": "Podstawowe zadanie filozofii w kulturze"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "9bca9fcc-6780-4f36-b684-0c87cf04dbee",
+      "title": "Kalila i Dimna. Edycje i rękopisy dzieła (Opowieść o sowach i krukach)"
+    },
+    {
+      "doi": [
+        "10.11649/sfps.2017.004"
+      ],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Radosław Kaleta | 2017 | O potrzebie ujednolicenia sposobów transliteracji imion i nazwisk białoruskich w Polsce",
+          "id": "biblioteka_nauki_articles_678651"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "9c3e3246-4dc7-4d48-b541-aaf800d18aa7",
+      "title": "O potrzebie ujednolicenia sposobów transliteracji  imion i nazwisk białoruskich w Polsce"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "9c64aa62-1714-48ed-84d4-6533d3fa1ef0",
+      "title": "Turystyka zrównoważona - przegląd literatury i charakterystyka"
+    },
+    {
+      "doi": [
+        "10.6084/m9.figshare.1335808"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "a00da632-20b2-433d-9da3-4e56e807a01b",
+      "title": "Hentai – erotyka z mangi i anime"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Maciej Szeląg | 2017 | Wpływ składu kompozytów cementowych na geometrię ich spękań termicznych | Politechnika Lubelska. Instytut Informatyki",
+          "id": "biblioteka_nauki_books_2028483"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "a0221582-b4ab-4a7f-9c7a-80a4b71eb625",
+      "title": "Wpływ składu kompozytów cementowych na geometrię ich spękań termicznych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "a06303dc-010c-4e93-8784-f733781fdb6e",
+      "title": "Miasta restytuowane w obliczu przemian demograficznych w XXI wieku"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "a874709f-88e5-435d-9135-c4f538d26813",
+      "title": "Człowiek pogryzł psa. Uwagi o „wolności twórczej” jako konsekwencji Kantowskiej figury geniusza"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "a89ebb40-9207-42a5-b0b9-3d6a7d05761c",
+      "title": "Atlas  Wyszehradzki,  P.  Śleszyński,  K.  Czapiewski  (red.),  Instytut  Współpracy  Polsko-Węgierskiej im. Wacława Felczaka, Polskie Towarzystwo Geograficzne, Warszawa, 2021, 303 ss."
+    },
+    {
+      "doi": [
+        "10.24131/3724.24ns06"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "a95ccc2b-4156-45c7-9a4d-ec154b9c1c2b",
+      "title": "Retoryka w szkole wobec edukacji klasycznej"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "aaec30a0-28cf-4d11-b913-d2d969f15660",
+      "title": "Kształtowanie strategii przedsiebiorstwa rodzinnego w korelacji do oczekiwań interesariuszy wewnetrznych i zewnetrznych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "acd678a4-5a30-4e7d-a142-7a0458f7c509",
+      "title": "Duchowość i religijność a wybór zawodu przez młodych ludzi"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "ad7ccc8e-1623-4bef-84ab-58e8b3984beb",
+      "title": "Dziedzictwo kulturalne i jego ochrona w kontekście infrastruktury krytycznej"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "ade27972-dc75-4646-88c8-34d39186efda",
+      "title": "Nie tylko praca socjalna: komunikacyjne aspekty funkcjonowania instytucji pomocy społecznej w przestrzeni publicznej; pomiędzy stereotypami i barierami a public relations i aktywną promocją"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "af211736-e429-466e-8722-ab11ee0438da",
+      "title": "Powoływanie, struktura i zadania Parafialnej Rady do spraw Ekonomicznych"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "b013047b-1b9a-4a3c-8540-a5a91972bd3b",
+      "title": "Sztuka w przestrzeni publicznej. Artystyczne wymiary wytwarzania kapitału społecznego i kulturowego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "b264af0a-92b3-4251-ae3a-5e8a31a739ea",
+      "title": "Walka byków oczami Sienkiewicza"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "bb7441c2-1107-4d52-865d-72f2201c3988",
+      "title": "Aktywizacja seniorów w procesie poradnictwa"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "bd474ffb-e622-49eb-a1c0-06c86d849884",
+      "title": "Naród wobec rewolucji"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "bd595f0f-4fcf-49cc-81a8-a4a0d967233c",
+      "title": "Szanse patriotyzmu w „globalnej wiosce”"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "bdb254e8-ab6d-466e-b1de-7c52157a9c35",
+      "title": "Rola kwasów tłuszczowych w żywieniu i zdrowiu człowieka"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "c47bc2d8-0214-4715-87e2-720492ee3da6",
+      "title": "Masaż Shantala – charakterystyka i sposób wykonania"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "c6fbda2c-6329-46d5-b7bb-d87277e3a108",
+      "title": "Retoryka rzeczy a proces zadomowienia w przestrzeniach pojałtańskiego okcydentu"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "c9f8df0c-d745-469c-a73d-574d875216fc",
+      "title": "Procesy logistyczne portowego terminalu kontenerowego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "cafebfde-e64a-4504-a76e-f388b552138d",
+      "title": "Projektowanie technologiczne oraz dobór deskowań stosowanych w budownictwie inżynieryjnym na przykładzie budownictwa mostowego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "cc630b22-1ef1-4cb2-8e9a-ec5e12dd2d3e",
+      "title": "Dlaczego uchodźcy „nie chcą” integrować się w Polsce? Ocena skuteczności programów integracyjnych z punktu widzenia uchodźców"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "d00d2c6c-d2c5-4735-bd56-1448427ca86c",
+      "title": "Ocena potencjału przedsiębiorczości i atrakcyjności gospodarczej obszarów wiejskich województwa świętokrzyskiego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "d9e454b5-1fb8-4aaf-9b14-76162ecd37e6",
+      "title": "Przemoc instytucjonalna w Jednostkach Organizacyjnych Pomocy Społecznej"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "de08c2f1-518b-4114-b7d3-b6c357e28852",
+      "title": "Kobieta w roli matki. Propozycje Anastazji z Jełowickich Dzieduszyckiej (1842–1890)"
+    },
+    {
+      "doi": [
+        "10.12657/czageo-94-30"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "e21f9705-9966-4828-8605-42a54e591aa9",
+      "title": "Blockchain i kryptowaluty – „nowa” odsłona w geograficznym podejściu do badania Internetu"
+    },
+    {
+      "doi": [
+        "10.61905/wwr/192723"
+      ],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "e346c430-9e67-4fae-97fa-463433a508c5",
+      "title": "Prawne i etyczne aspekty opieki nad niesamodzielnymi rodzicami"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "e38b08ae-4054-4a16-bacc-c5501579799a",
+      "title": "Systemy produkcyjne nowej generacji. Modele interdyscyplinarne"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "e3a11a86-7260-4b61-a043-56e418b05b7b",
+      "title": "Badanie efektywności ataków socjotechnicznych w jednostkach samorządu terytorialnego"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "ebd3c5a6-e7ac-4ac9-80f6-57cc9c6eea2e",
+      "title": "Wizualna konsumpcja a komunikacja wizualna. Perswazyjność obrazu w ramach współczesnej komunikacji marketingowej"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "ec993a0f-e213-49e2-8bf5-8aaa4009688b",
+      "title": "Imitacja w tworzeniu polskiej polityki integracji cudzoziemców"
+    },
+    {
+      "doi": [
+        "10.23734/wwr20172.075.088"
+      ],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Danuta Kowalczyk | 2017 | Pedagogiczny wymiar sądowego rozstrzygania konfliktów rodzinnych",
+          "id": "biblioteka_nauki_articles_2105383"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "ecbbe7bc-f0e8-465e-ab0b-f6cecd48980c",
+      "title": "Pedagogiczny wymiar sądowego rozstrzygania konfliktów rodzinnych"
+    },
+    {
+      "doi": [
+        "10.34616/wwr.2019.2.357.372"
+      ],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Dorota Rybczyńska Abdel-Kawy; Łukasz Rybczyński | 2019 | Podstawowe cechy osobowości a kompetencje społeczne uczestników programu terapii substytucyjnej i ich rola w procesie zmian",
+          "id": "biblioteka_nauki_articles_2127349"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "f8cc967a-0de5-41a4-9184-d10dcbc88e28",
+      "title": "Podstawowe cechy osobowości a kompetencje społeczne uczestników programu terapii substytucyjnej i ich rola w procesie zmian"
+    },
+    {
+      "doi": [],
+      "exact_title_or_doi_matches": [],
+      "fuzzy_title_matches": [],
+      "item_id": "fc0390e6-5fdb-4857-bcd9-225bdf41bd47",
+      "title": "Pseudo-Ksenofont"
+    },
+    {
+      "doi": [
+        "10.34616/wwr20182.053.068"
+      ],
+      "exact_title_or_doi_matches": [
+        {
+          "attribution": "Aneta Bołdyrew | 2018 | Przemoc fizyczna i seksualna wobec dzieci w środowisku rodzinnym jako problem społeczno-pedagogiczny w publicystyce Królestwa Polskiego na przełomie XIX i XX wieku",
+          "id": "biblioteka_nauki_articles_2123449"
+        }
+      ],
+      "fuzzy_title_matches": [],
+      "item_id": "fcb6b503-49bb-4e2d-b997-19d22904c6cd",
+      "title": "Przemoc fizyczna i seksualna wobec dzieci w środowisku rodzinnym jako problem społeczno-pedagogiczny w publicystyce Królestwa Polskiego na przełomie XIX i XX wieku"
+    }
+  ],
+  "target": "SlayerLab/polish-dynaword:data/biblioteka_nauki",
+  "target_revision": "863b8fa854fc9741e4791f1ad90ecaf6f2c0e29a",
+  "target_rows": 42071,
+  "text_overlap": "not tested; target-wide text dedup remains an integration gate"
+}

data/open_icm_pl/open_icm_pl.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11deaba8bd43159bc7d07fc022f84e1555809cb04f5f5ce4e0f3bde953dd5c31
+size 2584960

data/open_icm_pl/open_icm_pl.qa.json

@@ -0,0 +1,16 @@
+{
+  "benchmark_overlap": "pending",
+  "cross_source_text_dedup": "pending target integration",
+  "fuzzy_title_matches": 0,
+  "known_target_title_or_doi_overlap_excluded": 10,
+  "limitations": [
+    "repository-generated TEXT may inherit PDF extraction errors",
+    "figures are omitted and tables may be flattened",
+    "item-level license metadata does not prove every quoted third-party passage is reusable",
+    "pattern checks are not comprehensive de-identification",
+    "pilot yield and token counts must not be extrapolated to all 6183 eligible records"
+  ],
+  "scope": "DynaWord proposal derived from the deterministic OPEN ICM 100-record pilot",
+  "source_text_qa_rejected": 5,
+  "within_source_exact_and_near_dedup": true
+}

data/open_icm_pl/open_icm_pl.stats.json

@@ -0,0 +1,19 @@
+{
+  "added": "2026-09-10",
+  "author_coverage": 1.0,
+  "characters": 6219019,
+  "kept": 85,
+  "license_counts": {
+    "CC-BY-3.0-PL": 32,
+    "CC-BY-4.0": 7,
+    "CC-BY-SA-3.0": 1,
+    "CC-BY-SA-3.0-PL": 38,
+    "CC-BY-SA-4.0": 7
+  },
+  "sample_count": 12,
+  "source_pilot_acquired": 100,
+  "source_pilot_kept": 95,
+  "target_overlap_excluded": 10,
+  "target_revision": "fe44303ea2db290c7a195264226477f724973357",
+  "tokens": 2358174
+}

src/fetch_open_icm_pl.py

@@ -0,0 +1,909 @@
+#!/usr/bin/env python3
+"""Build an auditable pilot of permissively licensed Polish OPEN ICM texts."""
+from __future__ import annotations
+
+import argparse
+from collections import Counter
+from concurrent.futures import ThreadPoolExecutor, as_completed
+from datetime import datetime, timezone
+from difflib import SequenceMatcher, get_close_matches
+import gzip
+import hashlib
+import json
+import os
+from pathlib import Path
+import re
+import time
+import unicodedata
+
+import requests
+
+
+SOURCE = "open_icm_pl"
+OWN_REPO = "PiotrSty/open-icm-pl-publications"
+TARGET = "SlayerLab/polish-dynaword"
+API = "https://open.icm.edu.pl/server/api"
+SOURCE_URL = "https://open.icm.edu.pl/home"
+FIELDS = ["id", "text", "source", "added", "created", "token_count", "license", "author"]
+UA = "OpenICMPolishCorpusResearch/0.1 (PiotrSty; open research pilot)"
+MAX_TEXT_BYTES = 20 * 1024 * 1024
+MIN_TEXT_CHARS = 3_000
+POOL_MULTIPLIER = 2
+EMAIL_RE = re.compile(r"(?i)\b[A-Z0-9._%+-]+@[A-Z0-9.-]+\.[A-Z]{2,}\b")
+PHONE_RE = re.compile(r"(?i)(?:\btelefon|\btel\.|\bphone)\s*:?[ \t]*(?:\+48[ \t]*)?\d(?:[ .-]?\d){8}\b")
+
+LICENSES = {
+    "Uznanie autorstwa-Na tych samych warunkach 3.0 Polska": "CC-BY-SA-3.0-PL",
+    "Uznanie autorstwa 3.0 Polska": "CC-BY-3.0-PL",
+    "Uznanie autorstwa 4.0 Międzynarodowe": "CC-BY-4.0",
+    "Uznanie autorstwa-Na tych samych warunkach 4.0 Międzynarodowe": "CC-BY-SA-4.0",
+    "Uznanie autorstwa 3.0 Unported": "CC-BY-3.0",
+    "Uznanie autorstwa-Na tych samych warunkach 3.0 Unported": "CC-BY-SA-3.0",
+    "Uznanie autorstwa 2.0 Polska": "CC-BY-2.0-PL",
+    "Uznanie autorstwa-Na tych samych warunkach 3.0": "CC-BY-SA-3.0",
+}
+
+
+class AcquisitionBarrier(RuntimeError):
+    """Raised when OPEN ICM returns an interactive download barrier."""
+
+
+def now():
+    return datetime.now(timezone.utc).isoformat()
+
+
+def digest(value):
+    if not isinstance(value, bytes):
+        value = json.dumps(value, ensure_ascii=False, sort_keys=True, separators=(",", ":")).encode("utf-8")
+    return hashlib.sha256(value).hexdigest()
+
+
+def save(path, value):
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(json.dumps(value, ensure_ascii=False, sort_keys=True, indent=2) + "\n", encoding="utf-8")
+
+
+def write_lines(path, rows):
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text("".join(json.dumps(row, ensure_ascii=False, sort_keys=True) + "\n" for row in rows), encoding="utf-8")
+
+
+def read_lines(path):
+    # JSONL records may contain Unicode line-separator characters inside strings.
+    return [json.loads(line) for line in path.read_text(encoding="utf-8").split("\n") if line]
+
+
+def load(path):
+    return json.loads(path.read_text(encoding="utf-8"))
+
+
+def values(metadata, key):
+    return [str(item.get("value", "")).strip() for item in metadata.get(key, []) if str(item.get("value", "")).strip()]
+
+
+def request(url, params=None, attempts=5, timeout=(15, 60)):
+    response = None
+    for attempt in range(attempts):
+        response = requests.get(url, params=params, headers={"User-Agent": UA}, timeout=timeout)
+        if response.status_code not in (429, 500, 502, 503, 504):
+            response.raise_for_status()
+            return response
+        time.sleep(2 ** attempt)
+    response.raise_for_status()
+
+
+def request_json(url, params=None):
+    return request(url, params=params).json()
+
+
+def embedded(payload, name):
+    return payload.get("_embedded", {}).get(name, [])
+
+
+def item_objects(payload):
+    result = payload.get("_embedded", {}).get("searchResult", {})
+    objects = result.get("_embedded", {}).get("objects", [])
+    return [row.get("_embedded", {}).get("indexableObject", {}) for row in objects]
+
+
+def compact_item(item):
+    metadata = item.get("metadata", {})
+    keys = [
+        "dc.contributor.author", "dc.contributor.editor", "dc.date.issued", "dc.description.abstract",
+        "dc.identifier.doi", "dc.identifier.uri", "dc.language.iso", "dc.publisher", "dc.rights",
+        "dc.subject", "dc.title", "dc.type",
+    ]
+    return {
+        "item_id": item.get("uuid") or item.get("id"),
+        "title": item.get("name", ""),
+        "handle": item.get("handle"),
+        "last_modified": item.get("lastModified"),
+        "metadata": {key: metadata.get(key, []) for key in keys if metadata.get(key)},
+    }
+
+
+def is_polish_open_item(item):
+    metadata = item.get("metadata", {})
+    languages = {value.casefold() for value in values(metadata, "dc.language.iso")}
+    rights = set(values(metadata, "dc.rights"))
+    return any(language == "pl" or language.startswith("pl_") for language in languages) and bool(rights & LICENSES.keys())
+
+
+def deterministic_pool(items, target):
+    unique = {}
+    for item in items:
+        item_id = item["item_id"]
+        if item_id not in unique:
+            unique[item_id] = item
+        else:
+            old_rights = set(values(unique[item_id]["metadata"], "dc.rights"))
+            new_rights = set(values(item["metadata"], "dc.rights"))
+            unique[item_id]["metadata"]["dc.rights"] = [{"value": value} for value in sorted(old_rights | new_rights)]
+    ordered = sorted(unique.values(), key=lambda item: digest(("open-icm-pilot:" + item["item_id"]).encode("utf-8")))
+    chosen = []
+    for license_label in LICENSES:
+        match = next((item for item in ordered if license_label in values(item["metadata"], "dc.rights")), None)
+        if match and match["item_id"] not in {row["item_id"] for row in chosen}:
+            chosen.append(match)
+    wanted = min(len(ordered), target * POOL_MULTIPLIER)
+    chosen_ids = {row["item_id"] for row in chosen}
+    chosen.extend(item for item in ordered if item["item_id"] not in chosen_ids and len(chosen) < wanted)
+    return chosen[:wanted]
+
+
+def discover(out, target):
+    if (out / "source_manifest.jsonl").exists():
+        raise ValueError("source manifest exists; use a fresh directory for an immutable discovery run")
+    out.mkdir(parents=True, exist_ok=True)
+    raw_pages = out / "raw_api"
+    all_items, pages, counts = [], [], {}
+    for license_index, license_label in enumerate(LICENSES):
+        page = 0
+        while True:
+            params = [
+                ("size", 100), ("page", page), ("sort", "dc.date.issued,DESC"),
+                ("query", "dc.language.iso:pl"), ("f.license", f"{license_label},equals"),
+            ]
+            response = request(f"{API}/discover/search/objects", params=params)
+            raw = response.content
+            payload = response.json()
+            result = payload.get("_embedded", {}).get("searchResult", {})
+            page_info = result.get("page", {})
+            items = [compact_item(item) for item in item_objects(payload) if is_polish_open_item(item)]
+            compressed_path = raw_pages / f"license-{license_index:02d}" / f"page-{page:04d}.json.gz"
+            compressed_path.parent.mkdir(parents=True, exist_ok=True)
+            compressed_path.write_bytes(gzip.compress(raw, mtime=0))
+            pages.append({
+                "license": license_label, "page": page, "response_sha256": digest(raw),
+                "compressed_path": compressed_path.relative_to(out).as_posix(), "items": len(items),
+                "reported_total": int(page_info.get("totalElements", 0)),
+            })
+            all_items.extend(items)
+            counts[license_label] = int(page_info.get("totalElements", 0))
+            if page + 1 >= int(page_info.get("totalPages", 0)):
+                break
+            page += 1
+    unique = {item["item_id"]: item for item in all_items}
+    manifest = sorted(unique.values(), key=lambda item: item["item_id"])
+    pool = deterministic_pool(manifest, target)
+    write_lines(out / "source_manifest.jsonl", manifest)
+    write_lines(out / "selection_pool.jsonl", pool)
+    save(out / "api_pages.json", pages)
+    selection = {
+        "source": SOURCE, "source_url": SOURCE_URL, "observed_at": now(), "target": target,
+        "pool_size": len(pool), "eligible_records": len(manifest), "reported_counts_by_license": counts,
+        "accepted_licenses": LICENSES, "language_query": "dc.language.iso:pl",
+        "selection_method": "one deterministic representative per license, then lowest SHA-256 ranks",
+        "source_manifest_sha256": digest((out / "source_manifest.jsonl").read_bytes()),
+        "api_pages_sha256": digest(pages),
+    }
+    save(out / "selection.json", selection)
+    print(json.dumps({"eligible": len(manifest), "pool": len(pool), "target": target, "counts": counts}, ensure_ascii=False, indent=2))
+
+
+def bundle_map(item_id):
+    payload = request_json(f"{API}/core/items/{item_id}/bundles", {"size": 20})
+    return {bundle.get("name"): bundle for bundle in embedded(payload, "bundles")}
+
+
+def bundle_bitstreams(bundle):
+    if not bundle:
+        return []
+    return embedded(request_json(f"{API}/core/bundles/{bundle['uuid']}/bitstreams", {"size": 100}), "bitstreams")
+
+
+def bitstream_summary(bitstream):
+    return {
+        "id": bitstream.get("uuid") or bitstream.get("id"), "name": bitstream.get("name"),
+        "bytes": int(bitstream.get("sizeBytes", 0)), "checksum": bitstream.get("checkSum"),
+        "content_url": bitstream.get("_links", {}).get("content", {}).get("href"),
+        "metadata_rights": values(bitstream.get("metadata", {}), "dc.rights"),
+    }
+
+
+def choose_text(bitstreams):
+    candidates = [item for item in bitstreams if item.get("_links", {}).get("content", {}).get("href")]
+    if not candidates:
+        raise ValueError("missing downloadable TEXT bitstream")
+    return max(candidates, key=lambda item: (int(item.get("sizeBytes", 0)), item.get("name", "")))
+
+
+def verify_checksum(payload, metadata):
+    checksum = metadata.get("checkSum") or {}
+    algorithm = str(checksum.get("checkSumAlgorithm", "")).upper()
+    expected = checksum.get("value")
+    if algorithm == "MD5" and expected and hashlib.md5(payload).hexdigest() != expected:
+        raise ValueError("downloaded bitstream MD5 differs from repository metadata")
+
+
+def detect_download_barrier(response):
+    final_url = str(getattr(response, "url", ""))
+    content_type = str(getattr(response, "headers", {}).get("Content-Type", "")).casefold()
+    prefix = bytes(getattr(response, "content", b""))[:512].lower()
+    if "/captcha.html" in final_url or ("text/html" in content_type and b"<html" in prefix):
+        raise AcquisitionBarrier(
+            "OPEN ICM returned an interactive CAPTCHA page instead of the bitstream; "
+            "request an official bulk-download route before acquisition"
+        )
+
+
+def probe_download_access(item):
+    bundles = bundle_map(item["item_id"])
+    text_stream = choose_text(bundle_bitstreams(bundles.get("TEXT")))
+    response = request(text_stream["_links"]["content"]["href"], timeout=(15, 120))
+    detect_download_barrier(response)
+
+
+def text_stream_metadata(item):
+    bundles = bundle_map(item["item_id"])
+    return {
+        "text_stream": choose_text(bundle_bitstreams(bundles.get("TEXT"))),
+        "license_streams": bundle_bitstreams(bundles.get("LICENSE")),
+        "original_streams": bundle_bitstreams(bundles.get("ORIGINAL")),
+    }
+
+
+def acquire_one(out, item):
+    item_id = item["item_id"]
+    streams = text_stream_metadata(item)
+    text_stream = streams["text_stream"]
+    licenses = streams["license_streams"]
+    originals = streams["original_streams"]
+    text_bytes = int(text_stream.get("sizeBytes", 0))
+    if text_bytes <= 0 or text_bytes > MAX_TEXT_BYTES:
+        raise ValueError(f"TEXT bitstream outside 1..{MAX_TEXT_BYTES} byte pilot bound")
+    response = request(text_stream["_links"]["content"]["href"], timeout=(15, 120))
+    detect_download_barrier(response)
+    payload = response.content
+    if len(payload) != text_bytes:
+        raise ValueError("downloaded TEXT byte count differs from repository metadata")
+    verify_checksum(payload, text_stream)
+    raw_path = out / "raw_text" / f"{item_id}.txt"
+    raw_path.parent.mkdir(parents=True, exist_ok=True)
+    raw_path.write_bytes(payload)
+    license_evidence = []
+    for index, stream in enumerate(licenses):
+        content_url = stream.get("_links", {}).get("content", {}).get("href")
+        if not content_url or int(stream.get("sizeBytes", 0)) > 1_000_000:
+            continue
+        license_payload = request(content_url).content
+        verify_checksum(license_payload, stream)
+        license_path = out / "license_evidence" / f"{item_id}-{index:02d}.txt"
+        license_path.parent.mkdir(parents=True, exist_ok=True)
+        license_path.write_bytes(license_payload)
+        license_evidence.append({**bitstream_summary(stream), "sha256": digest(license_payload),
+                                 "path": license_path.relative_to(out).as_posix()})
+    return {
+        **item, "landing_url": f"https://open.icm.edu.pl/items/{item_id}",
+        "accepted_license_labels": sorted(set(values(item["metadata"], "dc.rights")) & LICENSES.keys()),
+        "text_bitstream": {**bitstream_summary(text_stream), "sha256": digest(payload),
+                           "path": raw_path.relative_to(out).as_posix()},
+        "license_evidence": license_evidence,
+        "original_bitstreams": [bitstream_summary(stream) for stream in originals],
+        "observed_at": now(),
+    }
+
+
+def manual_acquisition(out):
+    path = out / "manual_acquisition.json"
+    if path.exists():
+        return load(path)
+    selection = load(out / "selection.json")
+    return {
+        "source": SOURCE, "observed_at": now(), "target": int(selection["target"]),
+        "pool_size": int(selection["pool_size"]), "mode": "manual_browser_download",
+        "selected": [], "rejected": [],
+    }
+
+
+def sync_manual_acquisition(out, state):
+    state["observed_at"] = now()
+    state["attempted"] = len(state["selected"]) + len(state["rejected"])
+    state["target_reached"] = len(state["selected"]) >= int(state["target"])
+    state["manual_confirmation_required"] = not state["target_reached"]
+    save(out / "manual_acquisition.json", state)
+    save(out / "acquisition.json", state)
+
+
+def prepare_manual(out):
+    pool = read_lines(out / "selection_pool.jsonl")
+    state = manual_acquisition(out)
+    completed = {record["item_id"] for record in state["selected"]}
+    completed.update(record["item_id"] for record in state["rejected"])
+    for item in pool:
+        if item["item_id"] in completed:
+            continue
+        try:
+            streams = text_stream_metadata(item)
+            text_stream = streams["text_stream"]
+            text_bytes = int(text_stream.get("sizeBytes", 0))
+            if text_bytes <= 0 or text_bytes > MAX_TEXT_BYTES:
+                raise ValueError(f"TEXT bitstream outside 1..{MAX_TEXT_BYTES} byte pilot bound")
+            if text_bytes < MIN_TEXT_CHARS:
+                raise ValueError(f"TEXT bitstream has {text_bytes} bytes and cannot pass the {MIN_TEXT_CHARS}-character QA gate")
+        except Exception as error:
+            state["rejected"].append({"item_id": item["item_id"], "title": item["title"],
+                                      "reason": str(error), "observed_at": now()})
+            sync_manual_acquisition(out, state)
+            continue
+        pending = {
+            "source": SOURCE, "prepared_at": now(), "item": item,
+            "landing_url": f"https://open.icm.edu.pl/items/{item['item_id']}",
+            "accepted_license_labels": sorted(set(values(item["metadata"], "dc.rights")) & LICENSES.keys()),
+            "text_bitstream": bitstream_summary(text_stream),
+            "license_evidence": [{**bitstream_summary(stream), "downloaded": False}
+                                 for stream in streams["license_streams"]],
+            "original_bitstreams": [bitstream_summary(stream) for stream in streams["original_streams"]],
+            "progress": {"acquired": len(state["selected"]), "target": state["target"]},
+        }
+        save(out / "manual_pending.json", pending)
+        print(json.dumps({
+            "item_id": item["item_id"], "title": item["title"],
+            "content_url": pending["text_bitstream"]["content_url"],
+            "bytes": pending["text_bitstream"]["bytes"],
+            "acquired": len(state["selected"]), "target": state["target"],
+        }, ensure_ascii=False, indent=2))
+        return
+    raise RuntimeError("manual acquisition pool exhausted")
+
+
+def ingest_manual(out, download_file):
+    if not download_file:
+        raise ValueError("--download-file is required for ingest_manual")
+    pending = load(out / "manual_pending.json")
+    stream = pending["text_bitstream"]
+    payload = download_file.read_bytes()
+    prefix = payload[:512].lower()
+    if b"<html" in prefix or b"<!doctype html" in prefix:
+        raise AcquisitionBarrier("downloaded file is HTML rather than the declared TEXT bitstream")
+    if len(payload) != int(stream["bytes"]):
+        raise ValueError(f"downloaded file has {len(payload)} bytes; expected {stream['bytes']}")
+    checksum = stream.get("checksum") or {}
+    if str(checksum.get("checkSumAlgorithm", "")).upper() == "MD5":
+        actual_md5 = hashlib.md5(payload).hexdigest()
+        if actual_md5 != checksum.get("value"):
+            raise ValueError(f"downloaded file MD5 {actual_md5} differs from repository metadata")
+    state = manual_acquisition(out)
+    item = pending["item"]
+    if any(record["item_id"] == item["item_id"] for record in state["selected"]):
+        raise ValueError("pending item is already present in manual acquisition")
+    raw_path = out / "raw_text" / f"{item['item_id']}.txt"
+    raw_path.parent.mkdir(parents=True, exist_ok=True)
+    raw_path.write_bytes(payload)
+    record = {
+        **item, "landing_url": pending["landing_url"],
+        "accepted_license_labels": pending["accepted_license_labels"],
+        "text_bitstream": {**stream, "sha256": digest(payload),
+                           "path": raw_path.relative_to(out).as_posix()},
+        "license_evidence": pending["license_evidence"],
+        "original_bitstreams": pending["original_bitstreams"],
+        "observed_at": now(), "acquisition_mode": "manual_browser_download",
+        "download_filename": download_file.name,
+    }
+    state["selected"].append(record)
+    state["selected"].sort(key=lambda row: row["item_id"])
+    sync_manual_acquisition(out, state)
+    pending["ingested_at"] = now()
+    pending["download_sha256"] = digest(payload)
+    save(out / "manual_pending.json", pending)
+    print(json.dumps({"ingested": item["item_id"], "acquired": len(state["selected"]),
+                      "target": state["target"], "target_reached": state["target_reached"]},
+                     ensure_ascii=False, indent=2))
+
+
+def acquire(out, workers):
+    selection = load(out / "selection.json")
+    pool = read_lines(out / "selection_pool.jsonl")
+    target = int(selection["target"])
+    selected, rejected = [], []
+    try:
+        probe_download_access(pool[0])
+    except AcquisitionBarrier as error:
+        acquisition = {
+            "source": SOURCE, "observed_at": now(), "target": target, "pool_size": len(pool),
+            "attempted": 1, "selected": [],
+            "rejected": [{"item_id": pool[0]["item_id"], "title": pool[0]["title"], "reason": str(error)}],
+            "target_reached": False, "blocked": True,
+            "blocker": {
+                "type": "interactive_captcha",
+                "scope": "bitstream_content",
+                "resolution": "Obtain an official bulk-download route or written API access from OPEN ICM.",
+            },
+        }
+        save(out / "acquisition.json", acquisition)
+        print(json.dumps({"selected": 0, "rejected": 1, "target_reached": False,
+                          "blocked": True, "reason": str(error)}, ensure_ascii=False, indent=2))
+        return
+    next_index = 0
+    while len(selected) < target and next_index < len(pool):
+        batch = pool[next_index:min(len(pool), next_index + max(workers * 2, target - len(selected)))]
+        next_index += len(batch)
+        with ThreadPoolExecutor(max_workers=workers) as executor:
+            futures = {executor.submit(acquire_one, out, item): item for item in batch}
+            for future in as_completed(futures):
+                item = futures[future]
+                try:
+                    record = future.result()
+                    if len(selected) < target:
+                        selected.append(record)
+                        print(f"Acquired {len(selected)}/{target}: {record['title']}", flush=True)
+                    else:
+                        Path(out / record["text_bitstream"]["path"]).unlink(missing_ok=True)
+                        for evidence in record["license_evidence"]:
+                            Path(out / evidence["path"]).unlink(missing_ok=True)
+                except Exception as error:
+                    rejected.append({"item_id": item["item_id"], "title": item["title"], "reason": str(error)})
+        if len(selected) >= target:
+            break
+    selected.sort(key=lambda item: item["item_id"])
+    acquisition = {
+        "source": SOURCE, "observed_at": now(), "target": target, "pool_size": len(pool),
+        "attempted": len(selected) + len(rejected), "selected": selected, "rejected": rejected,
+        "target_reached": len(selected) == target,
+    }
+    save(out / "acquisition.json", acquisition)
+    print(json.dumps({"selected": len(selected), "rejected": len(rejected), "target_reached": acquisition["target_reached"]}, ensure_ascii=False, indent=2))
+
+
+def normalize(text):
+    text = unicodedata.normalize("NFKC", text or "").replace("\u00ad", "").replace("\u200b", "")
+    text = re.sub(r"[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]", "", text)
+    lines = [re.sub(r"[ \t\xa0]+", " ", line).strip() for line in text.splitlines()]
+    lines = [line for line in lines if not re.fullmatch(r"\d{1,4}", line)]
+    text = "\n".join(lines)
+    text = re.sub(r"(?<=\w)-\n(?=[a-ząćęłńóśźż])", "", text)
+    text = re.sub(r"(?<![.!?:;\n])\n(?!\n)(?=[a-ząćęłńóśźż])", " ", text)
+    return re.sub(r"\n{3,}", "\n\n", text).strip()
+
+
+def normalize_title(text):
+    text = unicodedata.normalize("NFKD", text or "").casefold()
+    text = "".join(character for character in text if not unicodedata.combining(character))
+    return " ".join(re.findall(r"\w+", text))
+
+
+def shingle_sketch(text, limit=5_000):
+    words = re.findall(r"\w+", text.casefold())
+    hashes = set()
+    for index in range(max(0, len(words) - 4)):
+        value = " ".join(words[index:index + 5]).encode("utf-8")
+        hashes.add(int.from_bytes(hashlib.blake2b(value, digest_size=8).digest(), "big"))
+    if len(hashes) > limit:
+        return set(sorted(hashes)[:limit])
+    return hashes
+
+
+def language_vote(identifier, text):
+    chunks = [text[:30_000], text[max(0, len(text) // 2 - 15_000):len(text) // 2 + 15_000], text[-30_000:]]
+    votes = [identifier.classify(chunk) for chunk in chunks if chunk.strip()]
+    languages = Counter(language for language, _ in votes)
+    return (languages.most_common(1)[0][0] if languages else "unknown", votes)
+
+
+def build(out):
+    import pyarrow as pa
+    import pyarrow.parquet as pq
+    import tiktoken
+    from langid.langid import LanguageIdentifier, model
+
+    acquisition = load(out / "acquisition.json")
+    selection = load(out / "selection.json")
+    encoder = tiktoken.get_encoding("cl100k_base")
+    identifier = LanguageIdentifier.from_modelstring(model, norm_probs=True)
+    identifier.set_languages(["pl", "en", "de", "uk", "ru"])
+    rows, attribution, decisions, exact_seen, sketches = [], [], [], {}, {}
+    pii = Counter()
+    added = acquisition["observed_at"][:10]
+    for record in acquisition["selected"]:
+        raw_path = out / record["text_bitstream"]["path"]
+        raw = raw_path.read_bytes()
+        if digest(raw) != record["text_bitstream"]["sha256"]:
+            raise ValueError("TEXT checksum mismatch: " + record["item_id"])
+        text = normalize(raw.decode("utf-8-sig", errors="replace"))
+        replacement_count = text.count("\ufffd")
+        letters = len(re.findall(r"[A-Za-zĄĆĘŁŃÓŚŹŻąćęłńóśźż]", text))
+        language, votes = language_vote(identifier, text)
+        reason = ""
+        if len(text) < MIN_TEXT_CHARS:
+            reason = "too_little_extractable_text"
+        elif letters / max(len(text), 1) < 0.55:
+            reason = "low_letter_ratio"
+        elif replacement_count > 20 or replacement_count / max(len(text), 1) > 0.0001:
+            reason = "excessive_replacement_characters"
+        elif language != "pl":
+            reason = "non_polish_text"
+        text = text.replace("\ufffd", "[UNREADABLE_GLYPH]")
+        text, emails = EMAIL_RE.subn("[REDACTED:EMAIL]", text)
+        text, phones = PHONE_RE.subn("[REDACTED:PHONE]", text)
+        pii.update(email=emails, labelled_phone=phones)
+        exact_key = digest(" ".join(text.casefold().split()).encode("utf-8"))
+        duplicate_of, duplicate_score = None, 0.0
+        if not reason and exact_key in exact_seen:
+            reason, duplicate_of, duplicate_score = "normalized_duplicate", exact_seen[exact_key], 1.0
+        sketch = shingle_sketch(text)
+        if not reason:
+            for other_id, other_sketch in sketches.items():
+                score = len(sketch & other_sketch) / max(len(sketch | other_sketch), 1)
+                if score >= 0.90:
+                    reason, duplicate_of, duplicate_score = "near_duplicate", other_id, score
+                    break
+        row_id = f"{SOURCE}_{record['item_id']}"
+        decision = {
+            "id": row_id, "selected": not bool(reason), "reason": reason or "include",
+            "characters": len(text), "letter_ratio": letters / max(len(text), 1),
+            "replacement_characters": replacement_count, "language": language,
+            "language_votes": [{"language": lang, "confidence": float(score)} for lang, score in votes],
+        }
+        if duplicate_of:
+            decision.update({"duplicate_of": duplicate_of, "jaccard": duplicate_score})
+        decisions.append(decision)
+        if reason:
+            continue
+        exact_seen[exact_key] = row_id
+        sketches[row_id] = sketch
+        metadata = record["metadata"]
+        authors = values(metadata, "dc.contributor.author") or values(metadata, "dc.contributor.editor")
+        author = "; ".join(authors or ["Unknown"])
+        issued = values(metadata, "dc.date.issued")
+        accepted_labels = record["accepted_license_labels"]
+        spdx = LICENSES[accepted_labels[0]]
+        row = {
+            "id": row_id, "text": text, "source": SOURCE, "added": added,
+            "created": issued[0] if issued else "unknown", "token_count": len(encoder.encode_ordinary(text)),
+            "license": spdx, "author": author,
+        }
+        rows.append(row)
+        attribution.append({
+            "id": row_id, "item_id": record["item_id"], "title": record["title"], "authors": authors,
+            "publisher": values(metadata, "dc.publisher"), "issued": issued,
+            "doi": values(metadata, "dc.identifier.doi"), "handle": record["handle"],
+            "landing_url": record["landing_url"], "repository_license_labels": accepted_labels,
+            "license": spdx, "text_bitstream": record["text_bitstream"],
+            "license_evidence": record["license_evidence"], "original_bitstreams": record["original_bitstreams"],
+            "text_sha256": digest(text.encode("utf-8")),
+            "transformations": ["repository TEXT bundle", "Unicode/whitespace normalization", "page-number-only removal",
+                                "line-wrap repair", "email and labelled-phone pattern redaction"],
+        })
+    root = out / "hf_repo"
+    (root / "data").mkdir(parents=True, exist_ok=True)
+    (root / "artifacts" / "source_pages").mkdir(parents=True, exist_ok=True)
+    schema = pa.schema([(field, pa.int64() if field == "token_count" else pa.string()) for field in FIELDS])
+    pq.write_table(pa.Table.from_pylist(rows, schema=schema), root / "data/train-00000-of-00001.parquet", compression="zstd")
+    write_lines(root / "artifacts/attribution.jsonl", attribution)
+    write_lines(root / "artifacts/decisions.jsonl", decisions)
+    write_lines(root / "artifacts/source_manifest.jsonl", read_lines(out / "source_manifest.jsonl"))
+    sample = sorted(rows, key=lambda row: digest(("sample:" + row["id"]).encode("utf-8")))[:12]
+    write_lines(root / "artifacts/sample.jsonl", sample)
+    save(root / "artifacts/selection.json", selection)
+    save(root / "artifacts/acquisition.json", acquisition)
+    save(root / "artifacts/api_pages.json", load(out / "api_pages.json"))
+    for source_page in sorted((out / "raw_api").rglob("*.json.gz")):
+        destination = root / "artifacts" / "source_pages" / source_page.relative_to(out / "raw_api")
+        destination.parent.mkdir(parents=True, exist_ok=True)
+        destination.write_bytes(source_page.read_bytes())
+    overlap = load(out / "overlap_audit.json") if (out / "overlap_audit.json").exists() else None
+    target_audit = load(out / "target_audit.json") if (out / "target_audit.json").exists() else None
+    if overlap:
+        save(root / "artifacts/overlap_audit.json", overlap)
+    if target_audit:
+        save(root / "artifacts/target_audit.json", target_audit)
+    stats = {
+        "eligible_source_records": selection["eligible_records"], "selection_pool": selection["pool_size"],
+        "pilot_target": acquisition["target"], "acquired": len(acquisition["selected"]),
+        "kept": len(rows), "rejected_by_text_qa": len(decisions) - len(rows),
+        "tokens": sum(row["token_count"] for row in rows), "characters": sum(len(row["text"]) for row in rows),
+        "author_coverage": sum(row["author"] != "Unknown" for row in rows) / len(rows) if rows else 0,
+        "license_counts": dict(Counter(row["license"] for row in rows)), "sample_count": len(sample), "added": added,
+    }
+    qa = {
+        "scope": "deterministic 100-record pilot, not a complete-corpus claim",
+        "license_gate": "repository item metadata is an allowlisted CC BY or CC BY-SA label",
+        "language_gate": "dc.language.iso:pl plus independent three-window langid vote",
+        "pii_pattern_matches": dict(pii), "exact_dedup": True,
+        "near_dedup": "deterministic capped 5-word-shingle hash Jaccard >= 0.90 within pilot",
+        "biblioteka_nauki_overlap": overlap or "pending", "cross_source_text_dedup": "pending target integration",
+        "benchmark_overlap": "pending", "limitations": [
+            "repository-generated TEXT may inherit PDF extraction errors",
+            "figures are omitted and tables may be flattened",
+            "item-level license metadata does not prove every quoted third-party passage is reusable",
+            "pattern checks are not comprehensive de-identification",
+            "pilot yield and token counts must not be extrapolated to all 6183 eligible records",
+        ],
+    }
+    save(root / "artifacts/stats.json", stats)
+    save(root / "artifacts/qa.json", qa)
+    protocol_id = "protocol:open-icm-pilot-v1"
+    run = {
+        "id": "run:" + digest({"script": digest(Path(__file__).read_bytes()), "selection": selection,
+                                  "acquisition": digest(acquisition)}),
+        "protocol": protocol_id, "started_at": selection["observed_at"], "finished_at": now(),
+        "success": True, "actor": "actor:codex", "stats": stats,
+    }
+    save(root / "artifacts/run.json", run)
+    excluded = {"README.md", "NOTICE.md", "artifacts/checksums.json", "artifacts/ontology.json"}
+    checks = {path.relative_to(root).as_posix(): digest(path.read_bytes()) for path in sorted(root.rglob("*"))
+              if path.is_file()
+              and path.relative_to(root).as_posix() not in excluded
+              and not path.relative_to(root).as_posix().startswith("src/")}
+    save(root / "artifacts/checksums.json", checks)
+    source_version = "version:source:" + digest({"manifest": selection["source_manifest_sha256"],
+                                                  "api_pages": selection["api_pages_sha256"]})
+    dataset_version = "version:dataset:" + digest(checks)
+    selection_evidence = "evidence:selection:" + digest(selection)
+    acquisition_evidence = "evidence:acquisition:" + digest(acquisition)
+    qa_evidence = "evidence:qa:" + digest(qa)
+    evidence = [
+        {"id": selection_evidence, "observation_type": "source_inventory_and_selection",
+         "artifact": "artifacts/selection.json", "content_address": digest(selection), "produced_by": run["id"]},
+        {"id": acquisition_evidence, "observation_type": "bitstream_acquisition",
+         "artifact": "artifacts/acquisition.json", "content_address": digest(acquisition), "produced_by": run["id"]},
+        {"id": qa_evidence, "observation_type": "pilot_qa", "artifact": "artifacts/qa.json",
+         "content_address": digest(qa), "produced_by": run["id"]},
+    ]
+    overlap_evidence = None
+    if overlap:
+        overlap_evidence = "evidence:overlap:" + digest(overlap)
+        evidence.append({"id": overlap_evidence, "observation_type": "metadata_overlap_audit",
+                         "artifact": "artifacts/overlap_audit.json", "content_address": digest(overlap),
+                         "produced_by": run["id"]})
+    target_evidence = None
+    if target_audit:
+        target_evidence = "evidence:target:" + digest(target_audit)
+        evidence.append({"id": target_evidence, "observation_type": "target_registry_audit",
+                         "artifact": "artifacts/target_audit.json", "content_address": digest(target_audit),
+                         "produced_by": run["id"]})
+    ontology = {
+        "schema": "slayer-research-ontology-profile-v1",
+        "objects": [{"id": "object:source:open-icm", "type": "Source"},
+                    {"id": "object:dataset:open-icm-pl-pilot", "type": "Dataset"}],
+        "versions": [{"id": source_version, "object": "object:source:open-icm",
+                      "content_address": source_version.rsplit(":", 1)[-1]},
+                     {"id": dataset_version, "object": "object:dataset:open-icm-pl-pilot",
+                      "content_address": dataset_version.rsplit(":", 1)[-1]}],
+        "protocols": [{"id": protocol_id, "procedure": "pinned discovery; allowlisted item license; Polish metadata and language vote; TEXT checksum; normalization; PII patterns; exact and near dedup"}],
+        "runs": [run], "evidence": evidence,
+        "claims": [
+            {"id": "claim:eligible-count-observed", "statement": f"The pinned discovery run observed {selection['eligible_records']} unique Polish records carrying an allowlisted CC BY or CC BY-SA label.",
+             "supported_by": [selection_evidence], "falsification_condition": "The preserved API pages and manifest do not reproduce the count and filters."},
+            {"id": "claim:pilot-retention", "statement": f"The pilot retained {stats['kept']} records after text QA and within-pilot deduplication.",
+             "supported_by": [acquisition_evidence, qa_evidence], "falsification_condition": "The decisions, Parquet rows, or checksums do not reproduce the retention count."},
+            {"id": "claim:source-absence-at-audit", "statement": "OPEN ICM was not registered as a source in the pinned DynaWord data tree or open pull-request list at audit time.",
+             "supported_by": [target_evidence] if target_evidence else [qa_evidence], "falsification_condition": "The pinned target evidence contains an OPEN ICM source or matching proposal."},
+            {"id": "claim:training-value-untested", "statement": "Net corpus novelty and training benefit remain untested hypotheses.",
+             "supported_by": [qa_evidence] + ([overlap_evidence] if overlap_evidence else []),
+             "falsification_condition": "Target-wide text deduplication and controlled ablations establish those properties."},
+        ],
+        "actors": [{"id": "actor:piotrsty", "type": "Contributor"},
+                   {"id": "actor:icm-uw", "type": "Organization"}, {"id": "actor:codex", "type": "Agent"}],
+        "relations": [{"source": dataset_version, "predicate": "DERIVED_FROM", "target": source_version},
+                      {"source": dataset_version, "predicate": "GENERATED_BY", "target": run["id"]}] +
+                     ([{"source": dataset_version, "predicate": "VALIDATED_AGAINST",
+                        "target": f"hf:dataset:{TARGET}@{target_audit['revision']}"}] if target_audit else []),
+        "pending": ["complete-corpus acquisition", "cross-source text deduplication", "benchmark contamination check",
+                    "review of third-party quoted text", "controlled training ablation"],
+    }
+    save(root / "artifacts/ontology.json", ontology)
+    card = f"""---
+license: other
+license_name: per-record-cc-by-or-cc-by-sa
+language:
+- pl
+task_categories:
+- text-generation
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-00000-of-00001.parquet
+---
+
+# OPEN ICM Polish publications pilot
+
+A deterministic 100-record research pilot of Polish full-text publications
+whose OPEN ICM item metadata declares an allowlisted CC BY or CC BY-SA license.
+This is not a complete-corpus release.
+
+- Acquired records: {stats['acquired']}
+- Retained after text QA and within-source deduplication: {stats['kept']}
+- Characters: {stats['characters']:,}
+- Tokens: {stats['tokens']:,} (`cl100k_base` proxy)
+- Author coverage: {stats['author_coverage']:.1%}
+- License field: preserved per record
+
+## Provenance and rights
+
+Every row is linked to its OPEN ICM item, authors, repository license label,
+normalized license identifier, selected TEXT bitstream metadata and checksums.
+The accepted source labels are limited to CC BY and CC BY-SA variants. License
+and attribution obligations remain applicable per record. Item-level licensing
+does not establish the status of every quoted third-party passage.
+
+## Processing and limitations
+
+The repository-generated TEXT bundle is normalized for Unicode, whitespace,
+page-number-only lines and wrapped lines. Email and labelled-phone patterns are
+redacted. Language is checked independently across three text windows. Exact
+and deterministic near deduplication are performed within the pilot.
+
+PDF-derived text may contain extraction errors; figures are omitted and tables
+may be flattened. Pattern checks are not comprehensive de-identification.
+Cross-source text deduplication, benchmark contamination checks, quoted-text
+review and controlled training ablations remain downstream gates.
+
+## Review artifacts
+
+See `artifacts/sample.jsonl`, `attribution.jsonl`, `decisions.jsonl`,
+`overlap_audit.json`, `stats.json`, `qa.json`, `checksums.json`, `run.json` and
+`ontology.json`. The ontology separates content-addressed Objects and Versions,
+Protocols, actual Runs, Evidence, falsifiable Claims, Actors and typed lineage.
+"""
+    (root / "README.md").write_text(card, encoding="utf-8")
+    (root / "NOTICE.md").write_text(
+        "# Attribution and license notice\n\n"
+        "Source: OPEN ICM, https://open.icm.edu.pl/home\n\n"
+        "Each record retains its item URL, authorship, repository license label, normalized license identifier, "
+        "bitstream metadata and license evidence where exposed by OPEN ICM. The applicable CC BY or CC BY-SA "
+        "terms and attribution obligations apply per record; consult `artifacts/attribution.jsonl`.\n\n"
+        "Preparation: Piotr Styla with OpenAI Codex. Changes: repository TEXT selection, Unicode and whitespace "
+        "normalization, page-number-only removal, line-wrap repair, limited email and labelled-phone redaction, "
+        "language/quality filtering and within-source deduplication. No endorsement by OPEN ICM or credited "
+        "authors is implied.\n",
+        encoding="utf-8",
+    )
+    print(json.dumps(stats, ensure_ascii=False, indent=2))
+
+
+def audit_target(out):
+    info = request_json(f"https://huggingface.co/api/datasets/{TARGET}")
+    revision = info["sha"]
+    tree = request_json(f"https://huggingface.co/api/datasets/{TARGET}/tree/{revision}",
+                        {"recursive": "true", "expand": "false"})
+    discussions = request_json(f"https://huggingface.co/api/datasets/{TARGET}/discussions",
+                               {"status": "open", "p": 0})
+    paths = sorted(item.get("path", "") for item in tree)
+    open_rows = [{"num": item.get("num"), "title": item.get("title"), "status": item.get("status"),
+                  "author": item.get("author", {}).get("name")} for item in discussions.get("discussions", [])]
+    matches = [path for path in paths if any(term in path.casefold() for term in ("open_icm", "ceon"))]
+    discussion_matches = [row for row in open_rows if any(term in (row.get("title") or "").casefold()
+                                                          for term in ("open icm", "ceon"))]
+    report = {
+        "target": TARGET, "revision": revision, "last_modified": info.get("lastModified"),
+        "tree_paths": len(paths), "source_path_matches": matches, "open_discussions": open_rows,
+        "matching_open_discussions": discussion_matches, "source_absent": not matches and not discussion_matches,
+        "observed_at": now(),
+    }
+    save(out / "target_audit.json", report)
+    print(json.dumps({"revision": revision, "source_absent": report["source_absent"],
+                      "tree_matches": matches, "discussion_matches": discussion_matches}, ensure_ascii=False, indent=2))
+
+
+def audit_overlap(out):
+    import pyarrow.parquet as pq
+    from huggingface_hub import HfApi, HfFileSystem
+
+    acquisition = load(out / "acquisition.json")
+    revision = HfApi().dataset_info(TARGET).sha
+    remote = f"datasets/{TARGET}@{revision}/data/biblioteka_nauki/biblioteka_nauki.parquet"
+    with HfFileSystem().open(remote, "rb") as handle:
+        table = pq.read_table(handle, columns=["id", "attribution"])
+    target = []
+    for row in table.to_pylist():
+        attribution = row.get("attribution") or ""
+        parts = attribution.split(" | ")
+        title = parts[2] if len(parts) >= 4 else attribution
+        target.append((row["id"], attribution, normalize_title(title)))
+    target_by_title = {}
+    for row_id, attribution, normalized in target:
+        if normalized:
+            target_by_title.setdefault(normalized, []).append((row_id, attribution))
+    target_titles = list(target_by_title)
+    results = []
+    for record in acquisition["selected"]:
+        title = normalize_title(record["title"])
+        dois = [value.casefold() for value in values(record["metadata"], "dc.identifier.doi")]
+        exact = [{"id": row_id, "attribution": attribution} for row_id, attribution, normalized in target
+                 if (title and title in normalized) or any(doi in attribution.casefold() for doi in dois)]
+        fuzzy = []
+        if not exact and title:
+            matches = get_close_matches(title, target_titles, n=3, cutoff=0.90)
+            for normalized in matches:
+                score = SequenceMatcher(None, title, normalized).ratio()
+                fuzzy.extend({"score": score, "id": row_id, "attribution": attribution}
+                             for row_id, attribution in target_by_title[normalized])
+            fuzzy = fuzzy[:3]
+        results.append({"item_id": record["item_id"], "title": record["title"], "doi": dois,
+                        "exact_title_or_doi_matches": exact, "fuzzy_title_matches": fuzzy})
+    report = {
+        "target": f"{TARGET}:data/biblioteka_nauki", "target_revision": revision,
+        "method": "DOI or normalized title substring; fallback SequenceMatcher >= 0.90 over attribution titles",
+        "target_rows": table.num_rows, "candidate_records": len(results),
+        "records_with_exact_match": sum(bool(row["exact_title_or_doi_matches"]) for row in results),
+        "records_with_fuzzy_match": sum(bool(row["fuzzy_title_matches"]) for row in results),
+        "text_overlap": "not tested; target-wide text dedup remains an integration gate",
+        "observed_at": now(), "results": results,
+    }
+    save(out / "overlap_audit.json", report)
+    print(json.dumps({key: report[key] for key in ("target_revision", "target_rows", "candidate_records",
+                                                    "records_with_exact_match", "records_with_fuzzy_match")},
+                     ensure_ascii=False, indent=2))
+
+
+def verify(out):
+    import pyarrow.parquet as pq
+
+    root = out / "hf_repo"
+    table = pq.read_table(root / "data/train-00000-of-00001.parquet")
+    rows = table.to_pylist()
+    stats = load(root / "artifacts/stats.json")
+    decisions = read_lines(root / "artifacts/decisions.jsonl")
+    attribution = read_lines(root / "artifacts/attribution.jsonl")
+    sample = read_lines(root / "artifacts/sample.jsonl")
+    assert table.column_names == FIELDS
+    assert len(rows) == stats["kept"] == len(attribution)
+    assert len(decisions) == stats["acquired"]
+    assert sum(item["selected"] for item in decisions) == len(rows)
+    assert sum(row["token_count"] for row in rows) == stats["tokens"]
+    assert all(row["source"] == SOURCE and row["license"] in LICENSES.values() for row in rows)
+    assert all(EMAIL_RE.search(row["text"]) is None for row in rows)
+    by_id = {row["id"]: row for row in rows}
+    assert len(sample) == stats["sample_count"] and all(by_id[row["id"]] == row for row in sample)
+    ontology = load(root / "artifacts/ontology.json")
+    evidence = {item["id"] for item in ontology["evidence"]}
+    assert all(item["falsification_condition"] and set(item["supported_by"]) <= evidence for item in ontology["claims"])
+    checks = load(root / "artifacts/checksums.json")
+    assert all((root / path).is_file() and digest((root / path).read_bytes()) == checksum for path, checksum in checks.items())
+    for item in ontology["evidence"]:
+        artifact = root / item["artifact"]
+        assert artifact.is_file() and digest(load(artifact)) == item["content_address"]
+    print(json.dumps({"verified": True, **stats}, ensure_ascii=False, indent=2))
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--output", type=Path, required=True)
+    parser.add_argument("--limit", type=int, default=100)
+    parser.add_argument("--workers", type=int, default=8)
+    parser.add_argument("--download-file", type=Path)
+    parser.add_argument("command", choices=["discover", "acquire", "prepare_manual", "ingest_manual",
+                                            "audit_target", "audit_overlap", "build", "verify"])
+    args = parser.parse_args()
+    if args.command == "discover":
+        discover(args.output, args.limit)
+    elif args.command == "acquire":
+        acquire(args.output, args.workers)
+    elif args.command == "prepare_manual":
+        prepare_manual(args.output)
+    elif args.command == "ingest_manual":
+        ingest_manual(args.output, args.download_file)
+    elif args.command == "audit_target":
+        audit_target(args.output)
+    elif args.command == "audit_overlap":
+        audit_overlap(args.output)
+    elif args.command == "build":
+        build(args.output)
+    else:
+        verify(args.output)
+
+
+if __name__ == "__main__":
+    main()

src/open_icm_requirements.txt

@@ -0,0 +1,6 @@
+requests==2.32.5
+pyarrow==23.0.1
+tiktoken==0.12.0
+langid==1.1.6
+huggingface_hub>=1.0,<2
+pytest>=8,<10

src/sources.py

@@ -601,6 +601,21 @@ SOURCES = {
                       "records should remain separately labeled or held until "
                       "legal review confirms their US status.",
     },
+    "open_icm_pl": {
+        "added": "2026-09-10",
+        "release": None,
+        "file_key": "open_icm_pl",
+        "pretty": "OPEN ICM Polish open-access publications pilot",
+        "license": "per-record CC BY or CC BY-SA",
+        "license_spdx": "MIXED",
+        "traceable": "Every retained record carries an allowlisted item-level CC BY or CC BY-SA label; original item URLs, authors, normalized license identifiers, TEXT bitstream checksums and available license evidence are preserved per record.",
+        "upstream": "https://open.icm.edu.pl/home",
+        "provenance": "Discovered directly through the official OPEN ICM DSpace API. Official repository-generated TEXT bitstreams were downloaded after the repository's interactive CAPTCHA and verified against declared sizes/checksums. Unicode/whitespace normalization, limited pattern redaction, language/quality filtering and within-source deduplication are documented in the committed artifacts.",
+        "domain": "academic/mixed",
+        "created": "1998-2024",
+        "is_ocr": True,
+        "custom_datasheet": True,
+    },
     "nkjp1m": {
         "added": "2026-06-15",
         "custom_datasheet": True,

src/test_open_icm_pl_contract.py

@@ -0,0 +1,116 @@
+import importlib.util
+from pathlib import Path
+
+
+SCRIPT = Path(__file__).with_name("fetch_open_icm_pl.py")
+SPEC = importlib.util.spec_from_file_location("fetch_open_icm_pl", SCRIPT)
+MODULE = importlib.util.module_from_spec(SPEC)
+SPEC.loader.exec_module(MODULE)
+
+
+def metadata(**entries):
+    return {key: [{"value": value}] for key, value in entries.items()}
+
+
+def test_license_gate_is_fail_closed():
+    valid = {"metadata": metadata(**{"dc.language.iso": "pl", "dc.rights": "Uznanie autorstwa 4.0 Międzynarodowe"})}
+    assert MODULE.is_polish_open_item(valid)
+    assert not MODULE.is_polish_open_item({"metadata": metadata(**{"dc.language.iso": "en", "dc.rights": "Uznanie autorstwa 4.0 Międzynarodowe"})})
+    assert not MODULE.is_polish_open_item({"metadata": metadata(**{"dc.language.iso": "pl", "dc.rights": "Dozwolony użytek"})})
+    assert not MODULE.is_polish_open_item({"metadata": metadata(**{"dc.language.iso": "pl", "dc.rights": "Uznanie autorstwa-Użycie niekomercyjne 4.0 Międzynarodowe"})})
+
+
+def test_deterministic_pool_covers_each_available_license():
+    rows = []
+    for index, label in enumerate(MODULE.LICENSES):
+        rows.append({"item_id": str(index), "title": str(index), "metadata": metadata(**{"dc.rights": label})})
+    rows.extend({"item_id": f"extra-{index}", "title": str(index),
+                 "metadata": metadata(**{"dc.rights": next(iter(MODULE.LICENSES))})} for index in range(20))
+    first = MODULE.deterministic_pool(rows, 10)
+    second = MODULE.deterministic_pool(list(reversed(rows)), 10)
+    assert [row["item_id"] for row in first] == [row["item_id"] for row in second]
+    covered = {label for row in first for label in MODULE.values(row["metadata"], "dc.rights")}
+    assert covered == set(MODULE.LICENSES)
+
+
+def test_normalize_repairs_hyphenation_and_page_numbers():
+    assert MODULE.normalize("Pierw-\nszy akapit.\n12\n\ndrugi wiersz") == "Pierwszy akapit.\n\ndrugi wiersz"
+
+
+def test_title_normalization_is_case_and_diacritic_insensitive():
+    assert MODULE.normalize_title("Ćwiczenia: Łódź") == "cwiczenia łodz"
+
+
+def test_text_bound_is_small_enough_for_pilot():
+    assert MODULE.MAX_TEXT_BYTES == 20 * 1024 * 1024
+    assert MODULE.MIN_TEXT_CHARS == 3_000
+
+
+def test_captcha_response_is_an_acquisition_barrier():
+    class Response:
+        url = "https://open.icm.edu.pl/captcha.html?protected=/bitstreams/example/download"
+        headers = {"Content-Type": "text/html; charset=UTF-8"}
+        content = b"<!doctype html><html><body>I'm not a robot</body></html>"
+
+    try:
+        MODULE.detect_download_barrier(Response())
+    except MODULE.AcquisitionBarrier as error:
+        assert "interactive CAPTCHA" in str(error)
+    else:
+        raise AssertionError("CAPTCHA response was not rejected")
+
+
+def test_plain_text_response_is_not_a_barrier():
+    class Response:
+        url = "https://open.icm.edu.pl/server/api/core/bitstreams/example/content"
+        headers = {"Content-Type": "text/plain"}
+        content = b"Tekst publikacji"
+
+    MODULE.detect_download_barrier(Response())
+
+
+def test_manual_ingest_verifies_and_records_download(tmp_path):
+    payload = "To jest poprawny tekst pilota.".encode("utf-8")
+    item = {
+        "item_id": "item-1", "title": "Tytul", "handle": "123/1",
+        "metadata": metadata(**{"dc.language.iso": "pl", "dc.rights": "Uznanie autorstwa 4.0 Międzynarodowe"}),
+    }
+    MODULE.save(tmp_path / "selection.json", {"target": 1, "pool_size": 2})
+    MODULE.save(tmp_path / "manual_pending.json", {
+        "item": item, "landing_url": "https://open.icm.edu.pl/items/item-1",
+        "accepted_license_labels": ["Uznanie autorstwa 4.0 Międzynarodowe"],
+        "text_bitstream": {
+            "id": "stream-1", "name": "text.txt", "bytes": len(payload),
+            "checksum": {"checkSumAlgorithm": "MD5", "value": MODULE.hashlib.md5(payload).hexdigest()},
+            "content_url": "https://open.icm.edu.pl/bitstreams/stream-1/download",
+        },
+        "license_evidence": [], "original_bitstreams": [],
+    })
+    download = tmp_path / "download.txt"
+    download.write_bytes(payload)
+
+    MODULE.ingest_manual(tmp_path, download)
+
+    acquisition = MODULE.load(tmp_path / "acquisition.json")
+    assert acquisition["target_reached"] is True
+    assert acquisition["selected"][0]["text_bitstream"]["sha256"] == MODULE.digest(payload)
+    assert (tmp_path / "raw_text/item-1.txt").read_bytes() == payload
+
+
+def test_manual_ingest_rejects_wrong_size(tmp_path):
+    MODULE.save(tmp_path / "selection.json", {"target": 1, "pool_size": 1})
+    MODULE.save(tmp_path / "manual_pending.json", {
+        "item": {"item_id": "item-1", "title": "Tytul", "handle": "123/1", "metadata": {}},
+        "landing_url": "https://open.icm.edu.pl/items/item-1", "accepted_license_labels": [],
+        "text_bitstream": {"bytes": 99, "checksum": {}, "content_url": "https://example.invalid"},
+        "license_evidence": [], "original_bitstreams": [],
+    })
+    download = tmp_path / "download.txt"
+    download.write_bytes(b"short")
+
+    try:
+        MODULE.ingest_manual(tmp_path, download)
+    except ValueError as error:
+        assert "expected 99" in str(error)
+    else:
+        raise AssertionError("wrong-size manual download was not rejected")